Firms' knowledge profiles: Mapping patent data with unsupervised learning

详细信息    查看全文

• 作者：Arho Suominen^a ; ^b ; ^{arho.suominen@vtt.fi}Author Vitae ; Hannes Toivanen^a ; ^b ; ^c ; ^{hannes.toivanen@teqmine.com}Author Vitae ; Marko Seppä ; nen^d ; ^{marko.seppanen@tut.fi}Author Vitae
• 关键词：Technology management ; Patent analysis ; Unsupervised learning ; Topic modelling ; Telecommunication industry
• 刊名：Technological Forecasting and Social Change
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：115
• 期：Complete
• 页码：131-142
• 全文大小：1437 K
• 卷排序：115

文摘

We show the applicability machine learning in industry level patent analysis. Data consists of nearly 160,000 full-text USPTO patents in telecommunication. A topic modelling approach, Latent Dirichlet Allocation (LDA), is employed. We identify topical and temporal dynamics in patenting within the sample. We discuss the benefits and limitations of the method in patent analysis.