If you are looking for an answer to the question What is Artificial Intelligence? and you only have a minute, then here's the definition the Association for the Advancement of Artificial Intelligence offers on its home page: "the scientific understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines."
However, if you are fortunate enough to have more than a minute, then please get ready to embark upon an exciting journey exploring AI (but beware, it could last a lifetime) …
Daniel Barbard Ping Chen George Mason University * Information and Software Engineering Department Fairfax, VA 22303 dbarbara, pchen}@gmu, edu Abstract dynamically fine tune the server to obtain better performance. As organizations accumulate data over time, the problem of tracking how patterns evolve becomes important. In this paper, we present an algorithm to track the evolution of cluster models in a stream of data. Our algorithm is based on the application of bounds derived using Chernoff's inequality and makes use of a clustering algorithm that was previously developed by us, namely Practal Clustering, which uses self-similarity as the property to group points together. Experiments show that our tracking algorithm is efficient and effective in finding changes on the patterns. Introduction Organizations today accumulate data at a astonishing rate.
Elements of the KDD Process Description in terms of view(scope) I) data selection 2) data preprocessing 3) data transformation 4) data mining 5) interpretation/evaluation 6) knowledge consolidation classification and filtering of entities according to a certain view, which decides whether the entity is used or not. Preprocessing of data can he expressed as a function over data, so naturally may he defined by a view(scope). Transformation can also he expressed as a function over data, so naturally may he defined by a view(scope). Data mining can he expressed as a generation of new view(scope). Hypothesis generation algorithms can he considered as view operators.
There have been efforts at combining expert systems and neural networks (connectionism) into hybrid systems, order to exploit their benefits (Medsker 1994). In some them, called embedded systems, a neural network is used in the inference engine of an expert system. For example, in NEULA (Tirri 1995) a neural network selects the next rule to fire. Also, LAM (Medsker 1994) uses two neural networks as partial problem solvers. However, the inference process in those systems, although gains efficiency, lacks the naturalness and the explanation capability of the symbolic component.
Our approach consists of four steps: concept definitions are automatically generated from the UMLS source, integrity checking of taxonomic and partonomic hierarchies is performed by the terminological classifier, cycles and inconsistencies are eliminated, and incremental refinement of the evolving knowledge base is performed by a domain expert. We report on experiments with a terminological knowledge base composed of 164,000 concepts and 76,000 relations.
A knowledge base is maintained by modifying its conceptual model and by using those modifications to specify changes to its implementation. The maintenance problem is to determine which parts of that model should be checked for correctness in response a change in the application. The maintenance problem is not computable for first-order knowledge bases. Two things in the conceptual model are joined by a maintenance link if a modification to one of them means that the other must be checked for correctness, and so possibly modified, if consistency of the model is to be preserved. In a unified conceptual model for first-order knowledge bases the data and knowledge are modelled formally in a uniform way. A characterisation is given of four different kinds of maintenance links in a unified conceptual model.
CIRCSlM-Tutor Version 3 includes a reactive planner, APE - the Atlas Planning Engine. By using this planner, we are able to effectively respond to dynamic changes in the environment of tutoring students to solve problems in cardiovascular physiology dealing with the regulation of blood pressure. Use of the reactive planner allows us to adapt to student initiatives or unpredictable answers to questions by replacing the planner's responses with newly computed better reactions and allows us to specify new goals to the ITS and have it alter its reaction so as to achieve these new goals. The result is a more cohesive dialog with the student.
The goal of this project is to increase students' interest in Artificial Intelligence, as well as to promote learning of the topics that comprise the subject. We describe the development of a web-based multimedia delivery method to accomplish this goal, and outline plans for continued development of tile tool. The highlight of the course material is aa integrated simulation environment that allows students to develop and test AI algorithms in a dynamic, uncertain, visual environment.
This paper presents a holonic coordination server which is made up of a recursive hierarchy of three different agent types: On the top stands an agent providing matchmaking services and representing the coordination server to the outside. It passes incoming requests to a subordinated agent type that is equipped with coordination mechanisms such as auctions, negotiations and coalition formation mechanisms. For each incoming request, this agent spawns an instance of a third agent type which executes the protocol of the coordination mechanism chosen in the request. The holonic structure of the coordination server helps to reduce complexity while allowing a high grade of adaptability and flexibility. In today's markets, business entities are forced to interact with other market participants flexibly in order to stay competitive.
Maria Fasli University of Essex Department of Computer Science Wivenhoe Park, Colchester, CO4 3SQ, UK mfasli @essex.ac.uk Abstract One of the most interesting issues that arises in agents based on the BDI (belief-desire-intention) formalism capturing notions of realism, that is constraints that describe possible interrelationships between the three attitudes. Three such sets of constraints have been considered in the literature: strong realism, realism and weak realism. In this paper we propose notions of realism for heterogeneous BDI agents and in particular we explore what we call bold agents. We interpret bold BDI agents as agents that are willing to take risks, and thus adopt intentions even though they may not believe in all respective accessible worlds that these are achievable. Keywords: Intelligent Agents, BDI Models, Agent Theories Introduction Agents are obviously highly complicated systems and formal theories that describe and explain their behaviour are of interest to the agent community. We consider agent theories as specifications, and we are mainly concerned with building a useful and expressive theory capable of capturing agents with a sufficient degree of rationality.
We consider plan based agents where the agents attempt to solve a problem by deriving a sequence of actions called a plan, and then execute the plan. In a static world (that is, no changes take place when the agent does not perform any actions), a plan based agent, hopefully, is certain to achieve its goal as long as it chooses the right sequence of actions to execute. In dynamic worlds however, even plans that are proved to be correct at the beginning may fail due to unexpected changes in the world. In this paper, we propose that agents which are autonomous, need to be guided by attitudes for effective problem solving in dynamic worlds.