Cobot: A Social Reinforcement Learning Agent (2002)

by Charles L. Isbell, Christian R. Shelton, Michael Kearns, Satinder Singh, and Peter Stone

Abstract: We report on our reinforcement learning with Cobot, a software agent residing in a the well-known online community LambdaMOO. Our initial work on Cobot (Isbell et al. 2000) provided him with the ability to collect social statistics and report them to users. Here we describe our application of RL allowing Cobot to take proactive actions in this complex social environment, and adapt this behavior from multiple sources of human reward. After 5 months of training, and 3171 reward and punishment events from 254 different LambdaMOO users, Cobot has learned nontrivial preferences for a number of users, modifying his behavior based on his current state. Here we describe LambdaMOO and the state and action spaces of Cobot, and report the statistical results of the learning experiment.

Download Information

Charles L. Isbell, Christian R. Shelton, Michael Kearns, Satinder Singh, and Peter Stone (2002). "Cobot: A Social Reinforcement Learning Agent." Advances in Neural Information Processing Systems 2001, vol 2 (pp. 1393-1400).

Bibtex citation

@inproceedings{IsbSheKeaSinSto02,
   author = "Charles L. Isbell and Christian R. Shelton and Michael Kearns and Satinder Singh and Peter Stone",
title = "Cobot: A Social Reinforcement Learning Agent",
   booktitle = "Advances in Neural Information Processing Systems 2001",
   booktitleabbr = "{NIPS}",
   pages = "1393--1400",
   volume = 2,
   year = 2002,
}