Linking innovations and patents - a machine learning assisted method

Research output: Working paper/PreprintPreprint (in preprint archive)


This paper describes the methodology behind the matching of patents and a literature-based innovation output indicator (LBIO) collected from trade journals covering the manufacturing and ICT service sectors in Sweden 1970-2015. A combination of manual processing and simple machine learning tools has enabled the identification, classification and linking of patents that otherwise would have been very difficult for either of the methods to detect on its own.
Data generated using this method can be used to assess many aspects of the relationship between patenting, knowledge accumulation and innovation activity.
Original languageEnglish
PublisherSocial Science Research Network (SSRN)
Number of pages17
Publication statusPublished - 2022

Publication series

NameSSRN:s working paper series
PublisherSocial Science Research Network (SSRN)

Subject classification (UKÄ)

  • Social Sciences Interdisciplinary

Free keywords

  • Innovation
  • Patents
  • Machine-learning
  • LBIO


Dive into the research topics of 'Linking innovations and patents - a machine learning assisted method'. Together they form a unique fingerprint.

Cite this