Title :
Web Potential Customer Classification Based on SVM
Author :
Sun, Lei ; Duan, Zhu
Author_Institution :
Sch. of Econ. & Manage., Xidian Univ., Xi´´an, China
Abstract :
Potential customers are the future sources of profits. The manager can make decisions and manage customer relationship specifically as soon as finding those people. In this paper, a novel support vector machine (SVM) algorithm is used in Web mining, in order to find potential customers who visit the Web sites. And those potential customers are divided into two classes. Support Vector Machine (SVM) constructs an optimal hyperplane utilizing a small set of vectors near boundary. However, when the two-class problem samples are very unbalanced, PSVM tends to fit better the class with more samples and has high error in the class with fewer samples. To address the problem, an improved SVM algorithm, DFP-PSVM is presented in this paper. Computational results indicate that the modified algorithm has a strong capability of classification for the unbalanced samples of the two-class problems.
Keywords :
Web sites; customer relationship management; decision making; electronic commerce; information retrieval; pattern classification; support vector machines; DFP-PSVM algorithm; Web mining; Web potential customer classification; customer relationship management; decision making; e-commerce Web sites; optimal hyperplane construction; profits; support vector machine algorithm; two-class problem samples; unbalanced sample classification; Classification algorithms; Companies; Equations; Support vector machines; Web mining; classification; potential customers; support vector machine; unbalanced data; web mining;
Conference_Titel :
Industrial Control and Electronics Engineering (ICICEE), 2012 International Conference on
Conference_Location :
Xi´an
Print_ISBN :
978-1-4673-1450-3
DOI :
10.1109/ICICEE.2012.155
