Idea Generation Support Using Conceptual Attributes
Masaki Murata · Information Engineering Express · 2023
We support idea generation from the conceptual attribute database.We make it possible to make analogies, such as "If we make a cat larger, it becomes a tiger," from the conceptual attribute database.We also confirmed the usefulness of the McRae dataset.Against the results of using the two analogy methods, we made a four-level evaluation ("very good," "good," "OK," and "bad").In the first analogy method, the accuracy rate of the proposed method was 0.84 when the cases other than "bad" were judged to be correct.The accuracy rate of using word2vec was 0.13.In the second analogy method, the accuracy rate of the proposed method was 0.67 when the cases other than "bad" were judged to be correct.The accuracy rate of using word2vec was 0.19.The performance of the proposed methods was reasonably high.It was also higher than using word2vec.Consequently, we obtained that the McRae database used in the proposed methods was useful.