|Title:||Personal recommender system based on CBR|
|Subject:||Hong Kong Polytechnic University -- Dissertations|
|Department:||Department of Computing|
|Pages:||viii, 48 leaves : ill. (some col.) ; 30 cm.|
|Abstract:||In this report, making use of customers' profile, consuming records and other information, construct the dining personal recommender system based on the theory case-based reasoning (referred to CBR). Recommender system is the hot topic in data mining and machine learning areas. The system learn about user preferences over time and automatically suggest items that more suitable for the user. On the other hand, CBR is one of the most practical machine learning methodologies that exploit a knowledge-rich representation of the application domain. In the framework of the CBR, a strange customer is viewed as "a new case", then find out (called retrieve) the similar one by different measures (comparison between customize metric and the Jaccard metric). "Similar customer" consumption records can be reused by recommendation directly in many occasions, but sometime it need to be adjusted for reuse, using different clustering algorithm (comparison between K-means and some convex clustering method, e.g. competitive learning, neural gas) and Bayesian belief network to adjust the results of retrieve. The practical results show that, some recommended results are praised by the customer in reality, and most of the rest recommended and actual preferences are divided into the same category on different clustering methods in most cases. Distinguish between different metrics and methods are not significant, test data may be sourced from the specific clusters naturally have impact on the results.|
Files in This Item:
|b26223090.pdf||For All Users (off-campus access for PolyU Staff & Students only)||1.87 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: