Automatic extraction of learning object metadata (LOM) from HTML web pages

Abstract

Obtaining learning resources from the Internet is common nowadays. However, locating relevant learning resources on the Internet for learning is difficult due to the loose structure of the web. Even with the help of search engines and keywords searching, the search results are too huge for manual selection and they are usually of poor relevancy. There is also no way to specify attributes, like keywords, author, type of media, etc. of a learning object for searching, not to mention its level of interactivity and difficulty. It is difficult to have a standard way for describing learning objects and allowing users to use these descriptions during searching. In order to solve these problems, learning technology standard such as IEEE LOM is emerged to provide a standard metadata set for describing learning resources, helping users to identify relevant learning objects easily. However, there are too many attributes in the standard which will make the metadata difficult to create and thus users reluctant to use. This thesis discusses the problems of locating relevant learning resources on the Internet; discusses the results of literature reviews on search engines and learning technology standards; analyzes the IEEE LOM and HTML information; introduces the methods for automating the extraction of LOM from HTML web pages; explains the design and implementation of the automatic extraction framework of LOM; reports the experiment for testing and evaluating the heuristic rules used in the automatic extraction framework of LOM; concludes the research project and explains the future extension of the automatic extraction framework of LOM.

Date of Award	16 Jul 2007
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Lam For KWOK (Supervisor)