Since my travel through Scientometrics started I felt that impact was measured in terms of citations within the academic world: articles citing articles, journals citing journals, metrics feeding other metrics. But what happens when the research work escapes its academic shell and appears in patents?
As I’m working to understand the state of the art regarding the science–technology interface (particularly how scientific publications migrate into industrial innovation) this paper offers exactly that perspective: a thorough comparison of Google Patents and Lens.org, two major open-access platforms that expose patent data and, crucially, the scientific literature cited within patents.
Velayos-Ortega, G., & López-Carreño, R. (2021). Google Patents versus Lens: citaciones de literatura científica en patentes. Revista general de información y documentación, 31(1).
https://doi.org/10.5209/rgid.72257
Linking Science and Technology through Patent Citations:
This patent begins by framing patents as both protective tools and drivers of innovation, serving researchers and technologists alike by providing access to prior technological knowledge. Crucially, patents increasingly include citations to scientific literature, forming a measurable bridge between academic research and technological develoment.
Foundational scientometric work (including Carpenter & Narin (1983), Narin & Norma (1985), Meyer (2000) among others) already showed that references to scientific works in patents can indicate the flow of knowledge from science to industry. The authors emphasise that such citations may eventually serve as indicators of technological impact for academic publications, complementing traditional citation-based metrics in scholarly evaluation.
The paper also clarifies hoy Non-Patent Literature (NPL) is incorporated during the patent examination process, and how differing national standards and lax normalisation lead to highly heterogeneous citation formats. This sets the stage for the comparative analysis of two key platforms that display such data: Google Patens and Lens.
The Method:
The paper has an exploratory comparative approach. The authors retrieve and visualise multiple types of patents, considering:
- Jurisdiction
- International applicability
- Patent Status
- Piublication date
- Keyword-based searches in high-patenting technological fields
For both platforms (Google Patens and Lens), they examine how it handles the citations, focusing on:
- Distinction between citing and cited documents
- Whether the author contribution (author vs examiner) is provided
- Search filters for citations
- Internal/External linking to scientific databases
- Treatment and structure of NPL
- Use of persistent identifiers
- Analytical tools (graphs, statistics, influence maps)
- Export options and formats
This methodological lens allows them to evaluate both data quality and usability for scientometric analysis.
Key Points:
Google Patents and Lens adopt contrasting approaches to citation handling:
- Google Patents emphasises transparency in citation authorship but provides limited structure, linking, and analytical functionality, particularly for NPL.
- Lens focuses on metadata-rich citation management, extensive use of persistent identifiers, and advanced analytical tools, although it does not distinguish between examiner- and applicant- added citations.
| Aspect Analysed | Google Patents | Lens |
|---|---|---|
| Separation of citation types | Yes, distinguishes patents, NPL, and citing documents. | Yes, distinguishes patents, NPL, and citing documents. |
| Citation authorship (examiner vs applicant) | Yes, indicates authorship (examiner, applicant, third parties). | No, does not distinguish authorship. |
| NPL normalisation | NPL presented in unstructured format; frequent duplication. | NPL structured where metadata are available; option to view the original citation. |
| Persistent identifiers (DOI, PMID, etc.) | Does not add PIDs. | Yes, uses DOI, PMID, PMCID, MAG ID, etc. |
| Internal and external linking | Links patents only; NPL without external links. | Comprehensive linking to patents and publications via metadata and PIDs. |
| Citation-based search filters | No, does not allow filtering by citations. | Yes, with advanced filters (DOI, ORCID, cited-by, etc.). |
| NPL document types | All types included, but not categorised. | Broad categories: articles, books, conference proceedings, standards, reports, theses, datasets. |
| Analysis and visualisation | Basic charts. | Advanced and configurable visualisations. |
| Influence maps | Not available. | Yes: PatCite and In4M. |
| Data export | CSV only, limited options. | CSV, RIS, BibTeX, JSON; flexible and shareable export. |
Limitations & What’s left to explore:
The authors acknowledge that, despite their strengths, both platforms still exhibit issues:
- Inconsistent or poorly normalised NPL citations due to upstream patent office practices.
- Duplicate entries in Lens caused by differing metadata from external providers.
- Google Patents’ lack of structured NPL data inhibits bibliometric usage.
- Absence of citation author attribution in Lens limits contextual interpretation.
These limitations highlight the need for standardised bibliographic practices in patent documentation and improved metadata harmonisation across platforms.
Conclusions:
The comparison reveals that Lens currently leads in terms of analytical depth, citation search functionality, NPL structuring, and integration with scholarly databases. Its influence maps and persistent identifiers make it particularly valuable for studying science-technology linkages.
Google Patents, while strong in global patent coverage and examiner attribution, still lacks robust NPL handling and citation-level linking.
The study underscores a growing recognition that citations to scientific literature in patents deserve a place in research assessment frameworks. If academia values societal and industrial impact, then tracking technological citations of scholarly work is not just desirable, it is necessary.
See you in the next paper =)