Ashok Goel

IJCAI Conference 2022 Conference Paper

Abstraction in Data-Sparse Task Transfer (Extended Abstract)

• Tesca Fitzgerald
• Ashok Goel
• Andrea Thomaz

When a robot adapts a learned task for a novel environment, any changes to objects in the novel environment have an unknown effect on its task execution. For example, replacing an object in a pick-and-place task affects where the robot should target its actions, but does not necessarily affect the underlying action model. In contrast, replacing a tool that the robot will use to complete a task will effectively alter its end-effector pose with respect to the robot's base coordinate system, and thus the robot's motion must be replanned accordingly. These examples highlight the relationship among (i) differences between the source and target environments, (ii) the level of abstraction at which a robot's task model should be represented to enable transfer to the target environment, and (iii) the information needed to ground the abstracted task representation in the target environment. In this abstract, summarizing our full article [Fitzgerald et al. , 2021], we present our taxonomy of transfer problems based on this relationship. We also describe a knowledge representation called the Tiered Task Abstraction (TTA) and demonstrate its applicability to a variety of transfer problems in the taxonomy. Our experimental results indicate a trade-off between the generality and data requirements of a task representation, and reinforce the need for multiple transfer methods that operate at different levels of abstraction.

AIJ Journal 2021 Journal Article

Abstraction in data-sparse task transfer

• Tesca Fitzgerald
• Ashok Goel
• Andrea Thomaz

When a robot adapts a learned task for a novel environment, any changes to objects in the novel environment have an unknown effect on its task execution. For example, replacing an object in a pick-and-place task affects where the robot should target its actions, but does not necessarily affect the underlying action model. In contrast, replacing a tool that the robot will use to complete a task will effectively alter its end-effector pose with respect to the robot's base coordinate system, and thus the robot's motion must be replanned accordingly. These examples highlight the relationship among (i) differences between the source and target environments, (ii) the level of abstraction at which a robot's task model should be represented to enable transfer to the target environment, and (iii) the information needed to ground the abstracted task representation in the target environment. In this article, we present a taxonomy of transfer problems based on this relationship. We also describe a knowledge representation called the Tiered Task Abstraction (TTA) and demonstrate its applicability to a variety of transfer problems in the taxonomy. Our experimental results indicate a trade-off between the generality and data requirements of a task representation, and reinforce the need for multiple transfer methods that operate at different levels of abstraction.

AAMAS Conference 2019 Conference Paper

Human-guided Trajectory Adaptation for Tool Transfer

• Tesca Fitzgerald
• Elaine Short
• Ashok Goel
• Andrea Thomaz

We introduce “transfer by correction": a method for transferring a robot’s tool-based task models to use unfamiliar tools. By having the robot receive corrections from a human teacher when repeating a known task with a new tool, it can learn the relationship between the two tools, allowing it to transfer additional tasks learned with the original tool to the new tool. The goal is to enable the robot to generalize its task knowledge to accommodate tool replacements and thus be more robust to changes in its environment. We demonstrate how the tool transform models learned from one episode of task corrections can be used to perform that task with ≥ 85% of maximum performance in 83% of tool/task combinations. Furthermore, these transformations generalize to unseen tool/task combinations in 27. 8% of our transfer evaluations, and up to 41% of transfer problems when the source and replacement tool share tooltip similarities. Overall, these results indicate that successful task adaptation for a new tool is dependent on the the tool’s usage within that task, and that the transform model learned from interactive corrections can be generalized to other tasks providing a similar context for the new tool.

AAAI Conference 2018 Conference Paper

The Structural Affinity Method for Solving the Raven’s Progressive Matrices Test for Intelligence

• Snejana Shegheva
• Ashok Goel

Graphical models offer techniques for capturing the structure of many problems in real-world domains and provide means for representation, interpretation, and inference. The modeling framework provides tools for discovering rules for solving problems by exploring structural relationships. We present the Structural Affinity method that uses graphical models for first learning and subsequently recognizing the pattern for solving problems on the Raven’s Progressive Matrices Test of general human intelligence. Recently there has been considerable work on computational models of addressing the Raven’s test using various representations ranging from fractals to symbolic structures. In contrast, our method uses Markov Random Fields parameterized by affinity factors to discover the structure in the geometric analogy problems and induce the rules of Carpenter et al. ’s cognitive model of problem-solving on the Raven’s Progressive Matrices Test. We provide a computational account that first learns the structure of a Raven’s problem and then predicts the solution by computing the probability of the correct answer by recognizing patterns corresponding to Carpenter et al. ’s rules. We demonstrate that the performance of our model on the Standard Raven Progressive Matrices is comparable with existing state of the art models.

AAAI Conference 2017 Conference Paper

What’s Hot in Case-Based Reasoning

• Ashok Goel
• Belen Diaz-Agudo

Case-based reasoning addresses new problems by remembering and adapting solutions previously used to solve similar problems. Pulled by the increasing number of applications and pushed by a growing interest in memory intensive techniques, research on case-based reasoning appears to be gaining momentum. In this article, we briefly summarize recent developments in research on case-based reasoning based partly on the recent Twenty Fourth International Conference on Case-Based Reasoning.

AAAI Conference 2016 Conference Paper

A Survey of Current Practice and Teaching of AI

• Michael Wollowski
• Robert Selkowitz
• Laura Brown
• Ashok Goel
• George Luger
• Jim Marshall
• Andrew Neel
• Todd Neller

The field of AI has changed significantly in the past couple of years and will likely continue to do so. Driven by a desire to expose our students to relevant and modern materials, we conducted two surveys, one of AI instructors and one of AI practitioners. The surveys were aimed at gathering information about the current state of the art of introducing AI as well as gathering input from practitioners in the field on techniques used in practice. In this paper, we present and briefly discuss the responses to those two surveys.

AAAI Conference 2016 Conference Paper

Design of an Online Course on Knowledge-Based AI

• Ashok Goel
• David Joyner

AAAI Conference 2014 Conference Paper

Confident Reasoning on Raven’s Progressive Matrices Tests

• Keith McGreggor
• Ashok Goel

We report a novel approach to addressing the Raven’s Progressive Matrices (RPM) tests, one based upon purely visual representations. Our technique introduces the calculation of confidence in an answer and the automatic adjustment of level of resolution if that confidence is insufficient. We first describe the nature of the visual analogies found on the RPM. We then exhibit our algorithm and work through a detailed example. Finally, we present the performance of our algorithm on the four major variants of the RPM tests, illustrating the impact of confidence. This is the first such account of any computational model against the entirety of the Raven’s.

AIJ Journal 2014 Journal Article

Fractals and Ravens

• Keith McGreggor
• Maithilee Kunda
• Ashok Goel

We report a novel approach to visual analogical reasoning, one afforded expressly by fractal representations. We first describe the nature of visual analogies and fractal representations. Next, we exhibit the Fractal Ravens algorithm through a detailed example, describe its performance on all major variants of the Raven's Progressive Matrices tests, and discuss the implications and next steps. In addition, we illustrate the importance of considering the confidence of the answers, and show how ambiguity may be used as a guide for the automatic adjustment of the problem representation. To our knowledge, this is the first published account of a computational model's attempt at the entire Raven's test suite.

AAAI Conference 2011 Conference Paper

Two Visual Strategies for Solving the Raven’s Progressive Matrices Intelligence Test

• Maithilee Kunda
• Keith McGreggor
• Ashok Goel

AIIM Journal 1991 Journal Article

The role of essential explanation in abduction

• Olivier Fischer
• Ashok Goel
• John R. Svirbely
• Jack W. Smith

The abduction task is to infer the best explanation for a given set of data. One common subtask of abduction is to synthesize the best composite explanatory hypothesis from elementary hypotheses retrieved from memory. The synthesis of best composite explanations, however, is computationally costly. One general approach to controlling the computational cost of synthesizing explanations is to decompose the synthesis search space into smaller spaces that can be searched more efficiently and effectively. The essential hypotheses, that is, the hypotheses that are the only available explanations for specific subsets of the data set, provide one such decomposition. In this method, first the essential hypotheses are included in the composite explanation, and, then, non-essential hypotheses are included to account for the remaining unexplained data elements. In addition to providing a more efficient method for synthesizing composite explanations, this decomposition leads to the formation of more parsimonious explanations. In this paper, we report on a set of experiments in the domain of medical data interpretation that demonstrates that the essential/non-essential decomposition of the abduction search space results in more efficient synthesis of more parsimonious composite explanations.

IJCAI Conference 1989 Conference Paper

Functional Representation of Designs and Redesign Problem Solving

• Ashok Goel
• B. Chandrasckaran

AAAI Conference 1987 Conference Paper

Complexity in Classificatory Reasoning

• Ashok Goel

Classificatory reasoning involves the tasks of concept evaluation and classification, which may be performed with use of the strategies of concept matching and concept activation, respectively. Different implementations of the strategies of concept matching and concept activation are possible, where an implementation is characterized by the organization of knowledge and the control of information processing it uses. In this paper we define the tasks of concept evaluation and classification, and describe the strategies of concept matching and concept activation. We then derive the computational complexity of the tasks using different implementations of the task-specific strategies. We show that the complexity of performing a task is determined by the organization. of knowledge used in performing it. Further, we suggest that the implementation that is computationally the most efficient for performing a task may be cognitively the most plausible as well.