*4.4. Limitations of Existing Approaches*

While the three approaches supported by cutting-edge computational and computer technologies are useful for social network analysis and matching optimal collaborators, none of them can be directly used to match TIC and TUC in transnational innovation ecosystems for three reasons. First, although the recent development of social network analysis has shifted attention to node identification and link prediction, which are essential in TIC and TUC synergy building, the related methods are still lacking efficiency especially in discovering hidden links and bridging missing links as well as processing unstructured data as the focus our research [107]. Second, while computational professional social matching has been more effectively used in practice [34], it is rarely used in cross-sectoral and transnational contexts. Third, while AI benefits business life, technology and industry [124], it is rare to see any efforts placed in using AI to facilitate understandings of the social system in which industrial businesses and technological innovations are embedded. A particular challenge in the case of TIC and TUC synergy building is finding both suitable algorithms and training data.

#### **5. Our Proposed Future Solution**

To further explore the three research questions, we aim to develop a transdisciplinary approach of integrating social sciences studies and AI technology. While achieving such an ambitious goal will be a long process, we will present our preliminary thoughts about the approach here.
