“Finding some direction on the opinions expressed in conversations that matter.” 4 April 2017, from 1000 to 1630 University of Reading, London Road Campus, Reading, UK Seminar led by Dr Christine Bernadas As an entrée to the analysis of Big Data, Sentiment Analysis is widely used in the social media arena to extract information from content on the Internet, including texts, tweets, blogs, news articles, reviews etc. which would otherwise be unmanageable. This approach to analysis allows the researcher to extract full value from interactions in conversations, product reviews on forums and other websites (where there are thousands, if not hundreds of thousands of data points) which could not be done by manual processing. Sentiment Analysis allows the researcher to understand feelings that are being expressed. By performing textual analysis of informants‘ opinions, It can be an interesting tool for academic research. A typical usage of Sentiment Analysis is to compare customers’ opinions on different brands or predict movements in the stock market. Thus Sentiment Analysis can help a researcher determine whether a piece of text that should be regarded, for example as positive, negative, or neutral. This workshop will allow participants to be in a position to understand the importance of Sentiment Analysis, investigate ways of performing Sentiment Analysis, and practice more specifically some Twitter Sentiment Analyses. No in-depth knowledge of statistics is required. In the afternoon, participants will obtain most value if they have a twitter account and install R (https://cran.r-project.org/mirrors.html ) on a portable computing device. They will use some packages (TM, Wordcloud, plyr, TwitterR, stringR) and Tableau public (https://public.tableau.com/en-us/s/ ). Seminar content includes
- In what contexts should Sentiment Analysis be used? Definitions required to understand the functioning of Sentiment Analysis.
- In which circumstances does it work best? Who does sentiment analysis?
- What do you need to be successful with Sentiment Analysis?
- Techniques required to get the most from this technique including software issues.
- Processes required to organize a project of sentiment analysis. R software rapid overview
- Create a workable data set. Doing the analysis. Analyzing results