Query-based legal information retrieval system

Clausehound为小型企业开发了一种基于云的法律文件起草产品，为他们提供了比传统法律服务提供商更低成本的选择。虽然Clausehound产品目前在市场上并产生收入，但提高从内部开发的法律知识库检索信息的准确性，并将其发送给用户进行审查和起草，对其用户（客户）基础的增长至关重要。为了使Clausehound在市场上更具竞争力，他们将与Seneca学院合作，特别是与Elnaz Delpisheh博士合作，以帮助进一步推进公司的提取和检索技术。本项目提出的工作将帮助Clausehound通过改进(1)从非结构化和半结构化文档中提取法律文本（例如从word文档中提取法律条款）的过程，以及(2)根据用户提交的查询检索信息的过程（例如搜索字符串“自动更新条款”），来改进和优化其现有的商业可用技术。这将提高目前的识别准确率和文档解析时间，从而增强Clausehound技术平台的适销性，并支持他们的业务增长。

Clausehound has developed a cloud-based, legal document drafting product for small businesses which provides them with alower-cost alternative to traditional legal providers. While the Clausehound product is currently in the market and generatingrevenues, improving the accuracy at which information is retrieved from their internally developed legal knowledge repository, androuted to the user for review and drafting, is crucial for the growth of their user (customer) base. In order for Clausehound tobecome more competitive in the market, they will engage with Seneca College, and in particular, with Dr. Elnaz Delpisheh, to helpfurther advance the companys extraction & retrieval technology. The work proposed in this project will help Clausehound refineand optimize its existing, commercially-available technology by improving (1) the process by which legal text is extracted (e.g.legal clauses from word documents) from unstructured and semi-structured documents, and (2) the process by which informationis retrieved based on queries submitted by a user (e.g. search string "automatic renewal clause"). This will improve on the currentidentification accuracy rates and document parsing times, and thereby enhance the marketability of the Clausehound technologyplatform and support the growth of their business.