Bibliometric analysis of patent infringement retrieval model based on self-organizing map neural network algorithm

• Published in: Library hi tech Vol. 38; no. 2; pp. 479 - 491
• Main Author: Zhu, Dimin
• Format: Journal Article
• Language: English
• Published: Bradford Emerald Publishing Limited 15.06.2020; Emerald Group Publishing Limited
• Subjects: Accuracy; Algorithms; Bibliometrics; Boolean; Clustering; Fuzzy logic; Fuzzy sets; Information retrieval; Infringement; Keywords; Methods; Neural networks; Ontology; Patent searches; Product development; R&D; Research & development; Self organizing maps; Data collection techniques; Data analysis; Databases; Hypertext; Data mining; Social sciences
• ISSN: 0737-8831; 2054-166X
• DOI: 10.1108/LHT-12-2018-0201

Purpose The purpose of this paper is to quickly retrieve the same or similar patents in a large patent database. Design/methodology/approach The research is carried out through the analysis of the issue of patent examination, the type of patent infringement search and theories related to patent infringement determination and text mining. Findings The results show that the model improves the speed of patent search. It can quickly, accurately and comprehensively retrieve the same or equivalent patents as the imported patent claims. Research limitations/implications The patent infringement detection mainly focuses on the measurement of patent similarity in the implementation method. It is not mature, and there is still much room for improvement in research. Practical implications The model improves the efficiency of patent infringement detection, increases the accuracy of detection and protects the interests of patent stakeholders. Originality/value This study has great significance for improving the efficiency of patent examiners.