Loading...

Community Detection in Social Networks by Using Information from Diffusion Network

Ramezani, Maryam | 2014

2235 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 46303 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Rabiee, Hamid Reza
  7. Abstract:
  8. Nowadays, Online Social Networks (OSNs) play an important role in the exchange of information among people. Some previous studies indicate that diffusion behavior and network structure are tightly related. Community structure is one of the most important features of OSNs. Access to the whole network topology is the necessary and prevalent requirement for most of community detection methods, so the limited access to full or partial topology can decrease their accuracy. Using traceable information over diffusion network is a solution to surmount this difficulty. In this work, we are concerned with the community detection by only using the diffusion information, while unlike the previous methods; the network topology is completely unknown.
    As the first step, we infer the underlying network topology while preserving their community structure by using the diffusion information. To this end, a random process called Markov chain is used to model the relation between the diffusion network and the community structure. The experimental results of the proposed method on synthetic and real datasets show that in contrast to other methods, it performs well on inferring underlying network and preserving the community structure in low running time. Unlike most of the algorithms, the scalable proposed method is independent of the length of cascades, which makes it suitable
    for real-world networks.
    Using the network inference method and then the community detection suffers from both complexities. Therefore, at the second step, we focus on detecting communities directly from diffusion information. The proposed method tries to extract network communities directly from diffusion information in a low running time. Furthermore, this method is able to identify the hidden communities that the network structure can not determine. Here, we utilize Conditional Random Field model constructed from diffusion information. It has no need to
    any prior knowledge about the network structure or the number of communities. The results indicate considerable improvements over the baseline methods for detecting communities in high accuracy and low running time
  9. Keywords:
  10. Social Networks ; Markov Chain ; Community Detection ; Diffusion Network ; Conditional Random Fields (CRF) ; Diffusion Information ; Network Inference ; Random Process Model

 Digital Object List

 Bookmark

...see more