Development of an improved collaborative filtering-based similarity model for User-Item Matrix
Computer-assisted information technology has a potential to provide knowledge about any specific theme in a given area. The access to information about any subject helps in making smart choices regarding various fields related to business endeavors, education, policy-making and other related subject matters. In order to streamline retrieval of information, there are a range of recommendation systems to execute effective dissemination of the knowledge. In this regard, collaborative filtering has been considered as the best approach for evolving effective and reliable recommendations.The major focus remains on the development of a similarity model that can deal with handling multi-tenancy, storing and processing a massive amount of data. Moreover, the envisaged model should also be successful in tackling safety concerns of the client's sensitive data on remote servers. In the present study, we proposed a new similarity model that overcomes the shortcomings of various existing similarity measure algorithms. The proposed model provides an affordable, personal and efficient recommendation system that facilitates simple and easy exploration of the related information.