Extracting Multimedia Information and Knowledge Discovery using Web Mining: Challenges and Research Directions

摘要

World Wide Web is an arrangement of interlinked hypertext records that are pervaded through the Internet. Internet browsers facilitate simple access to various sources of text content and multimedia effectively. Today billions of pages are indexed via various search engines as well as aquisite the desired data isn't so straightforward. This has provoked the requirement for creating programmed mining procedures on the WWW, by the new term known as - "Web mining". In the web mining scenario, the information material can be accumulated at the server side, proxy servers, customer side or can be obtained through the association's database. This obtained data is generally unstructured and massive in nature. Presently, the greater part of the mining procedures utilized are content driven and the calculations are arranged towards content mining system. Generally, the web oriented multimedia data is in the form of images, text, animations, audios and videos etc. Extraction and retrieval of significantly useful information and specific knowledge exploration from these multimedia forms is a great challenge. In this paper, we extensively surveyed significant developments occured over past years. We have listed some significant tools for performing the multimedia data mining and analysis, the data which is obtained from various web sources. We also given an argument on the various open challenges involved in this domain along with the problem identification and our possible research directions to overcome these challenges.