How team structure shapes technological differentiation: A semantic embedding-based assessment
A firm's research portfolio often faces a tension between reliance on dominant technologies and the pursuit of technological differentiation. This study examines how invention teams generate such differentiation by linking network structure to innovation outcomes. We develop a novel semantic distance measure based on patent text, capturing technological differentiation along a continuous space rather than discrete classification categories. Using this measure, we show that invention teams achieve higher levels of differentiation when they combine strong internal cohesion with diverse external network connections. Specifically, cohesive internal ties facilitate knowledge integration, while external ties spanning structural holes expand access to novel knowledge. These findings highlight the importance of balancing cohesion and brokerage in shaping technological differentiation and offer new insights into the organization of inventive activity.
Authors
- Bogang Jun (ORCID: https://orcid.org/0000-0003-4994-702X)
- Wonchang Hur (ORCID: https://orcid.org/0000-0001-6163-9148)
- Jonghyun Kim (ORCID: https://orcid.org/0000-0001-9432-2972)
Institutions
- Inha University (KR)
- IIT@MIT (US)
- Global Convergence (United States) (US)
Publication Details
- Journal
- Technovation
- Published
- 2026-10-01
- DOI
- https://doi.org/10.1016/j.technovation.2026.103736
- Primary Topic
- Innovation and Knowledge Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00