The Betty’s Brain computer-based learning system provides an open-ended and choice-rich environment for science learning. Using the learning-by-teaching paradigm paired with feedback and support provided by two pedagogical agents, the system also promotes the development of self-regulated learning strategies to support preparation for future learning. We apply metacognitive learning theories and experiential analysis to interpret the results from previous classroom studies. We propose an integrated cognitive and metacognitive model for effective, self-regulated student learning in the Betty’s Brain environment, and then apply this model to interpret and analyze common suboptimal learning strategies students apply during their learning. This comparison is used to derive feedback for helping learners overcome these difficulties and adopt more effective strategies for regulating their learning. Preliminary results demonstrate that students who were responsive to the feedback had better learning performance.

Our approach to mind-design derives from the view of language as a mirror of mind — a view compatible with the linguistic orientation of the Turing Test, and more concretely, with the remarkably tight coupling between linguistic structure and semantic entailment demonstrated by Richard Montague. Additional evidence for the power of this perspective comes from recent work in Natural Logic (NLog), in a sense a method of "reading off" certain obvious inferences directly from linguistic structure. Thus much of our past emphasis has been on developing a knowledge representation, Episodic Logic (EL), matching the expressivity of language, and inference machinery for this representation. More recently we have been striving to create broad bases of general world knowledge and lexical knowledge, while also adapting the latest version of our EPILOG inference engine to the kinds of obvious inferences that are the forte of NLog. At this point our knowledge collections range from sets of a few dozen core lexical axioms to millions of general "factoids" and quantified axioms derived from many of these, all expressed in EL. At the same time we have shown that EPILOG easily handles NLog-like inferences as well as ones beyond the scope of NLog.

Cognitive systems with human-level intelligence must dis­play a wide range of abilities, including reasoning about the beliefs of others, hypothetical and future situations, quanti­fiers, probabilities, and counterfactuals. While each of these deals in some way with reasoning about alternative states of reality, no single knowledge representation framework deals with them in a unified and scalable manner. As a conse­quence it is difficult to build cognitive systems for domains that require each of these abilities to be used together. To enable this integration we propose a representational framework based on synchronizing beliefs between worlds. Using this framework, each of these tasks can be reformu­lated into a reasoning problem involving worlds. This demonstrates that the notions of worlds and inheritance can bring significant parsimony and broad new abilities to knowledge representation.

This article proposes a suite of design decisions for the overall design of an Artificial Intelligence, i.e., a software system that exhibits intelligence in the spirit of the early days of A.I. research. The key aspects of the proposal are: (1) The identification of the A.I. system as a software individual that has the properties of integrity and persistence; (2) The construction of a software platform that integrates aspects of incremental programming languages and systems as well as of operating systems, with aspects that are intrinsic to knowledge-based artificial intelligence; (3) The use of a representation language that builds on essential aspects of S-expressions, Lisp, logic and extended set theory, but which is used both as a vehicle for software and as a publication language e.g. in lecture notes; (4) The identification of actions and aggregates of actions as first-class citizens in the representation language and as an important type of data object in the software system. The article also describes the Leonardo software platform, its representation language, its educational resources and its knowledgebase library which is one implementation of these proposed design decisions. Finally it makes a proposal concerning the research paradigm for this research area.

Desiderata for cognitive architectures that are to support the extent of human-level intelligence required in virtual humans imply the need to bridge a range of dichotomies faced by such architectures. The focus here is first on two general approaches to building such bridges — addition and reduction — and then on a pair of general tools – graphical models and piecewise continuous functions — that exploit the second approach towards developing such an architecture. Evaluation is in terms of the architecture’s demonstrated ability and future potential for bridging the dichotomies.

Historically, approaches to human-level intelligence have divided between those emphasizing the mechanisms involved, such as cognitive architectures, and those focusing on the knowledge content, such as ontologies. In this paper we argue that in order to build cognitive systems capable of human-level event-recognition, a comprehensive infrastructure of perceptual and cognitive mechanisms coupled with high-level knowledge representations is required. In particular, our contribution focuses on an integrated modeling framework (the “Cognitive Engine”), where the learning and knowledge retrieval mechanisms of the ACT-R cognitive architecture are combined with integrated semantic resources for the purpose of event interpretation.

Human wayfinding operates very very differently from traditional deterministic algorithms owing to a) restrictions in working memory resulting in subjective regionalized maps, and b)flexible adoption of different navigation strategies. While a number of cognitive strategies have been proposed for human wayfinding, these have been hard to evaluate thoroughly owing to a lack of computational simulation. In this work, we propose a stochastic approach for capturing these aspects, and argue for a memoryless, stationary implementation. In two longitudinal experiments on the same group of subjects, we first estimate the subjective regionalized maps for each subject on the same familiar spatial domain. Later, based on their wayfinding responses, we can estimate the stationary probabilities for different strategies. We apply this algorithm to evaluate three wayfinding strategies proposed in the literature, and repudiate the previously held suggestion that they are followed equiprobably.

This paper discusses the challenge of designing instructable agents that can learn through interaction with a human expert. Learning through instruction is a powerful paradigm for acquiring knowledge because it limits the complexity of the learning task in a variety of ways. To support learning through instruction, the agent must be able to effectively communicate its lack of knowledge to the human, comprehend instructions, and apply them to the ongoing task. Weidentify some problems of concern when designing instructable agents. We propose an agent design that addresses some of these problems. We instantiate this design in the Soar cognitive architecture and analyze its capabilities on a learning task.

For intelligent agents modeled to emulate people, reference resolution is memory management: when processing an object or event – whether it appears in language or in the simulated physical or cognitive experience of the agent – the agent must determine how that object or event correlates with known objects and events, and must store the new memory with semantically explicit links to related prior knowledge. This paper discusses eventualities for memory-based reference resolution and the modeling strategies used in the OntoAgent environment to permit agents to fully and automatically make reference decisions.

The Odd One Out test of intelligence consists of 3x3 matrix reasoning problems organized in 20 levels of difficulty. Addressing problems on this test appears to require integration of multiple cognitive abilities usually associated with creativity, including visual encoding, similarity assessment, pattern detection, and analogical transfer. We describe a novel fractal strategy for addressing visual analogy problems on the Odd One Out test. In our strategy, the relationship between images is encoded fractally, capturing important aspects of similarity as well as inherent self-similarity. The strategy starts with fractal representations encoded at a high level of resolution, but, if that is not sufficient to resolve ambiguity, it automatically adjusts itself to the right level of resolution for addressing a given problem. Similarly, the strategy starts with searching for fractally-derived similarity between simpler relationships, but, if that is not sufficient to resolve ambiguity, it automatically shifts to search for such similarity between higher-order relationships.  We present preliminary results and initial analysis from applying the fractal technique on nearly 3,000 problems from the Odd One Out test.