Automated Methods for the Media Analysis of Health News Coverage

健康新闻报道媒体分析的自动化方法

• 批准号: 7884853
• 负责人: Delano J McFarlane
• 金额: $ 3.87万
• 依托单位: COLUMBIA UNIVERSITY HEALTH SCIENCES
• 依托单位国家: 美国
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-09-20 至 2011-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7884853
• 关键词: Address; African American; Algorithms; American; Automation; Body of uterus; Code; Communities; Community Healthcare; Data; Disease; Environmental Risk Factor; Evaluation; Feasibility Studies; Health; Health Professional; Hispanics; Human; Informatics; Knowledge; Language; Low income; Manuals; Methods; Metric; Modeling; Obesity; Perception; Population; Publishing; Reader; Research; Research Personnel; Role; Sampling; Statistical Models; Structure; Time; Upper arm; Vocabulary; base; health disparity; improved; news; racial and ethnic

DESCRIPTION (provided by applicant): The role of the media analysis researcher is to study and describe the content and quality of news coverage. Unfortunately media analysis researchers typically use methods that are time consuming, subjective and produce data that is hard to replicate. These limitations often restrict the scope of the research that can be pursued to broad topics over limited spans of time. This means that a variety of topics, in particular health topics that impact populations with health disparities (e.g. Hispanics and African-Americans) are often not studied. Informatics methods like statistical language modeling and probabilistic content modeling can facilitate media analysis by automating certain required tasks. In this research statistical language modeling and probabilistic content modeling will be used to develop automated methods for use in media analysis studies. With these methods researchers will more quickly and efficiently identify any shortcomings in the media's coverage of health topics, in particular topics that impact populations with health disparities. This knowledge can then be made available so that the public can be more informed health news consumers, the media can have the information needed to correct problems that exist in their coverage of health news and the healthcare community can anticipate gaps in the public's perception and knowledge of health issues. The methods used to study and identify shortcomings in the media's coverage of health are typically very manual, subjective and time consuming. In this research automated media analysis methods will be developed and evaluated. If researchers are equipped with improved, more automated methods, then more timely information on the quality of news coverage of health issues, particularly health topics that are important to communities with health disparities can be acquired and shared with the journalists, healthcare professionals and the public.

描述（由申请人提供）：媒体分析研究员的作用是研究和描述新闻报道的内容和质量。不幸的是，媒体分析研究人员通常使用耗时，主观和难以复制的数据的方法。这些限制往往限制了研究的范围，可以追求广泛的主题在有限的时间跨度。这意味着，各种各样的主题，特别是影响健康差异人群（如西班牙裔和非洲裔美国人）的健康主题往往没有得到研究。统计语言建模和概率内容建模等信息学方法可以通过自动化某些必要的任务来促进媒体分析。在这项研究中，统计语言建模和概率内容建模将被用来开发自动化的方法，用于媒体分析研究。通过这些方法，研究人员将更快、更有效地发现媒体对健康话题报道的任何不足，特别是影响健康差异人群的话题。然后，可以提供这些知识，以便公众可以成为更知情的健康新闻消费者，媒体可以获得纠正其健康新闻报道中存在的问题所需的信息，医疗保健界可以预测公众对健康问题的认知和知识的差距。用于研究和查明媒体健康报道中的不足之处的方法通常是非常人工、主观和耗时的。在这项研究中，将开发和评估自动化媒体分析方法。如果研究人员配备了改进的，更自动化的方法，那么关于健康问题的新闻报道质量的更及时的信息，特别是对健康差异社区很重要的健康主题，可以获得并与记者，医疗保健专业人员和公众分享。