An Enhanced Web Mining Approach for Product Usability Evaluation in Feature Fatigue Analysis using LDA Model and Association Rule Mining with Fruit Fly Algorithm

Abstract

Objectives: The rapid growth of E-commerce emphasizes the customers' attention on product purchases through large number of websites. The customers tend to buy the products with more and best features, but those features fails to satisfy their expectations. This limitation is termed as Feature Fatigue (FF) in product usability analysis in web mining. The objective of this novel approach is to evaluate the product usability effectively and supports the designers to make decisions in future. Methods: The consumers' reviews are collected from various E-commerce websites using web crawler. These review sentences are preprocessed by removing the stop words and stemming. The synonym dictionary is created from the preprocessed sentences. Findings: In order to attain usability evaluation frequent item set is identified by improved Apriori algorithm and the association rule is generated. Finally the product capability is evaluated effectively in the FF analysis using analyzed features. Improvements: The product reviews are gathered from the E-commerce websites to analyze the feature usability of that product and obtain 95% accuracy. The feature analysis report helps the manufacture to alleviate the FF by balancing the capability and usability of the product.