主催: 一般社団法人 日本科学教育学会
会議名: 日本科学教育学会 第49回年会
回次: 49
開催地: 広島大学 東広島キャンパス
開催日: 2025/09/05 - 2025/09/07
This study analyzes the knowledge density and usage of scientific terminology in junior high school science textbooks from Taiwan and Japan, focusing on the topic of plant reproduction. The research examined the proportion of scientific terms appearing in the main body text of the textbooks, using both manual annotation and generative AI (ChatGPT) to identify scientific and non-scientific nouns. Results show that although both countries allocate only five pages to the plant reproduction topic, the Japanese curriculum spans both grades 7 and 9, while the Taiwanese content is concentrated solely in grade 7. The Taiwanese textbook contains fewer characters overall, yet has a higher proportion of scientific terms—46.75% of total characters compared to 29.60% in the Japanese textbook. Additionally, scientific terms in the Japanese textbook appear with greater repetition. These findings suggest that Taiwanese textbooks may impose a higher cognitive load on students. Statistical analysis revealed a high level of agreement between human and AI term identification, validating the feasibility of this hybrid method for textbook research. The study highlights the need for more balanced content design in Taiwanese textbooks and supports the application potential of generative AI in curriculum analysis.