Log in / Register
Home arrow Computer Science arrow Social Computing, Behavioral-Cultural Modeling and Prediction
< Prev   CONTENTS   Next >

5 Summary

Fig. 2. Accuracies when turning w1 : w2

We introduced a generalization of random walk graph kernel from the setting of homogeneous networks, i.e., networks consisting of only one type of nodes and one type of links, to the setting of heterogeneous networks, i.e., networks consisting of multiple types of nodes and links. We used the resulting kernel, HGK, to train an SVM classifier for labeling actors in heterogeneous social networks. The results of our experiments show that HGK classifiers often significantly outperform or are competitive with the state-of-the-art methods for labeling actors in social networks. Some promising directions for further research include: (i) combining multiple kernels [18] that capture different notions of similarity between nodes in heterogeneous networks; and (ii) using linking preferences directly estimated from the data to improve the accuracy of predicted labels.


1. Macskassy, S.A., Provost, F.: A simple relational classifier. In: MRDM Workshop at KDD 2003, pp. 64–76 (2003)

2. Macskassy, S.A., Provost, F.: Classification in networked data: A toolkit and a univariate case study. J. Mach. Learn. Res. 8, 935–983 (2007)

3. Sen, P., Namata, G., Bilgic, M., Getoor, L., Gallagher, B., Eliassi-Rad, T.: Collective classification in network data. AI Magazine, 93–106 (2008)

4. Lu, Q., Getoor, L.: Link-based classification. In: ICML, pp. 496–503 (2003)

5. Eldardiry, H., Neville, J.: Across-model collective ensemble classification. In: AAAI (2011)

6. Zhou, D., Bousquet, O., Lal, T.N., Weston, J., Scho¨lkopf, B.: Learning with local and global consistency. In: NIPS, pp. 321–328 (2004)

7. Wang, J., Jebara, T.: Graph transduction via alternating minimization. In: ICML (2008)

8. Angelova, R., Kasneci, G., Weikum, G.: Graffiti: graph-based classification in heterogeneous networks. World Wide Web Journal (2011)

9. Lin, F., Cohen, W.W.: Semi-supervised classification of network data using very few labels. In: Proceedings of the 2010 International Conference on Advances in Social Networks Analysis and Mining, pp. 192–199 (2010)

10. Ji, M., Han, J., Danilevsky, M.: Ranking-based classification of heterogeneous information networks. In: KDD, pp. 1298–1306 (2011)

11. Tang, L., Liu, H.: Scalable learning of collective behavior based on sparse social dimensions. In: CIKM, pp. 1107–1116 (2009)

12. Macskassy, S.A.: Improving learning in networked data by combining explicit and mined links. In: AAAI, pp. 590–595 (2007)

13. Macskassy, S.A.: Relational classifiers in a non-relational world: Using homophily to create relations. ICMLA (1), 406–411 (2011)

14. Kashima, H., Tsuda, K., Inokuchi, A.: Marginalized kernels between labeled graphs. In: ICML, pp. 321–328 (2003)

15. Vishwanathan, S.V.N., Schraudolph, N.N., Kondor, R., Borgwardt, K.M.: Graph kernels. Journal of Machine Learning Research 11, 1201–1242 (2010)

16. Kang, U., Tong, H., Sun, J.: Fast random walk graph kernel. In: SDM, pp. 828–838 (2012)

17. Li, X., Chen, H., Li, J., Zhang, Z.: Gene function prediction with gene interaction networks: a context graph kernel approach. Trans. Info. Tech. Biomed. 14(1), 119–128 (2010)

18. Cortes, C., Mohri, M., Rostamizadeh, A.: Multi-class classification with maximum margin multiple kernel. In: ICML 2013, vol. 28, pp. 46–54 (May 2013)

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science