Pauline Soutrenon , Aloïs De Valon , Romain Billet , Christophe Lecante , Denis Boulard , Jean-Yves Legendre
{"title":"Analyzing green science-related patents in cosmetics by using artificial intelligence","authors":"Pauline Soutrenon , Aloïs De Valon , Romain Billet , Christophe Lecante , Denis Boulard , Jean-Yves Legendre","doi":"10.1016/j.wpi.2023.102223","DOIUrl":null,"url":null,"abstract":"<div><p>A deep learning-based method was developed to retrieve and analyze patents related to green sciences in cosmetics. An exploratory phase was conducted with five transformer machine learning<span> models on a training set to screen out the most relevant one. A second phase was implemented to fine-tune the selected model by comparing the results generated by the algorithm to the manual scoring done by an expert on a naive set of data. The final model is based on a combination of BERT-like and Longformers models which successively analyze the abstract, the description and the first claim of each patent. A survey of the patent literature with the selected model shows that key information can be obtained regarding the patenting activity in green sciences in cosmetics.</span></p></div>","PeriodicalId":51794,"journal":{"name":"World Patent Information","volume":"74 ","pages":"Article 102223"},"PeriodicalIF":2.2000,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"World Patent Information","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0172219023000534","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"INFORMATION SCIENCE & LIBRARY SCIENCE","Score":null,"Total":0}
引用次数: 0
Abstract
A deep learning-based method was developed to retrieve and analyze patents related to green sciences in cosmetics. An exploratory phase was conducted with five transformer machine learning models on a training set to screen out the most relevant one. A second phase was implemented to fine-tune the selected model by comparing the results generated by the algorithm to the manual scoring done by an expert on a naive set of data. The final model is based on a combination of BERT-like and Longformers models which successively analyze the abstract, the description and the first claim of each patent. A survey of the patent literature with the selected model shows that key information can be obtained regarding the patenting activity in green sciences in cosmetics.
期刊介绍：
The aim of World Patent Information is to provide a worldwide forum for the exchange of information between people working professionally in the field of Industrial Property information and documentation and to promote the widest possible use of the associated literature. Regular features include: papers concerned with all aspects of Industrial Property information and documentation; new regulations pertinent to Industrial Property information and documentation; short reports on relevant meetings and conferences; bibliographies, together with book and literature reviews.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
