Identification of technological complementarity of enterprises based on the analysis of the patent database

Technological complementarity plays an increasingly important role in making strategic decisions regarding the choice of partners to work with as well as for expanding the diversity of innovation activities through the creation of alliances or mergers. Modern quantitative measurements of technology complementarity are largely based on research founded on patent classification codes. However, this approach ignores specific technologies in a particular field and, therefore, the results obtained are only general pointers. It is proposed to develop an approach to the quantitative measurement of the complementarity of enterprise technologies in reliance on statistical machine analysis of the patent array. The aim of the study is to develop a technology for clustering USPTO patent documents and identifying technological complementarity of enterprises based on the comparison of cluster information. The theoretical value lies in the following: the developed algorithms for patent parsing; clustering of the patent array based on the hLDA method; formation of "Enterprises – Clusters (Topics)" matrix; visualization of enterprise connections in a clustered patent space. The practical significance of the research lies in the developed software for identifying technological complementarity of enterprises, the effectiveness of which has been tested on a number of test examples.

1. 1. Guo Y. Research on the Risk Prevention of Overseas M & A of Chinese Enterprises. Proc. of the 2022 7th International Conference on Financial Innovation and Economic Development. 2022. DOI: 10.2991/aebmr.k.220307.085.

2. 2. Purba N. S., Nooraeni R. Using LDA for Innovation Topic of Technology : Quantum Dots Patent Analysis. Proc. of the 1st International Conference on Statistics and Analytics. 2020;1. DOI: 10.4108/eai.2-8-2019.2290336.

3. 3. Li H., Tan R., Xu В., Huo J. Patent design around method based on AFD for umbrella-type patent strategy. Journal of Mechanical Engineering. 2016;52(5):1–11. DOI: 10.3901/JME.2016.05.001.

4. 4. Wang M.-Y. Exploring potential R&D collaborators with complementary technologies: the case of biosensors. Technological Forecasting and Social Change. 2012;79(5):862–874. DOI: 10.1016/j.techfore.2011.11.002.

5. 5. Wang X., Li R., Liu Y., Lei M. A Morphology-Driven Method for Measuring Technology Complementarity: Empirical Study Involving Alzheimer's Disease. Journal of Data and Information Science. 2022;7(3):20–48. DOI: 10.2478/jdis-2022-001.

6. 6. Borodin N., Korobkin D., Bezruchenko A., Fomenkov S. The search for R&D partners based on patent data. Journal of Physics Conference Series. 2021;2060(1):012022. DOI: 10.1088/1742-6596/2060/1/012022.

7. 7. Sun L., Xu W. Enterprise Complementarity Based on Patent Information. Wireless Communications and Mobile Computing. 2022;2022(6):1–7. DOI: 10.1155/2022/5797285.

8. 8. Wang Х., Qiao Y., Hou Y., Zhang S., Han X. Measuring Technology Complementarity Between Enterprises With an hLDA Topic Model. IEEE Transactions on Engineering Management. 2021;68(5):1309–1320. DOI: 10.1109/TEM.2019.2958113.

9. 9. Patents USPTO. United States Patent and Trademark Office. Available from: https://www.uspto.gov/patents (accessed on: 07.11.2022).

10. 10. Korobkin D., Fomenkov S., Fomenkova M., Vayngolts I., Kravets A. The Software for Computation the Criteria-Based Assessments of the Morphological Features of Technical Systems. Cyber-Physical Systems. 2021;350:161–172. DOI: 10.1007/978-3-030-67892-0_14.

11. 11. Gibbs sampler for the Hierarchical Latent Dirichlet Allocation topic model. GitHub. Available from: github.com/joewandy/hlda (accessed on: 07.11.2022).

12. 12. Natural Language Toolkit documentation. NLTK Project. Available from: https://www.nltk.org (accessed on: 07.11.2022).