This paper provides a definition of automated negotiation within electronic commerce. It outlines two barriers to automated negotiation, the ontology issue and the strategy problem. State of the art overviews are given of automated negotiation, specifically Negotiation Support Systems, intelligent agents, the auction mechanism, and online marketspaces. Both academic research and currently functional systems are covered, and several World Wide Web addresses are given for readers who wish to investigate further on their own. 1 1 While every attempt is made to provide current URL locations, the Web changes more quickly than print media can ever capture. Hence, some of the URLs may not be current or correct by the time this article appears. We will try to keep our Negotiation Project web site, http://haas.berkeley.edu/~citm/nego-proj.html, current with respect to these addresses. 2 1.

We show how a system of artificial adaptive agents, using a genetic algorithm based learning technique, can learn strategies that enable it to effectively participate in stylized business negotiations. The negotiation policies learned are evaluated on several dimensions, including joint outcomes, nearness to the efficient frontier, and the similarity to outcomes of human negotiations. The results are promising for integrating such agents into practicable electronic commerce systems. We discuss what a system might look like and ways in which particular classes of business negotiations could be supported or even entirely automated. 1 Introduction Even in simple negotiations, people often reach sub-optimal negotiations thereby "leaving money on the table" [1] [2]. While many factors contribute to missing out on gains from trade---overconfidence, falsely assuming fixed pies, and the framing of the situation (e.g. [3])---the end result is that parties fail to find agreements which would ma...

Artificial agents, coevolving under a machine learning regime, offer a promising basis for modeling adaptive behavior in multilateral negotiations and developing practical applications in electronic commerce. Results from simulations of bargaining in two coalition games demonstrate that simple artificial agents, adapting to one another under a genetic programming protocol (Koza, 1992), formulate effective strategies for negotiating agreements that both approximate those prescribed by the theory of cooperative games and rival the performance of humans negotiating in a laboratory situation. 1. Introduction Despite substantial progress in getting artificial agents (AAs) to negotiate effectively in stylized settings, they are not yet ready for deployment in real-world applications. In such situations, there are both common and conflicting interests among the relevant players (Schelling, 1960: 5): ...the ability of one participant to gain his [or her] ends is dependent to an important deg...

Web search engines have become an integral part of the daily life of a knowledge worker, who depends on these search engines to retrieve relevant information from the Web or from the company's vast document databases. Current search engines are very fast in terms of their response time to a user query. But their usefulness to the user in terms of retrieval performance leaves a lot to be desired. Typically, the user has to sift through a lot of nonrelevant documents to get only a few relevant ones for the user's information needs. Ranking functions play a very important role in the search engine retrieval performance. In this paper, we describe a methodology using genetic programming to discover new ranking functions for the Web-based information-seeking task. We exploit the content as well as structural information in the Web documents in the discovery process. The discovery process is carried out for both the ad hoc task and the routing task in retrieval. For either of the retrieval tasks, the retrieval performance of these newly discovered ranking functions has been found to be superior to the performance obtained by well-known ranking strategies in the information retrieval literature.

this paper. The concept of intelligent agents and innovative multi-agent architectures for pattern recognition tasks is introduced for combining and elaborating the classification hypotheses of several classifiers. The architecture of a distributed digit-recognition system dispatching recognition tasks to a set of recognizers and combining their results is presented. This concept is being developed in the iart project, where intelligent agent architectures are built for pattern recognition tasks. 1 Introduction

This paper describes a new framework using intelligent agents for pattern recognition. A justification for using alternatives to current classifier systems is given. The use of the framework, called iart, is tested on a digit recognition system. 1 Introduction The use of multiple classifiers and the combination of their classification results has gained considerable interest in the last few years. This approach has a powerful potential because it may exploit the advantages of different feature representations and classification methods. In general two architectures can be distinguished, 1) individual classifiers with some combination scheme and 2) multi-stage or hierarchical classifiers. Several combination schemes are possible, like majority vote, max/min/median rule, BKS [8], the Dempster-Shafer rule or Borda count. These use either class labels, rank order or score combinations. Classifiers use one or more components in a "pattern recognition pipeline". Pattern recognition modules ...

This paper models adaptive learning behavior in a simple coordination game that Van Huyck, Cook and Battalio (1994) have investigated in a controlled laboratory setting with human subjects. We consider how populations of artificially intelligent players behave when playing the same game. We use the genetic programming paradigm, as developed by Koza (1992, 1994), to model how a population of players might learn over time. In genetic programming one seeks to breed and evolve highly fit computer programs that are capable of solving a given problem. In our application, each computer program in the population can be viewed as an individual agent's forecast rule. The various forecast rules (programs) then repeatedly take part in the coordination game evolving and adapting over time according to principles of natural selection and population genetics. We argue that the genetic programming paradigm that we use has certain advantages over other models of adaptive learning behavior in the context of the coordination game that we consider. We find that the pattern of behavior

In recent years, there has been increasing interest in automated e-business negotia-tions. The automation of negotiation requires a decision model to capture the negotia-tion knowledge of policymakers and negotiation experts so that the decision-making process can be carried out automatically. Current research on automated e-business negotiations has focused on defining low-level tactics (or negotiation rules) so that au-tomated negotiation systems can carry out automated negotiation processes. These low-level tactics are usually defined from a technical perspective, not from a business perspective. There is a gap between high-level business negotiation goals and low-level tactics. In this article, we distinguish the concepts of negotiation context, negotia-tion goals, negotiation strategy, and negotiation tactics and introduce a formal deci-sion model to show the relations among these concepts. We show how high-level nego-tiation goals can be formally mapped to low-level tactics that can be used to affect the behavior of a negotiation system during the negotiation process. In business, a busi-ness organization faces different negotiation situations (or contexts) and determines different sets of goals for different negotiation contexts. In our decision model, a busi-

Abstract. DNA is used to implement a simplified version of poker. Strategies are evolved that mix bluffing with telling the truth. The essential features are (1) to wait your turn, (2) to default to the most conservative course, (3) to probabilistically override the default in some cases, and (4) to learn from payoffs. Two players each use an independent population of strategies that adapt and learn from their experiences in competition. 1

A. I agree that the abstract of my thesis may be published by the University without further reference to me. Sig nature: In accordance with the