Context Aware Recommender System for Web Crawling by Similarity based Machine Learning Algorithms
Abstract
Recommender systems (RSs) learn the user preferences and relationships between the users and items and suggest ideas on the items on the basis of user interest. There are several major problems presented in the recommended systems, they are named as data sparsity, Cold-start problem, over-fitting, and big data problems. Based on the explicit feedback data such as votes and ratings, most of the recommender models are performed. Here, a new context aware recommender system for web crawling is going to propose, which is based on the similarity measures and using machine learning algorithms. On the basis of user opinion feedback reviews from the social networking services the online products are facilitated by this method. Finally, propose Hybrid Cascade Forward Neural Network with Elman Neural Network (HCFNN-ENN)for Prediction model to produce independent distributed depictions of contents of items and users and achieve good accuracy by using bald eagle search (BES) algorithm. The proposed approach is developed in the python platform, and the performance is evaluated because of various statistical metrics. The performance metrics, such as F-measure, Accuracy, Precision, Recall, and computation time, are evaluated in this research. The performance of the proposed approach is compared with several existing approaches such as Artificial Neural Network (ANN), ENN, K-Nearest Neighbour Network (KNN),and Recurrent Neural Network (RNN). Compared to the existing schemes, the proposed HCFNN-ENN prediction model gives the best outcomes.