Combining Machine Learning Techniques and Natural Language Processing to Infer Emotions Using Spanish Twitter Corpus

Proceedings of: 11th Conference on Practical Applications of Agents and Multi-Agent Systems (PAAMS 13). Salamanca, Spain, May 22-24, 2013.In the recent years, microblogging services, as Twitter, have become a popular tool for expressing feelings, opinions, broadcasting news, and communicating with friends. Twitter users produced more than 340 million tweets per day which may be consider a rich source of user information. We take a supervised approach to the problem, but leverage existing hashtags in Twitter for building our training data. Finally, we tested the Spanish emotional corpus applying two different machine learning algorithms for emotion identification reaching about 65% accuracy.This work was supported in part by Projects MEyC TEC2012-37832-C02-01, MEyC TEC2011-28626-C02-02 and CAM CONTEXTS (S2009/TIC-1485)Publicad