| 英文摘要 |
This study investigated the content of news articles pertaining to learning portfolios and subsequently analyzed and summarized their content orientation through text mining. Regarding the data source, news reports containing the keyword “learning portfolio” and published between November 1, 2014, and June 30, 2024 (116 months) were obtained from the Liberty Times, United Daily News, and China Times. In total, 1,094 relevant new reports were identified after a screening process. News content was processed using frequency distribution, structural topic modeling, and sentiment analysis, and annual quantities, trends, topic orientations, and emotional tendencies were thus determined. The analysis results are outlined as follows: 1. The year 2021 had the highest number of news reports on learning portfolios (304), followed by 2022 (177) and 2019 (170). Regarding monthly trends, September 2021 had the highest number of reports (90 reports), followed by March 2022 (49 reports). The reporting trend was closely related to admissions involving learning portfolios or significant public events. 2. In the “news headlines” section, two key issues were identified: “admissions system” and “highlighted events.” Among headlines, neutral sentiment was the most prevalent. In the “news content” section, six issues were identified: review materials,” “application for admission,” “multiple electives,” “concerns about advancement in education,” “file loss,” and “education policy.” The sentiment tendency was mostly positive. 3. A trend analysis of topics reported in the news content indicated that in early 2016, the focus was on application for admission issues. In 2019, emphasis shifted to review materials. In 2021, emphasis shifted to file loss, and in 2022, emphasis returned to review materials, multiple electives, and education policy. On the basis of its findings, this study provides recommendations for policy development. |