Assignment

Chapter 12Knowledge Systems: Expert Systems, Recommenders, Chatbots, Virtual Personal Assistants, and Robo Advisors

Learning Objectives

· Describe recommendation systems

· Describe expert systems

· Describe chatbots

· Understand the drivers and capabilities of chatbots and their use

· Describe virtual personal assistants and their benefits

· Describe the use of chatbots as advisors

· Discuss the major issues related to the implementation of chatbots

Advancement in artificial intelligence (AI) technologies and especially natural language processing (NLP), machine and deep learning and knowledge systems, coupled with the increased quality and functionalities of other intelligent systems, and mobile devices and their apps, have driven the development of chatbots (bots) for inexpensive and fast execution of many tasks related to communication, collaboration, and information retrieval. The use of chatbots in business is increasing rapidly, partly because of their fit with mobile systems and devices. As a matter of fact, sending messages is probably the major activity in the mobile world.

In the last two to three years, many thousands of bots have been placed into service worldwide by both organizations (private and public) and individuals. Many people refer to these phenomena as the chatbot revolution. Chatbots today are much more sophisticated than those of the past. They are extensively used, for example, in marketing; customer, government, and financial services; healthcare; and in manufacturing. Chatbots make communication more personal than faceless computers and excel in data gathering. Chatbots can stand alone or be parts of other knowledge systems.

We divide the applications in this chapter into four categories: expert systems, chatbots for communication and collaboration, virtual personal assistants (native products, such as Alexa), and chatbots that are used as professional advisors. Some implementation topics of intelligent systems are described last.

This chapter has the following sections:

1. 12.1 Opening Vignette: Sephora Excels with Chatbots 649

2. 12.2 Expert Systems and Recommenders 650

3. 12.3 Concepts, Drivers, and Benefits of Chatbots 660

4. 12.4 Enterprise Chatbots 664

5. 12.5 Virtual Personal Assistants 672

6. 12.6 Chatbots as Professional Advisors (Robo Advisors) 676

7. 12.7 Implementation Issues 680

12.1 Opening Vignette: Sephora Excels with Chatbots

The problem

Sephora is a French-based cosmetics/beauty products company doing business globally. It has its own stores and sells its goods in cosmetic and department stores. In addition, Sephora sells online on Amazon and on its online store. The company sells hundreds of brands, including many of its own. It operates in a very competitive market where customer care and advertising are critical. Sephora sells some products for men, but most beauty products are targeted to women.

The Solution

Sephora’s first use of chatbots occurred through messaging services. The purpose of the first bot was to search for information for the company’s resources such as videos, images, tips, and so on. This bot operates in a question-and-answer (Q&A) mode. It recommends relevant content based on customers’ interests. The company aims to appeal to young customers messaging on Kik.

Sephora researchers found that customers conversing with the Kikbot were engaged deeply in the dialog. Then the bot encouraged them to explore new products. Sephora’s newer bot called Reservation Assistant was placed on Facebook Messenger. It enables customers to book or reschedule makeover appointments.

Another Sephora bot delivered on Kik is Shade-Matching. It matches lips colors to photos (face and lips) uploaded by users and recommends the best match to them. The bot also lets users try on photos of recommended colors, using Sephora Virtual Artist that runs on Facebook Messenger. Bots are deployed as mobile apps. If users like the recommendation, they are directed to the company’s Web store to buy the products. Users can upload photos taken with selfies so that the program can do the matching. Over 4 million visitors tried 90 million shades in the first year of Virtual Artist’s operation.

The Q&A collection of the knowledge base was built by connecting it with store experts. Knowledge acquisition techniques ( Chapter 2 ) were used for this purpose. The company’s bots use NLPs that were trained to understand the typical vocabulary of users.

The Results

The company’s customers loved the bots. In addition, Sephora learned the importance of providing assistance and guidance to users who are motivated to return (at a reasonable cost!), happier, and more engaged.

Sephora’s bot asks users questions to find their tastes and preferences. Then it acts like a recommendation system ( Section 12.2 ), offering products. Kik and Messenger users can purchase items without leaving the messaging service.

Finally, the company has improved the bots’ knowledge over time and plans new bots for additional tasks.

NOTE:  Sephora was selected by Fast Company Magazine, March/April 2018, as one of the “World’s Most Innovative Companies.” Sephora is known for its digital transformation and innovation (Rayome, 2018). Also, Sephora’s bots are considered among the top marketing chatbots ( Quoc, 2017 ).

Sources:  Compiled from  Arthur (2016) , Rayome (2018), and  Taylor (2016) ,  theverge.com/2017/3/16/14946086/sephora-virtual-assistant-ios-app-update-ar-makeup/ , and  sephora.com/ .

Questions for the Opening Vignette

1. List and discuss the benefits of bots to the company.

2. List and discuss the benefits of bots to customers.

3. Why were the bots deployed via Messenger and Kik?

4. What would happen to Sephora if competitors use a similar approach?

What We Can Learn from This Vignette

In the highly competitive world of retail beauty products, customer care and marketing are critical. Using only live employees can be very expensive. In addition, customers are shopping 24/7,24/7, and physical stores are open during limited hours and days. In addition, there are large combinations of certain beauty products (e.g., many shades/colors) available. Sephora decided to use chatbots on Facebook Messenger and Kik to engage its customers. Chatbots, the subject of this chapter, are available 24/724/7 at a lower cost and are delivered via mobile devices. Bots deliver information to customers consistently and quickly direct customers to easy online shopping. Sephora placed its chatbots on messaging services. The logic was that people like to chat with friends on messaging services, and they may also like to chat with businesses.

In addition to several services to customers, using chatbots helps Sephora learn about customers. This type of chatbot is the most common type for customer care and marketing. In this chapter, we cover several other types of knowledge systems, including the pioneering expert systems, recommenders, virtual personal assistants offered by several large technology companies, and robo advisors.

12.2 Expert Systems and Recommenders

In Chapter 2 we introduced the reader to the concept of autonomous decision systems. An expert system is a category of autonomous decision systems and are considered the earliest applications of AI. Expert systems use started in research institutions in the early and mid-1960s (e.g., Stanford University, IBM) and was adopted commercially during the 1980s.

Basic Concepts of Expert Systems (ES)

The following are the major concepts related to ES technology.

Definitions

There are several definitions of expert systems. Our working definition is that an  expert system  is a computer-based system that emulates decision making and/or problem solving of human experts. These decisions and problems are in complex areas that require expertise to solve. The basic objective is to enable nonexperts to make decisions and solve problems that usually require expertise. This activity is usually performed in narrowly defined domains (e.g., making small loans, providing tax advice, analyzing reasons for machine failure). Classical ES use “what-if-then” rules for their reasoning.

Experts

An  expert  is a person who has the special knowledge, judgment, experience, and skills to provide sound advice and solve complex problems in a narrowly defined area. It is an expert’s job to provide the knowledge about how to perform a task so that a nonexpert will be able to do the same task assisted by ES. An expert knows which facts are important and understands and explains the dependent relationships among those facts. In diagnosing a problem with an automobile’s electrical system, for example, an expert car mechanic knows that a broken fan belt can be the cause for the battery to discharge.

There is no standard definition of expert, but decision performance and the level of knowledge a person has are typical criteria used to determine whether a particular person is an expert as related to ES. Typically, experts must be able to solve a problem and achieve a performance level that is significantly better than average. An expert at one time or in one region may not be an expert in another time or region. For example, a legal expert in New York may not be one in Beijing, China. A medical student may be an expert compared to the general public but not in making a diagnosis or performing surgery. Note that experts have expertise that can help solve problems and explain certain obscure phenomena only within a specific domain.

Typically, human experts are capable of doing the following:

· Recognizing and formulating a problem.

· Solving a problem quickly and correctly.

· Explaining a solution.

· Learning from experience.

· Restructuring knowledge.

· Breaking rules (i.e., going outside the general norms) if necessary.

· Determining relevance and associations.

Can a machine help a nonexpert perform like an expert? Can a machine make autonomous decisions that experts make? Let us see. But first, we need to explore what expertise is.

Expertise

An  expertise  is the extensive, task-specific knowledge that experts possess. The level of expertise determines the success of a decision made by an expert. Expertise is often acquired through training, learning, and experience in practice. It includes explicit knowledge, such as theories learned from a textbook or a classroom and implicit knowledge gained from experience. The following is a list of possible knowledge types used in ES applications:

· Theories about the problem domain.

· Rules and procedures regarding the general problem domain.

· Heuristics about what to do in a given problem situation.

· Global strategies for solving of problems amenable to expert systems.

· Meta knowledge (i.e., knowledge about knowledge).

· Facts about the problem area.

These types of knowledge enable experts to make better and faster decisions than nonexperts.

Expertise often includes the following characteristics:

· It is usually associated with a high degree of intelligence, but it is not always associated with the smartest person.

· It is usually associated with a vast quantity of knowledge.

· It is based on learning from past successes and mistakes.

· It is based on knowledge that is well stored, organized, and quickly retrievable from an expert who has excellent recall of patterns from previous experiences.

Characteristics and Benefits of ES

ES were used during the period 1980 to 2010 by hundreds of companies worldwide. However, since 2011, their use has declined rapidly, mostly due to the emergence of better knowledge systems, three types of which are described in this chapter. It is important, however, to understand the major characteristics and benefits of expert systems since many of them evolved evidenced newer knowledge systems.

The major objective of ES is the transfer of expertise to a machine. The expertise will be used by nonexperts. A typical example is a diagnosis. For example, many of us can use self-diagnosis to find (and correct) problems in our computers. Even more than that, computers can find and correct problems by themselves. One field in which such ability is practiced is medicine, as described in the following example:

Example: Are You Crazy?

A Web-based ES was developed in Korea for people to self-check their mental health status. Anyone in the world can access it and get a free evaluation. The knowledge for the system was collected from a survey of 3,235 Korean immigrants. The results of the survey were analyzed and then reviewed by experts via focus group discussions. For more information, see Bae (2013).

Benefits of ES

Depending on the mission and structure of ES, the following are their capabilities and potential benefits:

· Perform routine tasks (e.g., diagnosis, candidate screening, credit analysis) that require expertise much faster than humans.

· Reduce the cost of operations.

· Improve consistency and quality of work (e.g., reduce human errors).

· Speed up decision making and make consistent decisions.

· May motivate employees to increase productivity.

· Preserve scarce expertise of retiring employees.

· Help transfer and reuse knowledge.

· Reduce employee training cost by using self-training.

· Solve complex problems without experts and solve them faster.

· See things that even experts sometimes miss.

· Combine expertise of several experts.

· Centralize decision making (e.g., by using the “cloud”).

· Facilitate knowledge sharing.

These benefits can provide a significant competitive advantage to companies that use ES. Indeed, some companies have saved considerable amounts of money using them.

Despite these benefits, the use of ES is on the decline. The reasons for this and the related limitations are discussed later in this section.

Typical Areas for ES Applications

ES have been applied commercially in a number of areas, including the following:

· FINANCE. Finance ES include analysis of investments, credit, and financial reports; evaluation of insurance and performance; tax planning; fraud prevention; and financial planning.

· DATA PROCESSING. Data processing ES include system planning, equipment selection, equipment maintenance, vendor evaluation, and network management.

· MARKETING. Marketing ES include customer relationship management, market research and analysis, product planning, and market planning. Also, presale advice is provided for prospects.

· HUMAN RESOURCES. Examples of human resource ES are planning, performance evaluation, staff scheduling, pension management, regulatory advising, and design of questionnaires.

· MANUFACTURING. Manufacturing ES include production planning, complex product configuration, quality management, product design, plant site selection, and equipment maintenance and repair (including diagnosis).

· HOMELAND SECURITY. These ES include terrorist threat assessment and terrorist finance detection.

· BUSINESS PROCESS AUTOMATION. ES have been developed for desk automation, call center management, and regulation enforcement.

· HEALTHCARE MANAGEMENT. ES have been developed for bioinformatics and other healthcare management issues.

· REGULATORY AND COMPLIANCE REQUIREMENTS. Regulations can be complex. ES are using a stepwise process to ensure compliance.

· WEB SITE DESIGN. A good Web site design requires paying attention to many variables and ensures that performance is up to standard. ES can lead to a proper design process.

Now that you are familiar with the basic concepts of ES, it is time to look at the internal structure of ES and how their goals are achieved.

Structure and Process of ES

As you may recall from Section 2.5 and Figure 2.5, the process of knowledge extraction and its use is divided into two distinct parts. In ES we refer to these as the development environment and the consultation environment (see Figure 12.1). An ES builder builds the necessary ES components and loads the knowledge base with appropriate representation of expert knowledge in the development environment. A nonexpert uses the consultation environment to obtain advice and solve problems using the expert knowledge embedded into the system. These two environments are usually separated.

Major Components of ES

The major components in typical expert systems include:

· KNOWLEDGE ACQUISITION. Mostly from human experts, is usually obtained by knowledge engineers. This knowledge, which may derive from several sources, is integrated, validated, and verified.

· KNOWLEDGE BASE. This is a knowledge repository. The knowledge is divided into knowledge about the domain and knowledge about problem solving and solution procedures. Also, the input data provided by the users may be stored in the knowledge base.

· KNOWLEDGE REPRESENTATION. This is frequently organized as business rules (also known as production rules).

· INFERENCE ENGINE. Also known as the control structure or the rule interpreter, this is the “brain” of ES. It provides the reasoning capability, namely the ability to answer users’ questions, provide recommendations for solutions, generate predictions, and conduct other relevant tasks. The engine manipulates the rules by either forward chaining or backward chaining. In 1990s ES started to use other inference methods.

· USER INTERFACE. This component allows user inference engine interactions. In classical ES, this was done in writing or by using menus. In today’s knowledge systems, it is done by natural languages and voice.

These major components of ES generate useful solutions in many areas. Remember that these areas need to be well structured and in fairly narrow domains. Less common is a justifier/explanation subsystem that shows users of rule-based systems the chains of rules used to arrive at conclusions. Also, least common is a knowledge refining subsystem that helped to improve knowledge (e.g., rules) when new knowledge is added.

A major provider of expert systems technologies was Exsys Inc. While the company is no longer active in this business, its Web site (Exsys.com) is. It contains tutorials and a large number of cases related to its major software product, Exsys Corvid. Application Case 12.1 is one example.

Application Case 12.1 ES Aid in Identification of Chemical, Biological, and Radiological Agents

Terrorist attacks using chemical, biological, or radiological (CBR) agents are of great concern due to their potential for leading to large loss of life. The United States and other nations have spent billions of dollars on plans and protocols to defend against acts of terrorism that could involve CBR. However, CBR covers a wide range of input agents with many specific organisms that could be used in multiple ways. Timely response to such attacks requires rapid identification of the input agents involved. This can be a difficult process involving different methods and instruments.

The U.S. Environmental Protection Agency (EPA) along with Dr. Lawrence H. Keith, president of Instant Reference Sources Inc. and other consultants, have incorporated their knowledge, experience, and expertise as well as information in publicly available EPA documents to develop the CBR Advisor using Exsys Inc.’s Corvid software.

One of the most important parts of the CBR Advisor is providing advice in logical step-by-step procedures to determine the identity of a toxic agent when little or no information is available, which is typical at the beginning of a terrorist attack. The system helps response staff proceed according to a well-established action plan even in such a highly stressful environment. The system’s dual screens present three levels of information: (1) a top/executive level with brief answers, (2) an educational level with in-depth information, and (3) a research level with links to other documents, slide shows, forms, and Internet sites. CBR Advisor’s content includes:

· How to classify threat warnings.

· How to conduct an initial threat evaluation.

· What immediate response actions to take.

· How to perform site characterization.

· How to evaluate the initial site and safe entry to it.

· Where and how to best collect samples.

· How to package and ship samples for analysis.

Restricted content includes CBR agents and methods for analyzing them. The CBR Advisor can be used for incident response and/or training. It has two different menus, one for emergency response and another, longer menu for training. It is a restricted software program and is not publicly available.

Questions for Case 12.1

1. How can the CBR Advisor assist in making quick decisions?

2. What characteristics of the CBR Advisor make it an expert system?

3. What could be other situations in which similar expert systems can be employed?

Expert systems are also used in high-pressure situations in which human decision makers often need to take split-second actions involving both subjective as well as objective knowledge in responding to emergency situations.

Sources: www.exsys.com “Identification of Chemical, Biological and Radiological Agents” http://www.exsyssoftware.com/CaseStudySelector/casestudies.html. April 2018. (Publicly available information.) Used with permission.

Why the Classical Type of ES Is Disappearing

The large benefits described earlier drove the implementation of many ES worldwide. However, like many other technologies, the classical ES have been replaced by better systems. Let us first look at some of the limitations of ES that contributed to its declining use.

1. The acquisition of knowledge from human experts has proven to be very expensive due to the shortage of good knowledge engineers as well as the possible need to interview several experts for one application.

2. Any acquired knowledge needed to be updated frequently at a high cost.

1. The rule-based foundation was frequently not robust and not too reliable or flexible and could have too many exceptions to the rules. Improved knowledge systems use data-driven and statistical approaches to make the inferences with better success. In addition, case-based reasoning could work better only if a sufficient number of similar cases were available. So, usually it cannot support ES.

2. The rule-based user-interface needed to be supplemented (e.g., by voice communication, image maps). This could make ES too cumbersome.

3. The reasoning capability of rule-based technology is limited compared to use of newer mechanisms such as those used in machine learning.

New Generation of Expert SYSTEMS

Instead of using the old knowledge acquisition and representation system, newer ES based on machine learning algorithms and other AI technologies are deployed to create better systems. An example is provided in Application Case 12.2.

Application Case 12.2 VisiRule

VisiRule is an older ES company that remodeled its business over time. VisiRule (of the United Kingdom) provides easy-to-use diagramming tools to facilitate the construction of ES. Diagramming allows easier extraction and use of knowledge in expert systems.

The process of building the knowledge base can be seen on the left side of Figure 12.2. On the left-hand side, you can see the hybrid creation. Using a decision tree, the domain experts can create additional rules directly from relevant data (e.g., historical). In addition, rules can be created by machine learning (lower left side).

n). Using interactive questions and answers the system can generate advice. In addition, rules can be used to process data remotely and update the data repository. Note that the dual delivery option is based on machine learning’s ability to discover hidden patterns in data that can be used to form predictive decision models.

VisiRule also provides chatbots for improving the interactive part of the process and supplies an interactive map. According to the company’s Web site visirule.co.uk/, the major benefits of the product are:

· It is code-free; no programming is needed.

· The diagrams are drawn by human experts or induced automatically from data.

· It contains self-assessment tools with report generation and document production.

· The generated knowledge can be easily executed as XML code.

· It provides explanation and justification.

· The interactive expert advice attracts new customers.

· It can be used for training and advising employees.

· Companies can easily access the corporate knowledge repository.

· The charts to use VisiRule authoring tools are created with ease using flowcharting and decision trees.

· The charts allow creation of models that can be immediately executed and validated.

All-in-all, VisiRule provides a comprehensive AI-based expert system.

Source: Courtesy of VisiRule Corp. UK. Used with permission.

Questions for Case 12.2

1. Which of the limitations of early ES have been solved by the VisiRule system?

2. Compare Figures 12.2 and 12.1. What are the differences between the creation (Fig. 12.2) and the development (Fig. 12.1) subsystems?

3. Compare Figures 12.2 and 12.1. What are the differences between the delivery (Fig. 12.2) and the consultation (Fig. 12.1) subsystems?

4. Identify all AI technologies and list their contribution to the VisiRule system.

5. List some benefits of this ES to users.

Three major AI types of applications that overcome the earlier discussed limitations of RS are chatbots, virtual personal assistants, and robo advisors, which are presented next in this chapter. Other AI technologies that perform similar activities are presented in Chapters 4 to 9. Most notable is IBM Watson (Chapter 6); some of its advising capabilities are similar to those of ES but are much superior.

Another similar AI technology, the recommendation system, is presented next. Its newer variations use machine learning and IBM Watson Analytics.

Recommendation Systems

A heavily used knowledge system for recommending one-to-one targeted products or services is the  recommendation system , also known as recommender system or recommendation engine. Such a system tries to predict the importance (rating or preference) that a user will attach to a product or service. Once the rating is known, a vendor knows users’ tastes and preferences and can match and recommend a product or service to the user. For comprehensive coverage, see Aggarwal (2016). For a comprehensive tutorial and case study, see analyticsvidhya.com/blood/2015/10/recommendation-engines/.

Recommendation systems are very common and are used in many areas. Top applications include movies, music, and books. However, there are also systems for travel, restaurants, insurance, and online dating. The recommendations are typically given in rank order. Online recommendations are preferred by many people over regular searches, which are less personalized, slower, and sometimes less accurate.

Benefits of Recommendation Systems

Using these systems may result in substantial benefits both to buyers and sellers (see Makadia, 2018).

Benefits to customers are:

· PERSONALIZATION. They receive recommendations that are very close to fulfilling what they like or need. This depends, of course, on the quality of the method used.

· DISCOVERY. They may receive recommendations for products that they did not even know existed but were what they really need.

· CUSTOMER SATISFACTION. With repeated recommendations tends to increase.

· REPORTS. Some recommenders provide reports and others provide explanations about the selected products.

· INCREASED DIALOG WITH SELLERS. Because recommendations may come with explanations, buyers may want more interactions with the sellers.

Benefits to sellers are:

· HIGHER CONVERSION RATE. With personalized product recommendations, buyers tend to buy more.

· INCREASED CROSS-SELL. Recommendation systems can suggest additional products. Amazon.com, for example, shows other products that “people bought together with the product you ordered.”

· INCREASED CUSTOMER LOYALTY. As benefits to customers increase, their loyalty to the seller increases.

· ENABLING OF MASS CUSTOMIZATION. This provides more information on potential customized orders.

Several methods are (or were) used for building recommendation systems. Two classic methods are collaborative filtering and content-based filtering.

Collaborative Filtering

This method builds a model that summarizes the past behavior of shoppers, how they surf the Internet, what they were looking for, what they have purchased, and how much they like (rate) the products. Furthermore, collaborative filtering considers what shoppers with similar profiles bought and how they rated their purchases. From this, the method uses AI algorithms to predict the preference of both old and new customers. Then, the computer program makes a recommendation.

Content-Based Filtering

This technique allows vendors to identify preferences by the attributes of the product(s) that customers have bought or intend to buy. Knowing these preferences, the vendor recommends to customers products with similar attributes. For instance, the system may recommend a text-mining book to a customer who has shown interest in data mining, or action movies after a consumer has rented one in this category.

Each of these types has advantages and limitations (see example at en.wikipedia.org/wiki/Recommender_system). Sometimes the two are combined into a unified method.

Several other filtering methods exist. Examples include rule-based filtering and activity-based filtering. Newer methods include machine learning and other AI technologies, as illustrated in Application Case 12.3.

Application Case 12.3 Netflix Recommender: A Critical Success Factor

According to ir.netflix.com, Netflix is (Spring 2018 data) the world’s leading Internet television network with more than 118 million members in over 190 countries enjoying more than 150 million hours of TV shows and movies per day, including original series, documentaries, and feature films. Members can view unlimited shows without commercials for a monthly fee.

The Challenges

Netflix has several million titles and now produces its own shows. The large titles inventory often creates a problem for customers who have difficulty determining which offerings they want to watch. An additional challenge is that Netflix expanded its business from the United States and Canada to 190 other countries. Netflix operates in a very competitive environment in which large players such as Apple, Amazon.com, and Google operate. Netflix was looking for a way to distinguish itself from the competition by making useful recommendations to its customers.

The Original Recommendation Engine

Netflix originally was solely a mail-order business for DVDs. At that time, it encountered inventory problems due to its customers’ difficulties in determining which DVDs to rent. The solution was to develop a recommendation engine (called Cinematch) that told subscribers which titles they probably would like. Cinematch used data mining tools to sift through a database of billions of film ratings and customers’ rental histories. Using proprietary algorithms, it recommended rentals to customers. The recommendation was accomplished by comparing an individual’s likes, dislikes, and preferences against those of people with similar tastes, using a variant of collaborative filtering. Cinematch was like the geeky clerk at a small movie store who sets aside titles he knows you will like and suggests them to you when you visit the store.

To improve Cinematch’s accuracy, Netflix began a contest in October 2016, offering $1$1 million to the first person or team that will write a program that would increase Cinematch’s prediction accuracy by at least 10 percent. The company understood that this would take quite some time; therefore, it offered a $50,000$50,000 Progress Prize each year in which the contest was conducted. After more than two years of competition, the grand prize went to Bellkor’s Pragmatic Chaos team, a combination of two runner-up teams.

To learn how the movie recommendation algorithms work, see quora.com/How-does-the-Netflix-movie-recommendation-algorithm-work/.

The New Era

As time passed, Netflix moved to the streaming business and then to Internet TV. Also, the spread of cloud technology enabled improvement in the recommendation system. The new system stopped making recommendations based on what people have seen in the past. Instead, it is using Amazon’s cloud to mimic the human brain in order to find what people really like in their favorite movies and shows. The system is based on AI and its technology of deep learning. The company can now visualize Big Data and draw insights for the recommendations. The analysis is also used in creating the company’s productions. Another major change dealt with the transformation to the global arena. In the past, recommendations had been based on information collected in the country (or region) where users live. The recommendations were based on what other people in the same country enjoyed. This approach did not work well in the global environment due to cultural, political, and social differences. The modified system considers what people who live in many countries view and their viewing habits and likes.

Implementation of the new system was difficult, especially when a new country or region was added. Recommendations were initially made without knowing much about the new customers. It took 70 engineers and a year of work to modify the recommendation system. For details, see Popper (2016).

The Results

As a result of implementing its recommender system, Netflix has seen very fast growth in sales and membership. The benefits include the following:

· EFFECTIVE RECOMMENDATIONS. Many Netflix members select their movies based on recommendations tailored to their individual tastes.

· CUSTOMER SATISFACTION. More than 90 percent of Netflix members say they are so satisfied with the Netflix service that they recommend it to family members and friends.

· FINANCE. The number of Netflix members has grown from 10 million10 million in 2008 to 118 million in 2018. Its sales and profits are climbing steadily. In spring 2018, Netflix stock sold for over $400$400 per share compared with $140$140 a year earlier.

Sources: Based on Popper (2016), Arora (2016), and StartUp (2016).

Questions for Case 12.3

1. Why is the recommender system useful? (Relate it to one-to-one targeted marketing.)

2. Explain how recommendations are generated.

3. Amazon disclosed its recommendation algorithms to the public but Netflix did not. Why?

4. Research the research activities that attempt to “mimic the human brain.”

5. Explain the changes due to the globalization of the company.

Section 12.2 Review Questions

1. Define expert systems.

2. What is the major objective of ES?

3. Describe experts.

4. What is expertise?

5. List some areas especially amenable to ES.

6. List the major components of ES and describe each briefly.

7. Why is ES usage on the decline?

8. Define recommendation systems and describe their operations and benefits.

9. How do recommendation systems relate to AI?

12.3 Concepts, Drivers, and Benefits of Chatbots

The world is now infested with chatbots. According to 2017 data (Knight, 2017c), 60 percent of millennials have already used chatbots and 53 percent of those who have not used them are interested in doing so. Millennials are not the only generation using chatbots, although they may use them more than others. What chatbots are and what they do is the subject of this section.

What Is a Chatbot?

Short for chat robot, a  chatbot , also known as a “bot” or “robo,” is a computerized service that enables easy conversations between humans and humanlike computerized robots or image characters, sometimes over the Internet. The conversations can be in writing, and more and more are by voice and images. The conversations frequently involve short questions and answers and are executed in a natural language. More intelligent chatbots are equipped with NLPs, so the computer can understand unstructured dialog. Interactions also can occur by taking or uploading images (e.g., as is done by Samsung Bixby on the Samsung S8 and 8). Some companies experiment with learning chatbots, which gain more knowledge with their accumulated experience. The ability of the computer to converse with a human is provided by a knowledge system (e.g., rule-based) and a natural language understanding capability. The service is often available on messaging services such as Facebook Messenger or WeChat, and on Twitter.

Chatbot Evolution

Chatbots originated decades ago. They were simple ES that enabled machines to answer questions posted by users. The first known such machine was Eliza (en.wikipedia.org/wiki/ELIZA). Eliza and similar machines were developed to work in Q&A mode. The machine evaluated each question, usually to be found in a bank of FAQs, and generated an answer matched to each question. Obviously, if the question was not in the FAQ collection, the machine provided irrelevant answers. In addition, because the power of the natural language understanding was limited, some questions were misunderstood and the answers were at times at best entertaining. Therefore, many companies opted to use live chats, some with inexpensive labor, organized as call centers around the globe. For more about Eliza’s current generation, and how to build it, see search.cpan.org/dist/Chatbot-Eliza/Chatbot/Eliza.pm/. Chatbot use and reputation are rapidly increasing globally.

Example

Sophia is a chatbot created in Hong Kong and was awarded citizenship by Saudi Arabia in October 2017. Because she is not a Muslim, she is not wearing a hijab. She can answer many questions. For details, see newsweek.com/Saudi-arabia-robot-sophia-muslim-694152/.

Types of Bots

Bots can be classified by their capabilities; three classes follow:

1. REGULAR BOTS. These are essentially conversational intelligent agents (Chapter 2). They can do simple, usually repetitive, tasks for their owners, such as showing their bank’s debits, helping them to purchase goods online, and to sell or buy stocks online.

2. CHATBOTS. In this category, we include more capable bots, for example, those that can stimulate conversations with people. This chapter deals mainly with chatbots.

3. INTELLIGENT BOTS. These have a knowledge base that is improving with experience. That is, these bots can learn, for example, a customer’s preferences (e.g., like Alexa and some robo advisors).

A major limitation of the older types of bots was that updating their knowledge base was both slow and expensive. They were developed for specific narrow domains and/or specific users. It took many years to improve the supporting technology. NLP has become better and better. Knowledge bases are updated today in the “cloud” in a central location; the knowledge is shared by many users so the cost per user is reduced.

The stored knowledge is matched with questions asked by users. The answers by the machines have improved dramatically. Since 2000, we have seen more and more capable AI machines for Q&A dialogs. Around 2010, conversational AI machines were named chatbots and later were developed into virtual personal assistants, championed by Amazon’s Alexa.

Drivers of Chatbots

The major drivers are:

· Developers are creating powerful tools to build chatbots quickly and inexpensively with useful functionalities.

· The quality of chatbots is improving, so conversations are getting more useful to users.

· Demand for chatbots is growing due to their potential cost reduction and improved customer service and marketing services, which are provided 24/7.

· Use of chatbots allows rapid growth without the need to hire and train many customer service employees.

· Using chatbots, companies can utilize the messaging systems and related apps that are the darlings for consumers, especially younger ones.

Components of Chatbots and the Process of Their Use

The major components of chatbots are:

· A person (client).

· A computer, avatar, or robot (the AI machine).

· A knowledge base that can be embedded in the machine or available and connected to the “cloud.”

· A human-computer interface that provides the dialog for written or voice modes.

· An NLP that enables the machine to understand natural language.

Advanced chatbots can also understand human gestures, cues, and voice variations.

Person-Machine Interaction Process

The components just listed provide the framework for people-bot conversation. Figure 12.3 shows the conversation process.

· A person (left side of the figure) needs to find some information, or need some help.

· The person asks a related question from the bot by voice, texting, and so on.

· NLP translates the question to machine language.

· The chatbot transfers the question to cloud services.

· The cloud contains a knowledge base, business logic, and analytics (if appropriate) to craft a response to the question.

· The response is transferred to a natural language generation program and then to the person who asked the question in the preferred mode of dialog.

Drivers and Benefits

Chatbot use is driven by the following forces and benefits:

· The need to cut costs.

· The increasing capabilities of AI, especially NLP and voice technologies.

· The ability to converse in different languages (via machine translation).

· The increased quality and capability of captured knowledge.

· The push of devices by vendors (e.g., virtual personal assistants such as Alexa from Amazon and Google Assistant from Alphabet).

· Its use for providing superb and economic customer service and conducting market research.

· Its use for text and image recognition.

· Its use to facilitate shopping.

· Its support of decision making.

Chatbots and similar AI machines have been improved over time. Chatbots are beneficial to both users and organizations. For example, several hospitals employ robot receptionists to direct patients to their place of treatment. Zora Robotics created a robot named Nao to act as a chatting companion for people who are sick or elderly. The bot acts, for example, as a form of therapy for those suffering from dementia.

NOTE: For some limitations of chatbots, see Section 12.7.

Representative Chatbots from Around the World

For a chatbot directory of the more than 1,250 bots in 53 countries as of April 2018, see chatbots.org/ and at botlist.co/bots/. Examples of chatbots and what they can do from chatbot.org/ are provided here:

· ROBOCOKE. This is a party and music recommendation bot created for Coca-Cola in Hungary.

· KIP. This shopping helper is available on Slack (a messaging platform). Tell Kip what you want to buy, and Kip will find it and even buy it for you.

· WALNUT. This chatbot can discover skills relevant to you and help you learn them. It analyzes a large set of data points to discover the skills.

· RIDE SHARING BY TAXI BOT. If you are not sure whether Uber, Lyft, Grab, or Comfort DelGro is the cheapest service, you can ask this bot. In addition, you can get current promo codes.

· SHOPIIBOT. When you send a picture of a product to this bot, it will find similar ones in seconds. Alternatively, tell ShopiiBot what kind of product you are looking for at what price, and it will find the best one for you.

· CONCERNING DESIRED TRIPS. It can answer questions regarding events, restaurants, and attractions in major destinations.

· BO.T. The first Bolivian chatbot, it talks to you (in Spanish) and answers your questions about Bolivia, its culture, geography, society, and more.

· HAZIE. She is your digital assistant that aims to close the gap between you and your next career move. Job seekers can converse directly with Hazie just as they do with a job placement agent or friends.

· GREEN CARD. This Visabot product helps users to properly file requests for Green Cards in the United States.

· ZOOM. Zoom.ai (botlist.co/bots/369-zoomai), an automated virtual assistant, is for everyone in the workplace.

· AKITA. This chatbot (botlist.co/bots/1314-akita) can connect you to businesses in your area.

As you can see, chatbots can be used for many different tasks. Morgan (2017) classifies bots into the following categories: education, banking, insurance, retail, travel, healthcare, and customer experience.

Major Categories of Chatbots’ Applications

Chatbots are used today for many purposes and in many industries and countries. We divide the applications into the following categories:

· Chatbots for enterprise activities, including communication, collaboration, customer service, and sales (such as in the opening vignette). These are described in Section 12.4.

· Chatbots that act as personal assistants. These are presented in Section 12.5.

· Chatbots that act as advisors, mostly on finance-related topics (Section 12.6).

For a discussion of these categories, see Ferron (2017).

Section 12.3 Review Questions

1. Define chatbots and describe their use.

2. List the major components of chatbots.

3. What are the major drivers of chatbot technology?

4. How do chatbots work?

5. Why are chatbots considered AI machines?

12.4 Enterprise Chatbots

Chatbots play a major role in enterprises, both in external and internal applications. Some believe that chatbots can fundamentally change the way that business is done.

The Interest of Enterprises in Chatbots

The benefits of chatbots to enterprises are increasing rapidly, making dialog less expensive and more consistent. Chatbots can interact with customers and business partners more efficiently, are available anytime, and can be reached from anywhere. Businesses are clearly paying attention to the chatbot revolution. According to Beaver (2016), businesses should look at enterprise bots for the following reasons:

· “AI has reached a stage in which chatbots can have increasingly engaging and human conversations, allowing businesses to leverage the inexpensive and wide-reaching technology to engage with more consumers.

· Chatbots are particularly well suited for mobile-perhaps more so than apps. Messaging is at the heart of the mobile experience, as the rapid adoption of chat app demonstrates.

· The chatbot ecosystem is already robust, encompassing many different third-party chat bots, native bots, distribution channels, and enabling technology companies.

· Chatbots could be lucrative for messaging apps and the developers who build bots for these platforms, similar to how app stores have developed into moneymaking ecosystems.”

A study conducted in 2016 found that 80 percent of businesses want chatbots by 2020 businessinsider.com/80-of-businesses-want-chatbots-by-2020-2016-12. For more opportunities in marketing, see Knight (2017a).

Enterprise Chatbots: Marketing and Customer Experience

As we saw in the opening vignette to this chapter and will see in in the several examples later in this chapter, chatbots are very useful in providing marketing and customer service (e.g., Mah, 2016), obtaining sales leads, persuading customers to buy products and services, providing critical information to potential buyers, optimizing advertising campaigns (e.g., a bot named Baroj; see Radu, 2016), and much more. Customers want to do business on the app they are already in. For this reason, many bots are on Facebook Messenger, Snapchat, WhatsApp, Kik, and WeChat. Using voice and texting, it is possible to provide personalization as well as superb customer experience. Chatbots can enable vendors to improve personal relationships with customers.

In addition to the marketing areas, plenty of chatbots are in areas such as financial (e.g., banks) and HRM services as well as production and operation management for communication, collaboration, and other external and internal enterprise business processes. In general, enterprises use chatbots on messaging platforms to run marketing campaigns (e.g., see the opening vignette) and to provide superb customer experience.

Improving the Customer Experience

Enterprise chatbots create improved customer experience by providing a conversation platform for quick and 24/724/7 contact with enterprises. When customers benefit from the system, they are more inclined to buy and promote a specific brand. Chatbots can also supplement humans in providing improved customer experience.

Examples of Enterprise Chatbots

Schlicht (2016) provides a beginner’s guide to chatbots. He presents the following hypothetical example about today’s shopping at Nordstrom (a large department store) versus the use of chatbots.

If you wanted to buy shoes from Nordstrom online, you would go to their Web site, look around until you find the shoes you wanted, and then you would purchase them. If Nordstrom makes a bot, which I am sure they will, you would simply be able to message Nordstrom on Facebook. It would ask you what you are looking for and you would simply . . . tell it.

Instead of browsing a Web site, you will have a conversation with the Nordstrom bot, mirroring the type of experience you would get when you go into the retail store.

Three additional examples follow:

Example 1: LinkedIn

LinkedIn is introducing chatbots that conduct tasks such as comparing the calendars of people participating in meetings and suggesting meeting times and places. For details, see CBS News (2016).

Example 2: Mastercard

Mastercard has two bots based on massaging platforms, one bot for banks and another bot for merchants.

Example 3: Coca-Cola

Customers worldwide can chat with Coca-Cola bots via Facebook Messenger. The bots make users feel good with conversations that are increasingly becoming personalized. The bots collect customers’ data, including their interests, problems, local dialect, and attitudes and then can target advertisements tailored to each user.

A 5-min. video about Facebook is available at cnbc.com/2016/04/13/why-facebook-is-going-all-in-on-chatbots.html. It provides a Q&A session with David Marcus describing Facebook’s increasing interest in chatbots.

Why Use Messaging Services?

So far, we have noted that enterprises are using messaging services such as Facebook Messenger, WeChat, Kik, Skype, and WhatsApp. The reason is that in 2017, more than 2.6 billion2.6 billion people were chatting on messaging services. Messaging is becoming the most widespread digital behavior. WeChat of China was the first to commercialize its service by offering “chat with business” capabilities as illustrated in Application Case 12.4.

Application Case 12.4 WeChat’s Super Chatbot

WeChat is a very large comprehensive messaging service in China and other countries with about 1 billion1 billion members in early 2018. It pioneered the use of bots in 2013 (see mp.weixin.qq.com). Users can use the chatbot for activities such as the following:

· Hail a taxi.

· Order food to be delivered.

· Buy movie tickets and other items.

· Customize and order a pair of Nikes.

· Send an order to the nearest Starbucks.

· Track your daily fitness progress.

· Shop Burberry’s latest collection.

· Book doctor appointments.

· Pay your water bill.

· Host a business conference call.

· Send voice messages, emoticons, and snapshots to friends.

· Send voice messages to communicate with businesses.

· Communicate and engage with customers.

· Provide a framework for teamwork and collaboration.

· Conduct market research.

· Get information and recommendations on products and services.

· Launch a start-up on WeChat (you can make your own bot on WeChat for this purpose).

Griffiths (2016) has provided information concerning a Chinese online fashion flash sales company, Meici. The company used its WeChat account to gather information related to sales. Each time new users followed Meici’s account, a welcome message instructed them on how to trigger resources. WeChat is available in English and other languages worldwide due to its usefulness. Facebook installed similar capabilities in 2015.

Questions for Case 12.4

1. Find some recent activities that WeChat does.

2. What makes this chatbot so unique?

3. Compare the bot of WeChat to bots offered by Facebook.

Facebook’s Chatbots

Following the example of WeChat, Facebook launched users’ conversations with businesses’s chatbots on a large scale on Messenger, suggesting that users could message a business just the way they would message a friend. The service allows businesses to conduct text exchanges with users. In addition, the bots have a learning ability that enables them to accurately analyze people’s input and provide correct responses. Overall, as of early 2018, there were more than 30,000 company bots on Facebook Messenger. Some companies use Messenger bots to recognize faces in pictures, suggesting recipients for targeted ads. According to Guynn (2016), Facebook allows software developers access to its tools that build its personal assistant called “M,” which combines AI with a human touch for tasks such as ordering food or sending flowers. Using the M tools, developers can build applications for Messenger that can have an increased understanding of requests made in natural languages. A major benefit of these bots for Facebook is their collection of data and creation of profiles of users.

The following is another example of how the use of chatbots is facilitating customer service and marketing (Application Case 12.5).

Application Case 12.5 How Vera Gold Mark Uses Chatbots to Increase Sales

Vera Gold Mark is a real estate developer of luxury high rises in Punjab, India.

The Problem

Vera Gold Mark (VGM) is active in a very competitive market. As a developer of luxury apartments, which are usually expensive, it must try to attract many potential buyers and thus needs as many sales leads as possible at a reasonable cost. Chatting live with potential customers can be expensive since it requires very knowledgeable and courteous agents available 24/7.24/7. VGM has a large inventory of units that must be sold as soon as possible.

The Solution

VGM decided to use chatbots to supplement or replace expensive manual live chats. These work in the following ways. Buyers may click on the “chat with the robot” button on the company’s Facebook page, and receive any information they need. The chat helps VGM promote its available products. When they click, users are able to chat and get information about pricing, delivery dates, construction sites, and much more for VGM projects. Users can also tweet. The chatbots provide answers about the projects. Facebook provides VGM access to potential buyers’ profiles (with users’ permission), which VGM sales teams can use to refine sales strategies. The system is available 24/7.24/7. Voice communication is coming soon (2018).

The Results

VGM is now viewed in a very positive way and is considered to be very professional. VGM is getting good reviews for its customer service. The builder is considered more honest and unbiased because it provides written answers and promises to customers. Salespeople at VGM get an increased number of sales leads, and because they know more about prospective customers, they can better align them with units (optimal fit). The system is also able to attract international buyers without increasing cost. Because the system is available 24/7,24/7, global buyers can easily evaluate VGM’s available condominiums.

The chatbot is also used as a teaching tool for new employees. At the time that this case was written, no financial data were available.

The technology is available to other builders from Kenyt Technologies of India kenyt.com, which provides the smart real estate chatbot.

Sources: Based on Garg (2017) and facebook.com/veragoldmark/ (accessed April 2018).

Questions for Case 12.5

1. List the benefits to VGM.

2. List the benefits to buyers.

3. What is the role of Kenyt Technologies?

Chatbots Magazine provides a three-part overview on the use of chatbots for retail and e-commerce. For details, see chatbotsmagazine.com/chatbots-for-retail-and-e-commerce-part-three-c112a89c0b48.

Enterprise Chatbots: Financial Services

The second area in which enterprise bots are active is financial services. Here we briefly discuss their use in banking. In Section 12.6, we present the robo financial advisors for investment.

Banking

A 2017 survey (Morgan, 2017) found that most people in the United States will bank via chatbots by 2019. Chatbots can use predictive analytics and cognitive messaging to perform tasks such as making payments. They can inform customers about personalized deals. Banks’ credit cards can be advertised via chatbots on Facebook Messenger. It seems that customers prefer to deal with chatbots rather than with salespeople who can be pushy.

Examples

POSB of Singapore has an AI-driven bot on Facebook Messenger. The bot was created with the help of Kasisto, Inc. of the United States. Using actual Q&A sessions, it took IT workers 11,000 hours to create the bot. Its knowledge base was tested and verified. The bot can learn to improve its performance. Known as POSB digi-bank virtual assistant, the service is accessed via Messenger. Customers save time rather than waiting for human customer service. In the future, the service will be available on other messaging platforms. For details, see Nur (2017).

A similar application in Singapore is used by Citi Bank (by Citi Group). It can answer FAQs about people’s accounts in a natural language (English). The bank is adding progressively more capabilities to its bot.

A generic banking bot is Verbal Access (from North Side Co.) that provides recommendations for banking services (see Hunt, 2017).

Enterprise Chatbots: Service Industries

Chatbots are used extensively in many services. We provide several examples in the following sections.

Healthcare

Chatbots are extremely active in the healthcare area, helping millions of people worldwide (Larson, 2016). Here are a few examples:

· Robot receptionists direct patients to departments in hospitals. (Similar services are available at airports, hotels, universities, government offices, and private and other public organizations.)

· Several chatbots are chatty companions for people who are elderly and sick (e.g., Zora Robotics).

· Chatbots are used in telemedicine; patients converse with doctors and healthcare professionals who are in different locations. For example, the Chinese company Baidu developed the Melody chatbot for this purpose.

· Chatbots can connect patients quickly and easily with information they need.

· Important services in the healthcare field are currently provided by IBM Watson (Chapter 6).

For more on bots for healthcare, see the end of Section 12.6.

Education

Chatbot tutors are used in several countries to teach subjects ranging from English (in Korea) to mathematics (in Russia). One thing is certain: The chatbot treats all students equally. Students like the chatbots in online education as well. Machine translation of languages will enable students to take online classes in languages other than their own. Finally, chatbots can be used as private tutors.

Government

According to Lacheca (2017), chatbots are spreading in government as a new dialog tool for use by the public. The most popular use is in providing access to government information and answering government-related questions.

Travel and Hospitality

Chatbots are working as tour guides in several countries (e.g., Norway). They are not only cheaper (or free) but also may know more than some human guides. Chatbots work as guides in several hotels in Japan. In hotels, they act as concierges, providing information and personalized recommendations (e.g., about restaurants). Chatbots can arrange reservations for hotel rooms, meals, and events. In busy hotels, there is frequently a wait for human concierges; chatbots are available on smartphones all the time. As with other computer services, the chatbots are fast, inexpensive, easy to reach, and always nice. They give excellent customer experience.

An example of external travel service is given in Application Case 12.6.

Application Case 12.6 Transavia Airlines Uses Bots for Communication and Customer Care Delivery

Background

The air travel business is very competitive, especially in Europe. There is a clear trend for younger customers to use wireless devices as well as social media sites and chatting. Customers like to communicate with travel businesses by using their preferred technology via their preferred platforms. Most popular is Facebook Messenger, where over 1.2 billion people chat, many times via their smartphones. These users today interact not only among themselves but also with the business world.

Messaging platforms such as Messenger, WhatsApp, and WeChat are becoming the norm for this customer group. Vendors are building smart apps for the messaging platforms including bots.

Transavia’s Bot

Learning from other companies, Transavia decided to create a bot on Facebook Messenger. To do so, it hired the IT consultant Cognizant Digital Business unit, called Mirabean, which specializes in conversation interfaces, especially via bots. Transavia’s activities business processes, marketing, and customer care were combined with Mirabean’s technological experience to enable a quick deployment of the bot in weeks. It now enables real-time dialog with customers. The first application is Transavia Flight Search, which provides flight information as well as the ability to buy tickets. The system is now integrated with business processes that facilitate other transactions via the bot. Giving customers their digital tool of choice enables Transavia to increase market share and to drive growth.

Note that KLM, the owner of Transavia, was the first European airline that implemented a similar chatbot on Facebook Messenger in 2016.

Sources: Compiled from Cognizant (2017) and transavia.com.

Questions for Case 12.6

1. What drives consumer preference for mobile devices and chat?

2. Why was the bot placed on Facebook Messenger?

3. What were the benefits of using Cognizant?

4. What is the advantage of buying a ticket from a bot rather than from an online store?

Chatbot Platforms

Chatbots Inside Enterprises

So far we have seen chatbots that are working in the external side of enterprises, mostly in customer care and marketing (e.g., the opening vignette). However, companies lately have started to use chatbots to automate tasks for supporting internal communication, collaboration, and business processes. According to Hunt (2017), “Enterprise and internal chatbots are revolutionizing the way companies do business.” Chatbots in enterprises can do many tasks and support decision-making activities. For examples, see Newlands (2017a). Chatbots can cut costs, increase productivity, assist working groups, and foster relationships with business partners. Representative examples of chatbot tasks are:

· Help with project management.

· Handle data entry.

· Conduct scheduling.

· Streamline payments with partners.

· Advise on authorization of funds.

· Monitor work and workers.

· Analyze internal Big Data.

· Find discounted and less expensive products.

· Simplify interactions.

· Facilitate data-driven strategy.

· Use machine learning.

Facilitate and manage personal finance.

Given the large number of bots, it is not surprising that many developers started to offer tools and platforms to assist in building chatbots as discussed in Technology Insights 12.1.

Technology Insights 12.1 Chatbots’ Platform Providers

Several companies provide platforms for building enterprise chatbots. The companies can construct chatbots fairly easily using these tools for their entry into popular messaging platforms or for their Web sites. Some of the tools have machine-learning capability to ensure that the bots learn with every interaction. According to Hunt (2017), these are some popular vendors:

1. CHATTYPEOPLE. This chatbot builder assists in creating bots requiring minimal programming skills. It simply allows a business to link its social media pages to its ChattyPeople account. The created bot can:

· Arrange for payments to or from social media contacts.

· Use major payment providers such as Apple Pay and PayPal.

· Recognize variations in keywords.

· Support messaging.

2. KUDI. This financial helper allows people to make payments to vendors directly from their messaging apps, specifically, Messenger, Skype, and Telegram and through an Internet browser. Using the bot, users can:

· Pay bills.

· Set bill payment reminders.

· Transfer money by sending text messages.

The bot is safe and it protects users’ privacy. Vendors can easily install it for use.

3. TWYLA. This chatbot building platform is for improving existing customer care and offering live chats. It acts as a messaging platform for customers who prefer to use chatting. The major objective is to free humans in HR departments from routine tasks.

The most popular platforms are:

· IBM WATSON. This package uses a neural network of 1 billion words for excellent understanding of natural languages (e.g., English, Japanese). Watson provides free development tools, such as Java SDK, Node SDK, Pyton SDK, and iOS SDK.

· MICROSOFT’S BOT FRAMEWORK. Similar to IBM, Microsoft offers a variety of tools translatable into 30 languages. It is an open source. The system has three parts, Bot Connector, Developer Portal, and Bot Directory and is interconnected with Microsoft Language Understanding Intelligent Service (LUIS) that understands users’ intent. The system also includes active learning technology. A simplified tool is AZURE; see Section 12.7 and Afaq (2017). For a comparative table of 25 chatbots platforms, see Davydova (2017). For a list of other platforms, see Ismail (2017).

Sources: Compiled from Hunt (2017) and Davydova (2017).

Discussion Questions

1. What is the difference between a regular enterprise bot and a platform?

2. Discuss the benefits of ChattyPeople.

3. Discuss the need for Kudi.

4. Discuss the reasons for consumers to prefer messaging platforms.

For additional information about chatbot platforms for building enterprise chatbots, see entrepreneur.com/article/289788.

Industry-Specific Bots

As we have seen, bots can be specialists (e.g., for investment advice, customer service) or industry-specific experts (e.g., banking, airlines). An interesting bot for the waste industry is Alto (from Bio Hi Tech Global), which enables users to communicate intelligently with industrial equipment. This helps owners of the equipment make decisions that improve performance levels, smooth maintenance routines, and facilitate communication.

Knowledge for Enterprise Chatbots

Knowledge for chatbots depends on their tasks. Most marketing and customer care bots require proprietary knowledge, which is usually generated and maintained in-house. This knowledge is similar to that of ES; in many cases, enterprise chatbots operate very similarly to ES except that the interface occurs in a natural language and frequently by voice. For example, the knowledge of Sephora’s bot (opening vignette) is specific to that company and its products and is organized in a Q&A format.

On the other hand, chatbots that are used within the enterprise (e.g., to train employees or to provide advice on security or compliance with government regulations) may not be company specific. A company can buy this knowledge and modify it to fit local situations and its specific needs (as is done in ES; e.g., see Exsys Inc.). Newer chatbots use machine learning to extract knowledge from data.

Personal Assistants in the Enterprise

Enterprise chatbots can also be virtual personal assistants as will be described in Section 12.5. For example, these bots can answer work-related queries and help in increasing employees’ decision-making capabilities and productivity.

Section 12.4 Review Questions

1. Describe some marketing bots.

2. What can bots do for financial services?

3. How can bots assist shoppers?

4. List some benefits of enterprise chatbots.

5. Describe the sources of knowledge for enterprise chatbots.

12.5 Virtual Personal Assistants

In the previous section, we introduced enterprise chatbots that can be used to conduct conversations. In marketing and sales, they can facilitate customer relationship management (CRM, execute searches for customers, provide information, and execute many specific tasks in organizations for their customers and employees. For comprehensive coverage, including research issues, see Costa et al. (2018)).

An emerging type of chatbot is designed as a virtual personal assistant for both individuals and organizations. Known as a  virtual personal assistant (VPA) , this software agent helps people improve their work, assist in decision making, and facilitate their lifestyle. VPAs are basically extensions of intelligent software agents that interact with people. VPAs are chatbots whose major objective is to help people better perform certain tasks. At this time, millions of people are using Siri with their Apple products, Google Assistant, and Amazon’s Alexa. The assistants’ knowledge bases are usually universal, and they are maintained centrally in the “cloud,” which makes them economical for a large number of users. Users can get assistance and advice from their virtual assistants anytime. In this section, we provide some interesting applications. The first set of applications involves virtual personal assistants, notably Amazon’s Alexa and Apple’s Siri and Google Assistant. O’Brien (2016) provides a discussion of what personal assistant chatbots can do for business. The second set (presented in Section 12.6) is about computer programs that act mostly as advisors on specific topics (mostly investments).

Assistant for Information Search

A major task of virtual personal assistants is to help users conduct a search by voice for information. Without the assistant, users need to surf the Internet to find information and many times abandon the search. In business situations, users can call a live customer service agent for assistance. This may be an expensive service for the vendors. Delegating the search to a machine may save sellers considerable money and make customers happy by not having to wait for the service. For example, Lenovo uses the noHold assistant in its Single Point of Search service to help customers find answers to their questions.

If You Were Mark Zuckerberg, Facebook CEO

While Siri and Alexa were in development, Zuckerberg decided to develop his own personal assistant to help him run his home and his work as the CEO of Facebook. He viewed this assistant as Jarvis from Iron Man. Zuckerberg trained the bot to recognize his voice and understand basic commands related to home appliances. The assistant can recognize the faces of visitors and monitor the movement of Zuckerberg’s young daughter. For details, see Ulanoff (2016).

The essentials of this assistant can be seen in a 2:13 min.2:13 min. video at youtube.com/watch?v=vvimBPJ3XGQ and one (5:01 min.)(5:01 min.) at youtube.com/watch?v=vPoT2vdVkVc, with the narration by Morgan Freeman. Today, similar assistants are available for a minimal fee or even for free. The most well-known such assistant is Amazon’s Alexa.

Amazon’s Alexa and Echo

Of the several virtual personal assistants, the one considered the best in 2018 was Alexa. She was developed by Amazon to compete with Apple’s Siri and is a superior product. (See Figure 12.4.) Alexa works with a smart speaker, such as Amazon’s Echo (to be described later).

Figure 12.4 Amazon’s Echo and Alexa.

Source: McClatchy-Tribune/Tribune Content Agency LLC/Alamy Stock Photo

Amazon’s  Alexa  is a cloud-based virtual personal voice assistant that can do many things such as:

· Answer questions in several domains.

· Control smartphone operations with voice commands.

· Provide real-time weather and traffic updates.

· Control smart home appliances and other devices by using itself as a home automation hub.

· Make to-do lists.

· Arrange music in Playbox.

· Set alarms.

· Play audio books.

· Control home automation devices, as well as home appliances (e.g., a microwave).

· Analyze shopping lists.

· Control a car’s devices.

· Deliver proactive notification.

· Shop for its user.

· Make phone calls and send text messages.

Alexa has the ability to recognize different voices, so it can provide personalized responses. Also, she uses a mix of speech and touch to deliver news, hail an Uber, and play games. As time passes, her capabilities and skill grow. For more capabilities, which are ever-increasing, see Johnson (2017). For what Alexa can hear and remember and how she learns, see Oremus (2018).

Watch the 3:55 min. video of how Alexa works at youtube.com/watch?v=jCtfRdqPlbw. For more tasks, see cnet.com/pictures/what-can-amazon-echo-and-alexa-do-pictures/, Mangalindan (2017), and tomsguide.com/us/pictures-story/1012-alexa-tricks-and-easter-eggs.html.

Alexa’s Skills

In addition to the standard (native) capabilities listed, people can use Alexa apps (referred to as Skills) to download customized capabilities to Alexa (via your smartphone). Skills are intended to teach Alexa something new.

The following are examples of Alexa’s Skills (Apps):

· Call Uber and find the cost of a ride.

· Order a pizza.

· Order take-out meals.

· Obtain financial advice.

· Start a person’s Hyundai Genesis car from inside her or his house (Korosec, 2016).

These skills are provided by third-party vendors; they are required to activate invocation commands. There are tens of thousands of them.

For example, a person can say, “Alexa, call Uber to pick me up at my office at 4:30 p.m.” For more on Amazon’s Alexa, see Kelly (2018); for its benefits, see Reisinger (2016).

Alexa is equipped with NLP user interface, so it can be activated by providing a voice command. This is done by combining the Alexa software with Amazon’s intelligent speaker, Echo.

Alexa’s Voice Interface and Speakers

Amazon has a family of three speakers (or voice communication devices for Alexa: Echo, Dot, and Tag. Alexa can be accessed by a Fire TV line and some non-Amazon devices. For the relationship between Alexa and Echo, see Gikas (2016).

Amazon’s Echo

Echo  is a hands-free intelligent (or smart) wireless speaker that is controlled by voice. It is the hardware companion of Alexa (a software product), so the two operate hand in hand. Echo is always on, always listening. When Echo hears a question, command, or request, it sends the audio to Alexa and from there up to the cloud. Amazon’s servers match responses to the questions, delivering them to Alexa as “responses to questions” in a split second. Amazon’s Alexa/Echo is now available in some Ford vehicles.

Amazon Echo Dot

Amazon Echo Dot is the “little brother” of Echo. It offers full Alexa functionality but has only one very small speaker. It can be linked to any existing speaker systems to provide an Echo-like experience.

Amazon Echo Tap

Amazon Echo Tap is another “little brother” of Echo that can be used on the go. It is completely wireless and portable and can be charged via a charging dock.

Both Dot and Tap are less expensive than Echo, but they offer fewer functionalities and lower quality. However, people who already have good home speakers can use Dot with them. For a discussion about the three speakers, see Trusted Review at trustedreviews.com/news/amazon-echo-show-vs-echo-2948302.

Note: Non-Amazon speakers for Alexa are available now (e.g., Eufy Genie, from third-party vendors); some are inexpensive.

Note: Alexa was smart enough earlier to admit that she did not know an answer, but today, she will make references to third-party sources for an answer she cannot make. For details and examples, see uk.finance.yahoo.com/news/alexa-recommend-third-party-skills-192700876.html.

Alexa for the Enterprise

While the initial use of Alexa was for individual consumers, her use for business has increased. WeWork Corp. developed a platform for helping companies to integrate an Alexa skill in meeting rooms, for example. For details, see Crook (2017), and yahoo.com/news/destiny-2-alexa-skills-let-140946575.html/.

Apple’s Siri

Siri  (short for Speech Interpretation and Recognition Interface) is an intelligent virtual personal assistant and knowledge navigator. It is a part of Apple’s several operating systems. It can answer questions, make recommendations, and perform some actions by delegating requests to a set of Web services in the “cloud.” The software can adapt itself to the user’s individual language, search preferences with continuing use, and return personalized results. Siri is available for free to iPhone and iPad users.

Siri can be integrated into Apple’s Siri Remote. Using CarPlay, Siri is available in some auto brands where it can be controlled by iPhone (5 and higher). Siri 2 is the 2017–2018 model.

Viv

In 2016, Dag Kittlaus, the creator of Siri, introduced Viv, “an intelligent Interface for everything.” Viv is expected to be the next generation of intelligent virtual interactions (for details, see Matney, 2016). In contrast with other assistants, Viv is open to all developers (third-party ecosystem products). Viv is now a Samsung Company. In 2017, Samsung launched its own personal assistant for the Galaxy S8.

Google Assistant

Competition regarding virtual personal assistants is increasing with the improved capabilities of Google Assistant, which was developed as a competitor to Siri to fit Android smartphones. An interesting demonstration of it is available at youtube.com/watch?v=WTMbF0qYWVs; some advanced capabilities are illustrated in the video at youtube.com/watch?v=17rY2ogJQQs. For details, see Kelly (2016). The product improved dramatically in 2018 as shown in CES 2018 Conference.

Other Personal Assistants

Several other companies have virtual personal assistants. For example, Microsoft Cortana is well known. In September 2016, Microsoft combined Cortana and Bing (see Hachman, 2016). Alexa and Cortana now work together. Note that it is estimated that by the year 2022, voice-enabled personal assistants will reach 55 percent of all U.S. households. For this and the future of personal assistants, see Perez (2017).

Competition Among Large Tech Companies

Apple and Google have provided their personal assistants to hundreds of million users of their mobile devices. Microsoft has equipped over 250 million250 million PCs with its personal assistant. Amazon’s Alexa/Echo sells many more assistants than others. The competition is on voice-controlled chatbots. Their competitors view them as “the biggest thing since the iPhone.”

Knowledge for Virtual Personal Assistants

As indicated earlier, the knowledge for virtual personal assistants is kept in the “cloud.” The reason is that the assistants are commodities, available to millions of users, and need to provide dynamic, updated information (e.g., weather conditions, news, stock prices). When the knowledge base is centralized, its maintenance is performed in one place. This is in contrast with the knowledge of many enterprise bots, for which updating is decentralized. Thus, Siri on an iPhone will always be updated for its general knowledge by AAPL. Knowledge for the skills of Alexa has to be maintained locally or by the third-party vendors that create them.

Section 12.5 Review Questions

1. Describe an intelligent virtual personal assistant.

2. Describe the capabilities of Amazon’s Alexa.

3. Relate Amazon’s Alexa to Echo.

4. Describe Echo Dot and Tap.

5. Describe Apple’s Siri Google’s Assistant.

6. How is the knowledge of personal assistants maintained?

7. Explain the relationship between virtual personal assistants and chatbots.

12.6 Chatbots as Professional Advisors (Robo Advisors)

The personal assistants described in Section 12.5 can provide much information and rudimentary advice. A special category of virtual personal assistants is designed to provide personalized professional advice in specific domains. A major area for their activities is investment and portfolio management where robo advisors operate.

Robo Financial Advisors

It is known that the vast majority of “buy” and “sell” decisions of stock trading on the major exchanges, especially by financial institutions, are made by computers. However, computers can also manage an individual’s accounts in a personalized way.

According to an A. T. Kearney’s survey (reported by Regan, 2015), robo advisors are defined as online providers that offer automated, low-cost, personalized investment advisory services, usually through mobile platforms. These robo advisors use algorithms that allocate, deploy, rebalance, and trade investment products. Once enrolled for the robo service, individuals enter their investment objectives and preferences. Then, using advanced AI algorithms, the robo will offer alternative personalized investments for individuals to choose from funds or exchange-traded funds [ETFs]. By conducting a dialog with the robo advisor, an AI program will refine the investment portfolio. This is all done digitally without having to talk to a live person. For details, see Keppel (2016).

Evolution of Financial Robo Advisors

The pioneering emergence of Betterment Inc. in 2010 (described later) was followed by several other companies (Future Advisor and Hedgeable in 2010 and Personal Capital, Wealthfront, and SigFig in 2011 and 2012). Other well-known companies (Schwab Intelligent Portfolios, Acorns, Vanguard RAS, and Ally) joined the crowd in 2014 and 2015. In 2016 and 2017, the brokerage houses of E*Trade and TD Ameritrade joined, as did Fidelity and Merrill Edge. There is no question that robo advisors are game-changing phenomena for the wealth management business, even though their performance so far has not been much different from that of traditional, manual, and financial services.

Robo advising companies try to cut costs by using ETFs, whose commission fees are significantly lower than that of mutual funds. Annual fees vary as does the minimum amount of required assets. Premium services are more expensive since they offer the opportunity to consult human experts (advisors 2.0), which are described next.

Robo Advisors 2.0: Adding the Human Touch

As robo advisors matured, it became clear that sometimes they could not do an effective job by themselves. Therefore, in late 2016, several of the fully automated advisors started to add what they call the human touch (e.g., see Eule, 2017; Huang, 2017). Companies are adding a human option, or partner with another company. For example, UBS Wealth Management Americas has partnered with pure robo advisor SigFig.

Robo advisors with human additions vary in expertise. For example, Betterment (Plus and Permission options), Schwab Intelligent Advisory, and Vanguard Personal Advisor Service use certified financial planners (CFPs); other companies offer less expertise. For details, see Huang (2017).

Application Case 12.7 describes how Betterment has added the human touch.

Application Case 12.7 Betterment, the Pioneer of Financial Robo Advisors

As the pioneer of financial robo advisors in 2010, Betterment created an automated platform for wealth management. Since then, it has played a leading role in a growing industry. In 2017, the company controlled more than $9 billion$9 billion in assets, yielding over an 11 percent return to its 200,000 members. Like other robo advisors, Betterment appeals to investors who do not want to manage their portfolio by themselves or pay the 2 to 3 percent annual fee charged by human advisors.

The company advertises the following benefits:

· Provides unlimited professional expert advice (by the bot) anytime and anywhere.

· Provides advice from bots that contain the knowledge of human investment advisors.

· Assists investors in making decisions of how much to invest.

· Helps investors figure out how much risk to take.

· Helps in lowering investment-related tax.

· Provides actionable answers to questions.

· Advises on college savings.

· Helps plan for retirement.

· Assists in mortgage management (e.g., refinance).

· Provides personalized service via the use of investors’ goal-based analysis.

Betterment has no account minimum (competitors require up to $100,000).

Each investor’s portfolio is automatically adjusted to market conditions to meet his or her goals. All portfolios are built and managed by AI algorithms.

Premium Service—Adding the Human Touch

Like Amazon.com and Expedia, which started as pure online companies and later added physical commerce, in 2017 Betterment added what it calls a human touch; its Plus service is offered to customers with assets of over $100,000$100,000 who are willing to pay an annual fee of 0.4 percent for this service. Using it, customers can interact with human advisors in addition to the automated bot. An even better service is the company’s Premium level, which requires $250,000$250,000 in assets and charges 0.50.5 percent in fees.

While the quality of the automated service is getting better with added knowledge (machine learning), complex situations that require human intervention still remain. This is where the Plus and Premium services enter the picture. Several competitors also have added the human touch to their offering.

Sources: Compiled from O’Shea (2017), Eule (2017), and betterment.com (accessed April 2018).

Questions for Case 12.7

1. What are Betterment’s benefits to investors?

2. Compare Betterment to its major competitors (see Eule, 2017).

3. What are the benefits of adding the human touch (i.e., compared to pure automation and only human service)?

4. Find some new information about Betterment. Write a report.

Quality of Advice Provided by Robo Advisors

You may wonder how good the advice from robo advisors is. The answer is that it depends on their knowledge, the type of investments involved, the inference engine of the AI machine, and so on. However, remember that the robots are not biased and are consistent. They may prove to be even better than humans at one of the most important aspects in investment advising: know how to legally minimize the related tax. This implies that institutional-grade tax-loss harvesting is now within the reach of all investors. By contrast, some people believe that it is difficult to replace investment brokers with robots. De Aenlle (2018) believes that humans are still dominating advisory services (see the example of Nordea Bank by Pohjanpalo, 2017).

For a list of the best robo advisors, see Eule (2017), O’Shea (2016), and investorjunkie.com/35919/roboadvisors. For comprehensive coverage of robo advisors in finance and investment, including the major companies in the advisory industry, see McClellan (2016).

An emerging commercial robo advisor is being developed at Cornell University under the name Gsphere. In addition, robo advisors appear in countries other than the United States (e.g., Marvelstone Capital in Singapore).

Financial Institutions and Their Competition

Several large financial institutions and banks have reacted to robo advisors by creating their own or partnering with them. It is difficult to assess the winners and losers in this competition because there are no sufficient long-term data. So far it seems that customers like robo advisors, basically because they cost as little as 10 percent of full-service human advisors. For a discussion and data, see Marino (2016). Note that some observers point to the danger of using robo advisors in a declining stock market due to their use of ETFs.

Managing Mutual Funds Using AI

Many institutions and some individual investors buy stocks using AI algorithms. Some people prefer to buy a mutual fund that picks its holding with AI. EquBot is such a fund (its symbol is AIEQ). Its 2017 performance was above average.

The AI algorithms used by EquBot can process 1 million pieces of data each day. They follow 6,000 companies. For details, see Ell (2018).

Other Professional Advisors

In addition to investment advisors, there are several other types of robo advisors ranging from travel to medicine to legal areas.

The following are examples of noninvestment advisors:

· COMPUTER OPERATIONS. To cut costs, major computer vendors (hardware and software) try to provide users with self-guides to solve encountered problems. If users cannot get help from the guides, they can contact live customer service agents. This service may not be available in real time, which can upset customers. Live agents are expensive, especially when provided 24/724/7. Therefore, companies are using interactive virtual advisors (or assistants).

As an example, Lenovo Computers use a generic bot called noHold’s AI to provide assistance to customers as a single point of help for conducting a search.

· TRAVEL. Several companies provide advice on planning future national and international trips.

For example, Utrip (utrip.com) helps plan European trips. Based on their stated objectives, travelers get recommendations for what to visit in certain destinations. The service is different from others in that it customizes trips.

· MEDICAL AND HEALTH ADVISORS. A large number of health and medical care advisors operate in many countries. An example is Ad a Health of Germany. Founded in late 2017 as a chatbot, it assists people in activities such as deciphering their ailments and can connect patients to live physicians. This can be the future of health in adding bot-based patient-doctor collaboration.

A list of the top useful chatbots as of 2017 is provided by TalKing (2017). It includes:

· Health Tap acts like a medical doctor by providing a solution to common symptoms provided by patients.

· YourMd is similar to Health Tap.

· Florence is a personal nurse available on Facebook Messenger.

Other bots include OneStopHealth, HealthBot, GYANT, Buoy, Bouylon, and Mewhat.

· Bots are acting as companions (e.g., Endurance for dementia patients). In Japan, bots that look and feel like dogs are very popular companions for elderly people. Several bots are designed to increase patient engagement. For example, Lovett (2018) reports that a bot for patient engagement increased patients’ response rate to a flu shot campaign by 30 percent. Finally, the classic pioneering bot, ELIZA, acted as a very naïve psychologist.

· SHOPPING ADVISORS (SHOPBOTS). Shopbots can act as shopping advisors. An example is Shop Advisor (see shopadvisor.com/our-platform. It is a comprehensive platform that includes three components to help companies attract customers. The platform is a self-learning system that improves its operation over time. Its components are:

1. Product intelligence, which processes complex and diverse product data. It includes a competitive analysis.

2. Context intelligence, which collects and catalogs contextual data points about marketing facilities and inventories in different locations.

3. Shopper intelligence, which studies consumers’ actions related to different magazines, mobile apps, and Web sites.

There are thousands of other shopping advisors. Sephora (opening vignette) has several of them. There are chatbots for Mercedes cars and for top department stores such as Nordstrom, Saks, and DFS. The use of shopping chatbots is increasing rapidly due to the use of mobile shopping and mobile chatting on social networks. Marketers, as we stated earlier, can collect customer data and deliver targeted ads and customer service to specific customers.

Another trend that facilitates online shopping with the assistance of bots is the increase in the number of virtual personal shopping assistants. Users only have to tell Alexa by voice, for example, to buy something for them. Better than that, they can use their smartphones from anywhere to tell Alexa to go shopping. Ordering via voice directly from vendors (e.g., delivery of pizzas) is becoming popular. In addition to chatbots that operate by sellers, there are bots for providing advice on what and where to buy.

Example: Smart Assistant Shopping Bots

Shopping bots ask a few questions to understand what a customer needs and prefers. Then they recommend the best match for the customer. This makes customers feel they are receiving personalized service. The assistance simplifies the customer’s decision-making process. Smart assistants also offer advice on issues of concern to customers via Q&A conversations. For a guided test, go to a demo at smartassistant.com/advicebots. Note that these bots are essentially recommendation systems and that users need to ask for advice whereas other recommendation systems (e.g., that of Amazon.com) provide advice even when users do not ask for it.

A well-known global shopping assistant in the area of fashion is Alibaba’s Fashion AI. It helps customers who shop in stores. When shoppers enter a fitting room, the AI Fashion Consultant goes into action. For details of how this is done, see Sun (2017).

Another type of shopping advisor works as a virtual personal advisor to shoppers. This type was developed from traditional e-commerce intelligent agents, such as bizrate.com and pricegrabber.com.

IBM Watson

Probably the most knowledgeable virtual advisor is IBM Watson (see Chapter 6). Some examples of its use follow:

· Macy’s developed a service, Macy’s On Call, to help customers navigate its physical stores while they shop. Using location-based software, the app knows where they are in the store. By using smartphones, customers can ask questions regarding products and services in the stores and then receive a customized response from the chatbot.

· Watson can help physicians make a diagnosis (or verify one) quickly and suggest the best treatment. Watson’s Medical Advisor can analyze images very fast and look for things that physicians may miss. Watson already is used extensively in India where there is a large shortage of doctors.

· Deep Thunder provides accurate weather-forecasting service.

· Hilton Hotels are using Watson-based “Connie Robot” in their front desks. Connie did a superb job in experiments, and its service is improving.

Clark (2016) reports that 1 billion1 billion people will use Watson by 2018. This is in part because IBM Watson is coming to smartphones as an advisor. For more, see Noyes (2016).

Section 12.6 Review Questions

1. Define robo advisor.

2. Explain how robo advisors work for investments.

3. Discuss some of the shortcomings of robo advisors for investments.

4. Explain the people-machine collaboration in robo advising.

5. Describe IBM Watson as an advisor.

12.7 Implementation Issues

Several implementation issues are unique to chatbots and personal assistants. Examples of representative systems are described next.

Technology Issues

Many chatbots, including virtual personal assistants, have imperfect (but improving) voice recognition. There is no good feedback system yet for voice recognition systems to tell users, in real time, how well it understands them. In addition, voice recognition systems may not know when to do a current task and need to ask for human intervention.

Chatbots that are internal to organizations need to be connected to an NLP system. This may be a problem, but a bigger one may exist when chatbots are connected to the Internet, due to security and connectivity difficulties.

Some chatbots need to be multilingual. Therefore, they need to be connected to a machine language translator.

Disadvantages and Limitations of Bots

The following are points (which were observed at the time this book was written during 2017 and 2018) regarding bots’ disadvantages and limitations; some will disappear with time:

· Some bots provide inferior performance, at least during their initiation, making users frustrated.

· Some bots do not properly represent their brand. Poor design may result in poor representation.

· The quality of AI-based bots depends on the use of complex algorithms that are expensive to build and use.

· Some bots are not convenient to use.

· Some bots operate in an inconsistent manner.

· Enterprise chatbots pose great security and integration challenges.

For methods to eliminate some of the disadvantages and limitations, see Kaya 2017.

Virtual Assistants Under Attack

Cortana, Siri, Alexa, and Google Assistant are under attack by people who are enraged at machines in general, or just like to make fun of them. In some cases, the bots’ administrators try to compose a response to the attacks; in other cases, some machines provide senseless responses to the senseless attacks.

Quality of Chatbots

While the quality of most systems is not perfect, it is improving over time. However, the quality of those that retrieve information for users and are properly programmed can do a perfect job. Generally speaking, the more a company invests in acquiring or leasing a chatbot, the better its accuracy will be. In addition, bots that serve a large number of people, such as Alexa and Google Assistant, exhibit an increasing level of accuracy.

Quality of Robo Advisors

Given the short time since the emergence of robo advisors for financial services, it is difficult to assess the quality of their advice. Backend Benchmarking publishes a quarterly report (theroboreport.com) regarding robo advisor companies. Some reports are free. According to this service, Schwab’s Intelligent Portfolio Robot was the top performer in 2017. However, note that portfolio performance needs to be measured for the long run (e.g., 5 to 10 years).

A major issue when engaging bots is the potential loss of human touch. It is needed to build trust and answer complex questions so customers can understand bots’ answers. Also, bots cannot bring empathy or a sense of friendship. According to Knight (2017b), there is a solution to this. First, bots should perform only tasks that they are suited to do. Second, they should provide a visible benefit to the customer. Finally, because the bots face customers, the interactions must be fully planned to make sure the customers are happy.

In addition, note that robo advisors provide personalized advice. For information as to which robo may be best for you based on your objectives, see Eule (2017), who also provides a scorecard for the leading companies in the field. Finally, Gilani (2016) provides a guide for robo advisors as well as their possible dangers.

Microsoft’s Tay

Tay was a Twitter-based chatbot that failed and was discontinued by Microsoft. It collected information from the Internet, but Microsoft had not given the bot the knowledge of how to deal with some inappropriate material used on the Internet (e.g., trolls, fake news). Therefore, Tay’s output was useless and frequently offended its users. As a result, Microsoft discontinued the service of Tay.

Setting Up Alexa’s Smart Home System

Alexa is useful in controlling smart homes. Crist (2017) proposed a six-step process for how to use Alexa in smart homes:

1. Get a speaker (e.g., Echo).

2. Think about the location of the speaker.

3. Set up the smart home devices.

4. Sync related gadgets with Alexa.

5. Set up group and scene.

6. Fine-tune during the process.

These steps are demonstrated at cnet.com/uk/how-to/how-to-get-started-with-an-alexa-smart-home/.

Constructing Bots

Earlier, we presented some companies that provide development platforms for chatbots. In addition, several companies can build bots for users, so they can also build a simple bot by themselves. A step-by-step guide with the tools used is provided by Ignat (2017). The bot was constructed on Facebook Messenger. Another guide for creating a Facebook Messenger bot is provided by Newlands (2017b), who suggested the following steps:

1. Give it a unique name.

2. Give customers guides on how to build a bot and how to converse with it.

3. Experiment in making a natural conversation flow.

4. Make the bot sound smart, but use simple terminology.

5. Do not deploy all features at the same time.

6. Optimize and maintain the bot to constantly improve its performance.

There are several free sources for building chatbots. Most of them include “how-to” instructions. Several messaging services (e.g., Facebook Messenger, Telegraph) provide both chatbot platforms as well as their own chatbots. For a 2017 list of enterprise chatbot platforms and their capabilities, see entrepreneur.com/article/296504.

Using Microsoft’s Azure Bot Service

Azure is a comprehensive but not a very complex bot builder. Its Bot Service provides five templates for quick and easy creation of bots. According to docs.microsoft.com/en-us/bot-framework/azure-bot-service-overview/, any of the templates shown in Table 12.1 can be used.

Table 12.1 Azure’s Templates

Template	Description
Basic	Creates a bot that uses dialogues to respond to user input.
Form	Creates a bot that collects input from users via a guided conversation that is created using Form Flow.
Language understanding	Creates a bot that uses natural language models (LUIS) to understand user intent.
Proactive	Creates a bot that uses Azure Functions to alert users of events.
Question & Answer	Creates a bot that uses a knowledge base to answer users’ questions.

NOTE: Microsoft also provides a bot framework on which bots can be constructed (similar to that of Facebook Messenger). For Microsoft’s Bot and a tutorial, see Afaq (2017).

For a detailed tutorial for creating bots, see “Create a Bot with Azure Bot Service” at docs.microsoft.com/en-us/bot-framework/azure-bot-service-overview/.

Chapter Highlights

· Chatbots can save organizations money, provide a 24/724/7 link with customers and/or business partners, and are consistent in what they say.

· An expert system was the first commercially applied AI product.

· ES transfer knowledge from experts to machines so the machines can have the expertise needed for problem solving.

· Classical ES use business rules to represent knowledge and generate answers to users’ questions from it.

· The major components of ES are knowledge acquisition, knowledge representation, knowledge base, user interface, and interface engine. Additional components may include an explanation subsystem and a knowledge-refining system.

· ES help retain scarce knowledge in organizations.

· New types of knowledge systems are superior to classical ES, making ES disappear.

· We distinguish three major types of chatbots: enterprise, virtual personal assistants, and robo advisors.

· A relatively new application of knowledge systems is the virtual personal assistant. Major examples of such assistants are Amazon’s Alexa, Apple’s Siri, and Google’s Assistant.

· Knowledge for virtual personal assistants is centrally maintained in the “cloud” and it is usually disseminated via a Q&A dialog.

· Personal assistants can receive voice commands that they can execute.

· Personal assistants can provide personalized advice to their owners.

· Special breeds of assistants are personal advisors, such as robo advisors, that provide personalized advice to investors.

· Recommenders today use several AI technologies to provide personalized recommendations about products and services.

· People can communicate with chatbots via written messages, voice, and images.

· Chatbots contain a knowledge base and a natural language interface.

· Chatbots are used primarily for information search, communication and collaboration, and rendering advice in limited, specific domains.

· Chatbots can facilitate online shopping by providing information and customer service.

· Chatbots work very well with messaging systems (e.g., Facebook Messenger, WeChat).

· Enterprise chatbots serve customers of all types and can work with business partners. They can also serve organizational employees.

· Virtual personal assistants (VPAs) are designed to work with individuals and can be customized for them.

· VPAs are created as “native” products for the masses.

· A well-known VPA is Amazon’s Alexa that is accessed via a smart speaker called Echo (or other smart speakers).

· VPAs are available from several vendors. Well known are Amazon’s Alexa, Apple’s Siri, and Google’s Assistant.

· VPAs can specialize in specific domains and work as investment advisors.

· Robo advisors provide personalized online investment advice at a much lower cost than human advisors. So far, the quality seems to be comparable.

· Robo advisors can be combined with human advisors to handle special cases.

Questions for Discussion

1. Some people say that chatbots are inferior for chatting. Others disagree. Discuss.

2. Discuss the financial benefits of chatbots.

3. Discuss how IBM Watson will reach 1 billion people by 2018 and what the implications of that are.

4. Discuss the limitation of chatbots and how to overcome them.

5. Discuss what made ES popular for almost 30 years before their decline.

6. Summarize the difficulties in knowledge acquisition from experts (also consult Chapter 2).

7. Compare the ES knowledge-refining system with knowledge improvement in machine learning.

8. Discuss the difference of enterprises’ use of chatbots internally and externally.

9. Some people say that without a virtual personal assistant, a home cannot be smart. Why?

10. Compare Facebook Messenger virtual assistant project M with that of competitors.

11. Examine Alexa’s skill in ordering drinks from Starbucks.

12. Discuss the advantages of robo advisors over human advisors. What are the disadvantages?

13. Explain how marketers can reach more customers with bots.

14. Are robo advisors the future of finance? Debate; start with Demmissie (2017).

15. Research the potential impact of chatbots on work and write a summary.

Exercises

1. Compare the chatbots of Facebook and WeChat. Which has more functionalities?

2. Enter nuance.com and find information about Dragon Medical Advisor. Describe its benefits. Write a report.

3. Enter shopadvisor.com/our-platform and review the platform’s components. Examine the product’s capabilities and compare them with those of two other shopping advisors.

4. Enter chatbots.org/ and join a forum of your interest. Also explore research issues of your interest. Write a report.

5. There is intense competition between all major tech companies regarding their virtual personal assistants. New innovations and capabilities appear daily. Research the status of these assistants for Amazon, Apple, Microsoft, Google, and Samsung. Write a report.

6. Some people believe that chatbots will change how people interact with the Internet and browse online. Prepare a report regarding this.

7. Explain why is Amazon’s Echo needed to work with Alexa? Read howtogeek.com/253719/do-i-need-an-amazon-echo-to-use-alexa/. Write a report.

8. Find out how Simon Property Group is using chatbots across over 200 shopping malls. Write about the benefits to different types of users and to the company.

9. Read recent information about enterprise bots. Write a report.

10. Enter gravityinvestments.com/digital-advice-platform-demo. Would you invest in this project? Research and write a report.

11. Enter visirule.co.uk and find all products it has for expert systems. List them and write a short report.

12. Research the role of chatbots in helping patients with dementia.

13. Find information on the Baidu’s Melody chatbot and how it works with Baidu Doctor.

14. Pose a question related to a chatbot on quora.com. Summarize the answers received in a report.

15. Nina is an intelligent chatbot from Nuance Communication Inc. that works for Alexa Internet of Things (IoT), smart homes, and more. Find information and write a report about Nina’s capabilities and benefits.

16. Microsoft partners with the government of Singapore to develop chatbots for e-services. Find out how this is done.

17. Study the Tommy Hilfiger Facebook Messenger bot. Find out how it is (and was) used in the company’s marketing campaigns.

18. Two comprehensive building tools for chatbots are Botsify and Personality Forge (personalityforge.com). Compare the tools. Write a report.

19. Find information about the Alibaba-backed robo advisor Youyu by Yunfeng’s Investment. What is unique about this service? Start by visiting http://www.international-adviser.com/news/1035281/alibaba-backed-retail-robo-adviser-youyu-launches-honk-kong/.

20. Enter exsys.com. Select three case studies and explain why they were successful.

21. It is time now to build your own bot. Consult with your instructor about which software to use. Have several bots constructed in your class and compare their capabilities. Use Microsoft’s Azure if you have some programming experience.