NLP patent examination

NLP patent examinationexamine the Impact of Firm Knowledge Source Diversity and Technology Cycle on Patent Examination Duration in Natural Language Processing FieldAs a rapidly evolving frontier of artificial intelligence, natural language processing (NLP) is characterized by fast-paced iterations driven by software innovation and knowledge recombination, making patent examination duration a critical factor in sustaining its innovation momentum. This study employs the Cox proportional hazards model to explore the influence of firm knowledge source diversity and the technology cycle on patent examination duration in the NLP domain, using 200 firms with 30524 USPTO granted patents from 2001 to 2023. The findings reveal that a 1% increase in firm knowledge source diversity leads to a 1.51% increase in patent examination duration, and a 1% increase in technology cycle leads to a 0.19% decrease in patent examination duration. The AFT model confirms the robustness of these results. These findings underscore the distinct roles of knowledge source composition and temporal distance of technological knowledge in shaping patent examination duration.Keywords: patent examination duration; firm knowledge source diversity; technology cycle; natural language processing; emerging technologies; Cox proportional hazard model

自然语言处理专利审查：探究自然语言处理领域企业知识来源多样性与技术周期对专利审查时长的影响
作为人工智能领域快速演进的前沿方向，自然语言处理（Natural Language Processing，NLP）以软件创新与知识重组为核心驱动力，迭代节奏迅猛，使得专利审查时长成为维系其创新动能的关键要素。本研究采用Cox比例风险模型（Cox proportional hazards model），以2001年至2023年间200家企业的30524件美国专利商标局（United States Patent and Trademark Office，USPTO）授权专利为样本，探究企业知识来源多样性与技术周期对自然语言处理领域专利审查时长的影响。研究结果显示，企业知识来源多样性每提升1%，专利审查时长将增加1.51%；技术周期每提升1%，专利审查时长则降低0.19%。加速失效时间模型（Accelerated Failure Time Model，AFT）验证了上述结果的稳健性。上述研究结果凸显了知识来源构成与技术知识时间距离在塑造专利审查时长过程中的独特作用。
关键词：专利审查时长；企业知识来源多样性；技术周期；自然语言处理；新兴技术；Cox比例风险模型