Collective Behavior Learning on Heterogeneous Affiliation in OSN
KR.Akshara1, N.Srinivasan2
1KR. Akshara, P.G. Student of Computer Science and Engineering, Sathyabama University, Chennai, India.
2Dr.N. Srinivasan, Professor& Head Dept of Computer Applications, Sathyabama University, Chennai, India.
Manuscript received on March 26, 2014. | Revised Manuscript received on April 15, 2014. | Manuscript published on April 30, 2014. | PP: 45-47 | Volume-3, Issue-4, April 2014. | Retrieval Number: D2808043414/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Millions of real world data is generated by the social network like Facebook, twitter etc which gives us an opportunity to under or predict users behavior by means of collective learning. Many studies have been conducted already on this field. Existing algorithm uses connection or user relationship to understand their behavior but they lack the understanding of heterogeneity in their connection which reduces the effectiveness of their algorithm. In this project we proposed a mechanism to focus on these issues by introducing the concept of edge centric clustering for classification of users based on their heterogeneity affiliation and extracting the social dimension , relevant community detection is made and the social dimension is extracted by using chi square testing model. Here in this work it also reduce the computational problem by scaling the samples applying the scheme sparse social network
Keywords: Collective Learning, Social Dimension, Community Detection.