SBIR Phase I: Workforce Creative thinking Assessment

Sparkting, Berkeley CA

Investigators

Abstract

This SBIR Phase I project focuses on designing exercises and developing a set of computational methods to measure and enhance semantic-based creativity in the workforce. When hiring a prospect or training an existing employee, there is no efficient way to measure one's problem solving abilities. In order to be an efficient problem solver, an employee must possess the knowledge (e.g., an employee is given a problem whose solution she has witnessed or contributed to previously) and the employee needs to think creatively (e.g., the employee encounters a novel problem; in most scenarios, creative employees use brainstorming methods to generate a number of possible solutions and test possible solutions to arrive at the best answer). This project's unique approach is the design of models, where creativity and divergent thinking is measured and a CQ index is generated. Using advanced techniques in Natural Language Processing (NLP) and psychometric approach, the project evaluates the responses and generates an index called "Creativity Quotient," which represents users' creativity. The responses are evaluated based three factors: fluency (# of unique and relevant ideas), flexibility (# of categories of responses) & originality (infrequency and uniqueness of responses). The technique deploys a word or phrase-based associative weight where the algorithm measures the uniqueness of associations between words and phrases. Unique associations receive higher weighting.