Towards a Unifying Framework for Formal Theories of Novelty

Abstract

Managing inputs that are novel, unknown, or out-of-distribution is critical as an agent moves from the lab to the open world. Novelty-related problems include being tolerant to novel perturbations of the normal input, detecting when the input includes novel items, and adapting to novel inputs. While significant research has been undertaken in these areas, a noticeable gap exists in the lack of a formalized definition of novelty that transcends problem domains. As a team of researchers spanning multiple research groups and different domains, we have seen, first hand, the difficulties that arise from ill-specified novelty problems, as well as inconsistent definitions and terminology. Therefore, we present the first unified framework for formal theories of novelty and use the framework to formally define a family of novelty types. Our framework can be applied across a wide range of domains, from symbolic AI to reinforcement learning, and beyond to open world image recognition. Thus, it can be used to help kick-start new research efforts and accelerate ongoing work on these important novelty-related problems.

The anticipatory paradigm

“…One promising, environment‐based approach is to generate alternative world models and search for valuable novel problem situations; non‐valuable novelties, also called nuisance novelties (Boult et al. 2021), do not reveal an agent's ability to accommodate novelty. The domain‐independent novelty generator7 by Dannenhauer et al.…”

“…Theories of agent-based novelty define novelty as related to the agent's previous experience. Boult et al's framework provides functions to evaluate if a given input is novel from an agent's perspective, which is useful in developing novelty-robust AI agents (Boult et al 2021). Theories of environment-based novelty (Langley 2020) define novelty as a structured, deterministic transformation from one environment model to another, independent of any agent's internal beliefs, knowledge, or experience.…”

AI Autonomy: Self‐initiated Open‐world Continual Learning and Adaptation

“…Interested readers may also read (Boult et al. 2021) for a more nuanced and perception‐based study of novelty. There are other related concepts to novelty, for example, out‐of‐distribution (OOD) samples , outliers , and anomalies .…”

“…Novelty is not restricted to the perceivable physical world but also includes the agent's internal world, for example, novel interpretations of world states or internal cognitive states that have no correspondence to any physical world state. Interested readers may also read (Boult et al 2021) for a more nuanced and perception-based study of novelty. There are other related concepts to novelty, for example, out-of-distribution (OOD) samples, outliers, and anomalies.…”

PerceptSent - Exploring Subjectivity in a Novel Dataset for Visual Sentiment Analysis

“…Open Set Recognition(BOULT et al, 2019;BOULT et al, 2021).Regarding lessons learned from the current dataset design and to help future dataset generations, we recommend:• Consider employing an ordinal scale instead of a categorical one, as discussed in (MARTINEZ; YANNAKAKIS; HALLAM, 2014; MELHART et al, 2020a; YANNAKAKIS; COWIE; BUSSO, 2018; PARTHASARATHY; BUSSO, 2021). Treat ratings as ordinals rather than nominal or categorical values yields less biased datasets and, in turn, more reliable emotional models.…”