An Efficient User Centric Clustering Approach for Product Recommendation Based on Majority Voting: A Case Study on Wine Data Set

摘要

Analysis, design and implementation of software systems for online services, are a tedious and challenging. Amazon software provides product recommendations, Yahoo! dynamically recommends WebPages, afflux creates recommendations for movies, and Google creates advertisements on the Internet. Items are recommended based on the preferences, needs, characteristics and circumstances of users. The wine data set has been in use in research for several years and still it remains as the benchmark data set. Quality of wines is difficult to define as there are many factors that influence the perceived quality. This paper presents a critical review of research trends on Wine quality and user-centric similarity measures as well. A novel user centric similarity measure in product clustering is proposed to evaluate the popular Wine data set named Red Wine dataset. The experimental results obtained in this work are able to provide better recommendations to product buyers than the existing systems. The proposed approach is competent to group the Red wine dataset into ordered groups of preferred wine variants and can judge the wine quality based on these user preference groups.