SBIR Phase I: Automatic Editing Through Semantic Alignment with Deep Learning

这个小企业创新研究（SBIR）第一阶段项目的更广泛的影响/商业潜力是使繁琐、耗时和昂贵的合同审查和谈判的人工过程变得过时。拟议研究的结果将基于用户审查和谈判少量类似合同的历史，准确地审查和谈判绑定合同，并将导致(a)公司合同审查和谈判时间减少50-90%，(b)组织内所有合同的标准化风险。此外，拟议的研究将使中小型企业能够负担得起并获得与世界上最大和最成熟的公司相同质量的对绑定合同的法律审查。这个小企业创新研究第一阶段项目将开发第一个通过语义句子对齐和深度学习技术自动编辑合同的系统。拟议的研究将扩展最先进的方法，用于各种长度文本段的分布式表示的无监督学习，然后推出第一个也是唯一一个具有语义编辑功能的自动化合同审查和谈判的机器学习平台。

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to render the tedious, time-consuming, and expensive manual process of contract review and negotiation as archaic. The outcome from the proposed research will accurately review and negotiate in-bound contracts based on a user's history of reviewing and negotiating just a handful of similar contracts and will result in (a) a 50-90% reduction in companies' contract review and negotiation time and (b) standardized risk across all contracts within an organization. Furthermore, the proposed research will provide small and medium-sized businesses with the ability to afford and obtain the same quality of legal review of in-bound contracts as the largest and most sophisticated companies in the world. This Small Business Innovation Research Phase I Project will develop the first system to automate edits in contracts through semantic sentence alignment with deep learning techniques. The proposed research will expand upon state-of-the-art methodologies for unsupervised learning of distributed representation of various length text segments and then launch the first and only machine learning platform that automates contract review and negotiation with semantic editing capabilities.