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.

It is costly and time consuming to develop Multiple Choice Questions (MCQ) by hand. Using web-based resources to automate components of MCQ development would greatly benefit the education community through reducing reduplication of effort. Similar to many areas of Natural Language Processing (NLP), human-judged data is needed to train automated systems, but the majority of such data is proprietary. We present a graph-based representation for gathering training data from existing, web-based resources that increases access to such data and better directs the development of good questions.

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.

Reasoning with commonsense science knowledge is an important challenge for Artificial Intelligence. This paper presents a system that revises its knowledge in a commonsense science domain by constructing and evaluating explanations. Domain knowledge is represented using qualitative model fragments, which are used to explain phenomena via model formulation. Metareasoning is used to (1) score competing explanations numerically along several dimensions and (2) evaluate preferred explanations for global consistency. Inconsistencies cause the system to favor alternative explanations and thereby change its beliefs. We simulate the belief changes of several students during clinical interviews about how the seasons change. We show that qualitative models accurately represent student knowledge and that our system produces and revises a sequence of explanations similar those of the students.

This year, the Robotics Exhibition included two such robotics challenge problems: manipulation and learning by demonstration. In the Small-Scale Manipulation Challenge four teams demonstrated systems playing robotic chess. This exhibit was organized by David Touretzky and Monica D. Anderson. In the Learning by Demonstration Challenge, three teams demonstrated systems learning a block-sorting task. This exhibit was organized by Sonia Chernova. Additionally, this year marked another successful turnout for the Robotics Education Track, organized by Zachary Dodds, which highlights student-and educator-led robotics projects. In this article we give a summary of these three components of the exhibition.

The titles of the eight symposia were Artificial Intelligence and Health Communication, Artificial Intelligence and Sustainable Design, Artificial Intelligence for Business Agility, Computational Physiology, Help Me Help You: Bridging the Gaps in Human-Agent Collaboration, Logical Formalizations of Commonsense Reasoning, Multirobot Systems and Physical Data Structures, and Modeling Complex Adaptive Systems As If They Were Voting Processes. The goal of the Artificial Intelligence and Health Communication symposium was to advance the conceptual design of automated systems that provide health services to patients and consumers through interdisciplinary insight from artificial intelligence, health communication and related areas of communication studies, discourse studies, public health, and psychology. There is a large and growing interest in the development of automated systems to provide health services to patients and consumers. In the last two decades, applications informed by research in health communication have been developed, for example, for promoting healthy behavior and for managing chronic diseases. While the value that these types of applications can offer to the community in terms of cost, access, and convenience is clear, there are still major challenges facing design of effective health communication systems. Overall, the participants found the format of the symposium engaging and constructive, and they The symposium was organized around five main expressed the desire to continue this initiative in concepts: (1) Patient empowerment and education further events.

Most methods for decision-theoretic online learning are based on the Hedge algorithm, which takes a parameter called the learning rate. In most previous analyses the learning rate was carefully tuned to obtain optimal worst-case performance, leading to suboptimal performance on easy instances, for example when there exists an action that is significantly better than all others. We propose a new way of setting the learning rate, which adapts to the difficulty of the learning problem: in the worst case our procedure still guarantees optimal performance, but on easy instances it achieves much smaller regret. In particular, our adaptive method achieves constant regret in a probabilistic setting, when there exists an action that on average obtains strictly smaller loss than all other actions. We also provide a simulation study comparing our approach to existing methods.

We consider an adversarial online learning setting where a decision maker can choose an action in every stage of the game. In addition to observing the reward of the chosen action, the decision maker gets side observations on the reward he would have obtained had he chosen some of the other actions. The observation structure is encoded as a graph, where node i is linked to node j if sampling i provides information on the reward of j. This setting naturally interpolates between the well-known "experts" setting, where the decision maker can view all rewards, and the multi-armed bandits setting, where the decision maker can only view the reward of the chosen action. We develop practical algorithms with provable regret guarantees, which depend on non-trivial graph-theoretic properties of the information feedback structure. We also provide partially-matching lower bounds.

We consider a Gaussian process formulation of the multiple kernel learning problem. The goal is to select the convex combination of kernel matrices that best explains the data and by doing so improve the generalisation on unseen data. Sparsity in the kernel weights is obtained by adopting a hierarchical Bayesian approach: Gaussian process priors are imposed over the latent functions and generalised inverse Gaussians on their associated weights. This construction is equivalent to imposing a product of heavy-tailed process priors over function space. A variational inference algorithm is derived for regression and binary classification.

Using an interactive theorem prover to reason about programs involves a sequence of interactions where the user challenges the theorem prover with conjectures. Invariably, many of the conjectures posed are in fact false, and users often spend considerable effort examining the theorem prover's output before realizing this. We present a synergistic integration of testing with theorem proving, implemented in the ACL2 Sedan (ACL2s), for automatically generating concrete counterexamples. Our method uses the full power of the theorem prover and associated libraries to simplify conjectures; this simplification can transform conjectures for which finding counterexamples is hard into conjectures where finding counterexamples is trivial. In fact, our approach even leads to better theorem proving, e.g. if testing shows that a generalization step leads to a false conjecture, we force the theorem prover to backtrack, allowing it to pursue more fruitful options that may yield a proof. The focus of the paper is on the engineering of a synergistic integration of testing with interactive theorem proving; this includes extending ACL2 with new functionality that we expect to be of general interest. We also discuss our experience in using ACL2s to teach freshman students how to reason about their programs.