摘要
[目的/意义]关键技术问题研究对于辅助各创新主体把握创新突破方向、攻克关键核心技术、推进技术创新具有重要意义。[方法/过程]针对目前技术关键问题挖掘领域识别样本、流程、方法上的不足,提出多文本分析的领域关键技术问题挖掘框架,以专利说明书技术背景为主,综述性论文结束语和技术需求文档为辅,从数据获取及转化、技术问题提取、技术问题主题识别、关键技术问题发现4个环节展开分析。[结果/结论]对石墨烯领域案例研究证明了多文本分析的领域关键技术问题识别框架的有效性;专利说明书技术背景、综述论文结束语与技术需求文档在挖掘技术问题上的表现、描述极性、结果上具有不同的解释力度,可互相佐证补充。
[Purpose/significance] Key technical problems are of great significance for assisting innovation entities to grasp the direction of innovation breakthroughs, conquer key core technologies, and promote technological innovation.[Method/process] Aiming at the shortcomings in the identification samples, processes and methods of the current key technical problem areas, a multi-text analysis framework for identifying key technical problems in the field is proposed.It is mainly based on the technical background of the patent specification, supplemented by the conclusion of the review paper and the technical requirements document.The analysis is carried out from four links: data acquisition and transformation, technical problem extraction, technical problem topic identification, and key technical problem discovery.[Result/conclusion] The case study in the field of graphene proved the effectiveness of the multi-text analysis of the field’s key technical problem identification framework, and discovered the technical background of the patent specification, the conclusion of the review paper and the technical requirements document on the performance of mining technical problems, the description polarity, and the results have different interpretation strengths, which can be mutually corroborated and supplemented.
出处
《情报理论与实践》
CSSCI
北大核心
2023年第1期187-196,共10页
Information Studies:Theory & Application
关键词
关键技术问题
多文本分析
机器学习
主题模型
key technical problems
multi-text analysis
machine learning
topic model
