The literature on conceptual blending and metaphor-making has illustrations galore of how these mechanisms may support the creation and grounding of new concepts (or whole domains) in terms of a complex, integrated network of older ones. In spite of this, as of yet there is no general computational account of blending and metaphor-making that has proven powerful enough as to cover all the examples from the literature. This paper proposes a logic-based framework for blending and metaphor making and explores its applicability in settings as diverse as mathematical domain formation, classical rationality puzzles, and noun-noun combinations.

In creative industries such as design and research it is common to reason about ‘problem-finding’ before tasks or goals can be established. Problem-finding may also continue throughout the problem-solving process, so achieving goals may be an ongoing process of discovery as well as iterative improvement and refinement. This paper considers the design of cognitive systems with complementary processes for both problem-finding and problem-solving. We review a range of approaches that may complement goal-directed reasoning when an artificial system does not or cannot know precisely what it is looking for. We argue that there is a spectrum of approaches that can be used for reasoning in the absence of goals, which make progressively weaker assumptions about the definition and presence goals, and that goal-oriented behavior can be an intermediate result of problem-finding, rather than as a starting point for problem-solving. We demonstrate one such approach based on implicit motives and incentives.

In this paper we discuss an elaboration account of insight that provides answers to two of the main questions regarding insight problem solving: why insight problems are so difficult for humans and why insight is so rapid in nature. We claim that the difficulty in insight problems is due to misguided heuristic search and that this difficulty is overcome using a reformulation mechanism. Furthermore, we claim that search is carried out quickly when the heuristics are good--explaining the rapid nature of insight. We clarify our account by providing examples and initial empirical results. In conclusion, we review related work and discuss possible future work.

An important form of learning involves acquiring skills that let an agent achieve its goals. While there has been considerable work on learning in planning, most approaches have been sensitive to the representation of domain context, which hurts their generality. A learning mechanism that constructs skills effectively across different representations would suggest more robust behavior. In this paper, we present a novel approach to learning hierarchical task networks that acquires conceptual predicates as learning proceeds, making it less dependent on carefully crafted background knowledge. The representation acquisition procedure expands the system's knowledge about the world, and leads to more rapid learning. We show the effectiveness of the approach by comparing it with one that doesnot change domain representation.

The ability to delay intentions and remember them in the proper context is an important ability for general artificial agents. In this paper, we define the functional requirements of an agent capable of fulfilling delayed intentions with its long-term memories. We show that the long-term memories of different cognitive architec- tures share similar functional properties and that these mechanisms can be used to support delayed intentions. Finally, we do a preliminary evaluation of the different memories for fulfilling delayed intentions and show that there are trade-offs between memory types that warrant further research.

One challenge for cognitive architectures is to effectively use different forms of knowledge and learning. We present a case study of Soar agents that play a multiplayer dice game, in which probabilistic reasoning and heuristic symbolic knowledge appear to play a central role. We develop and evaluate a collection of agents that use different combinations of probabilistic decision making, heuristic symbolic reasoning, opponent modeling, and learning. We demonstrate agents that use Soar’s rule learning mechanism (chunking) to convert deliberate reasoning with probabilities into implicit reasoning, and then use reinforcement learning to further tune performance.

In order to collaborate with people in the real world, cognitive systems must be able to represent and reason about spatial regions in human environments. Consider the command "go to the front of the classroom". The spatial region mentioned (the front of the classroom) is not perceivable using geometry alone. Instead it is defined by its functional use, implied by nearby objects and their configuration. In this paper, we define such areas as context-dependent spatial regions and propose a method for a cognitive system to learn them incrementally by combining qualitative spatial representations, semantic labels, and analogy. Using data from a mobile robot, we generate a relational representation of semantically labeled objects and their configuration. Next, we show how the boundary of a context-dependent spatial region can be defined using anchor points. Finally, we demonstrate how an existing computational model of analogy can be used to transfer this region to a new situation.

Computationally hard problems, like the Traveling Salesperson Problem, can be solved remarkably well by humans. Results obtained by computers are usually closer to the optimum, but require high computational effort and often differ from the human solutions. This paper introduces Greedy Expert Search (GES) that strives to show the same flexibility and efficiency of human solutions, while producing results of similarly high quality. The Traveling Salesperson Problem serves as an example problem to illustrate and evaluate the approach.

This paper argues the position that an essential approach to the advancement of the state of the art in cognitive systems is to focus on systems that deeply integrate knowledge representations, cognitive capabilities, and knowledge content. Integration is the path to aggregating constraints in ways that improve the science of cognitive systems. However, evaluating the role of knowledge among these constraints has largely been ignored, in part because it is difficult to build and evaluate systems that incorporate large amounts of knowledge. We provide suggestions for evaluating such systems and argue that such evaluations will become easier as we come closer to applying usefully new, integrated learning mechanisms that are capable of acquiring large and effective knowledge bases.

One approach to bridging the historic divide between "symbolic" and "subsymbolic" AI is to incorporate a subsymbolic system and a symbolic system into a synergetic integrative cognitive architecture. Here we consider various issues related to incorporating (subsymbolic) compositional spatiotemporal deep learning networks (CSDLNs, a term introduced to denote the category including HTM, DeSTIN and other similar systems) into an integrative cognitive architecture including symbolic aspects. The core conclusion is that for such integration to be meaningful, it must involve dynamic and adaptive linkage and conversion between CSDLN attractors spanning sensory, motor and goal hierarchies, and analogous representations in the remainder of the integrative architecture. We suggest the mechanism of "semantic CSDLNs", which maintain the general structure of CSDLNs but contain more abstract patterns, similar to those represented in more explicitly symbolic AI systems. This notion is made concrete by describing a planned integration of the DeSTIN CSDLN into the OpenCog integrative cognitive system (which includes a probabilistic-logical symbolic component).