A1 Refereed original research article in a scientific journal
Analyzing multilingual knowledge innovation in patents
Authors: Segev A, Kantola J, Jung C, Lee J
Publisher: PERGAMON-ELSEVIER SCIENCE LTD
Publication year: 2013
Journal: Expert Systems with Applications
Journal name in source: EXPERT SYSTEMS WITH APPLICATIONS
Journal acronym: EXPERT SYST APPL
Volume: 40
Issue: 17
First page : 7010
Last page: 7023
Number of pages: 14
ISSN: 0957-4174
DOI: https://doi.org/10.1016/j.eswa.2013.06.013
Abstract
In the process of analyzing knowledge innovation, it is necessary to identify the existing boundaries of knowledge so as to determine whether knowledge is new - outside these boundaries. For a patent to be granted, all aspects of the patent request must be studied to determine the patent innovation. Knowledge innovation for patent requests depends on analyzing current state of the art in multiple languages. Currently the process is usually limited to the languages and search terms the patent seeker knows. The paper describes a model for representing the patent request by a set of concepts related to a multilingual knowledge ontology. The search for patent knowledge is based on Fuzzy Logic Decision Support and allows a multilingual search. The model was analyzed using a twofold approach: a total of 104,296 patents from the United States Patent and Trademark Office were used to analyze the patent extraction process, and patents from the Korean, US, and Chinese patent offices were used in the analysis of the multilingual decision process. The results display high recall and precision and suggest that increasing the number of languages used only has minor effects on the model results. (C) 2013 Elsevier Ltd. All rights reserved.
In the process of analyzing knowledge innovation, it is necessary to identify the existing boundaries of knowledge so as to determine whether knowledge is new - outside these boundaries. For a patent to be granted, all aspects of the patent request must be studied to determine the patent innovation. Knowledge innovation for patent requests depends on analyzing current state of the art in multiple languages. Currently the process is usually limited to the languages and search terms the patent seeker knows. The paper describes a model for representing the patent request by a set of concepts related to a multilingual knowledge ontology. The search for patent knowledge is based on Fuzzy Logic Decision Support and allows a multilingual search. The model was analyzed using a twofold approach: a total of 104,296 patents from the United States Patent and Trademark Office were used to analyze the patent extraction process, and patents from the Korean, US, and Chinese patent offices were used in the analysis of the multilingual decision process. The results display high recall and precision and suggest that increasing the number of languages used only has minor effects on the model results. (C) 2013 Elsevier Ltd. All rights reserved.
UTU Research Portal