VIETNAM NATIONAL UNIVERSITY, HANOI UNIVERSITY OF ENGINEERING AND TECHNOLOGY VU TIEN THANH A FEATURE-BASED OPINION MINING MODEL ON PRODUCT REVIEWS IN VIETNAMESE MASTER THESIS OF INFORMATION TECHNOLOGY Hanoi – 2012 TIEU LUAN MOI download : skknchat@gmail.com VIETNAM NATIONAL UNIVERSITY, HANOI UNIVERSITY OF ENGINEERING AND TECHNOLOGY VU TIEN THANH A FEATURE-BASED OPINION MINING MODEL ON PRODUCT REVIEWS IN VIETNAMESE Major : Computer Science Code : 60 48 01 MASTER THESIS OF INFORMATION TECHNOLOGY Supervisor: Assoc. Ha QuangThuy Hanoi – 2012 TIEU LUAN MOI download : skknchat@gmail.com ORIGINALITY STATEMENT ‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substan- tial proportions of material which have been accepted for the award of any other degree or diploma at University of Engineering and Technology (UET/Coltech) or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UET/Coltech or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project’s design and conception or in style, presentation and linguistic expression is acknowledged.’ Hanoi, November 25th , 2012 Signed.
i TIEU LUAN MOI download : skknchat@gmail.com ii TIEU LUAN MOI download : skknchat@gmail.com iii ABSTRACT Feature-based opinion mining and summarizing (FOMS) of reviews is a very in- teresting and attracting issue in the opinion mining field. With the development of e-commerce in Vietnam, there are more and more commercial sites and technical forums where people can review or express their opinions on the products which they have used. As a result, the number of reviews has been increasing rapidly to hun- dreds or even thousands for a hot-product in recent years. Not only is it difficult for the customer to read in order to make a decision whether to buy product but hard for the producer to handle customer opinions to improve their products as well.
In this thesis, we describe a Feature-based opinion mining and summarizing model on Vietnamese product reviews. Our model performs four following steps:(1)Pre- processing the input customer reviews by standardizing reviews, segmenting Token, and POS tagging(2) extracting explicit product features and opinion-words by using Vietnamese syntax rules, identifying implicit product features by using relationships with opinion words,and automatically grouping synonym product features by combin- ing HAC clustering method and semi-supervised SVM-kNN classification method; (3) identifying opinion sentences in each review and deciding whether each opinion sen- tence is positive, negative or neutral by using a VietSentiWordNet extended from an initial SentiWordNet 3.0; (4) summarizing the results which is different from the tra- ditional text summarization because we only focus on product-features on which the customers reviewed and whether opinions are positive, negative or neutral. Experi- mental results on Vietnamese reviews of mobile phone product domain demonstrate the effectiveness of the model. Publications: ? Huyen-Trang Pham, Tien-Thanh Vu, Mai-Vu Tran and Quang-Thuy Ha.
A Solution for Grouping Vietnamese Synonym Feature Words in Product Reviews. In Proceedings of the 6th international conference on Asia-Pacific Services Computing (APSCC 2011). ? Quang-Thuy Ha, Tien-Thanh Vu, Huyen-Trang Pham and Cong-To Luu. An Upgrading Feature- based Opinion Mining Model on Vietnamese Product Reviews.
In Proceedings of the 7th interna- tional conference on Active media technology (AMT 2011), pp. ? Tien-Thanh Vu, Huyen-Trang Pham, Cong-To Luu and Quang-Thuy Ha. A Feature-Based Opin- ion Mining Model on Product Reviews in Vietnamese. In Semantic Methods for Knowledge Man- agement and Communication (SCI 381), pp.
TIEU LUAN MOI download : skknchat@gmail.com ACKNOWLEDGEMENTS First and foremost, I would like to express my deepest gratitude to my supervi- sor, Assoc. Ha Quang Thuy, for his patient guidance and continuous support throughout the years. He always appears when I need help, and responds to queries so helpfully and promptly. I would like to give my honest appreciation to my col- leagues at the Knowledge and Technology laboratory for their great support.
I also would like to thank my friend, Nguyen Quoc Dat, for his kindly help. I sincerely acknowledge the Vietnam National University, Hanoi, NAFOSTED Vietnam and especially, QG.TN04/11-15 projects for supporting finance to my master study. Finally, this thesis would not have been possible without the support and love of my parents and my wife. Thank you! iv TIEU LUAN MOI download : skknchat@gmail.com To my family ♥ v TIEU LUAN MOI download : skknchat@gmail.com Table of Contents 1 Introduction 1 2 Literature review 4 2.1 The demand of opinion mining .2 The basic concepts in the opinion mining field .3 Opinion mining problems .2 Feature-based Opinion Mining .3 Opinion Orientation Identification .4 Feature-based Opinion Mining System on Vietnamese Product Reviews.
14 3 Our Feature-based Opinion Mining Model 15 3.2 Phase 1: Pre-processing .2 Token Segmenting and POS Tagging .3 Phase 2: Product Features and Opinion Words Extraction .1 Explicit Product Features Extraction .2 Opinion word Extraction .3 Implicit Features identification .4 Grouping Synonym Features .5 Frequent Features Identification .4 Phase 3: Determining the opinion orientation. 28 vi TIEU LUAN MOI download : skknchat@gmail.com TABLE OF CONTENTS vii 4 Evaluation 29 4.1 Environment and Experimental Data .2 Product Features Extraction Evaluation .3 Opinion Words Extraction Evaluation .4 The Whole System Evaluation. 32 5 Conclusion 36 TIEU LUAN MOI download : skknchat@gmail.com List of Figures 1.1 An example summarization of Samsung Galaxy Tab.1 OM documents on Google Scholars (In title) .2 OM documents on Google Scholars (In anywhere) .3 The tree of Nokia N72 object .1 Model for Feature-based Opinion Mining and Summarizing in Viet- namese Product Reviews.1 (Precision values (%))A comparison between our method in (Vu et al., 2011) and in this thesis .2 (Recall values (%))A comparison between our method in (Vu et al., 2011) and in this thesis .3 (F1 values (%))A comparison between our method in (Vu et al., 2011) and in this thesis .4 A summarization of Nokia C5-03 .5 A summarization of LG Wink Touch T300. 35 viii TIEU LUAN MOI download : skknchat@gmail.com List of Tables 3.1 Some examples of using opinion words to identify implicit features .2 Experimental result on HAC algorithm with 5 α threshold value .3 Samples in VietSentiWordnet .1 Total of crawled reviews .2 Results of frequent product features extraction (MF: Number of man- ual product feature; SF: Number of product features found by the system; CSF: Number of correct product features found by the system) 30 4.3 Results of opinion words extraction (MO: Number of manual opinion words; SO: Number of opinion words found by the system; CSO: Number of correct opinion words found by the system) .4 Precision, Recall and F1 of Feature-based Opinion Mining Model on Vietnamese mobile phones Reviews.
33 ix TIEU LUAN MOI download : skknchat@gmail.com List of Abbreviations OM Opinion Mining FOM Feature-based Opinion Mining FOMS Feature-based Opinion Mining and Summarizing NLP Natural Language Processing PMI Pointwise Mutual Information SVM Support Vector Machine HAC Hierarchical Agglomerative Clustering kNN k-Nearest Neighbor VNNIG Vietnam Internet Center POS Part Of Speech NP Noun Phrase N Noun A Adjective V Verb O Object prep Preposition x TIEU LUAN MOI download : skknchat@gmail.com Chapter 1 Introduction With rapid development of e-commerce in the world in general and Vietnam in particular, there are more and more commercial websites, technical forums, etc not only bringing to their potential customers a new way of purchasing products online, but also enabling their customers to review and to express their opinions on the products that they have purchased. So the number of customer reviews has been increasing sharply in recent years. Especially, some hot-products can get hundreds or even thousands of reviews at some popular commercial websites. It is not only difficult for customers to read in order to make a decision whether to buy product but also hard for product manufactures to handle customer opinions to improve their products.
As a results, feature-based opinion mining and summarizing (FOMS) of customer reviews of product sold online is a very interesting and attracting issue in the opinion mining field (Hu and Liu, 2004; Popescu and Etzioni, 2005; Qiu et al., 2011; Stoyanov and Cardie, 2008; Zhai et al., 2010; Zhang et al., 2010; Vu et al., 2011; Ha et al. There are many research have done to improve FOMS systems (Scaffidi et al., 2007; Kim and Hovy, 2006; Qiu et al., 2011, 2009; Vu et al., 2011; Stoyanov and Cardie, 2008; Zhai et al., 2010; Zhang et al. There are two important tasks which are extracting product features and opinion words task and grouping synonym product features task to improve FOMS systems. The former task depends on syntax rules, which means that different languages have different ways to resolve the task.
Most of researches using to resolve the later task focus on using dictionaries, supervised or semi-supervised machine learning methods which require users to build features synonym dictionaries, and to annotate training set. Sentiment lexical resources are very useful for Opinion Mining, especially for 1 TIEU LUAN MOI download : skknchat@gmail.com 2 FOMS system (Baccianella et al., 2010; Das and Bandyopadhyay, 2010; Esuli and Sebastiani, 2006; Esuli, 2008) to decide the opinion orientation. Stefano Baccianella et al (Baccianella et al., 2010) described the evolution of SentiWordNet, a remarkable English sentiment lexical resource. In this thesis, we propose a Feature-based opinion mining and summarizing model on Vietnamese product reviews (customer reviews)overcoming some drawbacks of the recent FOMS systems.
With an input customer reviews set of products, our model performs four following steps:(1)Pre-processing the input customer reviews by standardizing reviews, segmenting Token, and POS tagging(2) extracting ex- plicit product features and opinion-words as well by using Vietnamese syntax rules, identifying implicit product features by using relationships with opinion words,and automatically grouping synonym product features by combining HAC clustering method and semi-supervised SVM-kNN classification method; (3) identifying opin- ion sentences in each review and deciding whether each opinion sentence is positive, negative or neutral by using a VietSentiWordNet extended from an initial Senti- WordNet 3.0; (4) summarizing the results which integrates the results of previous steps and presents them in the column diagram; figure 1.1 is an example to illustrate a summarizing the results of FOMS system on Samsung Galaxy Tab. The rest of this thesis is organized as following. In the second chapter, we pro- vide some literature reviews. In the next chapter, the FOMS model with four steps is described.
Experiment results and remarks are described in the fourth chapter. Conclusions are showed in the last chapter. TIEU LUAN MOI download : skknchat@gmail.1: An example summarization of Samsung Galaxy Tab. TIEU LUAN MOI download : skknchat@gmail.com Chapter 2 Literature review In this chapter, we introduce opinion mining overview in section 2.1 and feature- based opinion mining (FOM) in section 2.
In the first section (2.1), firstly, we introduce the demand of opinion mining in 2. Secondly, the basic concepts in the opinion mining field such as Object, Opinion passage on a feature, etc. are described in 2. Finally, opinion mining problems are defined in 2.
In the second section (2.2), firstly, we define FOM problem in 2. Secondly, some related works extracting features are described in 2.3, we introduce related works extracting opinion words and aggregating opinions. Finally, some FOM systems in Vietnamese are introduced in 2.1 The demand of opinion mining In (Liu, 2010), textual information in the world can be split into two main cate- gories: facts and opinions. Facts are objective expressions about entities, events and their properties.
Opinions are usually subjective expressions that describe people’s sentiments, assessments or feelings toward entities, events and their prop- erties. The concept of opinion is very broad. In this thesis, we only focus on opinion expressions expressing people’s positive, negative sentiments, or neutrals.