Review of Data Mining for Twitter Data Analysis

Abstract

In today’s life Twitter, Facebook, Google are well known social sites that many uses for different purposes. Social sites are the fastest medium which delivers news to user as compared to the new paper and television. One among the online social networks like Twitter, has quickly gained fame as it provides people with the opportunity to communicate and share messages known as “tweets”. Tremendous value lies in automated analysis and data mining of such vast and diverse data to derive meaningful insights, which carries potential opportunities for businesses, consumers, product survey and political survey. In the proposed system of analysis of the tweets, a search query for topic is provided to extract required data using ‘clustering algorithm’ in machine learning. The unique advantage of using machine learning is that once an algorithm knows what to do with the information, it can do its job automatically. To deduce for a search query, the proposed system extracts feature from the tweets like keywords in tweets, number of words in tweets. The output tweet list can further be used for analysis for business improvement or for surveys.