This paper presents a twostep approach to generating comprehensive abstractive overviews for biomedical topics.It starts with a sensitivitymaximizing search of MEDLINE/PubMed and MeSHbased filtering of the results that arethen processed using NLP methods to extract relations between entities of interest. We evaluate this approach in acase study based on the IOM report on the role of vitamin D in human health. The report defines disorders thatserve as health indicators for the role of vitamin D. We evaluate the abstractive overviews generated using MeSHindexing and the extracted relations using the disorders listed in the IOM report as reference standard. We concludethat MeSHbased aggregation and filtering of the results is a useful and easy step in the generation of abstractiveoverviews. Although our relation extraction achieved 83.6% recall and 92.8% precision, only half of the disorders
【 预 览 】
附件列表
Files
Size
Format
View
Mining MEDLINE for problems associated with vitamin D