A Robust Multi-Factor Recommender System for Online Libraries and E-book Portals

The online libraries or book portals have gained significant popularity among the readers in the recent years. The online libraries or book portals receive a number of users and are programmed to show the popular books among similar reader and user profile (collaborative filtering) and the books similar to the reading history of reader (content based filtering). These methods are incorporated to estimate the appropriate suggestions, recommendations and new arrivals according to user interests and history. The recommender system calculates the recommendations on the basis of various online and local features which include the e-book/book popularity, user rating, trust factor, page rank, total access, etc. The recommendations are calculate for the user convenience to find the related stuff easily over the populated book libraries and e-book shopping portals. In this paper, the entire has been based upon solving the problem of correct recommendations according to the users interest, profile contents and similar users history and interests. The proposed model will be evaluating the global factors such as Alexa Recommendation, Google page rank and trust factor for appropriate recommendations. The local factors of total access, rating and downloads or purchases, number of readers and user opinion will be exercised to calculate the correct recommendations. The major area of concern will be the institutional online libraries and shared libraries such as Google Books. The expected results would be collected in the form of accuracy, precision, recall and F1-error.