Comparison of Multiple Book Recommendation Algorithms After Analysis of User Characteristics Using Big Data

ABSTRACTThrough big data analysis of borrowing information, it is possible to find the characteristics of different users. This paper applied K-means clustering to classify users. User feature vectors were extracted, and several algorithms were utilized to recommend books to different users. The hybrid algorithm achieved a precision of 96.29%, a recall rate of 92.13%, and an score of 94.16%, significantly higher than other algorithms. The findings prove that it is feasible to discover the characteristics of different users through big data analysis for book recommendation, and the hybrid algorithm performs the best.KEYWORDS: clustering analysisuser characteristicsbook recommendationrecommendation algorithm Disclosure statementNo potential conflict of interest was reported by the author(s).