A Study on an Assessment Framework for the Novelty of Ideas Generated by Analogical Thinking

Although there have been many educational programs focusing on the creation of new ideas, the assessment of novelty is still a controversial issue. As an ideation tool, analogical thinking enables conceptual change, which is seen as a crucial aspect of creativity. In this regard, the use of analogy can be an important instrument to facilitate novel idea generation. Analogies are generated by superficial or structural similarities from the memory. For creating a new idea by analogy, this study regards novelty as the domain-changing influenced by structural consistency with the source ideas. Consequently, we designed an assessment framework based on the latent semantic analysis of the domains and the consistency of the underlying mechanism between the source and the new ideas. Data was collected from the 14 subjects who participated in the workshop for this study. The workshop consists of three tasks: 1) Pre-task: All subjects were asked to read the 25 cases of the collective intelligence services, which is a business model creating value from large and loosely organized groups of people working together electronically e.g. Amazon.com, Google Japanese input; 2) Categorization task: Subjects were asked to categorize each case based on the underlying mechanism of the business through group discussion; 3) Generation task: Subjects were asked to create a new service idea individually using analogical thinking. As a result, 12 ideas were created, 6 of which were assessed as novel according to our assessment framework. Among the remaining 6 ideas, 4 were assessed as having high superficial similarity in terms of the idea domain, and 2 as having neither superficial nor structural similarity with the source ideas. Although our findings suggest that the proposed assessment framework for novelty evaluation is unable to provide a ‘one-size fits-all method’, it does enable us to overcome some of the limitations of current evaluation methods which depend on subjective judgements for rating.