Abstract: Social media services, including social networking services (SNSs) and microblogging services, are gaining prominence. SNSs have a variety of information on products and services, such as product introductions, utilization methods, and reviews. It is important for companies to utilize SNSs to understand the various ways of engaging with them. Against this backdrop, numerous studies have focused on marketing activities (e.g., consumer behavior and sales promotion) using information on the internet from sources such as SNSs, blogs, and news sites. In particular, to understand the dissemination of information on the Internet, various researchers have undertaken studies pertaining to the diffusion phenomenon occurring in the real world. Here, topic diffusion is a phenomenon whereby a certain topic is shared with several other users. In this study, we aimed to evaluate the diffusion phenomenon on Twitter. In particular, we focused on the state of a targeted topic and analyzed the estimation of the topic using natural language processing (NLP) and time series analysis. First, we collected tweets containing four titles of animation broadcasts using hashtags. Approximately 250,000 tweets were posted on Twitter in a month. Second, we used NLP methods such as morphological analysis and N-gram analysis to characterize the contents of each title. Third, using the time series data for the tweets, we created a mixture model that replicated the diffusion phenomenon. We clustered the diffusion phenomenon using this model. Finally, we combined the features related to the content of the tweets and the results of the clustering of the diffusion phenomenon and evaluated them.

Kohei Otake and Takashi Namatame, “Evaluation of the Diffusion Phenomenon using Information from Twitter” International Journal of Advanced Computer Science and Applications(IJACSA), 11(5), 2020. http://dx.doi.org/10.14569/IJACSA.2020.0110517

@article{Otake2020,
title = {Evaluation of the Diffusion Phenomenon using Information from Twitter},
journal = {International Journal of Advanced Computer Science and Applications},
doi = {10.14569/IJACSA.2020.0110517},
url = {http://dx.doi.org/10.14569/IJACSA.2020.0110517},
year = {2020},
publisher = {The Science and Information Organization},
volume = {11},
number = {5},
author = {Kohei Otake and Takashi Namatame}
}