Reputation - Ming2 Sheng1
 

Publications

Readings & Reviews

Log

Courses


Mike Smith

Trust Heart/Mind/Intelligence/Soul (Intention)

Department of Computer Science
University of Maryland Baltimore County
1000 Hill Top Circle
Baltimore, MD 21250
msmith27@cs.umbc.edu
http://www.cs.umbc.edu/~msmith27
Introduction

I'm a Ph.D. Student, affiliated with the Maple Lab at UMBC. I'm currently working on a research thesis proposal for trust and reputation modeling and learning.

Motivation, Research Area, and Interests

The issue of trust and reputation rise almost immediately in any social transaction, whether between humans, machines, or some combination. Trust and its social expression as reputation are tightly coupled with notions of honesty versus deception, commitment versus convenience, and good versus evil. Whether agents must choose between being "ethical" or utilitarian or generally able to integrate the two desiderata is an open question. It seems that the quest for effective trust and reputation mechanism may be viewed in one sense as a search for a "machine morality" to go with emerging machine intelligence.

My current research focus dwells on the relationship between reputation, trust, and competence in Multi-Agent Systems (MAS). The aim is to synthesize and extend existing frameworks for the management of commitment and competence risks posed by agents to one another while interacting in a heterogeneous MAS environment. The key element will be the study of individual decision making as the foundation of the framework. The current expectation is that a general solution to this problem can be built upon a base of decision theory, combining elements of bounded rationality, probability, and utilitarianism. The framework will be evaluated by experiments.

More generally, my interests include knowledge representation and machine learning and means to combine the two. The abstraction of agents acting in an environment and more particularly agents interacting in multi-agent societies provide an arena with great opportunitites for evaluating the utility of various representation and learning techniques.

Trust, Cooperation, and Reputation Links
Other Links