Adaptive social network construction using Gaussian mixture model

作者：Guo Xin^*; Xiang Yang; Chen Qian; Wei Wei

来源：Information Technology Journal, 2011, 10(12): 2252-2259.

DOI：10.3923/itj.2011.2252.2259

• Adaptability
• Dynamic information
• Gaussian Mixture Model
• Performance measure
• Social environment
• Social interactions
• Social Networks
• Social relationships
• Vector space models

A Social network graph shows social interactions and relationships between individuals in a specific social environment, which is very helpful for analyzing social relationships, activities, structures, etc. The author quantized the strengths of social objects' relationships in social environment using an improved vector space model. Gaussian mixture model was employed to set the threshold for identifying social relationships adaptively and divide social subgroups automatically. According to the threshold, social network graph would be constructed based on performance measures. It is concluded that hidden social relationships can be discovered effectively by using this approach which is very flexible and adaptive for dynamic information feedback mechanism.

• 出版日期2011
• 单位同济大学