MENTORSHIP
收藏资源简介:
MENTORSHIP是一个由雪城大学信息学院创建的大规模数据集,包含743,176条跨112个科学领域的导师关系。该数据集通过众包方式收集,避免了传统数据集的局限性,如领域狭窄和忽视早期职业及非出版相关互动。数据集内容丰富,包括科学家的出版数据和使用深度学习内容分析的‘语义’研究表示。此外,考虑到性别和种族在科学分析中的重要性,数据集还提供了这些因素的估计。MENTORSHIP旨在促进科学领域导师关系的研究,加深我们对导师在科学家职业成果中作用的理。
MENTORSHIP is a large-scale dataset developed by the School of Information Studies at Syracuse University, containing 743,176 mentor-mentee relationships spanning 112 scientific disciplines. Collected via crowdsourcing, this dataset circumvents the limitations of traditional datasets, such as narrow domain coverage, neglect of early-career researchers and exclusion of non-publication-related interactions. The dataset features rich content, including scientists' publication records and "semantic" research representations derived through deep learning-based content analysis. Furthermore, given the critical role of gender and race in scientific analyses, the dataset also provides estimates of these two demographic factors. MENTORSHIP aims to advance research on mentor-mentee relationships in scientific fields and deepen our understanding of the role mentors play in scientists' career outcomes.

- 1A dataset of mentorship in science with semantic and demographic estimations信息学院,雪城大学,纽约州雪城 · 2021年



