A system to analyze Twitter data for social science study

Twitter data has been becoming more interesting in social science study since it can effectively reflect a nature of human behavior. Unfortunately, it is complicated to analyze Twitter data, and the existing tools are not suitable for this domain. In this paper, we present a system that is tailored to analyze Twitter data for the social science research. The system comprises four main functions including: (i) case study management, (ii) user/keyword search, (iii) interest group customization, and (iv) user-friendly analysis and visualization. Furthermore, three kinds of measures: connectivity, reciprocity, and mentioning, are presented to support the analysis process. Some of them are selectively employed from other domains, while others are invented in this work. The experiments were conducted on more than two millions Twitter activities related to the political situation in Thailand during May-June 2014. The results showed that our proposed measures can reveal useful knowledge in Twitter social group with the aid of the system that can provide scenario-based analysis and capture interactions among user groups.

I. Introduction

The number of people who use social network is continually increasing. The statistics in [1] report that there are 1.79 billion users in 2014 and it will reach 2.44 billion users in 2018. Consequently, social science researchers can extract useful knowledge from these networks to understand their structure and behavior. This makes them gain new knowledge in specific areas, e.g., politics, linguistics, and communication arts. In some countries, more than a half of their people are social network users. For instance, around sixty percent of entire adults in the United States use Facebook [2]. So, the trend of social network analysis in social science is also going up like the number of social network users because the more users use, the higher similarity between the social network and the real society is.