2019 Fiscal Year Research-status Report
Domestic and International Online Discourse on Polarized Issues: The Case of Japanese Whaling
| Project/Area Number |
19K01442
|
|---|
| Research Institution | Hitotsubashi University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2019-04-01 – 2022-03-31
|
| Keywords | social media / Twitter / whaling / Japanese politics / database / network analysis |
|---|
| Outline of Annual Research Achievements |
The first year of this research project has focused on building two databases. The first is a database of tweets containing particular keywords. The main case study is whaling, so I am collecting whaling-related tweets in Japanese, English and other languages (currently approx. 1.4 million). This database will be the starting point for examining how online exchanges regarding polarized issues are structured in relation to linguistic and ideological borders. I am also collecting tweets about the “comfort women” issue as a possible second case.
The second is a general database of the Japanese political Twittersphere. It contains the accounts of Japanese political parties, party leaders and politicians, plus their followers (currently approx. 3.1 million). It also contains the tweets sent from those political accounts, and also by the most active of their followers. This database will allow me to understand which parts of the Japanese online political spectrum are tweeting about a given issue.
I have carried out a survey of previous research on Japanese whaling, and have also read widely on historical and contemporary political activism in Japan. I have consulted with experts on Japanese whaling, which has given me a better understanding of how whaling actors themselves use (or do not use) social media.
Finally, I have worked on relevant theoretical approaches and research methods. In particular I have investigated work that applies gatekeeping theory to social media, and written code to identify gatekeepers in online networks.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
Data collection has proceeded smoothly, and I have been able to refine the collection code to make it more reliable and efficient.
I have been learning about the importance of planning when dealing with large datasets. Twitter imposes rate limits on access to its API, so for example it could take two months to obtain additional data on 3 million users. This makes it important to plan which analyses to carry out when, and to collect all the data likely to be required in future analyses at one time.
The same applies to some social network analyses, for example removing a single node from a network and then recalculating all the shortest paths in the network. It may simply not be realistic to carry out such analyses on the complete networks of users. I have understood the importance of (a) making a realistic estimate of how long an analysis will take; (b) carrying out multiple pilot runs on smaller datasets to make sure everything runs smoothly to the end, (c) being doubly sure of the dataset being analyzed, and (d) thinking of ways to reduce the dataset without sacrificing integrity.
Attending the European Political Science Association annual meeting in June 2019 was very useful to this project, in particular hearing a paper by and talking with a researcher using politicians’ Twitter follower data to map party spaces.
My plans to meet with experts on whaling and research methods in early 2020 were disrupted by COVID-19. I adapted by switching my focus to improving data collection and analysis.
|
| Strategy for Future Research Activity |
The research is now moving from the data collection to the data analysis stage. In order to do the analysis I will have to collect some additional data about users from the Twitter API, which will take about two months.
In 2020 my analysis will focus on Japanese Twitter users. My first aim is to map the Twitter political space using party leaders’ followers, and to see which parts of this space are engaged in discussions of controversial issues such as whaling. Second, I will conduct a robustness check on this mapping using an alternative method that analyses the followers of a wider set of national and local politicians; this will require additional data, which may take two or three months to collect.
In order to capitalize on the effort of constructing the database of politicians and their followers, I will upgrade it to track changes in politicians and their followers over time, which will allow an investigation into the dynamics of follower networks, for example during spikes in interest in particular issues, during election campaigns or when new parties are formed.
I will continue to work on network analysis, and plan to use online computing capacity to speed up analysis. I will also embark on text analysis of the Japanese tweets on whaling, using topic modelling.
It is becoming clear that the dataset and method can have wider applications and significance beyond whaling or polarized issues.
|
| Causes of Carryover |
I had planned two overseas trips for February and March 2020 but had to cancel them both at the last minute due to COVID-19.
|
