WCP-Nets: A Weighted Extension to CP-Nets for Web Service Selection

User preference often plays a key role in personalized applications such as web service selection. CP-nets is a compact and intuitive formalism for representing and reasoning with conditional preferences. However, the original CP-nets does not support fine-grained preferences, which results in the inability to compare certain preference combinations (service patterns). In this paper, we propose a weighted extension to CPnets called WCP-nets by allowing users to specify the relative importance (weights) between attribute values and between attributes. Both linear and nonlinear methods are proposed to adjust the attribute weights when conflicts between users’ explicit preferences and their actual behaviors of service selection occur. Experimental results based on two real datasets show that our method can effectively enhance the expressiveness of user preference and select more accurate services than other counterparts.