Linking Text Mining with Ontology and Systems Biology

乔治城大学获得了一项资助，用于开展一系列生物创新挑战评估和生物固化研讨会，以解决目前在生物学领域使用文本挖掘工具的障碍。具体地说，将组织三个讲习班，汇集生物研究界和文本挖掘工具的开发人员，用于用户需求分析、基于用户的评价和工具集成的标准开发。讲习班的具体目标是：(i)确定要求和评价标准，以便广泛的生物用户群体最大限度地利用文本挖掘工具；（ii）提供基于系统和用户的评估，以及测量精确度和召回率的指标（基于系统），以及文本挖掘对生物定位和知识发现的影响，如吞吐量和质量（基于用户）；（iii）采用、开发和推荐社区标准，以改进用于数据交换和工具集成的文本挖掘工具的互操作性。结果将是通过关注数据库管理和知识发现的生物学用例来弥合文献与知识之间的差距。研讨会的可交付成果将包括：(i)文本挖掘工具，包括交互式和/或集成文本挖掘系统，由BioCreative进行基准测试和评估，（ii） BioCreative评估中使用的文献语料库，以及（iii）这些研讨会在高影响力期刊特刊上发表的科学出版物，其中包含结果、评估和关键文章，包括对文本挖掘社区标准的建议。乔治敦大学将与Mitre公司合作举办这一系列讲习班。讲习班将把文本挖掘和生物界联系起来，以制定共同的标准和用户需求，其广泛影响超出了本项目中具体的文本挖掘应用。该项目将为参与项目或参加研讨会的学生和研究人员提供跨学科的研究经验。该项目将为广泛的生物学领域提供研究和教育基础设施，使文本挖掘系统成为生物定位和知识发现的有利基础设施。有关生物创意研讨会活动的进一步信息可在http://biocreative.sourceforge.net/上找到

Georgetown University is awarded a grant to conduct a series of BioCreative Challenge Evaluations and Biocuration Workshops to address the current barriers in using text mining tools in the biology domain. Specifically, three workshops will be organized to bring together the biological research community and developers of text mining tools for user requirement analysis, user-based evaluations and standard development for tool integration. The specific aims of the workshops are to: (i) define requirements and evaluation criteria that will maximize utilization of text mining tools by the broad biological user community; (ii) provide both system- and user-based evaluations, with metrics that measure precision and recall (system-based), as well as the effect of text mining on biocuration and knowledge discovery such as throughput and quality (user-based); and (iii) adopt, develop and recommend community standards to improve interoperability of text mining tools for data exchange and tool integration. The outcome will be to bridge the gap in linking literature to knowledge by focusing on biological use cases for database curation and knowledge discovery. The deliverables from the workshops will consist of: (i) text mining tools, including interactive and/or integrated text mining systems, that are benchmarked and evaluated by BioCreative, (ii) literature corpora used in the BioCreative evaluations, and (iii) scientific publications from these workshops in special issues of high impact journals containing results, evaluations and critical articles, including recommendations for community standards for text mining. Georgetown will collaborate with the Mitre Corporation in carrying out this workshop series.The workshops will connect the text mining and biological communities to develop common standards and user requirements, with broad impact beyond the specific text mining applications in this project. The project will provide interdisciplinary research experience for students and researchers involved in the project or participate in the workshops. The project will provide a research and educational infrastructure for broad areas of biology, allowing text mining systems to become an enabling infrastructure for biocuration and knowledge discovery. Further information on the Biocreative workshop activities may be found at http://biocreative.sourceforge.net/