I-Corps: A Trustworthy, Interactive, Up to Date COVID-19 Knowledge Graph

I-Corps项目的更广泛影响/商业潜力在于开发一个易于使用的交互式知识图谱，其中包含最新公布的COVID-19医学发现的可靠信息。容易获得经过审查的医学发现可能会促使人们做出明智的决定，预计这将导致更好的卫生做法。这一解决方案可能会挽救美国乃至全世界许多人的生命。此外，该知识图谱可以扩展到其他疾病，并为衰老、癌症、心血管疾病、糖尿病等创建可访问的、易于使用的、值得信赖的医疗实践。I-Corps项目基于交互式知识图谱（KG）的开发，其中填充了最新发表的关于COVID-19的医学发现的可靠信息。目前，社会维护的现有KGs缺乏COVID-19医学发现和可扩展机制来保持图表的最新状态。提出的解决方案包括设计和评估新的可扩展算法和抽象。该技术在具有不同结构和元数据的非关系表中包含了COVID症状和可能的疫苗副作用。该团队已经构建了图形的初始“骨架”，并建议自动处理来自最近出版物的表格，以丰富KG。虽然大多数医疗表格都是复杂的，表现出分层的垂直/水平元数据，但该技术通过提供一种新颖的、可扩展的混合图，结合新的抽象来处理复杂的表格数据，开发具有新的二维表格嵌入层的多层深度/机器学习网络，以及为图表和医疗表格设计一个新的搜索引擎，解决了基本的挑战。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact/commercial potential of this I-Corps project is the development of an interactive, easy to use knowledge graph populated with trustworthy information from the latest published medical findings on COVID-19. Easy access to the vetted medical findings may motivate people to make informed decisions, which is expected to lead to better health practices. The solution may save many lives in the US and worldwide. Additionally, this knowledge graph can extend to other diseases and create accessible, easy to use, trustworthy medical practices for aging, cancer, cardiovascular diseases, diabetes, etc.This I-Corps project is based on the development of an interactive Knowledge Graph (KG) populated with trustworthy information from the latest published medical findings on COVID-19. Currently existing, socially maintained KGs lack COVID-19 medical findings and scalable mechanisms to keep the graphs up to date. The proposed solution includes the design and evaluation of new scalable algorithms and abstractions. The technology incorporates COVID symptoms and possible vaccine side-effects in non-relational tables having different structures and metadata. The team has constructed the initial “skeleton” of the graph and proposes to automatically process tables from recent publications in order to enrich the KG. While most medical tables are complex, exhibiting hierarchical vertical/horizontal metadata, this technology addresses the fundamental challenges by providing a novel, scalable hybrid graph, incorporating new abstractions to handle complex tabular data, developing a multi-layer deep-/machine-Learning network with new 2D tabular embedding layer, and designing a new search engine for graph and medical tables.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.