摘要
伴随着工业4.0、大数据、云计算等新兴技术的凸现和发展，越来越多的国家和企业认识到推动经济、技术结构转型的重要性，因此如何掌握一个产业或行业的技术发展信息并未来的机遇，成为需要关注的问题。而专利作为一种重要的科技、产品、经济及法律信息资源，是科技创新成果传播的知识载体，承载着丰富的技术发展数据源。专利技术已经成为衡量一个国家或者产业科技竞争力的重要指标，科学技术的发明创造成为增强国家创新能力、提髙企业竞争优势的核心问题。因此，研究如何从专利中挖掘技术信息并形成知识图谱，从而识别核心专利，探寻技术的发展现状及路径，寻发展机遇，对企业探索如何保持核心竞争力具有重要的指导意义。 目前国内外对专利知识图谱的研究大多是对专利数据进行定量研究形成，包含的技术信息不足。本文在对专利文献进行分析的基础上，结合定性分析和定量分析方法，提出基于专利知识图谱的技术发展研究方法。本文提出基于SAO-VSM模型形成本文专利相似度计算方法，首先对专利摘要进行SAO结构提取，并计算SAO结构之间的相似度并计算得到基于SAO的专利相似度；选取技术关键词准确地提取专利说明书的文本特征构建VSM模型，通过基于相似系数的方法计算VSM专利相似度，最后两种方法进行结合得到本文的专利相似度，并通过RFID行业的专利数据验证了该方法的有效性。进一步地，通过对专利外部特征项进行定量分析，构建专利质量评估方法并将其运用于知识图谱的可视化及分析过程中，形成本文基于专利知识图谱的技术发展研究方法，并通过实验证明了专利质量对有价值的异常专利识别的有效性。之后将本文
的方法用于智能家居行业的技术发展研究中，通过各个发展时期的专利地图分析智能家居的核心专利，揭示该行业技术发展的情况；通过专利地图分析智能家居的异常专利及空位，指出未来发展的方向和机遇。最后提出本文研究的管理启示、存在不足及未来的研究方向。
关键词：专利相似度；专利质量；知识图谱；技术机遇；智能家居
Abstract
With the development of emerging technologies such as Industry 4.0. , Big Data and cloud computing, more and more countries and enterprises recognize the importance of promoting economic and technological transformation. Therefore, how to grasp the development information of an industry, and seize the opportunities of the future become a concern. As the important resources about technology, product, economic and legal information, patent is the knowledge carrier of the dissemination of scientific and technological innovation. Patent has become an important indicator to measure the competitiveness of a country or industry, and the invention and creation of science and technology become the core problem of strengthening the national innovation ability and improving the competitive advantage of the enterprise. Thus, it has important guiding significance for the enterprise to explore how to keep the core competitiveness to study how to dig out the technical information and f
orm the knowledge map from the patent. It can identify the core patents, explore the development status and the path of the technology, and find the development opportunity.
At present, the domestic and foreign research on the patent knowledge map is mostly based on the quantitative analysis of the patent data, which contains insufficient technical information. Based on the analysis of patent literature, this paper presents the research methods for technology development based on patent knowledge map combined with qualitative analysis and quantitative analysis. A patent similarity calculation method based on SAO-VSM model is proposed. Firstly, the SAO structure is extracted from the patent abstract, and calculates the similarity between the SAO structures to obtain the SAO - based patent similarity. Next, select the technical key words to accurately extract the patent instructions text structure to build VSM model, and calculate the VSM-based patent similarity by using the similarity coefficient. Then we obtain the patent similarity in this paper with SAO-VSM model, and verifies the validity of the method through the patent data about RFID industry. Furthermore, through the quantitative analysis of the external features of the patent, the patent quality evaluation method is proposed to constitute the research methods for technology development based on patent knowledge map. It is applied to the visualization and analysis of knowledge map, and this paper proves the
validity of patent quality to the identification of valuable anomaly patents through experiments. Moreov
er, the method of this paper is used in the research for technology development of intelligent home industry. Through the analysis of core patents in patent networks at different stages of development, this paper reveals the technology development of smart home industry. And by analyzing the abnormal patents and vacancies in patent map, this paper points out the direction and opportunities for future development. Finally, this paper puts forward the management implications, existing problems and future research directions.
Keywords: patent similarity; patent quality; knowledge map; technical opportunity; intelligent home
目录
摘要......................................................................................................................................... I A BSTRACT ............................................................................................................................. II 第一章绪论 .. (1)
1.1 研究背景及意义 (1)
1.1.1 研究背景 (1)
1.1.2 研究意义 (1)
1.2 国内外研究现状 (2)
1.2.1 专利相似度研究评述 (3)
1.2.2 专利知识图谱研究评述 (5)
1.3 研究内容 (7)
1.3.1 研究目标 (7)
1.3.2 研究内容 (8)
1.3.3 研究方法 (8)
1.3.4 主要创新点 (10)
1.4 论文安排 (11)
第二章专利数据的采集与处理 (13)
2.1 专利的组成及特征 (13)
2.1.1 专利数据的组成 (13)
2.1.2 专利信息的特征 (13)
2.2 专利数据的采集与处理 (14)
2.2.1 确定行业主题和范围 (15)
2.2.2 选择专利数据来源 (15)
2.2.3 编写检索表达式 (17)
2.2.4 建立数据库表格 (18)
2.2.5 专利数据清洗及信息抽取 (18)
2.3 本章小结 (20)
第三章基于SAO-VSM的专利相似度计算方法 (21)
3.1 文本挖掘的概念 (21)
3.2 专利相似度算法设计 (21)
3.3 基于SAO结构的专相似度 (23)
3.3.1 SAO结构概述 (23)
3.3.2 SAO结构提取 (23)
3.3.3 基于SAO的专利相似度计算 (25)
3.4 基于VSM模型的专利相似度 (28)
3.4.1 VSM模型概述 (28)
3.4.2 构建特征向量 (29)
3.4.3 基于VSM模型的专利相似度 (30)
3.5 基于SAO-VSM的专利相似度 (31)
3.6 实验分析 (31)
3.7 本章小结 (33)
第四章基于专利质量的知识图谱可视化及分析方法 (35)
4.1 专利知识图谱的构建流程 (35)
4.2 专利质量的评估方法 (36)
4.3 基于专利质量的专利网络 (41)
4.4 基于专利质量的专利地图 (44)
4.5 实验分析 (46)
4.6 本章小结 (48)
第五章基于专利知识图谱的智能家居技术发展研究 (49)
5.1 智能家居概述 (49)
5.1.1 智能家居的概念 (49)
5.1.2 智能家居的发展 (49)
5.2 智能家居专利数据获取及处理 (50)
5.3 智能家居专利网络分析 (51)
5.4 智能家居专利地图分析 (58)
5.5 我国智能家居发展的对策建议 (65)
豫公网安备41160202000603
站长QQ:729038198
关于我们
投诉建议