基于第三方点评数据的科技馆社会评价指标体系构建和实证研究

Empirical Study and Construction of Social Evaluation Index System for Science and Technology Museums Based on Third-Party Comment Data

中文关键词: 科技馆 社会评价 第三方点评数据 文本情感分析

英文关键词: science and technology museum  social evaluation  third party comment data  text sentiment analysis

中文摘要:

      随着科技馆在公众科学素质提升和科普教育中的作用日益凸显，如何科学、客观地评价科技馆的社会影响成为重要课题。本研究基于收集的全国400余家科技馆第三方点评数据，利用基于自然语言处理技术的主题建模、层次聚类方法，构建了一个包含9个一级指标、24个二级指标在内的科技馆社会评价指标体系，并利用机器学习算法对大众点评、美团、携程等第三方点评平台的133 371条评论数据进行实证分析。结果显示，各种规模科技馆的前3个一级指标“儿童娱乐与科学互动”“科学展示与文化体验”“互动学习与互动科技”得分都很高；各种规模科技馆在交通与环境满意度方面仍有较大提升空间。研究建议，特大型科技馆应突出科教展示优势，补足服务短板；大型科技馆要均衡发展核心功能，优化基础设施；中小型科技馆可打造“专题+特色”差异化发展路径，精细化运营弥补资源不足。

英文摘要:

      With the increasing impact of science and technology museums on enhancing public scientific literacy and science education，how to scientifically and objectively evaluate their social impact has become a critical issue. Based on third-party comments collected from over 400 science and technology museums nationwide，this paper employs natural language processing techniques such as topic modeling and hierarchical clustering to establish a social evaluation index system for science and technology museums. The system includes 9 primary indicators and 24 secondary indicators. Machine learning algorithms are used to conduct empirical analysis on 133 371 valid comments from third-party platforms such as Dianping，Meituan，and Xiecheng. The results show that the three primary indicators—“children’s entertainment and ecience interaction”“science exhibitions and cultural experiences，”and“interactive learning and technology”—score high across all types of science and technology museums. However，there remain significant improvement possibilities in transportation and environmental satisfaction. The study recommends that mega-sized science and technology museums should leverage their strengths in science education exhibitions while addressing service shortcomings；large-sized ones should balance core functions and optimize infrastructure；while medium and small-sized ones can pursue a differentiated development path of“specialization+features”and refine operations to compensate for resource limitations.