Standard Grant: ToxicDocs Research Infrastructure Project

该奖项支持一个名为ToxicDoc的研究基础设施项目。它将使人们能够方便地访问和使用数百万份文件进行多学科研究。最近的集体诉讼侵权诉讼已经从负责大规模引入有毒物质（如石棉和多氯联苯）的私营公司中发现了数百万份以前的秘密内部记录。这些文件包括高管之间的备忘录、未发表的科学研究、公关活动的规划文件、致政策制定者的信函以及行业协会会议记录。虽然从技术上讲，这些材料是公共记录的一部分，但即使是科学研究人员也很难获得。毒理学将为多个领域的学者提供重要的知识资源，研究健康，有毒物质和企业行动。预期用户包括环境和健康史、关键法律的研究、危害地理学、环境健康和环境正义研究等领域的学者。Toxicological将使用高性能计算方法，快速将数百万份文档转换为全文可搜索格式，从而使研究人员能够对制造商、监管机构和科学界之间的相互作用进行新的了解，并显示各利益攸关方如何对潜在健康风险的知识作出反应。它利用了数据库管理和Web开发方面的最新创新，创造了高效快速的用户体验。该项目包括开发一系列文本分析工具：毒理学关系矿工，使研究人员能够从数据库中提取频繁出现的名称，组织和地点，并映射其网络关系;自动文档分类器，对文档进行分类;以及时间序列机器，使用户能够跟踪短语的频率。ToxicDocs将为基于关键历史记录的知情政策辩论提供经验基础，不会偏袒任何特定一方。该奖项反映了NSF的法定使命，并且通过使用基金会的知识价值和更广泛的知识价值进行评估，被认为值得支持影响审查标准。

This award supports a research infrastructure project called ToxicDoc. It will enable easy access and use of millions of documents for multi-disciplinary research. Recent class-action tort litigation has unearthed millions of previously secret internal records from private firms responsible for wide-scale introduction of toxic substances, such as asbestos and polychlorinated biphenyls (PCBs). These documents include memos between executives, unpublished scientific studies, planning documents for public relations campaigns, letters to policymakers, and trade association meeting minutes. Although technically part of the public record, this material is extremely hard to access, even for scientific researchers. ToxicDocs will provide a critical intellectual resource for scholars in multiple fields examining health, toxics, and corporate action. Expected users include scholars in environmental and health history, critical legal studies, hazards geography, environmental health and environmental justice research.ToxicDocs will use high performance computing methods that quickly renders millions of documents into a full-text searchable format and thereby enable researchers to shed new light on interactions among manufacturers, regulatory agencies, and the scientific community, and to show how various stakeholders responded to knowledge of potential health risks. It leverages recent innovations in database management and web development to create an efficient and fast user experience. The project includes the development of a host of tools for textual analysis: a ToxicDocs Relationship Miner that enables researchers to extract frequently occurring names, organizations, and places from the database and map their network relations; an automated document classifier that categorizes documents; and a Time-Series Machine that enables users to trace the frequency of phrases across time. Without taking any particular side, ToxicDocs will provide an empirical base for informed policy debates that are grounded in key historical records.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.