CreaCogs is a 3 year project funded by the German Research Foundation (DFG).

The project was proposed and is run by Ana-Maria Olteteanu (Principal Investigator).

You can follow CreaCogs on ResearchGate.

Short Project Synopsis:

The aims of this project are twofold. One is the study of creative problem solving in humans. The second is the implementation of, study and experimentation with artificial cognitive systems which yield similar performance as human participants and can be evaluated with human creativity tests. These systems will be built on the bases of cognitive knowledge acquisition and cognitively inspired knowledge-organization and processes.

Three main task classes used for assessing creativity in humans will be considered: (i) the Remote AssociatesTest; (ii) creative object replacement and composition and (iii) empirical insight problems with objects. For each task class, normative data will be gathered from human participants to be used as a comparability point for artificial cognitive systems solving the same tasks. New tasks will be developed in each class, allowing the deeper study of processing particularities.

The artificial cognitive systems developed to solve similar tasks will be based on a unified framework of knowledge organization and processing, which has been established in preliminary work. Cognitive knowledge acquisition will be performed for each of these systems, collecting human data on visual associates, object properties and problem templates which will be used for populating the knowledge base of the systems. Some such systems will also have generative power, being able to generate new items for human creativity tests, which will allow better control over variables.

This research will contribute to: a) promote a unified framework for the study of creative problem-solving; b) elucidate cognitive principles of knowledge organization and processes/algorithms for creativity; c) develop variants of classical creative problem solving tests; d) extend and build new cognitive creative systems comparable to human performance, which can be used in cognitive modeling and e) establish a paradigm for artificial creative cognitive inference.