Optimized Computational Intelligence with Association Rules for a Collaborative Consumer Product Recommendation System

Abstract

In essence, a manufactured goods suggestion is a filtering method that tries to predict and exhibit the goods that an end user is likely to want to purchase. Product suggestions are product listings that are tailored to each individual visitor to a website based on information about them, their preferences, and/or the preferences of other like buyers. Recommendation systems have gained favor recently as a solution to the problem of information overload. They achieve this by showing consumers the products that are most pertinent to them based on a huge quantity of information. Through the identification of exact match neighbors connecting individuals or goods based on their mutual evaluations in the precedent, online collaborative product suggestions seek to assist users in discovering their preferred items. However, the little data makes neighbor selection more difficult if the number of objects and users increases quickly. This study suggests a hybrid model-based approach for product recommendations that uses better Apriori algorithms to partition transformed user space. It makes use of the principal component analysis data reduction technique to classify products densely, potentially reducing the computational complexity of intelligent product suggestions. The experiment results indicate that, in addition to providing good accuracy performance, the suggested strategy might generate more personalized and reliable product suggestions than the present methods.