Patrick J. Hayes

OWL-QL-a language for deductive query answering on the Semantic Web

This paper discusses the issues involved in designing a query language for the Semantic Web and presents the OWL query language (OWL-QL) as a candidate standard language and protocol for query-answering dialogues among Semantic Web computational agents using knowledge represented in the W3Cs ontology web language (OWL). OWL-QL is a formal language and precisely specifies the semantic relationships among a query, a query answer, and the knowledge base(s) used to produce the answer. Unlike standard database and Web query languages, OWL-QL supports query-answering dialogues in which the answering agent may use automated reasoning methods to derive answers to queries, as well as dialogues in which the knowledge to be used in answering a query may be in multiple knowledge bases on the Semantic Web, and/or where those knowledge bases are not specified by the querying agent. In this setting, the set of answers to a query may be of unpredictable size and may require an unpredictable amount of time to compute.

Modeling Cyclic Change

Database support of time-varying phenomena typically assumes that entities change in a linear fashion. Many phenomena, however, change cyclically over time. Examples include monsoons, tides, and travel to the workplace. In such cases, entities may appear and disappear on a regular basis or their attributes or location may change with periodic regularity. This paper introduces an approach for modeling cycles based on cyclic intervals. Intervals are an important abstraction of time, and the consideration of cyclic intervals reveals characteristics about these intervals that are unique from the linear case. This work examines binary cyclic relations, distinguishing sixteen cyclic interval relations. We identify their conceptual neighborhood graph, showing which relations are most similar and demonstrating that this set of sixteen relations is complete. The results of this investigation provide the basis for extended data models and query languages that address cyclically varying phenomena.

Artificial intelligence — Where are we?

ed f rom experts or any o ther principles, to do as well as experts who, according to Fe igenbaum, are never satisfied with general principles but prefer to think of their field of expertise as a huge set of special cases. Compe ten t systems are certainly valuable, but by being misnamed and over-sold they may contr ibute to an impending compute r backlash . 5In Putting Computers in Their Place (Free Press/Macmillan, New York, 1985), my brother, Stuart, and I argue that beginners in a domain arc given principles which they know and use, but that experts can do much better than beginners because they know how to deal with thousands of special cases. Thus everyday expertise or know-how does not consist of the principles and rules of thumb thought to comprise common-sense knowledge. We fall back on procedures and principles only in situations which are so unlike any situation we have dealt with that our everyday skills fail

Why Go¨del's theorem cannot refute computationalism

Abstract Godels theorem is consistent with the computationalist hypothesis. Roger Penrose, however, claims to prove that Godels theorem implies that human thought cannot be mechanized. We review his arguments and show how they are flawed. Penroses arguments depend crucially on ambiguities between precise and imprecise senses of key terms. We show that these ambiguities cause the Godel/Turing diagonalization argument to lead from apparently intuitive claims about human abilities to paradoxical or highly idiosyncratic conclusions, and conclude that any similar argument will also fail in the same ways.

The triples rule

A fundamental stance taken in human-centered computing is that information processing devices must be thought of in systems terms. At first blush, this seems self-evident. However, the notion has a long history, and not just in systems engineering. In this new age of symbiosis, machines are made for specific humans for use in specific contexts. The unit of analysis for cognitive engineering and computer science is a triple: person, machine and context The triples rule asserts that system development must take this triple as the unit of analysis, which has strong implications, including a mandate that the engineering of complex systems should include detailed cognitive work analysis. It also has implications for the meaning of intelligence, including artificial intelligence.

Primitive Intervals versus Point-Based Intervals: Rivals or Allies?

The notion of time is a very interesting and exciting subject both in science and everyday life. One of the fundamental questions is: what is time composed of? While the traditional time structure is based on a set of points, a notion that has been prevalently adopted in classical physics and mathematics, it has also been noticed that intervals have been widely adopted for expression of commonsense temporal knowledge, especially in the domain of artificial intelligence. However, there has been a long-standing debate on whether intervals should be treated as primitive or not, leading to two different approaches to the treatment of intervals. In the first, intervals are modelled as derived objects constructed from points, e.g. sets of points, or pairs of points. In the second, intervals are taken as primitive themselves. This article provides a critical examination of these two approaches. By means of proposing a definition of intervals in terms of points and types, we shall demonstrate that, while the two different approaches have been viewed as rivals in the literature, they are actually reducible to logically equivalent expressions under some requisite interpretations, and therefore they can also be viewed as allies.

Human-centered computing: thinking in and out of the box

The motion picture AI lets the AI profession squirm in the glory of misrepresentation. Its not fun, especially when ones field suffers from waves of innovation/hype/backlash. The problem is that the film AI reinforces the dream of the android just when many who work toward truly intelligent technologies are cutting loose from the dreams more surrealistic aspects. People are asking new questions: Is the Turing test really the right kind of standard? If not, what is better? Must we define intelligence in reference to humans? Must intelligent technology be boxes chock-full of this thing we call intelligence, or should it operate as a cognitive prosthesis to amplify or extend human perceptual, cognitive and collaborative capabilities? Must intelligence always be in some individual thing - either a headbone or a box - or is intelligence a system property that is definable only in terms of the triple of humans-machines-contexts?.

Robots in the AI Classroom: What Smart Machines Can Teach Smart Students

The design and implementation of autonomous robots provides experience with engineering problems as well as with hard problems in artificial intelligence. Our aims in this article are fourfold: First, we describe how the use of robots complements traditional classroom lectures for the teaching of important concepts in a graduate‐level AI class. Second, we discuss the use of robots as a multimedia tool for capturing student interest and facilitating the self‐guided exploration of AI concepts. Third, we relate readings and robots in the curriculum and show how the robots aid understanding of philosophical issues in AI. Finally, we describe the robot kits and the organization of the class, present several completed projects by student teams, and report some of the feedback given by students who completed the course. We also analyze the successes and failures of the first two courses offered and describe the changes made for the recently completed second course.

Alternative conceptions of the time-line

Summary form only given. It is often assumed that if time is continuous, then the standard mathematical account of the continuum, the real line, is the appropriate model of the local structure of the timeline. However, another popular account of the timeline which provides the basis for Allens famous 13 interval relations does not seem naturally compatible with the real line. This paper sharpens the differences between these two accounts and offers a unified account which retains some advantages of both, at the cost of making the continuum a slightly more complicated structure. The end result, which allows intervals to have negative and zero lengths, might be seen as a re-invention of arithmetic.