|Author:||Kan, Patrick T. Y|
|Title:||User profiling to information discovery over WWW|
|Subject:||World Wide Web|
Hong Kong Polytechnic University -- Dissertations
Department of Computing
|Pages:||vi, 87 leaves : ill. ; 30 cm|
|Abstract:||The development of the Internet and the World Wide Web presents a significant challenge to information retrieval systems everywhere. The explosive growth has lead to the growth of the amount of information resources available over the Web and increasing difficult to search for information: User Profiling system which can help user to search the relevant information over the Web as well as filter out those unwanted by making use of a profile. The system is special, adaptive, exploratory and reusable. That means the system is specialized to user interests, adaptive to interests changes, exploratory to new information domains and reusable to initialize the profile. The profile contains information about the user interests in form of keywords and their weightings. Profile searches for information that has the highest scores or weightings. After finishing ranking processing, top-scoring information is retrieved for presentation to user in the order of its scores. Besides, the system also has a Profile Learning unit to allow user providing positive or negative feedback of the documents so as to modify the profile based on the relevance feedback. Actually, other than providing feedback to the retrieved documents, the system can use the concept of feedback by examples as an additional method of providing feedback to the system. In short, our goal is making use of the user profile to supplement browsing with keyword search and brings user closer to potentially relevant information and further to explore interesting information. The user can use the web browser such as Netscape Navigator or Microsoft Internet Explorer to view the user profile and documents index information, to create documents index and to search for document information. The user interface is friendly and accessible to both naive as well as expert users. Finally, we will discuss the relationship of profile size and the term occurrences. Then, the system is evaluated with performance tests and the experimental results show the result of using relevance feedback to specializing to user interests, the relationship of different learning rates to adapt to user interests change, and the explored words from the system after successive queries. Besides, a test on feedback by examples by varying the number of training documents shows that larger number of examples to train the system have more accurate result.|
|Rights:||All rights reserved|
Files in This Item:
|b14605181.pdf||For All Users (off-campus access for PolyU Staff & Students only)||3.19 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: