Show simple item record

dc.contributor.authorHettikankanama
dc.contributor.authorHKSK
dc.contributor.authorVasanthapriyan
dc.contributor.authorS
dc.contributor.authorRathnayake
dc.contributor.authorRMTK
dc.date.accessioned2019-11-22T12:48:34Z
dc.date.available2019-11-22T12:48:34Z
dc.date.issued2019
dc.identifier.urihttp://ir.kdu.ac.lk/handle/345/2287
dc.description.abstractWith the evolution of Internet and ecommerce physical stores and businesses were moved to web breaking geographical barriers. Most online businesses use recommender systems to find right product for right customer at right time to increase customer satisfaction. Product recommendation systems are filtering tools which use data mining and machine learning algorithms to suggest the most relevant items to a particular user. This study illustrate how recommender systems increase the quality of the decisions that customers make while selecting a product by reducing the information overload and complexity. The goal of this study is to propose a novel product recommendation algorithm considering user reviews which provide multiuser recommendation. In this research a data set was taken through some different supervised and unsupervised learning methods, available recommendation systems and finally through proposed recommendation system. New recommendation model and its workflow is illustrated here which analyse review text and provide rating value for reviews through sentiment analysis and polarity estimation. This paper presents the methodology and techniques used in novel recommendation algorithm and its evaluation.
dc.language.isoenen_US
dc.subjectData Miningen_US
dc.subjectProduct Recommendationen_US
dc.subjectSentiment Analysisen_US
dc.subjectUser Reviewsen_US
dc.titleData Mining and Machine Learning Approach for Online Product Recommendation System Using Sentiment Analysisen_US
dc.typeArticle Full Texten_US
dc.identifier.journalKDUIRC-2019en_US
dc.identifier.pgnos443-448en_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record