No abstract

No abstract

No abstract

In this era of information overload, Recommender Systems have become increasingly important to assist internet users in finding the right choice from umpteen numbers of choices. Especially, in the case of book recommender systems, suggesting an appropriate book by considering user preferences can increase the number of book readers in turn having an aftereffect on the users’ lifestyle by reducing stress, stimulating imagination, improving vocabulary, and making readers smarter. The majority of book recommender systems in the literature have used Collaborative Filtering (CF) and Content-Based Filtering (CBF) methods. Even though CBF methods have shown better performance than CF methods, they are mostly confined to shallow linguistic features. The present work proposed an aggrandized framework having three concurrent modules to improve the recommendation process. NER module extracts the Named Entities from the entire book content which are the key semantic units in providing clues on the possible choices of reading other related books. The Visual feature extraction module analyzes the book front cover to detect objects and text on the cover as well as the description of the cover which can bestow a clue for the genre of that book. The Stylometry module enhances the feature set used in the literature to analyze the author’s literary style for identifying similar authors to the present author of the book. These three modules conjointly improved the overall recommendation accuracy by 18% over the baseline CBF method that indicates the effectiveness of the present framework.

Due to the massive memory and computational resources required to build complex machine learning models on large datasets, many researchers are employing distributed environments for training the models on large datasets. The parallel implementations of Extreme Learning Machine (ELM) with many variants have been developed using MapReduce and Spark frameworks in the recent years. However, these approaches have severe limitations in terms of Input-Output (I/O) cost, memory, etc. From the literature, it is known that the complexity of ELM is directly propositional to the computation of Moore-Penrose pseudo inverse of hidden layer matrix in ELM. Most of the ELM variants developed on Spark framework have employed Singular Value Decomposition (SVD) to compute the Moore-Penrose pseudo inverse. But, SVD has severe memory limitations when experimenting with large datasets. In this paper, a method that uses Recursive Block LU Decomposition to compute the Moore-Penrose generalized inverse over the Spark cluster has been proposed to reduce the computational complexity. This method enhances the ELM algorithm to be efficient in handling the scalability and also having faster execution of the model. The experimental results have shown that the proposed method is efficient than the existing algorithms available in the literature.