[1]ZHAO Guanzhe,QI Jianpeng,YU Yanwei,et al.Online check-in outlier detection method in mobile social networks[J].CAAI Transactions on Intelligent Systems,2017,12(5):752-759.[doi:10.11992/tis.201706027]
Copy
CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume:
12
Number of periods:
2017 5
Page number:
752-759
Column:
学术论文—机器学习
Public date:
2017-10-25
- Title:
-
Online check-in outlier detection method in mobile social networks
- Author(s):
-
ZHAO Guanzhe; QI Jianpeng; YU Yanwei; LIU Zhaowei; SONG Peng
-
School of Computer and Control Engineering, Yantai University, Yantai 264005, China
-
- Keywords:
-
location-based social networks; outlier detection; check-in location; distance-based outlier; friend circle; status of check-in; neighbor searching; time-triggered detection
- CLC:
-
TP391
- DOI:
-
10.11992/tis.201706027
- Abstract:
-
With the increasing popularization of smartphone, Pads and other smart mobile devices, the use of mobile social networks has also developed rapidly. In this paper, we propose an online method for detecting check-in outliers based on user mobility behavior in mobile social networks. First, based on a distance-based outlier model, we propose two check-in outlier models with respect to historical location (H-Outlier) and friend circle (F-Outlier), respectively. Second, for the H-Outlier, we propose an optimized detection algorithm called H-Opt, which utilizes the proposed check-in status model and an optimized neighbor searching mechanism to reduce computation time. For the F-Outlier, we propose a trigger-based optimized detection algorithm called F-Opt, which transforms continuous online outlier detection into trigger-based outlier detection. Lastly, we present our experimental results, based on a real-world check-in dataset, which demonstrate the effectiveness of the proposed algorithm. Our experimental results show that F-Opt significantly reduces the error rate of H-Opt outlier detection. In addition, compared with the LUE algorithm, the F-Opt and H-Opt algorithms improved efficiency by 2.34 and 2.45 times, respectively.