面向情报学课程设置的数据科学技能素养自动抽取及分析研究

Automatic Extraction and Analysis of Data Scientific and Technical Literacy for Informatics Curriculum

情报课程是情报学教学体系的核心和灵魂,是新的情报学人才培养模式的基础。在大数据、数据科学、人工智能发展的大环境下,数据学科与情报学之间存在很多共同点和交叉点,尤其是在对相应工作者所具有的技能素养上。文章通过对数据科学招聘信息中出现的人才技能素养实体的抽取,探究指导情报学学科紧跟时代发展潮流的课程设计内容。通过对招聘网站中数据科学相关工作岗位公告的抓取,经人工标注10534条公告数据,构建了基于预训练字嵌入的BI-LSTM-CRF神经网络的技能素养实体自动抽取模型,并在开发集中取得最高调和平均值85. 04%的效果。文章利用最优模型在11508条招聘公告中进行实体自动抽取,分析抽取结果,并围绕数据科学技能素养要求为情报学课程发展提出了适当建议。

Informatics curriculum is the core and soul of Informatics teaching system,which is the basis of the new mode of information science talents cultivation.In the environment of the development of big data,data science and artificial intelligence, there are many intersections between data science and information science,especially in the competency and skills of relative workers.This paper explores the teaching mode that guides the discipline of Informatics to keep up with the development trend of the times through the extraction of talents competency and literacy in recruitment information of data science.Based on the crawling of job advertisements related to data science in the recruiting websites,10534bulletin boards are manually tagged to construct an automatic extraction model of competency and literacy entities based on pre-training words embedded in BI-LSTM-CRF neural network, which achieves the highest.F-score of 85.04.By using the optimal model to extract entities automatically in 11508recruitment announcements and analyzing the results,the paper puts forward some appropriate suggestions for the development of Informatics curriculum based on the requirements of data science competency and literacy.