Abstract
Purpose The purpose of this study is to compare and analyze technology convergence trend analysis and technology convergence prediction performance of ‘SW original R&D' and `HW(device) original R&D.' Methods QAP(Quadratic Assignment Procedure) for network relevance measurement and graph auto-encoder model for link prediction have been used in this study. Results ‘SW original R&D' shows greater structural variability in the technology convergence network than ‘HW(device) original R&D.' In terms of link characteristics between IPC nodes, ‘HW(device) original R&D' shows more significant variability in the technology convergence network than ‘SW original R&D.' The link prediction performance using graph auto-encoder is higher for ‘SW' than ‘Device.' Conclusion The variability of node-to-node connection characteristics has a more significant impact on link prediction performance than the variability of network structure. In addition, this study shows that for technology convergence analysis and link prediction, it is necessary to examine not only the structural characteristics of the network but also the connection status and characteristics based on the interrelationships between entities in the network.
| Translated title of the contribution | A study on Technology Convergence Structure Variability and Convergence Prediction : Focused on R&D patent performance in ICT Original Technology |
|---|---|
| Original language | Korean |
| Pages (from-to) | 47-59 |
| Number of pages | 13 |
| Journal | 한국경영공학회지 |
| Volume | 28 |
| Issue number | 3 |
| DOIs | |
| State | Published - Sep 2023 |