Twitter Data Classification Using Side Information

Abstract

The important aspect of information gathering is to know what people think about particular topic. With the growth of IT and Internet, resource of opinion-rich text goes on increasing and thus the importance of text classification is increased in order to determine opinion about a subject. The main goal of sentiment analysis is to define sentiment or polarity of a sentence or a piece of text. Polarity means to check whether a particular piece text is positive, negative or neutral. The goal is to extract useful data and present it in a clear way so that one can get a general idea about people’s opinions. While doing Text analytics on posts or tweets obtained from social networking site, along with text some useful information is available know as Side Data. Such side data may be found in the form of access behavior of user from web logs, referred links to sites, special symbols, emotions, etc. This side data is helpful in predicting the polarity of the text. Further, the process of classification can be done using SparkR to improve processing speed and to analyze huge amount of data.