Clustering to Support Users Finding Unexpected Perspectives in Brainstorming

We propose a clustering method to support users in discovering unexpected perspectives of ideas. The proposed method performs principal component analysis on a set of ideas and clusters ideas with unexpected perspectives by reducing specific dimensions. By presenting the clustered results to users, the proposed method encourages discovery of unexpected perspectives. We conducted user experiments to verify the effectiveness of the proposed method, where participants were asked to name the groups of ideas output by the proposed method. We then asked the participants to evaluate the unexpectedness of the group names. The results suggest that the proposed method can cluster a set of ideas from unexpected perspectives compared to existing clustering methods.