Developing computational design creativity systems

Abstract

The underlying hypothesis of this paper is that the rigor needed to build computational design creativity systems will produce insights that will increase our understanding of creativity. The paper points to a variety of factors that will need to be studied and understood in detail in order to build such systems. It examines selected technical literature about creativity to determine what key characteristics have been proposed as underlying creative production. In addition, as the existence of creativity is a judgment relative to personal or group norms, it is helpful to understand what factors inform that judgment. Besemer's creative product semantic scale for product assessment provides three dimensions, with nine attributes that can be used for judgments. This paper discusses how creative product assessment scales might provide important clues about how to build computational systems in order to produce creative products.

Computing the creativeness of amusing advertisements: A Bayesian model of Burma-Shave's muse

“…Another insight is that the evaluation of creativity is a judgment (Brown, 2013) and hence governed by psychological processes. According to EVE 0 , the processes of arousal and appraisal combine to produce a feeling of pleasure, which is how a person gauges the creativeness of an artifact.…”

“…In that case, the feedback loop from evaluation to generation lies in the head and hand of the developer, so it is not clear how much of an artifact's creativeness can be attributed to the computer (Boden, 2009; Burns & Maybury, 2010; Jennings, 2010). It is also not clear how creativity is evaluated in the minds of developers or other critics (Gero, 2010; Brown, 2013). What is clear is that artificial intelligence will require a capability to evaluate creativity, such as that embodied in natural intelligence, if systems are to credibly generate artifacts that humans will agree are creative.…”

Developing and evaluating computational models of musical style

“…), and sources (have too many consecutive events been sampled from the same database piece?) act as critics, recognizing potential mistakes and controlling the number of candidate passages generated (Thurston, 1991;Minsky, 2006;Brown, 2013). 2.…”

“…According to Gero and Maher (1993), artificial intelligence has generally been concerned with addressing routine design activities, whereas the field of computational creativity has emerged more recently, out of a need to study nonroutine design activities. These sentiments are echoed by an observation that although artificial intelligence algorithms “are able to produce remarkable results (Spector, 2008), it appears unlikely they will tell us much about creativity” (Brown, 2013, p. 52).…”

“…), and sources (have too many consecutive events been sampled from the same database piece?) act as critics, recognizing potential mistakes and controlling the number of candidate passages generated (Thurston, 1991; Minsky, 2006; Brown, 2013). Final evaluation of our systems' designs consists of a listening study that uses Amabile's (1996) consensual assessment technique (CAT), adapted to assessment of music-stylistic success.…”