Abstract Semantic relations are fundamental to human conceptual knowledge and underpin cognitive processes such as categorization, analogical reasoning, and language comprehension. Despite extensive research on semantic relations in English, a systematic Chinese dataset aligned with established taxonomies remains unavailable, limiting cross-linguistic and developmental research. Here, we present a large-scale Chinese semantic relation dataset based on the ten-category taxonomy proposed by Bejar et al ., comprising 1,000 typical word pairs that were behaviorally validated. Initially, 3,758 candidate word pairs were generated by 32 college students and subsequently validated via a semantic relation judgment task with 5,898 participants aged 9–18 years. The final dataset includes relation labels, concreteness ratings for 2,295 unique words, and behavioral measures across developmental stages. Developmental analyses revealed significant age-related increases in accuracy for most semantic relations, with performance in most relations approaching a plateau at ages 15–17. This dataset provides validated materials and developmental performance patterns for investigating semantic cognition in Chinese-speaking populations, with applications in cognitive psychology, developmental science, computational linguistics, and natural language processing.
Fu et al. (2026) studied this question.