Analysing Purchase Satisfaction Using Opinion Mining

Please enable javascript in your browser.

Derakhshan, Ali | 2016

1772 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 49411 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Beigy, Hamid
Abstract:
Opinions and experiences of others give us valuable information in making decisions. Recently, with the expansion of using social networks and websites, people can easily share their opinions about miscellaneous things. This huge amount of information cannot be analyzed by individuals, so a system that automatically analyzes opinions is needed. This need invokes new field of research that is called opinion mining. User’s viewpoints could change during the time, and this is an important issue for companies. One of the most challenging sub-problems of opinion mining is model-based opinion mining, which aims to model the generation of words by modeling their probabilities. In this thesis, we address the problem of model-based opinion mining by introducing a part-of-speech graphical model to extract user’s opinions in two different datasets in English and Persian, and in the prediction of the market by this model, we could achieve the accuracy close to methods that using explicit sentiment labels for comments
Keywords:
Opinion Mining ; Sentiment Analysis ; Satisfaction ; Probabilistic Graphical Models ; User Reviews ; Market Prediction ; Model-Based Opinion Mining