The objective of this week's deliverable is to provide Robert M. Lopez with an executive summary of your findings. In a 2-3 page paper, summarize your findings. Give proof for your analysis. Within your paper, explain the process you did to come up with

4

Running head: Sentiment analysis

Sentiment Analysis

Lisa Garay

Rasmussen College

Authors Note

This paper is being submitted for Anastashia Rashtcian’s B288 Business Analytics course.

Sentiment analysis has played a significant role in the concurrent marketing field, specifically in product marketing. According to Somasundaran, Swapna, (2010), the process’ operational module is structured on a data mining sequence, whereby the end users of given particulars the feedback pertaining a used product, or any experience. The feedback primarily comprises the feelings, attitudes, views, and satisfactory compliments about the same implement at hand (Wan, S. & Angryk, R. A., 2007).

The most common channel for obtaining this data is through the use of online portals. These portals are deemed to be apt for the task based on its diversity around the globe, extensive connectivity among a wide number of users, easy access, and fast transfer of information.

Sentiment analysis is necessary for evaluating the viability of an asset, service or aspect based on the ideas and opinions of other parties on the same implement (Wan, S. & Angryk, R. A., 2007). The ideas can in turn influence different responses on the recipients centered on the analysis made. The views can impact the viability of the element at hand either positively or negatively (T. Mullen and N. Collier, s., 2004). For example, the data about the online based components can be collected through online reviews and testimonies, along with the ratings awarded.

Various tools are available to facilitate operation of sentiment analysis. For instance, technological enhancements have led to the unleashing of the Web 2.0, which is highly

compatible to various data mining widgets. The mining tools include support for podcasting, tagging, blogging, RSS support, social networking and social bookmarking. In line with the tools, there are various computational strategies that can be opted for as well. These include data-driven strategies among them being Maximum Entropy, Naïve Byes, Voted perceptions, and SVN. Another common technique is the use of Cognitive Psychology (Wan, S. & Angryk, R. A., 2007).

A popular media in use today for Sentiment analysis is Twitter. The online media allows sharing of short texts, media links and videos. Steps used for its operation flow between the use of POS-tagged n-gram components, hash tags, text normalization, tackle spam and lastly entity specific sentiment analysis (Somasundaran, Swapna., 2010).

Sentiment analysis is a potent assessing technique that is used to gauge both the strengths and weaknesses of various tools by the use of online platforms. In the end, it aids in coming up with efficient output that can be used for either usability decisions by end users or formation decisions by the producer (T. Mullen and N. Collier, s., 2004).

References

1. T. Mullen and N. Collier, Sentiment analysis using support vector machines with diverse information sources, In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 412–418, 2004.

2. Somasundaran, Swapna, Discourse-level relations for Opinion Analysis, PhD Thesis, University of Pittsburgh, 2010.

3. Wan, S. & Angryk, R. A., Measuring semantic similarity using worldnet-based context vectors., in ‘SMC’07’, pp. 908–913, 2007