An automated book recommendation system has becoming a necessary role in increasing the efficacy and user experience of college library. The number of books in the library is very large, and it is difficult for students to choose the appropriate book from each department efficiently. The manual selection of books is time-consuming, and an automatic library book recommendation system is highly required. Therefore, this research proposes a novel ranking-based hybrid recommendation system for assisting each student in different department to select required books with minimal time. Initially, pre-processing is done to label each department name, which can effectively reduce the time complexity. Then, weighting is provided for each book using a timestamp based on the days the book was issued between the date of issue and return. After, Pearson correlation coefficient (PCC) is computed to find the similarity between the department and the corresponding books. Finally, the Hidden Markov-assisted Chaotic Artificial Humming Bird with Discriminant Analysis and weighted fuzzy ranking (HMCAHB_DA-WFR) model is proposed to classify the books according to the department accurately based on the generated ranks and provides better recommendation to the students. The proposed method is implemented in PYTHON software, and a real-time dataset is utilized from Bhilai Institute of Technology, Durg. The proposed method obtains a better accuracy of 99.2%, Kappa of 97.6%, MAE of 0.074, NMCC of 98.7%, FPR of 0.0085% and FNR of 0.034%. The attained simulation results prove the efficiency of proposed method over other existing methods.
Read full abstract