학술논문
Hybrid Automatic Multistage Recommendation Techniques for Social Media Networks
Document Type
Conference
Author
Source
2023 International Conference on Computational Intelligence, Networks and Security (ICCINS) Computational Intelligence, Networks and Security (ICCINS), 2023 International Conference on. :1-6 Dec, 2023
Subject
Language
Abstract
The researchers on web recommendations on social media networks only consider similarity of web user's behavior and similarity of resources used by web user but not consider web user tagging similarity index. Collaborative filtering method facing sparsity problem, content filtering method not able to predict future web user interested web documents while recommending web pages, and web usage mining also not able to recommend newly added web documents to web users. In this research paper we propose hybrid multistage recommendation techniques for social media networks to resolve problems faced by previous recommendation techniques. Similarities matrix are measured using various performance evaluation metrics. Our proposed method outperforming when compared with traditional content filtering or collaborative filtering methods.