SBIR Phase I: Artificial Intelligence Tool for Analysis of Legal Documents

SBIR 第一阶段：用于分析法律文件的人工智能工具

• 批准号: 2112315
• 负责人: Vinicius Silvestrin Pantoja
• 金额: $ 25.6万
• 依托单位: CLAUDIUS LEGAL INTELLIGENCE INC
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-08-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2112315&HistoricalAwards=false
• 关键词: SBIR; Phase; Artificial; Intelligence; Tool

The broader impact of this Small Business Innovation Research (SBIR) Phase I project will be to establish an artificial intelligence (AI) system capable of providing data-driven insights for attorneys. The legal community currently lacks data analysis tools to help with civil case preparation, which can lead to suboptimal trial outcomes. The proposed technology can help lower costs through document analysis. The technology is designed to both automate and improve the decision-making process and enable attorneys to expand their case load, as well as enabling cost-effective representation. This Small Business Innovation Research (SBIR) Phase I project will use federated learning techniques to train the technology’s algorithm across multiple decentralized databases without exchanging data samples, thus keeping information private and confidential. This approach overcomes the lack of access problem in applying AI to legal cases, without compromising data confidentiality. The proposed research will include two major objectives: 1) improve and verify the accuracy of the platform, and 2) create internal checks to ensure that the model does not propagate bias. Computational outputs will be assessed using data and records from randomly selected cases with known outcomes to demonstrate system accuracy; moreover, the model will explicitly account for potential sources of bias.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.

小企业创新研究 (SBIR) 第一阶段项目的更广泛影响将是建立一个能够为律师提供数据驱动见解的人工智能 (AI) 系统。法律界目前缺乏帮助民事案件准备的数据分析工具，这可能导致审判结果不佳。所提出的技术可以通过文档分析帮助降低成本。该技术旨在自动化和改进决策过程，使律师能够扩大案件量，并实现具有成本效益的代理。 这个小型企业创新研究（SBIR）第一阶段项目将使用联邦学习技术在多个去中心化数据库中训练该技术的算法，而无需交换数据样本，从而保持信息的私密性和机密性。这种方法克服了人工智能应用于法律案件时缺乏访问的问题，且不影响数据机密性。拟议的研究将包括两个主要目标：1）改进和验证平台的准确性，2）创建内部检查以确保模型不会传播偏差。将使用随机选择的具有已知结果的案例中的数据和记录来评估计算输出，以证明系统的准确性；此外，该模型将明确考虑潜在的偏见来源。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。