Advertisement

Link prediction in signed networks based on connection degree

  • Xiao Chen
  • Jing-Feng GuoEmail author
  • Xiao Pan
  • Chunying Zhang
Original Research

Abstract

Link prediction has recently received considerable attention in signed networks. Most of the existing methods assume that the signed network topology is certain, such as network structure, entities relations and entities attributes. However, the assumption is not applicable, since the signed network is uncertain. As a result, the prediction accuracy cannot be ensured if the uncertainty of the signed networks is ignored. In this paper, we regard the signed network as an identical-discrepancy-contrary system employing the set pair theory, and propose a new link prediction measure SNCD which integrates both the certain and uncertain relations, local and global information at the same time. After a series of experiment, the experimental results show that our proposed method provides better prediction accuracy and correctness.

Keywords

Connection degree Identical-discrepancy-contrary system Signed networks Link prediction 

Notes

Acknowledgements

This work is supported by the National Youth Science Foundation of Hebei (No. F2017209070), the National Science Foundation of China, (No. 61472340, No. 61303017), the National Youth Science Foundation of China (No. 61602401), the Natural Science Foundation of Hebei Province (No. F2014210068), and the Fourth Outstanding Youth Foundation of Shijiazhuang Tiedao University.

References

  1. Cartwright D, Harary F (1956) Structural balance: a generalization of Heider’s theory [J]. Psychol Rev 63(5):277.  https://doi.org/10.1016/B978-0-12-442450-0.50008-0 CrossRefGoogle Scholar
  2. Chen K, Luesukprasert L, Chou ST (2007) Hot topic extraction based on timeline analysis and multidimen-sional sentence modeling [J]. IEEE Trans Knowl Data Eng 19(8):1016–1025.  https://doi.org/10.1109/TKDE.2007.1040 CrossRefGoogle Scholar
  3. Chen X, Du X, Yu J et al (2014) Research of community mining in signed social network based on game theory [J]. Int J Innov Comput Info Control 10(6):2221–2235Google Scholar
  4. Chen X, Guo JF, Liu FC et al (2017) Study on similarity based on connection degree in social network [J]. Clust Comput 20(1):167–178.  https://doi.org/10.1007/s10586-017-0743-5 CrossRefGoogle Scholar
  5. Chiang K-Y, Natarajan N, Tewari A (2011) Exploiting longer cycles for link prediction in signed networks [C]. In: Proc of the 20th ACM intconf on information and knowledge management. ACM, New York, pp 157–1162Google Scholar
  6. Chiang K-Y, Hsieh C-J, Nagarajan N, Dhillon IS, Tewari A (2014) Prediction and clustering in signed networks: a local to global perspective [J]. J Mach Learn Res, 15(1):1177–1213MathSciNetzbMATHGoogle Scholar
  7. Esley D, Kleinberg J (2010) Networks, crowds, and markets [M]. Cambridge University Press, Cambridge, pp 49–50CrossRefGoogle Scholar
  8. Fuyong Y, Lin M, Shunpan L (2015) Algorithm of reconstructing trust matrix by integrating user similarity and trust propagation [J]. J Yanshan Univ, 39(6):535–540 (in Chinese).  https://doi.org/10.3969/j.issn.1007-791X.2015.06.011 Google Scholar
  9. Guha R, Kumar R, Raghavan P et al (2004) Propagation of trust and distrust [C]. Proc of the 13th int confon world wide web. ACM, New York, pp 403–412Google Scholar
  10. Hanely JA, Mcneil BJ (1982) The meaning and use of the area under a receiver operating characteristic curve [J]. Radiology 143:29–36.  https://doi.org/10.1148/radiology.143.1.7063747 CrossRefGoogle Scholar
  11. Hsieh C-J, Chiang K-Y, Dhillon IS (2012) Low rank modeling of signed networks [C]. In: Proc of the 18th ACM SIGKDD int conf on knowledge discovery and data mining. ACM, New York, pp 507–515Google Scholar
  12. Jingfeng G, Xiao C, Chaozhi F, Xinzhuan H (2015) A Novel hierarchical clustering method in signed social network [J]. J Comput Inf Syst 11(21):7865–7872.  https://doi.org/10.12733/jcis16047 Google Scholar
  13. Katz L (1953) A new status index derived from sociometric analysis [J]. Psychometrika 18(1):39–43.  https://doi.org/10.1007/BF02289026 CrossRefzbMATHGoogle Scholar
  14. Kunegis J, Lommatzsch A, Bauckhage C (2009) The Slashdot Zoo: mining a social network with negative edges [C]. In: Proceedings of the 18th international conference on World wide web.  https://doi.org/10.1145/1526709.1526809
  15. Lan M, Li C, Wang S (2015) Survey of sign perdition algorithms in signed social networks [J]. J Comput Res Dev, 52(2):410–422 (in Chinese).  https://doi.org/10.7544/issn1000-1239.2015.20140210 Google Scholar
  16. Leskovec J, Huttenlocher D, Kleinberg J (2010) Predicting positive and negative links in online social networks [C]. In: Proc of the 19th intconf on world wide web. ACM, New York, pp 641–6508Google Scholar
  17. Liu C, Liu J, Jiang Z (2014) A multiobjective evolutionary algorithm based on similarity for community detection from signed social networks [M]. IEEE Trans Cybern.  https://doi.org/10.1109/TCYB.2014.2305974 Google Scholar
  18. Lü L, Pan L, Zhou T (2015) Toward link predictability of complex networks [J]. PNAS 112(8):2325–2330.  https://doi.org/10.1073/pnas.1424644112 MathSciNetCrossRefzbMATHGoogle Scholar
  19. Priyanka A, Garg VK, Narayanam R (2013) Link label prediction in signed social networks [C]. In: Proc of the 23rd int joint conf on artificial intelligence. AAAI Press, California, pp 2591–2597Google Scholar
  20. She H, Hu M (2015) Link prediction based on signed network [J]. J Wut (Info Manag Eng) 37(5):602–606.  https://doi.org/10.3963/j.issn.2095-3852.2015.05.017 Google Scholar
  21. Shen H-W, Cheng S-Q, Zhang G-Q, Cheng X-Q (2014) Survey of signed network research [J]. J Softw 25(1):1–15 (in Chinese).  https://doi.org/10.13328/j.cnki.jos.004503 MathSciNetzbMATHGoogle Scholar
  22. Symeonidis P, Tiakas E (2013) Transitive node similarity: predicting and recommending links in signed social networks [J]. World Wide Web.  https://doi.org/10.1007/s11280-013-0228-2 Google Scholar
  23. Tian X-x, Song Y-l, Zhu T, Wang X-l (2014) An efficient user similarity metric based community leader electing approach [J]. J Yanshan Univ, 38(6):516–522 (in Chinese).  https://doi.org/10.3969/j.issn.1007-791X.2014.06.006 Google Scholar
  24. Xinzhuan H, Guo J, Chen X, Zhao X (2016) Research of signed networks community detection based on the tightness of common neighbors [C]. In: The 6th international conference on digital home. IEEE, Washington, pp 155–159Google Scholar
  25. Zhang C, Guo J (2013) The α relation communities of set pair social networks and its dynamic mining algorithms [J]. Chin J Comput 36(8):1682–1692 (in Chinese).  https://doi.org/10.3724/SP.J.1016.2013.01682 CrossRefGoogle Scholar
  26. Zhang C, Guo J (2014) The attribute graph model of social networks and its application [M]. Beijing University of Posts and Telecommunications Press, Beijing, pp 186–198 (in Chinese) Google Scholar
  27. Zhang CY, Liang R, Liu L (2011) Set pair social network analysis is model and its application [J]. J Hebei Polytech Univ (Nat Sci Ed), 33(3):99–103 (in Chinese).  https://doi.org/10.3969/j.issn.1674-0262.2011.03.023 Google Scholar
  28. Zhao K (2000) Set pair analysis and its preliminary application [M]. Zhejiang Science and Technology Press, Hang Zhou, (in Chinese) Google Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  • Xiao Chen
    • 1
    • 4
  • Jing-Feng Guo
    • 1
    • 4
    Email author
  • Xiao Pan
    • 2
  • Chunying Zhang
    • 3
  1. 1.College of Information Science and EngineeringYanShan UniversityQinhuangdaoChina
  2. 2.College of Economic and ManagementShijiazhuang Tiedao UniversityShijiazhuangChina
  3. 3.Science CollegeNorth China University of Science and TechnologyTangshanChina
  4. 4.The Key Laboratory for Computer Virtual Technology and System Integration of Hebei ProvinceQinhuangdaoChina

Personalised recommendations