Developers Saif Mohammad and Tony Yang used 'sentiment analysis' to monitor emotions in more than 32,000 emails, love letter and suicide notes.
Based on their analysis, the developers were able to detect the 'emotional temperature' of the emails by predicting joy, sadness, anger, fear, trust, disgust, surprise and anticipation through the tone of the emails, news.com.au reports.
The developers found that there were distinct differences between the emotional tone of emails between men and women, as women used more words from the joy-sadness axis, while men prefer using words from the fear-trust axis.
However, the tool does not take into account slang, sarcasm or the relationship shared between two people having the email exchanges.
The report said that the developers are now working with Google apps to roll out the app for the emotion prediction app and will soon make a call for public volunteers.
--ANI (Posted on 18-10-2013)