A Hybrid Approach to Building Peer Groups for Technology Evaluation to Generate Reference Information

Technology evaluation facilitates successful implementation of a technology financing system. It is usually done with a combined approach of data-driven quantitative evaluation and expert-driven qualitative evaluation due to the labyrinthine nature of the technology, but largely relies on the qualitative judgement of experts. It naturally causes a problem that the evaluation result for the same technology varies considerably depending on who evaluates it. To enable consistent technology evaluation, a few methods have been presented to generate reference information for technology evaluation by identifying relevant firms. However, they cannot explore the detailed properties of individual technologies when building peer groups with the identified relevant firms. Since technology is the ultimate subject of evaluation, not only the similarity between firms, but also the similarity between technologies should be investigated in the process of identifying relevant firms. Therefore, this study proposes a hybrid approach, which builds peer groups by measuring both the similarities between firms and between technologies and generates reference information that enables efficient and consistent technology evaluation. It is quite common for several evaluators to be involved in technology evaluation. If the evaluation results are inconsistent, additional manual work should be performed to reconcile these inconsistent results. Therefore, the proposed approach can improve the efficiency of the technology evaluation activities by avoiding unnecessary manual work. Furthermore, by providing useful reference information to the evaluator in an automated way, it will help maintain the consistency of the evaluation result so that the result does not vary greatly depending on the evaluators.