SENTIMENT CLASSIFICATION OF ONLINE CUSTOMER REVIEWS AND BLOGS USING SENTENCE-LEVEL LEXICAL BASED SEMANTIC ORIENTATION METHOD

KHAN, AURANGZEB (2011) SENTIMENT CLASSIFICATION OF ONLINE CUSTOMER REVIEWS AND BLOGS USING SENTENCE-LEVEL LEXICAL BASED SEMANTIC ORIENTATION METHOD. PhD. thesis, UNIVERSITI TEKNOLOGI PETRONAS.

[thumbnail of final_copy1.pdf]
Preview
PDF
final_copy1.pdf

Download (2MB)

Abstract

Sentiment analysis is the process of extracting knowledge from the peoples’ opinions, appraisals and emotions toward entities, events and their attributes. These opinions greatly impact on customers to ease their choices regarding online shopping, choosing events, products and entities. With the rapid growth of online resources, a vast amount of new data in the form of customer reviews and opinions are being generated progressively. Hence, sentiment analysis methods are desirable for developing efficient and effective analyses and classification of customer reviews, blogs and comments.
The main inspiration for this thesis is to develop high performance domain independent sentiment classification method. This study focuses on sentiment analysis at the sentence level using lexical based method for different type data such as reviews and blogs. The proposed method is based on general lexicons i.e. WordNet, SentiWordNet and user defined lexical dictionaries for sentiment orientation. The relations and glosses of these dictionaries provide solution to the domain portability problem.
The experiments are performed on various datasets such as customer reviews and blogs comments. The results show that the proposed method with sentence contextual information is effective for sentiment classification. The proposed method performs better than word and text level corpus based machine learning methods for semantic orientation. The results highlight that the proposed method achieves an average accuracy of 86% at sentence-level and 97% at feedback level for customer reviews. Similarly, it achieves an average accuracy of 83% at sentence level and 86% at feedback level for blog comments.

Item Type: Thesis (PhD.)
Departments / MOR / COE: Sciences and Information Technology
Depositing User: Users 5 not found.
Date Deposited: 05 Jun 2012 10:20
Last Modified: 25 Jan 2017 09:42
URI: http://utpedia.utp.edu.my/id/eprint/3028

Actions (login required)

View Item
View Item