STTR Phase I: Simplification AI for Workforce Upskilling

这个STTR第一阶段项目通过开发新颖的个性化内容简化技术来解决提高劳动力技能的挑战。它将建立一个具有人工智能授权的内容选择和简化能力的新型数字粘合剂工具。劳动力技能提升问题的严重性和紧迫性需要立即和强有力的解决方案。企业迫切需要找到并培养具备适当技能的工人来填补新兴岗位。除了大型高科技公司，贸易和制造业领域对再培训的需求也在扩大。从现在到2030年，估计将有3亿至6亿人需要接受再培训。提出的突破性解决方案利用最先进的深度学习和自然语言处理算法，自动生成适合学员对所需内容和技能熟悉程度的培训材料。三个核心技术促进了这一点：一个内容简化解决方案，能够通过识别关键概念搜索和识别数字资源集合。应用的算法考虑了学习器的当前背景，并包含了理解这些关键概念所需的文档。第二种技术通过提出一种传统文本简化技术（如词替换）和一种新的信息检索方法（识别对高级文档的可理解性至关重要的概念）的混合方法来解决文本简化的开放问题。第三项技术是针对每个学员的需求和能力量身定制的个性化内容简化引擎，这些学员可以继续使用该技术进行持续的再培训、技能提升和终身学习。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。

This STTR Phase I project addresses the challenge of upskilling the workforce by developing novel personalized content simplification technology. It will build a novel digital binder tool with an AI empowered content selection and simplification capability. The magnitude and urgency of the workforce upskilling problem require an immediate and robust solution. Corporations are desperate to find and nurture appropriately skilled workers to fill emerging roles. Beyond big high-tech corporations, the need for retraining is expanding in the trades and manufacturing space. An estimated 300-600 million people will need to be retrained between now and 2030. The proposed breakthrough solution utilizes state-of-the-art deep learning and natural language processing algorithms to automatically generate training materials appropriate for the level of familiarity that trainees have with the content and skills they need to acquire. This is facilitated by three core technologies: a content simplification solution with the capability of searching and identifying collections of digital resources by identifying key concepts. The applied algorithms consider the current background of the learner and include documents needed to comprehend those key concepts. The second technology addresses the open problem of text simplification by proposing a hybrid approach of traditional text simplification techniques like word substitution and a novel information retrieval approach that identifies concepts critical for the comprehensibility of advanced documents. The third technology is a personalized content simplification engine tailored to the needs and capacities of each trainee who can continue to use the technology for continuous retraining, upskilling and lifelong learning.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.