|
Title:
|
A NEW MATHEMATIC MODEL-BASED RECOMMENDER SYSTEM FOR E-COMMERCE |
|
Author(s):
|
Yadong Liu |
|
ISBN:
|
978-972-8939-24-3 |
|
Editors:
|
Sandeep Krishnamurthy, Gurmit Singh and Maggie McPherson |
|
Year:
|
2010 |
|
Edition:
|
Single |
|
Keywords:
|
Double integral, hybrid information, mathematic model, three-dimension |
|
Type:
|
Short Paper |
|
First Page:
|
118 |
|
Last Page:
|
122 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Recommender systems can provide valuable services in a digital library environment, as demonstrated by its commercial success in book, movie, and music industries. Many recommender systems rely on a single approach-content based approach or a collaborative approach and most based on collaborative filtering do not consider the hidden theory in the similarity values. We argue that the mathematical logic in the calculation of similarity is appreciable. We also suggest that a recommender system should further utilize the product, customer and transaction information. In part of our work, we investigate the information in relevance with both users and products from a new perspective. The survey discloses customers underlying preference and products properties. Then, a new mathematic model based on solid analytic geometry and Uncertainty theory has been developed to recalculate various kinds of information such as similarity, a users preference, etc. by combining with the approach of the survey and Flooding algorithm. After a comparison and discussion of significant factors advancing the quality of predictions based on our preliminary experimental results, we conclude that the mathematic method improves the performance of recommender systems substantially. |
|
|
|
|
|
|
