Analysis of Similarity of Twitter Topic Categories among Regions

Abstract

Twitter can spread and share all kinds of information such as facts, opinions, and ideas in real time. In this paper, we empirically compare and analyze the topic categories in Twitter with all top 100 users in each of geographic region. We mainly consider the relationships among regions and selected four regions: Global, Seoul, Tokyo, and Beijing. Each of the top 100 users in Twitter is classified into a specific category and then statistical analysis is conducted. Among eight topic categories, the "Arts" category is the largest and the second is "Life". The correlation between global and Seoul groups has the lowest value among the six pairs of relationships between regional groups, and this difference is statistically significant. We find that the Seoul, Tokyo, and Beijing regional Twitter groups, all in East Asia, have high topical similarity. Based on the correlation analysis, Seoul and Tokyo saliently show a sticky trend. The correlation coefficient presents very a strong positive correlation between Seoul and Tokyo. The correlation between the global group and the East Asian groups is relatively lower than that among the East Asian groups.

Topical evolution patterns and temporal trends of microblogs on public health emergencies

“…Topical analysis of microblogs Common methods of topical analysis of microblogs include statistical techniques and manual classification. Studies on microblogs involving general topics, e.g., Yun (2012), were usually based on fairly rough topic classifications. More fine-grained topic categories were identified in the topical analysis of microblogs focusing on a specific topic, such as classifying main topics of public concern toward Ebola into several categories (Lazard et al, 2015).…”

An Efficient Analysis Model for Growth Environment Information System using Multi Regression and Modified Kmeans Algorithms

“…Big Data, which is produced across various fields today, is reproduced as new processed data to serve the goal of usability [1][2]. Reproduced data mainly include the goal of finding meanings and patterns hidden in them based on classification and analysis.…”