Output details
11 - Computer Science and Informatics
Open University
A hybrid model for automatic emotion recognition in suicide notes
<22>The paper describes the winning entry to an open NLP challenge set by an NIH-funded National Center for Biomedical Computing, and is a development of the paper which won the Best Paper award at the associated workshop. It describes an original approach to sentiment analysis by training different machine learning algorithms on each of several different emotions, and combining them with a voting system. The result shows that this architecture can be important for sentiment recognition in text. Sentiment analysis is becoming increasingly commercially important in market analysis on social media text and understanding how online communities work.