Framework for Hybrid Book Recommender System based on Opinion Mining

Abstract

Recommender system plays an important role in automatic filtering out the important and personalized information for the intended user from a large amount of available information on internet. Recommender systems for books provide personalized recommendations to the readers for reading and to the librarians for book acquisition process. The objective of this research paper is four folds. Firstly, it conducts an extensive literature review pertaining to book recommender systems, secondly it specifies the popular recommendation techniques being used in specific application area of books, thirdly the paper reflects on the methodology followed and evaluation techniques being used based on the techniques discussed. Lastly, the paper proposes a framework for a book recommender system using best-suited recommendation techniques.

An Approach to Incorporate Textual Reviews in Book Recommender System

“…Existing recommender systems use recommendation techniques like collaborative filtering, contentbased, ontology-based, association rule mining, and hybrid techniques. These available techniques can be applied to ratings, ranking, and implicit information to calculate similarity among users or items to generate recommendations [3]. Among these techniques, the hybrid technique is the most popular in recommender systems due to its feature of utilizing the strength of different techniques which results in improved performance [4].…”