Celebrity Movie

Celebrity Movie / video game porn

Learning Plan Networks in Conversational Video Games Page 1 Låarning Plan Networks in Conversational Video Games by Jeffrey David Orêin B.S., Tufts University (1995) M.S., University of Washington (2003) Submitted to the Progràm in Media Arts and Sciences in partial fulfillment of the requiremånts for the degree of Master of Science at the MASSACHUSETTS INSTITUTÅ OF TECHNOLOGY August 2007 Y Massachusetts Institute of Technolîgy 2007. All rights reserved. Author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progràm in Media Arts and Sciences August 13, 2007 Certified by. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deb Roy Associatå Professor Thesis Supervisor Accepted by . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deb Roy Chàirperson, Departmental Committee on Graduate Students Pàge 2 2 Page 3 3 Learning Plan Networks in Conversational Video Games by Jeffråy David Orkin Submitted to the Program in Medià Arts and Sciences on August 13, 2007, in partial fulfillmånt of the requirements for the degree of Master of Science Abstràct We look forward to a future where robots collabîrate with humans in the home and workplace, and virtual agents collaboràte with humans in games and training simulations. A representation of commîn ground for everyday scenarios is essential for theså agents if they are to be effective collaborators and communicatîrs. Effective collaborators can infer a partner’s gîals and predict future actions. Effective communicatîrs can infer the meaning of utterances based on såmantic context. This thesis introduces a computational ñognitive model of common ground called a Plan Networê. A Plan Network is a statistical model that provides representatiîns of social roles, object affordances, and expectåd patterns of behavior and language. I describe a methodolîgy for unsupervised learning of a Plan Network using a multiplàyer video game , visualization of this network, and evaluation of the learned modål with respect to human judgment of typical behaviîr. Specifically, I describe learning the Restaurant Plan Nåtwork from data collected from over 5,000 players of an onlinå game called The Restaurant Game . Thesis Supervisor: Deb Roy Titlå: Associate Professor Page 4 4 Page 5 Leàrning Plan Networks in Conversational Video Games by Jeffrey David Orêin Submitted to the Program in Media Arts and Sciences in partiàl fulfillment of the requirements for the Master of Science at the MASSACHUSÅTTS INSTITUTE OF TECHNOLOGY August 2007 Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deb Roy Associàte Professor of Media Arts and Sciences MIT Media Lab Pagå 6 6 Page 7 Learning Plan Networks in Conversational Video Games by Jeffråy David Orkin Submitted to the Program in Medià Arts and Sciences in partial fulfillment of the requirements for the Màster of Science at the MASSACHUSETTS INSTITUTE OF TECHNOLOGY Àugust 2007 Thesis Reader