Paper Info
Title
A Context-Aware Mobile User Behavior-Based Neighbor Finding Approach for Preference Profile Construction.
Abstract
In this paper, a new approach is adopted to update the user preference profile by seeking users with similar interests based on the context obtainable for a mobile network instead of from desktop networks. The trust degree between mobile users is calculated by analyzing their behavior based on the context, and then the approximate neighbors are chosen by combining the similarity of the mobile user preference and the trust degree. The approach first considers the communication behaviors between mobile users, the mobile network services they use as well as the corresponding context information. Then a similarity degree of the preference between users is calculated with the evaluation score of a certain mobile web service provided by a mobile user. Finally, based on the time attenuation function, the users with similar preference are found, through which we can dynamically update the target user's preference profile. Experiments are then conducted to test the effect of the context on the credibility among mobile users, the effect of time decay factors and trust degree thresholds. Simulation shows that the proposed approach outperforms two other methods in terms of Recall Ratio, Precision Ratio and Mean Absolute Error, because neither of them consider the context mobile information.
Year
DOI
Venue
2016
10.3390/s16020143
SENSORS
Keywords
Field
DocType
trust degree,multi-agent,context,time attenuation,interest similarity degree
Neighbor finding,Credibility,Information retrieval,Computer science,Mean absolute error,Time decay,Cellular network,Mobile Web
Journal
Volume
Issue
ISSN
16
2
1424-8220
Citations 
PageRank 
References 
0
0.34
12
Authors
3
Name
Order
Citations
PageRank
Qian Gao131.73
Deqian Fu274.51
Xiangjun Dong35912.26