Sentiment Analysis is a growing feature of several applications that we see today. The trend is not surprising given the fact that the billions of messages per day are flowing in and out of several services on the web. Twitter, being one of the largest generators of data, is seeing its data mined in various ways to provide everything from movie reviews, predicting the result of a sporting event and much more.
TweetSentiments.com is a website that is joining a growing number of services that analyze Twitter data and provide sentiment analysis about a particular Twitter user or a topic, in addition to several other analytical trends. It is based on a machine learning method called Support Vector Machines(SVM). TweetSentiments.com uses an implementation of SVM developed at Taiwan National University called LIBSVM. You can use the website to analyze Tweets from a particular user or choose to enter a Topic and get an analysis done. The results are surprising at times and tweets get marked as negative, when the user did not really intend it that way.
TweetSentiments.com also has an API and currently exposes 3 methods. The API is REST-like and supports JSON.
The 3 sample methods as discussed at the blog are given below:
- Sentiment on tweets http://data.tweetsentiments.com:8080/api/analyze.json?q=<text to analyze>
- Sentiment on topics http://data.tweetsentiments.com:8080/api/search.json?topic=<topic to analyze>
- Sentiment on users http://data.tweetsentiments.com:8080/api/search.json?user=<twitter user to analyze>
Data mining is becoming important by the day. In the past we had covered Mombo API that analyzes tweets about movies and provides recommendations. What is your experience about current Sentiment Analysis engines and sites out there?