Progress in research of textual quality evaluation of health-related media reports
10.3760/cma.j.cn112338-20241012-00629
- VernacularTitle:健康相关新闻文本质量评价研究进展
- Author:
Lei YANG
1
;
Min ZHAO
;
Shuying ZHAO
;
Wangxin XIAO
;
Peixia CHENG
;
Guoqing HU
Author Information
1. 中南大学湘雅公共卫生学院流行病与卫生统计学系，长沙 750013
- Publication Type:Journal Article
- Keywords:
Infodemiology;
Health-related media reports;
Textual big data;
Quality evaluation;
Social network
- From:
Chinese Journal of Epidemiology
2025;46(7):1269-1275
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To summarize the progress in the research of textual quality evaluation of health-related media reports.Methods:A systematic literature retrieval about textual quality evaluation of health-related media reports based on PubMed, Web of Science, Embase, Wanfang database, and China National Knowledge Infrastructure was conducted. Information regarding the textual quality definition, evaluation dimensions, indicators and methods of included papers was extracted.Results:A total of 29 study papers were included in this analysis, in which 26 were about retrospective textual quality evaluation of health-related media reports, and 3 were about the model or tool development for textual quality evaluation of health-related media reports. The topics of news reports included: 16 studies on injury, 3 on general health, 3 on infectious disease, 3 on cancer screening and treatment, 3 on chronic non-communicable disease, and 1 on medication risk. The definition of textual quality of health-related media reports and the dimensions of the quality evaluation varied across the studies. The quality evaluation indicators of media reports can be divided into three categories: availability of surveillance information, availability of professional information, and adherence to principles of media reporting. Most studies conducted the quality evaluation manually, with only 2 studies employing semi-automated or automated evaluation methods.Conclusions:No unified definition, set of dimensions, indicators, or automated algorithms exist for evaluating the textual quality of health-related media reports, which limits assessing massive news data effectively. It is necessary to conduct methodological studies on the textual quality evaluation of health-related media reports based on journalism and communication theory, infodemiology, deep learning, natural language processing, text mining, as well as specific disease and injury prevention theory.
