CRII: III: A Socio-technical Approach for Biomedical Content Authoring and Structured Web Publishing

从世界上可用的海量信息中高效地检索与上下文相关的、准确的和有价值的信息是任何搜索引擎成功的基石。例如，在图书馆，图书馆员使用复杂的策略，在不同程度上成功地从众多可用图书中检索适合读者的书。同样，由于生物医学研究和临床实践的空前增长，近年来产生了大量的数字生物医学文本内容，如研究论文、临床笔记、生物医学报告等。访问这些生物医学内容的有效策略对于允许将正确的信息从科学研究界及时传输到同行研究人员和感兴趣的个人至关重要。尽管有价值的信息被嵌入到在线生物医学内容中，但对于信息检索和知识提取搜索引擎来说，这些信息仍然是不透明的。这主要是因为大多数生物医学内容是非结构化的，带有少量或没有显式的机器可解释的语义(上下文感知)标记或注释。例如，谷歌搜索引擎算法需要元数据来以上下文感知(语义)的方式适当地标记生物医学内容，从而实现更精确的搜索。因此，开发一个APT技术基础设施，使内容用户能够向生物医学内容添加上下文感知标记(语义注释)，并在以后分享它们，以提高准确性促进其可访问性，将是游戏规则的改变者，也是当今的需要。该项目通过开发最先进的可自由访问的交互系统来促进科学和技术的进步，该系统使生物医学领域中不同专业水平的个人能够合作编写和发布生物医学语义内容。因此，这项研究使目前对精英用户可用的语义内容创作过程大众化。该项目旨在让不同社会经济背景的学习者参与获取实践经验，实现点对点的知识共享，并打开许多职业发展机会。这项研究不仅将产生公共教育材料、书目和可重复使用的经验数据，以支持广泛的生物医学利益相关者，而且还将通过将研究成果整合到生物医学信息学课程中，为课程改进提供机会。本研究引入了一种开箱即用的社会技术方法和一套新颖的语义内容模型来开发开放式交互系统，以克服在语义内容创作和结构化发布过程中平衡速度和准确性以及语义保留的研究挑战。平衡速度和准确性是关键的研究挑战；在内容创作过程中实时找到正确的语义注释是极其困难的，因为通常一个语义注释在具有不同文本或内涵的多个生物医学本体中可用。本研究开发了一个生物医学语义内容创作系统，通过在整个过程中保持原始作者在循环中，来平衡现有生物医学注释器的速度和准确性。像AuthorsLikeMe和RecommendMe这样的智能算法是为了从从事类似生物医学内容的其他作者那里获得实时帮助而开发的。语义咖啡馆：虚拟协作环境的串联使新手作者能够向专家寻求帮助，专家是基于帮助他人的激励模型而被吸引的。保留内容级语义仍然是在线出版面临的关键挑战和关注的焦点。该团队通过开发新的语义模型来解决这个问题，该模型扩展了研究引擎认可的受控元数据，以在网络上发布生物医学内容。开发了多个用户友好的、可定制的界面，用于创作、发布、概念搜索和与同行共享知识。这项研究的副产品之一将是独一无二的生物医学注释语料库，它将通过API公开提供，供任何潜在用途使用。从用户的角度了解系统的效率和内容的质量将通过用户研究和系统评估进行研究。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Efficient retrieval of context-appropriate, accurate and valuable information from the myriad of information available in the world is the bedrock of any successful search engine. For instance, in a library, librarians employ complex strategies with varying degrees of success to retrieve the appropriate book for a reader out of the many available titles. Similarly, a large volume of digital biomedical textual content such as research papers, clinical notes, biomedical reports, etc., has been produced in recent years due to the unprecedented growth of biomedical research and clinical practices. Efficient strategies for accessing these biomedical contents are crucial for allowing a timely transfer of correct information from the scientific research community to peer investigators and interested individuals. Even though valuable information is embedded in online biomedical contents, it remains opaque to information retrieval and knowledge extraction search engines. It is mainly because most biomedical contents are unstructured with minor or no explicit machine-interpretable semantic (context-aware) markups or annotations. For example, Google search engine algorithms require metadata to properly tag biomedical contents in a context-aware (semantic) fashion enabling more precise searches. Thus, developing an apt technology infrastructure to empower content users to add context-aware tagging (semantic annotations) to the biomedical content and later share them to promote their accessibility with improved accuracy would be a game-changer and the need of the hour. This project promotes the progress of science and technology by developing state-of-the-art freely accessible interactive systems that enable individuals at different expertise levels in the biomedical domain without any technical background to collaboratively author and publish biomedical semantic content. Hence, this research democratizes the process of semantic content authoring currently available to elite users. This project intends to involve a socio-economically diverse pool of learners in getting hands-on experience, enabling peer-to-peer knowledge sharing and opening up many career progression opportunities. The research will not only generate public educational material, bibliographies, and reusable empirical data to support wide-ranging biomedical stakeholders, but it will also provide opportunities for curriculum enhancement by integrating research results in biomedical informatics courses. This research introduces an out-of-the-box socio-technical approach and a novel set of semantic content models to develop open interactive systems overcoming the research challenges of balancing speed and accuracy and semantic preservation during semantic content authoring and structured publishing. Balancing speed and accuracy is the key research challenge; finding the right semantic annotations in real-time during content authoring is extremely difficult since often one semantic annotation is available in multiple biomedical ontologies with different text or connotations. This research develops a biomedical semantic content authoring system to balance the speed and accuracy of available biomedical annotators by keeping the original author in the loop during the entire process. Intelligent algorithms such as AuthorsLikeMe and RecommendMe are developed to get real-time help from other fellow authors working on similar biomedical contents. Semantically Cafe: A virtual collaboration environment in tandem enables novice authors to seek help from experts where the experts are attracted based on an incentive model for helping others. Preserving content-level semantics remains the key challenge and center of attention of online publishing. The team addresses this by developing novel semantic models extending the research engines endorsed controlled metadata to publish the biomedical contents on the web. Multiple user-friendly, customizable interfaces are developed for authoring, publishing, concept searching, and knowledge sharing with peers. One of the by-products of this research will be one-of-its-own kind of biomedical annotations corpus that will be openly available through the API for any potential usage. Understanding the system's efficiency and the quality of the content from the users' perspective will be studied through user studies and system evaluations.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.