Data mining in recommendation and prediction modeling as sentiment and emotion analysis through keyword extraction

Abstract

This paper proposes a methodology for analyzing raw data operating in the text input of online users, with keyword extraction that focuses on “emotion-laden” words as the framework from which to launch the useful processing of information, with predictive value, for the benefit of different industries. In recent years, "big data", "data mining", and "data sciences" are some of the terms circulating in popular currency in the field of computer technology. For instance, with the advent of various social media platforms such as Facebook, Twitter, Instagram, Snapchat, and YouTube, an exhaustive amount of data has been stored online without adequate filtering. Various industries that make use of big data generated by social media are determined to gain a competitive advantage in the market. A clear example points to the global music industry. At present, we live at a time in which we can listen to music without downloading. Further, users can have an opportunity to get hold of other songs under their preferred genre through app features like "music recommendation" generated online by their collected music data. The academic world is also impacted. Nowadays, we utilize an online database to access academic papers. We can collect a sizable number of works, the data content of which generates more related reference materials online. The same principle may also apply to news data. Since most people rely on the Internet for their news information instead of the traditional print media, they can readily access multiple sources on a particular issue from different countries with variable perspectives. My overriding goal is to collect data generated by my proposed method of text extraction and analysis and, in turn, convert that data into usable information to help industries determine and predict consumer emotional sensibilities and trends.