A Framework for Aspect and Sentiment Extraction for Online Review

カテゴリ: 国際会議

論文No: MS3-4

グループ名: ACIS2015

発行日: 2015/10/15

著者名(英語): Noor Rizvana Binti Ahamed Kabeer(Universiti Sains Malaysia),Gan Keng Hoon(Universiti Sains Malaysia)

キーワード: Aspect Extraction, Opinion Mining,\nOnline Review, Sentiment Analysis, Rule-based

要約(英語): Proliferation of Internet and Web 2.0 social media ease millions of people to express opinions globally. To guide people to decide about product purchases, reviews on internet sites are often referred. However, as days goes on, amounts of reviews increases. It is laborious to manually analyze every review to extract opinions. Thus, there is a need to extract useful information from it to enhance decision making. Since reviews are written in natural language, the opinions about the aspects of a product or service can be expressed in different ways. Hence, the detection of opinion may not be straightforward, e.g. mixture of aspects and opinions within a single sentence. In this work, our objective is to improve the extraction of aspect and its sentiments by detecting correct association between them. We propose a framework to capture common sentence structure on how aspect and its sentiments are associated. The framework will be evaluated using benchmarked data by domain experts in review analysis.

原稿種別: 英語

PDFファイルサイズ: 907 Kバイト