Unsupervised and knowledge-poor approaches to sentiment analysis

Zagibalov, Taras (2010) Unsupervised and knowledge-poor approaches to sentiment analysis. Doctoral thesis (DPhil), University of Sussex.

[img]
Preview
PDF - Published Version
Download (1MB) | Preview

Abstract

Sentiment analysis focuses upon automatic classiffication of a document's sentiment (and more generally extraction of opinion from text). Ways of expressing sentiment have been
shown to be dependent on what a document is about (domain-dependency). This complicates supervised methods for sentiment analysis which rely on extensive use of training data or linguistic resources that are usually either domain-specific or generic. Both kinds of resources prevent classiffiers from performing well across a range of domains, as this requires appropriate in-domain (domain-specific) data.

This thesis presents a novel unsupervised, knowledge-poor approach to sentiment analysis aimed at creating a domain-independent and multilingual sentiment analysis system.
The approach extracts domain-specific resources from documents that are to be processed, and uses them for sentiment analysis. This approach does not require any training corpora, large sets of rules or generic sentiment lexicons, which makes it domain- and languageindependent but at the same time able to utilise domain- and language-specific information.

The thesis describes and tests the approach, which is applied to diffeerent data, including customer reviews of various types of products, reviews of films and books, and news items; and to four languages: Chinese, English, Russian and Japanese. The approach is applied not only to binary sentiment classiffication, but also to three-way sentiment classiffication (positive, negative and neutral), subjectivity classifiation of documents and sentences, and to the extraction of opinion holders and opinion targets. Experimental results suggest that the approach is often a viable alternative to supervised systems, especially when applied to large document collections.

Item Type: Thesis (Doctoral)
Schools and Departments: School of Engineering and Informatics > Informatics
Subjects: Q Science > QZ Psychology
Depositing User: Library Cataloguing
Date Deposited: 26 Jan 2011 07:59
Last Modified: 14 Aug 2015 11:33
URI: http://srodev.sussex.ac.uk/id/eprint/6297
Google Scholar:1 Citations

View download statistics for this item

📧 Request an update