FUZZY SOCIAL NETWORK ANALYSIS (FSNA): A NEW FRAMEWORK TO MODELING A HUMAN SOCIETY BY FUZZY RELATIONS

摘要

Social Network Analysis is a method to analyze interactions amongst members in a society. Under the realm of social network analysis, interacting records of each member-pair can be collected to determine the structure of a given society. That kind of records is called social network data. The data is classified into three types, namely dichotomous relation, value relation and sign relation. This paper proposes a new framework to analyze social network of a human society based on opinions of its members. Key words are used to identify the opinions and their associated degrees. The new framework employs fuzzy relations as an alternative data type for representing the opinions. The society of Thai Members of Parliament is used to test the validity of the proposed framework. Opinions of each MP are collected from a news online program of Reuters (Thailand) Ltd. Two new fuzzy relational products have been invented to address the weak points of the classical products. They are named Similarity Product and Square Product. The results from the applications of the two new products together with two existing products help to determine two kinds of social networks: Agree network and Conflict network.