Abstract:Purpose This study aims to analyse the relationship between a patent’s network position in a knowledge search network and the likelihood and speed of patent transactions. Additionally, it explores whether patent scope moderates these relationships. Design/methodology/approach In this empirical study, the authors collected a sample of patents in the artificial intelligence industry over the period of 1985–2018. Then, the authors examined the direct roles of degree centrality, betweenness centrality and closeness centrality on the likelihood and speed of patent transactions and the moderating role of patent scope in the knowledge search network using the logit and accelerated failure time models. Findings The findings reveal that degree centrality positively affects both the likelihood and speed of patent transactions, while betweenness centrality enhances the likelihood, and closeness centrality significantly boosts both. However, regarding the speed of patent transactions, closeness centrality is the most impactful, followed by degree centrality, with no significant influence of betweenness centrality. Additionally, the patent scope moderates how betweenness centrality affects the likelihood of transactions. Research limitations/implications This study has limitations owing to its exclusive use of data from the Chinese Intellectual Property Office, lack of visibility of the confidential terms of most patent transactions, omission of transaction directionality and focus on a single industry, potentially restricting the breadth and applicability of the findings. In the future, expanding the data set and industries and combining qualitative research methods may be considered to further explore the content of this study. Practical implications This study has practical implications for developing a better understanding of how network structure in the knowledge search network affects the likelihood and speed of patent transactions as well as the identification of high-value patents. These findings suggest future directions for patent holders and policymakers to manage and optimise patent portfolios. Originality/value This study expands the application boundaries of social network theory and the knowledge-based view by conducting an in-depth analysis of how the position characteristics of patents within the knowledge search network influence their potential and speed of transactions in the technology market. Moreover, it provides a theoretical reference for evaluating patent value and identifying high-quality patents by quantifying network positions. Furthermore, the authors construct three centrality measures and explore the development of patent transactions, particularly within the context of the developing country.

Recommendation Mechanism for Patent Trading Empowered by Heterogeneous Information Networks.

A personalized recommendation system for high-quality patent trading by leveraging hybrid patent analysis

Patent collaborative filtering recommendation approach based on patent similarity

A Trust-Enhanced Patent Recommendation Approach to University-Industry Technology Transfer

Leveraging Online Behaviors for Interpretable Knowledge-Aware Patent Recommendation

Sequential patent trading recommendation using knowledge-aware attentional bidirectional long short-term memory network (KBiLSTM)

A Tri-Attention Neural Network Model-BasedRecommendation.

Semantic Based Heterogeneous Information Network Embedding for Patent Citation Recommendation

Recommendation in Heterogeneous Information Networks Based on Generalized Random Walk Model and Bayesian Personalized Ranking

Patent Partner Recommendation in Enterprise Social Networks

Interpretable patent recommendation with knowledge graph and deep learning

HRET: Heterogeneous Information Network for Recommendation in Testing and Inspection.

Citation Recommendation As Edge Prediction in Heterogeneous Bibliographic Network: A Network Representation Approach.

Newly Published Scientific Papers Recommendation in Heterogeneous Information Networks

Promoting Knowledge Recommendation in Innovative Engineering Design: a BERT-GAT-based Patent Representation Learning Approach

A Recommendation Method for Fusion of Multiple HIN Meta-Paths

Meta-Path Based Service Recommendation In Heterogeneous Information Networks

Predicting Patent Transaction Behaviour Based on Embedded Features of Knowledge Search Networks

Sequential Recommendation on Dynamic Heterogeneous Information Network

PatentRank: an Ontology-Based Approach to Patent Search

A Recommendation Approach Based on Heterogeneous Network and Dynamic Knowledge Graph