Faceted Search on Amazon Data using Naïve Bayes Approach

Abstract

Facets can allow the user to interactively search in online webshops, e-commerce websites, and digital libraries. Faceted search in E-commerce websites is a challenging task. After successful identification of facets, the quality of information retrieval results improves drastically. Experiments about faceted search on E-commerce domain by taking into account research done on Amazon webpages corpus data words. The main task is to detect facets using text classification methods namely the Naïve Bayes method. In this paper, the primary focus is on accomplish information retrieval using faceted search for boosting the searching results to find the desired product with the help of large data set of words using K-Means, Term Frequency, Naïve Bayes approach collectively as a hybrid approach. In this work Naïve Bayes technique are used on Amazon text words to develop and detect faceted search results using Python language. To classify the text words random forest, machine learning algorithms have been used for faceted search. Results of experiments show that the use of this hybrid approach of faceted search in Amazon webpages data words using Naïve Bayes will enhance the accuracy by 90.55% on corpus data irrespective of Random Forest model results.