Sentiment Analysis Challenge
The Sentiment Analysis challenge posed by the Language Computer Corporation is a text analytics problem aimed at engineers and researchers ranging from having no previous exposure to NLP, and all the way up to extensive research and experience in the field. The task is to design a software approach to assess textual movie ratings and assign each a likely rating of 0-5. These system scores (ratings) are then compared against actual ones (provided by the reviewer) to minimize the Mean Square Error (MSE).
Sentiment analysis systems are used to understand how masses of individuals feel about various companies, products, decisions, factors, and various other things. More complex variants can extract concerns about, as well as differentiating variables between competing services. Given the exploding rate at which people are producing text on the web (e.g. blogs, posts, documentation, tweets, classifieds, discussions, books, instructables), text analytics is quickly becoming an important field of study. The challenge posed here is an introduction to a very interesting problem space. If you'd like to learn more about what we're doing, and how you can help, please email us at email@example.com.
Fill out the form below with your name and email address (giving us a way to update you in case there are any bug fixes or updates to the project), and download the 'challenge.sentiment.zip' project on the following page. Extract the project to your disk, and use your favorite IDE or editor to familiarize yourself with the code and data. See the README in the project for further details.