期刊
COMPUTERS IN INDUSTRY
卷 142, 期 -, 页码 -出版社
ELSEVIER
DOI: 10.1016/j.compind.2022.103749
关键词
Design knowledge; Natural language processing; Chinese patent text; Knowledge graph; Design for patentability
资金
- National Innovative Methods Work Projects of China [2019IM020200]
- Hebei Province Innovative Ability Uplifting Project [21567657H]
- Advanced Postdoc Project of Hebei Province [B2021002049]
- Center for Information Retrieval at Harbin Institute of Technology
Patent documents are crucial sources of knowledge for engineering design. However, existing patent analysis methods face challenges in extracting design knowledge from patent databases due to language differences. This paper introduces a new approach that uses natural language processing techniques to build a patent design knowledge graph for extracting useful design information from Chinese patent texts. The proposed method is validated using randomly selected patent texts and tested for the design of a new storage device for patentability.
Patent document is one of the most important sources of knowledge for engineering design. Design for patentability has become increasingly important for the success of product in the competitive market. However, language differences hinder the effective extraction of design knowledge from patent databases of different countries using existing patent analysis methods. This paper proposes a new approach to build the patent design knowledge graph to extract useful design information from Chinese patent texts by using natural language processing techniques. A meta-model of the patent design knowledge graph is built based on the language characteristics of Chinese. The proposed method is verified using randomly selected patent text for the effective extraction of Chinese patent design knowledge. The feasibility of the proposed method is further tested to support design of a new storage device for the patentability.(c) 2022 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据