Increasing quality of automated patent search based on distributional semantics and bibliographic data

The authors describe the approach to increasing quality of automated patent search and emphasize the inefficiency of existing systems. The approach is based on automated formulation of query terminological vector out from the application text with its further extension with quasisynonyms from the distributional thesaurus built on the body of patent documents, and with further enrichment with bibliographic data – IPC (international Patent Classification) codes. Mathematical formalization of acquiring and extending query vector is provided; acquisition of patent distributional thesaurus is discussed. The authors propose original indicators of retrieval quality assessment with account to patent document specific character, i. e. the so-called “patent families”, which enables to evaluate the system’s capability to find the relevant inventions beyond the corresponding numbers. The experiments with Russian and English-language collections demonstrate the S@20 indicator increase by 10% as compared to standard search by keywords. The authors conclude on the effect of inclusion of patent families on the evaluation of search results efficiency. The independent expertise of search performance in the Russian language patent collection confirmed that the system delivered at least one relevant document in 96.25% cases. The developed algorithms have been implemented into the Rospatent search platform.  

1. Lahorte P. Inside the mind of an EPO examiner // World Patent Information. 2018. Vol. 54, Suppl. P. S18–S22. DOI: 10.1016/j.wpi.2017.03.005.

2. Grazhdanskii` kodeks Rossii`skoi` Federatcii (chast` chetvertaia) ot 18.12.2006 № 230-FZ. St. 1350.

3. Ukaz Prezidenta Rossii`skoi` Federatcii ot 28.02.2024 № 145 «O Strategii nauchno-tekhnologicheskogo razvitiia Rossii`skoi` Federatcii» // Sobranie zakonodatel`stva RF. 2024. № 9. St. 1123.

4. Lupu M., Hanbury A. Patent retrieval // Foundations and Trends in Information Retrieval. 2013. Vol. 7, № 1. P. 1–97.

5. Mahdabi P., Crestani F. Learning-based pseudo-relevance feedback for patent retrieval // Information Retrieval Facility Conference. Springer, 2012. P. 1–11.

6. Hain, Daniel & Jurowetzki, Roman & Buchmann, Tobias & Wolf, Patrick. (2022). A textembedding-based approach to measuring patent-to-patent technological similarity. Technological Forecasting and Social Change. 177. 121559. DOI: 10.1016/j.techfore.2022.121559.

7. Haghighian Roudsari, Arousha & Afshar, Jafar & Lee, Suan & Lee, Wookey. (2021). Comparison and Analysis of Embedding Methods for Patent Documents. 152–155. DOI: 10.1109/BigComp51126.2021.00037.

8. Ali, Amna & Tufail, Ali & De Silva, Liyanage & Abas, Pg Emeroylariffion. (2024). Innovating Patent Retrieval: A Comprehensive Review of Techniques, Trends, and Challenges in Prior Art Searches. Applied System Innovation. 7. 91.DOI: 10.3390/asi7050091.

9. Gorbunov A. V. Clasterny`i` podhod k formirovaniiu naborov patentny`kh danny`kh i ocenivanie kachestva poiska «urovnia tekhniki» // Nauchny`e i tekhnicheskie biblioteki. 2025. № 5. S. 58–80. DOI: 10.33186/1027-3689-2025-5-58-80.

10. Manning C. D., Raghavan P., Schütze H. Introduction to Information Retrieval. Cambridge University Press, 2008. 496 p.

11. Harris C. G., Arens R., Srinivasan P. Using classification code hierarchies for patent prior art searches // Current Challenges in Patent Information Retrieval. Springer, 2017. P. 287–304.

12. Mikolov T. et al. Efficient Estimation of Word Representations in Vector Space // arXiv preprint arXiv:1301.3781. 2013.

13. WIPO. Handbook on Industrial Property Information and Documentation. Glossary of Terms. URL: http://www.wipo.int/standards/en/pdf/08-01-01.pdf.

14. Sakai T. Metrics, Statistics, Tests: An IR Researcher’s Guide to Analysing and Comparing Systems. Springer Nature Singapore, 2021. 335 p.