Extrapolation for Positioning and Navigation Focused Content - Based Recommendation Mathematical Framework

Abstract

The exponential development of the global website and the advent of electronic commerce have made the processing of information more than powerful analyzes in order to acquire the most useful results. It can be a challenging and time consuming method to search information on such large websites. Recommended programs can allow users to access knowledge by giving custom advice to them. It helps to make decisions without adequate personal knowledge. In this article we increase the scalability, sparsity and precision of the device suggested by solving the issue of sparseness and cold start. The correlations between items are usually determined by means of techniques such as the similarity between items and cosines. The 'Coldstart' problem cannot be solved with these techniques by using a statistical approximation principle that allows to improve improved rating estimating methods The dataset is the 'Film Lens' dataset. Mean absolute error and error ratio are the metrics used (ER). Some students at our college checked the method and it was found that the issue of 'Cold Start' and 'Sparsity' was successfully solved.