Method for forming ontology “Patent representation of technical systems” for creating innovative technical systems

In this work, one of the most pressing problems of the synthesis of new technical solutions was solved - the automated generation of information support based on the analysis of USPTO patents. As concepts of the ontology of the subject area "Patent representation of technical systems", the structural elements of a technical object (TO) and the relationship between them, as well as descriptions of the problems solved by the invention were considered. The first claim of the patent document acted as the main source of information. The unit of extraction was the semantic structures SAO (Subject-Action-Object). The main linguistic features of patent documents were identified. Methods for preprocessing the patent array, extracting SAO from the patent formula, exporting extracted SAOs to the domain ontology have been formed. The developed methods have been tested on US patent documents. The average time for parsing one patent by an automated system is 1.72316 seconds, the accuracy of extracting information from the text of a patent is over 70%.

1. Korobkin D.M., Fomenkov S.A., Kolesnikov S.A. Metod sinteza funkcional'noj struktury novyh tekhnicheskih reshenij na osnove dannyh patentnyh massivov. Modelirovanie, optimizaciya i informacionnye tekhnologii. 2019;7(2):135-148.

2. Korobkin D.M., Fomenkov S.A., Kolesnikov S.G. Avtomatizaciya processa formirovaniya informacionnogo obespecheniya bazy dannyh fizicheskih effektov. Vestnik komp'yuternyh i informacionnyh tekhnologij. 2005;3(9):22-25.

3. Kharitonov A., Korobkin D., Fomenkov S., Kolesnikov S. Extraction of morphological features of technical systems from russian patent. V sbornike: CEUR Workshop Proceedings. IS 2019 - Proceedings of the 14th International Conference on Interactive Systems: Problems of Human-Computer Interaction. 2019:205-213.

4. Korobkin D.M., Vasiliev S.S., Fomenkov S.A., Lobeyko V.I. Extraction of structural elements of inventions from russian-language patents. V sbornike: Multi Conference on Computer Science and Information Systems, MCCSIS 2019 - Proceedings of the International Conferences on Big Data Analytics, Data Mining and Computational Intelligence 2019 and Theory and Practice in Modern Computing 2019. 4. 2019:159-166.

5. Vasil'ev S.S., Korobkin D.M., Fomenkov S.A. Metod izvlecheniya elementov konstrukcii izobretenij iz russkoyazychnyh patentov. Matematicheskie metody v tekhnike i tekhnologiyah - MMTT. 2019;7:105-110.

6. Choi, S. et al, 2011. SAO network analysis of patents for technology trends identification: A case study of polymer electrolyte membrane technology in proton exchange membrane fuel cells. Scientometrics, 2011:863-883. DOI: 10.1007/s11192-011-0420-z.

7. Stanza, 2020. URL: https://stanfordnlp.github.io/stanza/.

8. Kravets A.G., Korobkin D.M., Dykov M.A.E-patent examiner: two-steps approach for patents prior-art retrieval. V sbornike: IISA 2015 - 6th International Conference on Information, Intelligence, Systems and Applications. 2015. DOI: 10.1109/IISA.2015.7388074.