Peer-to-Peer Networking and Applications

, Volume 11, Issue 4, pp 778–792 | Cite as

Towards fast and lightweight spam account detection in mobile social networks through fog computing

  • Jiahao Zhang
  • Qiang Li
  • Xiaoqi Wang
  • Bo Feng
  • Dong Guo
Part of the following topical collections:
  1. Special Issue on Fog Computing on Wheels


Now, mobile devices play an increasingly important role in social networks by sharing information quickly, such as mobile phones and wearable health surveillance devices. Mobile social networks are vulnerable to spammers because of the fragile security policies of mobile operating systems. Especially, social networks on mobile devices face many difficulties in defending against spammers due to their low computing power, poor network quality and long response time. Since graph-based algorithms require huge computing power, machine-learning classifiers require very short response time, and existing PC-based research is not suitable for mobile devices, we need a lightweight and fast response method for mobile devices to detect spammers in mobile social networks. Regarded as the extension of cloud computing, fog computing puts the data, data processing and applications in the devices that are at the edge of the Internet (without storing all of them in the cloud), which leads to a better real-time performance, adapts to the wide geographical distribution and the high mobility of mobile devices. In this paper, we propose COLOR + , a method based on fog computing that performs most computations at terminal (mobile devices). It only uses the interaction between the account and its neighbors, which makes it easy to store and calculate a local graph on a mobile device. Each interaction value can be applied to any request. COLOR + detects spammers based on a threshold of the suspicion degree. We collect 50 million normal accounts and about 40,000 spammers from Twitter. Experiments show that the accuracy of COLOR + is about 85.95%, whose average time to detect an account is 0.01s. Therefore, COLOR + is an effective detection method that can be quickly applied.


Spammer Interactive relation Graph-based algorithm Fog computing 



This work is supported by the National Natural Science Foundation of China under Grant No.61472162.


  1. 1.
    Dong M, Ota K, Rmer AL (2016) Reliable and energy-efficient data collection for large-scale wireless sensor networks. IEEE Int Things J 3(4):511–519CrossRefGoogle Scholar
  2. 2.
    (2016). Digital, social and mobile worldwide in 2015.
  3. 3.
    (2016). Mobile twitter: 164m+ (75%) access from handheld devices monthly, 65% of ad sales come from mobile.
  4. 4.
    Lee S, Warningbird JK (2012) Detecting suspicious urls in twitter stream NDSSGoogle Scholar
  5. 5.
    (2016). Fighting spam with botmaker.
  6. 6.
    Ghosh S, Viswanath B, Kooti F, Sharma NK, Korlam G, Benevenuto F, Ganguly N, Gummadi KP (2012) Understanding and combating link farming in the twitter social network Proceedings of the 21st international conference on world wide web, pp 61–70CrossRefGoogle Scholar
  7. 7.
    Cao Q, Sirivianos M, Yang X, Pregueiro T (2012) Aiding the detection of fake accounts in large scale social online services Presented as part of the 9th USENIX symposium on networked systems design and implementation (NSDI 12), pp 197–210Google Scholar
  8. 8.
    Boshmaf Y, Logothetis D, Siganos G, Jorge L, Lorenzo J, Ripeanu M, Beznosov K (2015) Integro: Leveraging victim prediction for robust fake account detection in osns NDSS, vol 15. Citeseer, pp 8–11Google Scholar
  9. 9.
    Wei W, Fengyuan X, Tan CC, Sybildefender QL (2012) Defend against sybil attacks in large social networks INFOCOM, 2012 Proceedings IEEE, pp 1951–1959CrossRefGoogle Scholar
  10. 10.
    Haifeng Y, Kaminsky M, Gibbons PB, Flaxman A (2006) Sybilguard: defending against sybil attacks via social networks. ACM SIGCOMM Comput Commun Rev 36(4):267–278CrossRefGoogle Scholar
  11. 11.
    Haifeng Y, Gibbons PB, Kaminsky M, Sybillimit FX (2008) A near-optimal social network defense against sybil attacks Security and privacy, 2008. SP 2008. IEEE symposium on, pp 3–17Google Scholar
  12. 12.
    Danezis G, Mittal P Sybilinfer: Detecting sybil nodes using social networks NDSS. San Diego, CA, p 2009Google Scholar
  13. 13.
    Tran N, Li J, Subramanian L, Chow SSM (2011) Optimal sybil-resilient node admission control INFOCOM, 2011 Proceedings IEEE, pp 3218–3226CrossRefGoogle Scholar
  14. 14.
    Cao Qiang, Yang Xiaowei, Jieqi Y u, Palow Christopher (2014) Uncovering large groups of active Malicious accounts in online social networks Proceedings of the 2014 ACM SIGSAC conference on computer and communications security, pp 477–488Google Scholar
  15. 15.
    Huang J, Xie Y, Fang Y, Ke Q, Abadi M, Gillum E, Morley Mao Z (2013) Socialwatch: detection of online service abuse via large-scale social graphs Proceedings of the 8th ACM SIGSAC symposium on information, computer and communications security, pp 143–148Google Scholar
  16. 16.
    Wang G, Konolige T, Wilson C, Wang X, Zheng H, Zhao BY (2013) You are how you click: Clickstream analysis for sybil detection Proc. USENIX security. Citeseer, pp 1–15Google Scholar
  17. 17.
    Viswanath B, Ahmad Bashir M, Crovella M, Guha S, Gummadi KP, Krishnamurthy B, Mislove A (2014) Towards detecting anomalous user behavior in online social networks 23Rd USENIX security symposium (USENIX security 14), pp 223–238Google Scholar
  18. 18.
    Gao H, Chen Y, Lee K, Palsetia D, Choudhary AN (2012) Towards online spam filtering in social networks NDSSGoogle Scholar
  19. 19.
    Zhang Y, Ruan X, Wang H, Wang H (2014) What scale of audience a campaign can reach in what price on twitter? INFOCOM, 2014 Proceedings IEEE, pp 1168–1176Google Scholar
  20. 20.
    Stein T, Chen E, Mangla K (2011) Facebook immune system Proceedings of the 4th workshop on social network systems, p 8Google Scholar
  21. 21.
    Yang C, Harkreader R, Gu G (2013) Empirical evaluation and new design for fighting evolving twitter spammers. IEEE Trans Inf Forensics Secur 8(8):1280–1293CrossRefGoogle Scholar
  22. 22.
    Lee S, Jong K (2012) Warningbird: Detecting suspicious urls in twitter stream NDSSGoogle Scholar
  23. 23.
    Egele M, Stringhini G, Kruegel C, Giovanni V (2013) Compa: Detecting compromised accounts on social networks NDSSGoogle Scholar
  24. 24.
    De W, Navathe SB, Liu L, Irani D, Tamersoy A, Calton P u (2013) Click traffic analysis of short url spam on twitter Collaborative computing: networking, Applications and Worksharing (Collaboratecom), 2013 9th International Conference Conference on, pp 250–259Google Scholar
  25. 25.
    Thomas K, Grier C, Ma J, Paxson V, Song D (2011) Design and evaluation of a real-time url spam filtering service Security and privacy (SP), 2011 IEEE symposium on, pp 447–462CrossRefGoogle Scholar
  26. 26.
    Yang Z, Wilson C, Wang X, Gao T, Zhao BY, Dai Y (2014) Uncovering social network sybils in the wild. ACM Trans Knowl Discov Data (TKDD) 8(1):2Google Scholar
  27. 27.
    Yang C, Zhang J, Guofei G (2014) A taste of tweets: reverse engineering twitter spammers Proceedings of the 30th annual computer security applications conference, pp 86–95Google Scholar
  28. 28.
    Cha M, Haddadi H, Benevenuto F, Gummadi PK (2010) Measuring user influence in twitter The million follower fallacy. ICWSM 10(10–17):30Google Scholar

Copyright information

© Springer Science+Business Media New York 2017

Authors and Affiliations

  • Jiahao Zhang
    • 1
    • 2
  • Qiang Li
    • 1
    • 2
  • Xiaoqi Wang
    • 1
    • 2
  • Bo Feng
    • 1
    • 2
  • Dong Guo
    • 1
    • 2
  1. 1.College of Computer Science and TechnologyJilin UniversityChangchunChina
  2. 2.Symbol Computation and Knowledge Engineer of Ministry of EducationJilin UniversityChangchunChina

Personalised recommendations