Content based Recommender System on Customer Reviews using Sentiment Classification Algorithms

the paper proposes RecoProd a recommender system which uses sentiment analysis techniques to provide the best products for the customers. The system uses the existing product reviews upon which sentiment classification is carried out. RecoProd consists of an Information Retrieval component which extracts the reviews from the ecommerce websites using the product names as queries. Sentiment Analysis algorithms like Naive Bayes and SVM are used to categorize the reviews and opinion scores are assigned to the reviews. A comparative study on the accuracy of the sentiment analysis algorithms used is also carried out. Aspect based summary of opinions for each product is carried out and visually compared. The products are then clustered and the optimal product along with the recommended products is displayed to the user.

[1]  Henry Anaya-Sánchez,et al.  Combining Probabilistic Language Models for Aspect-Based Sentiment Retrieval , 2012, ECIR.

[2]  Yu Zhang,et al.  Extracting implicit features in online customer reviews for opinion mining , 2013, WWW '13 Companion.

[3]  Henry Anaya-Sánchez,et al.  Retrieving Product Features and Opinions from Customer Reviews , 2013, IEEE Intelligent Systems.

[4]  David M. Pennock,et al.  Mining the peanut gallery: opinion extraction and semantic classification of product reviews , 2003, WWW '03.

[5]  Peter D. Turney.  Thumbs Up or Thumbs Down? Semantic Orientation Applied to Unsupervised Classification of Reviews , 2002, ACL.

[6]  Bo Pang,et al.  Using Very Simple Statistics for Review Search: An Exploration , 2008, COLING.

[7]  Maite Taboada,et al.  Lexicon-Based Methods for Sentiment Analysis , 2011, CL.

[8]  Renata Vieira,et al.  Ontology based feature level opinion mining for portuguese reviews , 2013, WWW '13 Companion.

[9]  Jure Leskovec,et al.  Hidden factors and hidden topics: understanding rating dimensions with review text , 2013, RecSys.

[10]  Koji Eguchi,et al.  Opinion Retrieval Experiments Using Generative Models: Experiments for the TREC 2006 Blog Track , 2006, TREC.

[11]  Kevin P. Murphy.  Performance evaluation of binary classifiers , 2007 .

[12]  Hinrich Schütze,et al.  Book Reviews: Foundations of Statistical Natural Language Processing , 1999, CL.

[13]  Satoshi Morinaga,et al.  Mining product reputations on the Web , 2002, KDD.

[14]  Wei Zhang,et al.  Improve the effectiveness of the opinion retrieval and opinion polarity classification , 2008, CIKM '08.

[15]  Li Chen,et al.  Recommendation based on Mining Product Reviewers' Preference Similarity Network , 2012 .

[16]  Clement T. Yu,et al.  An effective approach to document retrieval via utilizing WordNet and recognizing phrases , 2004, SIGIR '04.

[17]  Suk Hwan Lim,et al.  Extracting and Ranking Product Features in Opinion Documents , 2010, COLING.

[18]  Kam-Fai Wong,et al.  An efficient approach for sentence-based opinion retrieval , 2010, ICMLC.

[19]  Kam-Fai Wong,et al.  An efficient approach for sentence-based opinion retrieval , 2010, 2010 International Conference on Machine Learning and Cybernetics.

[20]  Ewan Klein,et al.  Natural Language Processing with Python , 2009 .