Geopinion
For a second-year course, the assignment was to create a web application using state-of-the-art technologies. As Web Science students interested in web analytics, we decided to focus on sentiment analysis (determining how people feel about a certain topic), in the form of analysing Tweets about a user-entered keyword. Thinking of how to present these results to the user in a clear and fun way, we ended up creating Geopinion.
Geopinion uses the location of all Tweets on the topic, and for each location, it analyses the Tweets’ content using lists of positive and negative words to determine whether users in that location feel positively or negatively (or neutral) about the topic. The results for each location are then shown to the user on a map, using markers with happy, green emoticons for positive sentiments, and unhappy, red emoticons for negative sentiments. Geopinion also offers a Twitter search by location and an API. Want to try it yourself? Take a look: http://sanderboelders.nl:5013/