Abstracts

Jackson, P. C. Toward Human-Level Artificial Intelligence – Representation and Computation of Meaning in Natural Language. Doctoral Thesis, Tilburg University, 2014.

This doctoral thesis presents a novel research approach toward human-level artificial intelligence.

The approach involves developing an AI system using a language of thought based on the unconstrained syntax of a natural language; designing this system as a collection of concepts that can create and modify concepts, expressed in the language of thought, to behave intelligently in an environment; and using methods from cognitive linguistics such as mental spaces and conceptual blends for multiple levels of mental representation and computation. Proposing a design inspection alternative to the Turing Test, these pages discuss ‘higher-level mentalities’ of human intelligence, which include natural language understanding, higher-level forms of learning and reasoning, imagination, and consciousness.

This thesis endeavors to address all the major theoretical issues and objections that might be raised against its approach, or against the possibility of achieving human-level AI in principle. No insurmountable objections are identified, and arguments refuting several objections are presented.

This thesis describes the design of a prototype demonstration system, and discusses processing within the system that illustrates the potential of the research approach to achieve human-level AI.

This thesis cannot claim to actually achieve human-level AI, it can only present an approach that may eventually reach this goal.

Thesis: http://www.philjackson.prohosting.com/PCJacksonPhDThesis20140422.pdf

Slides: http://www.philjackson.prohosting.com/TowardHumanLevelAI20140422.pdf

Jackson, P. C. Proving unsatisfiability for problems with constant cubic sparsity, Artificial Intelligence, Vol. 57, 1992, pp. 125-137.

This paper presents a perspective on a previously studied heuristic tree search algorithm for the NP-complete "3SAT" satisfiability problem, which synthesizes results by Brown, Franco, Purdom, and others. It is shown that when an a priori calculation indicates a random, nontrivial 3SAT problem is very likely unsatisfiable, then the heuristic algorithm is expected to prove unsatisfiability in a search tree with size essentially independent of the problem size, and bounded by a constant that depends only on the ratio of the number of clauses to the cube of the number of variables.

Vora, L. S., Veres, R. E., Jackson, P. C., and Klahr, P. TIES: An engineering design methodology and system, Proceedings of 1990 AAAI Conference on Innovative Applications of Artificial Intelligence, pp. 59-65.

Technical Information Engineering System (TIES) provides a framework in which cross functional product / process design teams can collect and store relevant engineering information. TIES incorporates several AI techniques (contexts / frames with multiple inheritance, user-defined knowledge representation, user-defined rules / "symbolic spreadsheet", "conceptual zoom", searches for direct and indirect linkages).

Jackson, P. C. Meta-belief intervals, Proceedings of IASTED 1987 Conference on Expert Systems, Geneva, Switzerland, pp. 144-146.

This paper discusses the semantics of Dempster-Shafer (DS) belief intervals for missing and conflicting evidence. An argument is given against using the DS normalization convention in the aggregation of conflicting evidence. This argument is extended to study certain kinds of missing evidence. The use of "meta-belief vectors" is introduced as a pragmatic method of representing missing and anomalous evidence, from the standpoint of constructing expert systems that emulate human judgment.

Jackson, P. C. Elements of variable precision logic based on ART, Proceedings of Society of Automotive Engineers 1987 Conference, SP-695, pp. 21-26, SAE Technical Paper # 870113.

Variable precision logic (VPL) is a formalism for reasoning with uncertain information, in situations where resources constrain the amount of reasoning which can take place. The conclusions developed using VPL may vary in certainty according to the computational resources that are available. This paper describes implementation methods for embedding certain elements of VPL on top of ART, an AI programming language developed by Inference Corporation. The elements of VPL discussed in this paper include certainties of facts and rules, censored rules, and time resource constraints on logical derivations.

Jackson, P. C. Introduction to Artificial Intelligence, Second Edition, Dover Publications, New York, 1985. First Edition, Mason-Charter Publishers, New York, 1974.

Are we intelligent enough to understand intelligence? One approach to answering this question is "artificial intelligence", the field of computer science that studies how machines can be made to act intelligently. This book is intended to be a general introduction to the field of artificial intelligence (AI). The subjects for discussion are machines that can solve problems, play games, recognize patterns, prove mathematical theorems, understand English, and even demonstrate learning by changing their own behavior to perform such tasks more successfully. In general this book is addressed to all persons who are interested in studying the nature of thought, and hopefully much of it can be read without previous formal exposure to computers.

DeRemer, F., Levy, P., Hanson, S., Jackson, P., Julig, R., Pittman, T. Summary of the characteristics of several “modern” programming languages. ACM SIGPLAN Notices, Vol. 14, Issue 5, May 1979, pp. 28-45. Some characteristics of nine recently developed or historically significant languages are discussed. Abstraction capabilities of the languages are noted in particular.

Jackson, P. C. Concept - A Context for High-Level Descriptions of Systems Which Develop Concepts. Master's Thesis, University Microfilms, 1979, # 1322722.

This thesis presents research in artificial intelligence directed toward the construction of systems which develop their own concepts. The ultimate goal of this research is to construct a system of initial concepts which is so flexible and general that it can develop concepts to understand any real-world environment. This thesis will present a context for high-level discussions of fully general artificial intelligence systems, and will present partial results on a variety of subjects within this context.