Sentiment Analysis
Sentiment Analysis
Sentiment Analysis is a form of text analysis that determines the emotional polarity or attitude of a speaker, though their text, on a specific topic. This is an important topic because it allows searches to be more specific. For example, because of Sentiment Analysis, if a consumer is searching a product online they can search for good reviews and see the positive qualities of the product, or they can search for bad reviews and see the negative qualities of the product. Sentiment Analysis allows for this because it can determine the emotional polarity or the attitude by adding up the sentiment values of the key words found in the text of the speaker and determine if the review is positive, negative, or neutral.
Sentiment Analysis relates the material we have covered in MIS2502 because it is another way of sorting data and turning it into information. With Sentiment Analysis attitude can be used as another filter when sorting data. This could be useful when dealing with a database of customers as we did with the Pivot Table exercises. Because you can tell the attitude of a review without manually reading it you can add to a customer profile if they were satisfied or not.
An example of Sentiment Analysis in use is a case study that set out to determine if Twitter could be used to predict an election. In the study 104,003 published tweets, that mentioned any of the political parties or figures in the 2009 federal election of the national parliament of Germany, were analyzed using LIWC text analysis software. The study found that Twitter was in fact used to voice political opinions and could be a valuable tool in predicating elections. Although Twitter was used to voice political opinions the Sentiment Analysis did have limitations such as the same person being counted twice, not being able to filter by specific topic, and the software was not based on the particular type of text used in tweets such as short hand language.
References
Li, Nan, and Desheng Dash Wu. “Using Text Mining and Sentiment Analysis for Online Forums Hotspot Detection and Forecast.” Using Text Mining and Sentiment Analysis for Online Forums Hotspot Detection and Forecast. Elsevier, Jan. 2010. Web. 15 Apr. 2014.
Chakraborty, Goutam, Murali Pagolu, and Satish Garla. “Pages 1-34.” Text Mining and Analysis: Practical Methods, Examples, and Case Studies Using SAS. Cary, NC: SAS Institute, 2013. N. pag. Print.
Tumasjan, Andranik, et al. “Predicting Elections with Twitter: What 140 Characters Reveal about Political Sentiment.” ICWSM 10 (2010): 178-185.