Title: An effective movie recommender system enhanced with time series analysis of user rating behaviour

Authors: Bam Bahadur Sinha; R. Dhanalakshmi

Addresses: Computer Science and Engineering, National Institute of Technology Nagaland, Dimapur, India ' Computer Science and Engineering, Indian Institute of Information Technology Tiruchirappalli, Tiruchirappalli, India

Abstract: Recommender system aims at improvising user satisfaction by taking decision on what movie or item to recommend next. Over time though, learners and learning behaviours shift regularly. This paper introduces a novel behaviour-inspired suggestion algorithm named the TimeFly-PPSE algorithm, which operates on the concept of changing user's motives around time. The suggested model takes temporal knowledge into account and monitors the progression of consumers and items that are useful in providing adequate recommendations. The latter outlines a framework that enrolls the user's shifting behaviour to include guidance for personalisation. TimeFly's findings are contrasted with those of other well-known algorithms. Simulation test on 100K MovieLens dataset shows that utilising TimeFly contributes to recommendations that are exceptionally efficient and reliable.

Keywords: recommendation; prediction; MovieLens; time-aware; genres; mean absolute error; MAE.

DOI: 10.1504/IJMOR.2021.116966

International Journal of Mathematics in Operational Research, 2021 Vol.19 No.3, pp.317 - 331

Received: 06 Sep 2019
Accepted: 11 Apr 2020
Published online: 10 Aug 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article