Chapter

Advanced Techniques in Web Intelligence-2

Volume 452 of the series Studies in Computational Intelligence pp 105-126

Web Opinion Mining and Sentimental Analysis

  • Edison Marrese TaylorAffiliated withWeb Intelligence Consortium Chile Research Centre, Department of Industrial Engineering School of Engineering and Science, University of Chile Email author 
  • , Cristián Rodríguez O.Affiliated withWeb Intelligence Consortium Chile Research Centre, Department of Industrial Engineering School of Engineering and Science, University of Chile
  • , Juan D. VelásquezAffiliated withWeb Intelligence Consortium Chile Research Centre, Department of Industrial Engineering School of Engineering and Science, University of Chile
  • , Goldina GhoshAffiliated withDepartment of Computer Science, Birla Institute of Technology
  • , Soumya BanerjeeAffiliated withDepartment of Computer Science, Birla Institute of Technology

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Web Opinion Mining (WOM) is a new concept in Web Intelligence. It embraces the problem of extracting, analyzing and aggregating web data about opinions. Studying users’ opinions is relevant because through them it is possible to determine how people feel about a product or service and know how it was received by the market. In this chapter, we show an overview about what Opinion Mining is and give some approaches about how to do it. Also, we distinguish and discuss four resources from where opinions can be extracted from, analyzing in each case the main issues that could alter the mining process. One last interesting topic related to WOM and discussed in this chapter is the summarization and visualization of the WOM results.We consider these techniques to be important because they offer a real chance to understand and find a real value for a huge set of heterogeneous opinions collected. Finally, having given enough conceptual background, a practical example is presented using Twitter as a platform for Web Opinion Mining. Results show how an opinion is spread through the network and describes how users influence each other.