| 研究課題/領域番号 |
23K01591
|
|---|
| 研究機関 | 日本大学 |
|---|
研究代表者 |
Joe Geluso 日本大学, 法学部, 准教授 (20938055)
|
|---|
| 研究分担者 |
臼井 哲也 学習院大学, 国際社会科学部, 教授 (60409422)
|
|---|
| 研究期間 (年度) |
2023-04-01 – 2026-03-31
|
| キーワード | In progress |
|---|
| 研究実績の概要 |
The 2023 fiscal year was spent collecting, cleaning, and annotating data. The data is Annual Reports published by multinational automotive companies (e.g., Ford, Toyota, BMW). To date, we have collected, converted, cleaned, and annotated over 5,000 pages and 2.25 million words of text from PDF files into cleaned and annotated utf-8 .txt files. Preliminary data analysis in 2023 focused on BMW, Volkswagen, GM, and Ford. Findings include that BMW and Volkswagen have longer reports with higher word counts on average than their American counterparts, and are more consistent with word counts and the number of figures and tables that are used per ten pages of report. Meanwhile GM and Ford show more variation between years in terms of word counts and the number of tables and figures used.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
3: やや遅れている
理由
The main reason for the slight delay revolves around the cleaning of text data. Specifically, how quickly students workers clean and annotate the converted PDF files, and how many hours students can work. To the first point, the students work slightly slower than expected. A second point is that principal researcher assumed that the student workers could work year round, but it seems that it is hard to employ students between semesters as rules dictate that the student must work in the presence of one of the investigators. Balancing schedules when students and the researcher are available at the same time has limited the amount of time that students can work. Nevertheless, we continue to progress and I am confident that we will produce meaningful research results, if slightly delayed.
|
| 今後の研究の推進方策 |
We plan to continue collecting, converting, cleaning, and annotating data. This year we will also begin data analysis in earnest and start writing. Our first set of analyses will revolve around RQ1 focusing on the statistical side of linguistic analysis (e.g., keyword analysis, topic modeling). The second step will involve more qualitative analyses of images and text organization.
Both of these analyses are aimed at exploring how multinational automotive companies represent themselves in the documents entitled "Annual Reports" with an eye toward whether the Annual Reports use language that will resonate more with shareholders or broader stakeholders in the general public.
|
| 次年度使用額が生じた理由 |
The grant was afforded to (1) pay student research assistants to aid in converting, cleaning, and annotating texts, (2) purchase corpus analysis tools for text analysis, (3) pay a programming consultant to expedite the creation complex scripts in programming languages for natural language processing, and (4) attend conferences to disseminate research results. The reason we will use grant money in the coming fiscal year is for reasons (1), (2), and (3). We must continue paying student research assistants to prepare Annual Reports for linguistic analysis. We will renew the subscription to the corpus tool SketchEngine using the miscellaneous funding category for data analysis. We will also consult with an expert computer programmer to ensure our computer programs are achieving the intended result for analyses such as Topic Modeling and Multi-Dimensional Analysis.
|
