An Integrated Sentiment Topic Mixture Model for Short Text in Identification of User Preference Emotion and Sarcasm

Abstract

Social Media is a forum designed to allow people to create content, interact with society, share ideas, information, and lead a new way to socializing. Social media analytics is an art or practice of gathering data from blogs, social media websites, etc to get insights for arriving at a business decision. Social sentiment topics are sources that apply to all facets of the social analytics. Recently, researchers experienced the demand and challenges of analyzing short text. The first and the foremost challenge is the sparsity of the data and its reliance on the context. Secondly, there exists the presence of non-formal terms and noises such as slang words, misspellings, grammatical errors, abbreviations, or even foul languages. Thirdly, extracting sentiment feature, topic distribution feature and jointly capturing sentiment based topics is of great challenge. newlineThe Integrated Sentiment Topic Mixture (ISTM) framework was devised to address the three challenges in its design newline