Monday, September 14, 2009


TweetSentiments provides Sentiment Analysis on tweets using Natural Language Processing (NLP) and Machine Learning (ML) technologies.

The service analyzes tweets content in a chosen topic or tweets of a specified user, and divide them to positive, neutral, or negative tweets. The formal used is to determine the Sentiment Index is [100*((positive-negative)/total/2+0.5)] and the result categories the tweets to:
0 -> most negative
1..49 -> negative
50 -> neutral
51..99 -> positive
100 -> most positive

The sentiment results are calculated based on OpenAmplify’s Natural Language Processing application and LibSVM/LibLINEAR (Support Vector Machines) machine learning tools.

No comments:

Post a Comment