JUSTIFICATION ARCHITECTURE FOR IAAS RECOMMENDATION SYSTEM USING USER REVIEWS

Abstract

Providing justifications for recommendation systems is crucial to inform users about the basis of the suggested items and increase user satisfaction with the recommendations. However, the IAAS recommendation system currently does not provide justifications for its suggestions, which are based on user advisories. This study proposes a new architecture for IAAS that involves item filtration and justification extraction during recommendation generation, as well as incorporating user reviews on recommended items. The proposed architecture is compared to other justification approaches in recommender systems, including techniques such as feature extraction, term ranking, sentence filtering, and text summarization. The proposed approach involves the weight of comments, relevance of recommendations, and justification of recommendation architecture for IAAS. It includes the extraction of relevant and distinctive terms of the items that are discussed in the reviews, ranking the extracted terms, filtering out unnecessary sentences, and summarizing the main contents of the item's reviews while avoiding redundancy. The study highlights the importance of providing justifications for recommendation systems and provides a potential solution for IAAS users to justify their recommendations.