IT Client Last/Final Report

1

ITI220 [Section 09] [Fall 2015] [10/28/15] – [Nikita Abraham], [ICP4 – Final Report] Executive Summary

My client, Shibu Daniel, currently works at Bank of New York Mellon (BNY Mellon) as

a Vice President of User Experience (UX). Bank of New York Mellon provides investment

services to various individuals and institutions. As an organization, they focus on asset and

wealth management in the forms of stocks, real estate, bonds, and other similar securities. BNY

Mellon also provides trade execution services for their clients. Essentially, this means that if an

institution would like to purchase an investment vehicle, the organization can act as brokers on

their behalf and consolidate the trade. In addition to the aforementioned points, the company also

delivers necessary financial advice and provides credit offerings when appropriate.

The UX team currently works under “Client Technology Solutions,” a special branch

under the company’s massive IT department. The “Client Technology Solutions” sector supports

multiple lines of business by providing technology solutions to internal and external clients.

Although UX is a broad field, spanning a wide variety of topics, my client’s responsibilities

include creating conceptualizations for technology solutions, designing low and high fidelity

mockups and prototypes, running usability tests, and conducting and analyzing user research.

As an investment company with global reach, a significant amount of data is constantly

being collected and stored. Understanding, organizing, and using that data is synonymous with

the company’s own success. Under the “Client Technology Solutions” sector, real insights can

be made into technological solutions if individuals can make educated predictions. As a result of

this influx of data, my client’s information need is most closely related to Big Data. For a

preliminary understanding, he wanted me to research the following topics within the scope of

Big Data: what is Big Data and why it is important, how can companies use visualizations to

2

organize data, how have companies done so (real-world examples or case studies), and what

frameworks or software are available for data processing in this way.

I was pleased to see that there was no shortage of relevant and reputable resources for my

client’s information need. A brief breakdown of the most pertinent information is listed below:

• Information dashboards can be used to display visualizations (Choudhary, 2014).

• There are a number of modern data analytics and visualization tools available to my

client, such as Hadoop (Ishwarappa & Anuradha, 2015; Minelli, Dhiraj, & Chambers,

2013) and VisIT (Childs, 2013).

• There are some visualization methods that work better than other depending on the user

need and the kind of information being displayed (Keahey, 2013).

• Challenges with Big Data are related to the characteristics of Big Data – namely, its

volume, its variety, and its velocity (Intel, 2013; Franks, 2014; LaValle, Lesser,

Shockley, Hopkins, & Krushwitz, 2011).

The information sources will be used to help my client to represent data sets more

effectively, to make use of BNY Mellon’s current data analytics software more efficiently, and to

organize data when designing technology solutions. The insights gained will also be relevant to

consumer behavior, to current business models that support data analytics, and to mitigate asset

risk.

Discussion of Search Technologies and Search Strategies Used

Summary of technologies used to access information and conduct research.

For my client’s information need, I used a combination of search engines and databases

to find relevant resources. I used Google’s search engine to find a majority of my non-scholarly

material. Big Data, as a whole, is a relatively recent topic and from my preliminary search, I saw

that much of the material was produced within the last five years. I felt that search engines

provided more business-oriented articles. Many of the sources were found in either online IT-

3

related magazines or information sheets provided by companies who specialize in data analytics.

I attempted to use Google Scholar to search for peer-reviewed material. Unfortunately, while the

articles themselves would have fulfilled my client’s information need, for many of them, only the

abstract was provided. Some of the sources on Google Scholar were also very outdated, even

preceding the topic of modern Big Data analytics.

I also used Rutgers University Libraries Articles+, a resource provided by Rutgers

University that searches multiple academic databases and aggregates the results into an

organized list. I primarily used Business Source Premier database and ScienceDirect database to

find relevant scholarly work pertaining to my topic. I found that while the search engines pulled

up a business-oriented perspective, the databases brought up more technologies that specifically

work with Big Data. For example, a number of my sources detailed Hadoop, a software

tool being used by many large organizations for data analytics.

Ultimately both methods provided relevant and reputable material for my client’s

information need. While using the search engine, I had to be more cautious of how reputable a

source was, whereas for databases, I had to be more careful about how relevant a source was.

Summary of search strategies used to access information and conduct research.

While using the search engine, I used parentheses to refine my searches. For example, the

search query ((big data) business AND management) brought up results about big data first and

then limited the search by searching for the words Business and Management within those

results. In Google, I used the related words feature at the bottom of the SERP to find other

relevant keyword combinations. In doing so, I learned that “Big Data” could also be represented

as “data analytics.” When I kept getting redundant searches for my database query, I used the

other keywords I learned from my search engine query to facilitate my information seeking.

4

On the databases, once the articles were pulled up on the results page, I refined my

searches in several ways. I limited my articles to “Full Text” and “Scholarly/Peer Reviewed” -

this would assure that my searches would only pull up scholarly sources that allowed students to

access the full text. I also limited the “Publication Date” to only bring up articles within the last

ten years. Considering the ever-changing nature of the IT industry, and the recent emergence of

Big Data as a topic, this was very important to the success of my results. The Rutgers Libraries

source also pulls up entries from databases that are outside of my client’s industry. For example,

initially I was receiving results related to analyzing Big Data within Science. By limiting the

providers to just Business Source Premier, and other IT-related databases, my results were

significantly improved. I also used the keywords within relevant articles (under the Abstract) to

broaden my searches.

Discussion of Communications with and Feedback from IT Professional (Client)

Summary of communications

My client and I began communications during the second week of September. For

the “Client Interview” portion of the IC Project, I met with my client on September 25, 2015.

The entire interview took about 30-40 minutes to complete. During that time, we discussed the

following topics: background information about BNY Mellon as a company, background

information about my client’s position such as the department he works under and the

responsibilities he must attend to, and finally, background information about the client’s topic in

relation to his information need.

Between the Client Interview and the Annotated Bibliography due date, I had two follow-

up discussions. On September 28, 2015 my client sent me a more detailed explanation of BNY

Mellon’s key services via email. The document he shared is one that he gives to his own clients

5

when they would like a general overview of what BNY Mellon does. On October 5, 2015, we

also spoke briefly on the phone to clarify the scope of the information need.

Summary of the IT client’s overall feedback and recommendations

Overall, my client seemed very pleased by the resources that I found for his information

need. In the IT Client Feedback form, he mentioned that all of the references provided a detailed

enough description to get an overall idea of the article. He was particularly happy to see that all

of the articles were written within the last couple of years and that I included a

detailed description of the authors’ credentials. In a follow-up conversation I had earlier this

week, my client mentioned that he was already looking through some of the articles again for a

point of reference.

The client did not give me any suggestions for improvement, even in the follow-up phone

call. I did receive some minor revisions regarding APA format from the Professor - the needed

changes are reflected in the revised Annotated Bibliography shown below.

Final Annotated Bibliography Childs, H. (2013). VisIT: An end-user tool for visualizing and analyzing very large data.

eScholarship University of Alabama. Retrieved from http://escholarship.org/uc/item/69r5m58v#page-7.

This particular article discusses the tool, VisIT that was originally developed in response to Big Data processing. It is an open source tool that focuses on three main functions: enabling data understanding, providing scalable support for large sets of data, and creating a robust product that is simple for the end user. In addition, the paper outlines the software’s design, architecture, and user interface concepts. In addition to the software’s initial successes, the author also outlines companies who have made use of VisIT. The article was found on eScholarship, an open-access repository of scholarly sources under California Digital Library. Even though it is not in a traditional database, the material itself is scholarly and the original study was published in a scholarly journal. Childs’ and his colleague’s work was also supported by the U.S. Department of Energy; the document was prepared as an account of the work sponsored by the United States Government. In addition to the aforementioned points, the comprehensive list of references provided at the end emphasizes the

6

author’s credibility. While the source does not cover a wide variety of topics as some of the other articles do, this particular case study fulfills my client’s need for real world examples of companies who have implemented visualization software, or have made data processing software. Depending on the long-term goals for my client’s team, perhaps the insights gained from this case study could provide recommendations for BNY Mellon’s data processing tools. Choudhury, S. (2014). The future of information dashboards. UX Magazine. Retrieved from

https://uxmag.com/articles/the-future-of-information-dashboards The entire article focuses on information dashboards as a means of visualizing large data sets. There are five primary categories that she divides the article into: (1) dashboards offered by independent software vendors as part of their data analytics tools, (2) the dashboard for the user on the go (mobile smartphones and tablets), (3) dashboards with high location intelligence quotients, (4) dashboards for real time data and (5) dashboards future predictive data. Shilpa Choudhury has published a number of articles in other well-known blogs such as Wired, ReadWrite, and Visual.ly. In addition, UX Magazine is a well-known online magazine that focuses on experience design. Practitioners and industry leaders who are well versed in the UX field write the majority of their articles. At one point Choudhury mentioned nine analytic predictions made by IIA (The International Institute for Analytics) for 2014. While the information has not been updated, many of the elements from the previous predictions were repeated or relevant for the following year. This article is relevant for my client because it provides examples of different kinds of dashboards (a specific visualization), along with real- world examples of companies that have implemented them. By analyzing this article, my client will be able to assess what his user needs are, what his team is trying to accomplish, how to execute it, and additional resources. Franks, B. (2014). Making big data actionable: How data visualization and other tools change the

game [Webinar]. Harvard Business Review. Retrieved from https://hbr.org/2014/05/making-big-data-actionable-how-data-visualization-and-other- tools-change-the-game

In this webinar, hosted by Harvard Business Review (HBR), Franks begins by defining big data and discussing how visualizations were used in the past to aid cognition. He transitions to modern day, outlining how data visualizations have changed according to the needs of the modern user as well as advances in technology. Today, customized visualizations are tied to a specific analysis – in that way, modern visualization tools must be interactive, interconnected, collaborative, flexible and enabling. Franks’ greater point is essentially that businesses need to use data to support decision making, rather than just reporting what occurred – this calls for a change in methodology. Perhaps it is the nature of a webinar session, but there were no direct links or references provided at the end of the presentation. Nonetheless, Franks does an excellent job maintaining currency by discussing contemporary software vendors, companies who have implemented visualizations, and other relevant anecdotes. He also works for Teradata, a company that provides services in data warehousing, big data and analytics, and marketing applications. His experience within the

7

field of big data can clearly be seen. In addition, the magazine that hosted the webinar, HBR, is a reputable organization under Harvard University that provides a bridge between academia and various enterprises. As such, the articles are reliable but also relevant to the industry. I would recommend this webinar primarily because it emphasizes a change in the current business model (the need to make decisions, not just analyze them). Franks provides practical advice as an industry professional that would be relevant to my client’s need for integrating visualizations into big data processing.

Hoffer, D. (2014). What does Big Data look like? Visualization is key for humans. Wired

Magazine. Retrieved from http://www.wired.com/insights/2014/01/big-data-look-like- visualization-key-humans/

The basis of Hoffer’s entire article rests on the idea that big data needs to be more human. In developing his points, he establishes that information visualization is a means of wayfinding, not simply a means of organizing data. In doing so, he also reiterates the need for data visualizations to be comprehensive (displaying multiple levels of information according to user needs), to be scalable, and to be simple. Hoffer provides several examples of companies who are successfully humanizing big data and how they have implemented the considerations mentioned above. At the end of the article, his call to action seems to be that since big data itself is constantly evolving, the technology we use to organize that data must be robust enough to process it – just like every other piece of software the IT industry uses. The article was written in 2014 and is still very current considering the real world examples that Hoffer mentions (such as Google Maps). That being said, each of the sources that he mentioned had a link to the original article or a related source that provided further information. When the article was first written, the author was the head of User Experience at Declara, but is currently a Design Director at McKinsey & Company. His experience is well translated in the article itself. This would be a very relevant source for my client considering that Hoffer has a background in User Experience and the article seems to focus on data processing within that particular subfield of IT. The article as a whole is a good baseline for the technology necessary to implement data processing through information visualizations. Intel. (2013). Big data visualization: Turning big data into big insights [Web document]. Intel.

Retrieved from http://www.intel.com/content/dam/www/public/us/en/documents/white- papers/big-data-visualization-turning-big-data-into-big-insights.pdf

The paper starts off with an overview of the current IT landscape as well as the problems that are leading industry professionals to data discovery tools such as information visualization. Essentially this new form of business intelligence involves integrating data from a variety of user-based sources that are then displayed in an interactive and easily understood format. The author(s) describe(s) how challenges with big data have to do with its volume, variety and velocity (the three V's of big data) as well as the increasing availability of mobile devices. The paper then outlines key features of visualization-based data discovery, how it addresses the aforementioned challenges, how to protect data quality, and how to establish data governance policies.

8

While no author is provided, the source document is from Intel – a well-known company that has always stood at the forefront of technological innovation. In addition to providing a comprehensive list of endnotes, the article mentions the Gartner BI (business intelligence) Summit often. Gartner is a leading information technology research and advisory company; the Summit meeting is an important event within the industry. The inclusion of such sources reiterates the document’s reliability and credibility. This is relevant to my client because it provides a list of necessary conditions from a business perspective, including security. Considering that my client works in the banking industry, security is an important part of data processing and risk mitigation. Ishwarappa, & Anuradha, J. (2015). A brief introduction on big data: 5Vs characteristics and

Hadoop technology. Procedia Computer Science, 48. Retrieved from ScienceDirect database, through Rutgers University Libraries Articles+.

This paper discusses the 5V’s of big data (volume, velocity, value, veracity, and variety) in detail. Moving forward, the paper also looks at challenges that companies face when handling big data, namely, challenges involves capturing, analyzing, storing, searching, sharing, visualization, and transferring data. The article primarily features Hadoop, an open source distributed data processing tool that is being used by many industry professionals. This article was found on the ScienceDirect database, making it a reliable and accurate source. Considering that Apache’s Hadoop is still a popular solution to companies for their information needs, the article is also timely and current. This would benefit my client by going into detail about Hadoop and its data processing frameworks. His own team can implement similar frameworks whether they use in-house software or a third party source. Keahey, A.T. (2013). Using visualization to understand big data. IBM Software Business

Analytics. Retrieved from https://tdwi.org/~/media/E3362B4A0E184F75AB29403 676C4C3CD.pdf.

This article discusses how businesses can make use of visualizations when making sense of the following five categories of user needs: (1) simple customer data (2) customer data involving time (3) customer sentiment, which is essentially how people feel about the company (4) measuring customer relationships and (5) customers at different levels. In each of these five categories, the author gives one or two examples of the most effective kind of visualization for that particular need. For example, when measuring customers at different levels, a hierarchal visualization conveys the information better than other forms. In addition to the simple data visualizations that the author advises his reader to use, Keahey also promotes IBM’s data analytical services. As far as accuracy and credibility, this particular article is one of several resources that IBM shows to potential clients who are interested in IBM’s data processing services. As such, I can see that the information is well researched and up-to-date to fit the current IT environment; it even has a list of references at the end of the article. In addition, the author is one of IBM’s Visualization Science and Systems experts, making him a reliable and knowledgeable source. I can see this being useful for my client in addressing his information need for what kinds of

9

visualizations can be used to display data. Not only does Keahey showcase several visualization techniques, he also clearly addresses which situations each of them would be appropriate for. LaValle, S., Lesser, E., Shockley, R., Hopkins, M.S., & Krushwitz, N. (2011). Big Data,

analytics and the path from insights to value. MIT Sloan Management Review, 52(2). Retrieved from Business Source Premier database, through Rutgers University Libraries Articles+.

The purpose of this article is to answer the question, how are organizations using analytics to gain insight and make decisions from big data? Essentially, there were three main findings from the report: that top organizations are more likely to apply analytics to activities, that managerial and cultural aspects are the biggest challenges to implementing analytics, and that being able to visualize data in different ways is increasing in overall value for the company. The article goes into detail about a new methodology for adopting analytics within an organization, along with several recommendations from a business perspective. Although the article was not found on a database, MIT Sloan Management Review is a reliable and dependable resource, known for their well-researched articles and relevant topics. In order to better understand the challenges that companies have faced and have overcome for analytics, MIT Sloan Management Review along with IBM Institute for Business Value conducted a survey with more than 3000 industry professionals (business managers, analysts, and executives). In addition both organizations interviewed a number of academic experts and subject matter experts from various related disciplines. In doing so, the article provides a holistic but specific viewpoint. All of the authors have a significant amount of experience in the Business Analytics realm of IT. It would be relevant to my client’s information need for business models that work with analytics implementation. The recommendations would provide practical business advice in how to overcome common problems with using visualizations in big data processing. Minelli, M., Dhiraj, A., & Chambers, M. (2013). Big data, big analytics: Emerging business

intelligence and analytic trends for today's businesses [ebook]. Wiley CIO Series. Retrieved from eBook Collection database, through Rutgers University Libraries Articles+.

This particular eBook provides a holistic approach to big data. The first couple of chapters consider what big data is and why it is relevant, along with some industry examples of it in action. Considering my client’s needs, the third chapter on “Big Data Technology” would be the most beneficial. This particular chapter goes into detail about Hadoop, a well-known data analytics company (including their business model, critical components, and overall goals). The chapter also considers the importance of data discovery when compared to more traditional data processing models. The book is written for business and IT professionals with the purpose of helping such individuals integrate big data analytics within their own organization. Combined, the authors have years of experience related to business analytics solutions, technology-based solutions, and related decision sciences. In addition to their expertise, the notes or references at the end of each chapter, and its addition in the EBSCOHost database all reiterate the eBook’s credibility. The article will be relevant in addressing the client’s need for different

10

ways to visualize data, as well has the need to outline business processes (necessary parties, technologies, and management models) involved when implementing a data visualizing software. Not only are the topics expansive, they are explained in detail. Visualizations make big data meaningful. (2014). Communications of the ACM, 57(6), 19-21.

Retrieved from Business Source Premier database, through Rutgers University Libraries Articles+.

The author of this article provides a unique perspective by describing data visualization as an expressive medium, where ultimately the “artist” or data analyst can choose what pieces of information to show or withhold from the viewer. Moving forward, the article as a whole discusses advances in technology that set the stage for more effective data visualization practices such as the availability of storage, the efficiency of cloud computing, and the rise of software tools such as Apache Hadoop – all of which, simplify data processing. Other important aspects that the article mentions include the importance of “humanizing” these visualizations, and looking for new and more dynamic ways to display data outside of simple charts and graphs.

While an author is not provided, the article was found on the Business Source Premier database and published in the academic journal Communications of the ACM, making it highly reputable. The article itself was written in 2014 and none of the technologies or techniques mentioned are outdated - which is to be expected considering the recency of big data in general. Even though it does not address the business aspect of big data processing, the article will be relevant in describing how data visualizations have changed to fit the current IT environment and what advances are now possible. The details mentioned in this article are not expansive in its coverage, but it will fulfill the client’s information need for general information about data visualizations, what can be accomplished with current technologies, as well as the future of big data. IC Project Reflection

Overall, I really enjoyed the Information Consultant Project. Because it was divided over

the course of a semester, looking back at all that I had done was somewhat fulfilling. As I was

going over the Annotated Bibliography and the positive feedback that I received from my client,

I was happy to see that I really understood his information need. Selecting an information need

that was neither too broad nor too narrow was probably the most challenging for me. But I

realized that spending more time refining the information need actually helped me determine

which sources were relevant much faster later in the process.

11

I had the privilege of being able to interview my father - an IT professional who has been

serving in the industry for years and is also working in the field (UX) that I would eventually like

to go into. Learning more about his role and how it fits in with the needs of BNY Mellon was

fascinating, because I have never actually seen how User Experience fits in on a corporate level.

I definitely underestimated what I can learn from him and am surprised that I haven’t used him

more as a resource - that, of course, has changed since the start of this project. It is one thing to

study how to satisfy our information needs, but quite another to actually fulfill the information

need of a client in our intended field of interest. As is usually the case, I learned so much more

being able to put into practice what we have been discussing in class.

Between the information consultant and the information client, I realized that one thing is

essential to fulfilling one’s information need - communication. I had to communicate with my

client several times during the course of this assignment. I found it interesting that as I took more

time to follow-up and refine the information need, the client also realized that his information

need was too broad. From this particular instance, I can see that as information seekers, we have

difficulty pinpointing exactly what kinds of information we’re looking for. Like Bates’ Berry

Picking model states, through a series of searches, our query is constantly subject to change, and

as a result, it is not uncommon for our information need to change somewhere along that process.

When I realized from my preliminary searches that the information need was too broad, I asked

my client more questions to help him develop a clearer idea of what he was looking for.

Fortunately, with the extra communication, we were able to break up the broad topic of Big Data

into smaller, measurable information needs.

I have a better appreciation for how to refine searches on both search engines and

databases. Previously, I was placing a lot of reliance of the system itself to find what I needed.

12

But now I know how to take control of the querying and retrieval process so that I can find

resources that match exactly what I’m looking for. Evaluating information sources thoroughly

takes a significant amount of time. When I was preparing the Annotated Bibliography, I used the

chart we made during the practice exercises to fill in each of the necessary measures. But

because each source was thoroughly developed, my client knew exactly what each of the articles

discussed and even what the drawbacks of the articles might have been. When I followed up with

my client, one of the questions that I asked him was whether his team-members do something

similar when they are researching information. Essentially, he mentioned that information

seeking is a never-ending process in his line of work - they are constantly researching new

technologies. The only difference is that they take that information a step further, and use it to

make technology solutions. I think that the skills that I’ve gained in retrieval will give me a

significant advantage if in the future I find myself in a similar field of work.