Paper on Increasing Diversity in Crowd Ideation with Language Model Embedding published in ACM CHI 2021

Crowd sourcing can obtain many creative ideas, but without careful coordination, many ideators may generate the same ideas. This leads to redundancy and limits the diversity of ideas from the crowd.
To mitigate this redundancy, we propose Directed Diversity (DD), that selects diverse phrase prompts based on a diversity maximization algorithm on the language embedding of phrases.
We further propose the Diversity Prompt Evaluation Framework (DPEF) to evaluate the diversification process with various measurement methods and metrics.
We show that Directed Diversity improves diversity compared to Random or No prompts.

Congratulations to team members Yunlong Wang, Sam Cox, Ashraf Abdul and collaborator Christopher von der Werth!

Directed Diversity technique


Diversity Prompt Evaluation Framework


Video Preview:

CHI presentation:

Samuel R. Cox, Yunlong Wang, Ashraf Abdul, Christian von der Werth, Brian Y. Lim. 2021. Directed Diversity: Leveraging Language Embedding Distances for Collective Creativity in Crowd Ideation.
In Proceedings of the international Conference on Human Factors in Computing Systems. CHI ’21.