Tagged images browsing system

• Main Author: Nguyen Tran Nam, Khanh.
• Other Authors: Sun Aixin
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Computer science and engineering::Information systems::Information storage and retrieval
• Online Access: http://hdl.handle.net/10356/42450
• Institution: Nanyang Technological University

Nowadays, tagging systems have been integrated into many websites, especially for social media websites. By integrating a tagging system with a search engine, the accessing of users to media contents or even documents can be easier. However, retrieving the contents which are most relevant to a tag is still challenging and attracting numerous of research effort. Since the content-related searching is still not scalable, in this paper we propose various methods to improve the purely tag-based search on tagged image system. The proposed methods are: Tf-Idf weight and similarity between tags’ association and tags’ global weight. We also proposed 5 different methods to compute the association of tags and 3 methods to compute tags’ global weight. The above methods are integrated in to the existing image browsing system named TagViz. After conducting the experiments on the proposed methods, we found out that: generally the method “similarity between tags’ Pointwise KL and tags’ Idf weight” performs the best and can provide good results for searching 25 or 50 images.