| 研究課題/領域番号 |
19K23181
|
|---|
| 研究機関 | 早稲田大学 |
|---|
研究代表者 |
|
|---|
| 研究期間 (年度) |
2019-08-30 – 2021-03-31
|
| キーワード | polarisation / social polarisation / comparative politics / cross-national / social networks / social media / network analysis |
|---|
| 研究実績の概要 |
This research project aims to create and test an empirical method for measuring social polarisation - as distinct from the more conventionally measured political polarisation - across different countries by using social media data. Since the project began, rapid progress has been made on identifying the initial targets for study and gathering the large-scale social media data required for analysis. The first four countries to be studied are Japan, Italy, Belgium and Ireland - a set of advanced liberal democracies which all have high social media usage and multi-party democratic systems, but otherwise offer a significant range in terms of demographics, existing social cleavages, and the influence of political populism. Data collection for each country has also been completed, using a "snowball" collection process which begins from the followers of political Twitter accounts, randomly sampling 10,000 users in each nation, and then works outwards to identify the other (non-political) accounts which they follow.
Initial analysis of this data has used network graph analysis based on the Jaccard Distance between users in each data set. This has allowed the identification of clusters of parties whose followers also follow similar non-political accounts, and possible lines of social polarisation between those clusters. A chapter based on this preliminary analysis is set to appear in the book "Values in European and Japanese Politics" later in 2020, and a conference presentation based on this analysis has been accepted for the APSA 2020 General Meeting in September 2020.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
Data collection for the project has proceeded smoothly and coders (research assistants) from each of the target countries were recruited without difficulty. The analysis process has also been smooth so far, although the next stage of the analysis will be more challenging and will also require more input from the research assistants. At present the project is progressing as expected and should deliver several more research outputs in the coming year.
|
| 今後の研究の推進方策 |
In addition to completing work on the book chapter and conference presentation, the next stage for this project will involve the development of a new analysis approach. At present, SNS users are being clustered according to their party leaning, but a completely "a posteriori" approach in which SNS users are clustered based on individual characteristics would be preferable. Moreover, a new algorithm - most likely based on the topic modelling algorithm presently popular in text analysis - needs to be developed to identify sub-networks of accounts within the data set.
While this new analysis is being prepared, data for several more countries will also be collected. At present the list of possible target countries includes Taiwan, South Korea, Germany, Finland, the Netherlands, Spain and Portugal, but a final decision on which countries to target must be made with reference to the results of the first round of analysis.
This research project should yield at least one more conference presentation and at least two peer-reviewed journal articles in the coming year. If the results are sufficiently good, there is also the possibility of a book based on the work, with chapters devoted to each of the national examples.
|
| 次年度使用額が生じた理由 |
The sum to be used next fiscal year was carried over due to being unable to confirm bookings (travel, accommodation and registration fees) for overseas conferences in 2020 due to the COVID-19 pandemic. As most of these conferences are now being held online, this money will instead be used to expand the number of countries covered by the project (which involves more data collection and hiring more research assistants), as well as to purchase equipment required to effectively participate in overseas virtual conference events.
|
