Submission + - Machine learning system detects emotions and suicidal behavior (thestack.com)
An anonymous reader writes: A new machine learning technology is being developed by Israeli scientists which can identify emotion in text messages and email, such as sarcasm, irony and even antisocial or suicidal thoughts. The new computerised system, created by Eden Saig a computer science student at the Technion-Israel Institute of Technology, is described in a paper titled ‘Sentiment Classification of Texts in Social Networks.’ The system works by recognising repeated word patterns and was developed by Saig after he studied a course in artificial intelligence (AI) supervised by Professor Shaul Markovich. Saig explains that voice tone and vocal inflections are so crucial for conveying feelings in verbal communication, while with text and email messages these characteristics are lost – recently encouraging users to illustrate sentiment through superficial smileys or emoticons. Applying machine learning algorithms to popular opinion Facebook pages, Saig was able to use the results to pick out stereotypical habits in social network conversations. “Now, the system can recognise patterns that are either condescending or caring sentiments and can even send a text message to the user if the system thinks the post may be arrogant,” said Saig.