Title
Modeling Agent Interactions using Common Ground Knowledge from a Joint Activity Theory Perspective
Abstract
Believability is necessary for agents to establish intimate, real-time collaborations with humans in an interactive game environment. In this paper, the authors model sophisticated interaction patterns to improve believability by adapting Herbert Clark's joint activity theory. The authors use virtual basketball as an environment, where many communicative scenarios occur and common ground knowledge of collaborative actions such as passing is necessary. The authors have completed preliminary implementation of the agents through the use of a unique computational model based on common ground, and evaluations of basketball simulations have been made with respect to believability scenarios.
Year
DOI
Venue
2013
10.4018/ijssci.2013100101
IJSSCI
Keywords
Field
DocType
collaboration,virtual environment,agents,joint activity theory,believability,human-computer interaction
Virtual machine,Computer science,Human–computer interaction,Artificial intelligence,Common ground,Machine learning,Basketball
Journal
Volume
Issue
Citations 
5
4
0
PageRank 
References 
Authors
0.34
16
2
Name
Order
Citations
PageRank
Divesh Lala15213.91
Toyoaki Nishida21097196.19