Big Data Customization through Data Science with Artificial Intelligence to Reduce Cold Start Problem in Recommendation System

Abstract

Recommender systems commonly provide recommendations to users based on newlinetheir preferences needs and or tastes Because there is a never ending supply of newlineinformation available online recommender systems have been shown to be a newlinesuccessful method for preventing information overload The importance of using newlinerecommender systems cannot be overstated given its ability to assist in resolving a newlinenumber of over choice problems Recommendation systems come in many newlinedifferent varieties each with their own unique ideas and methods Numerous newlineapplications have adopted recommendation systems including those in various newlinecommercial sectors such as the health industry online e commerce websites the newlinetravel and tourism industry the agriculture industry and online music movie and newlineother media newlineCold start issues are referred to new items that are added to any dataset as well as newlinenew users that are entering the big data world The issue emerges when businesses newlineor decision makers lack knowledge about new users or items in product newlinerecommendations A recommendation system displays content based newlinecollaborative based or hybrid approaches to a user who is curious in a certain newlineproduct The recommender system is currently experiencing a cold start issue newlinewhich means it is having trouble identifying new users or products. To put it another newlineway the system lacks the knowledge necessary to offer recommendations. newlineIn this study we propose a method HMCAC Hydride Method of Content Based newlineAssociation Clustering for overcoming the cold-start issue by fusing clustering and newlineassociation rule approaches. The research is founded on excerpts from several newlinexxv approaches to studies This study also explains the models that have been suggested newlineas a solution to the cold-start issue focuses on the cutting edge of recommender newlinesystems and predicts the future of a variety of software applications The quality newlineof a recommendation system is assessed using a recommendation architecture and newlinequalitative assessment technique We have found several machine learning newlineideologies and using