Title: Metaheuristic enabled hot event detection and product recommendation in social media data streams

Authors: Manu G. Thomas; S. Senthil

Addresses: School of Computing and Information Technology, REVA University, India ' School of Computer Science and Applications, REVA University, India

Abstract: In this study, a unique system that recognises hot events and reliably integrates them with product recommendation systems is modelled. The correlation between the detected hot event and the user interestingness identifies the respective product that influenced the discussion. This results in the hottest action recognition being carried out through pre-processing, extraction of features and weight, text data modelling and cluster dependent topic identification. The keywords are initially extracted from each tweet. The evaluation of the different feature space is then completed and submitted to determine the design of the text model. Finally, clustering dependent hot topic identification takes place, where the optimisation logic plays its major role. Two different clustering processes take place: micro and macro-based clustering, where the selection of optimal centroid is made by a new enhanced monarch butterfly optimisation with three-level butterfly adjusting operator (EMBO-3BAR).

Keywords: hot topic detection; HTD; pre-processing; feature selection; user interestingness; optimisation; product recommendation.

DOI: 10.1504/IJCNDS.2023.133909

International Journal of Communication Networks and Distributed Systems, 2023 Vol.29 No.6, pp.573 - 597

Received: 04 Aug 2021
Accepted: 06 Jul 2022
Published online: 05 Oct 2023 *

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