Domain independent goal recognition

D. Pattison, D. Long, T. Agnotes (Editor)

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

4 Citations (Scopus)

Abstract

Goal recognition is generally considered to follow plan recognition. The plan recognition problem is typically defined to be that of identifying which plan in a given library of plans is being executed, given a sequence of observed actions. Once a plan has been identified, the goal of the plan can be assumed to follow. In this work, we address the problem of goal recognition directly, without assuming a plan library. Instead, we start with a domain description, just as is used for plan construction, and a sequence of action observations. The task, then, is to identify which possible goal state is the ultimate destination of the trajectory being observed. We present a formalisation of the problem and motivate its interest, before describing some simplifying assumptions we have made to arrive at a first implementation of a goal recognition system, AUTOGRAPH. We discuss the techniques employed in AUTOGRAPH to arrive at a tractable approximation of the goal recognition problem and show results for the system we have implemented.
LanguageEnglish
Title of host publicationSTAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium
Pages238 - 250
Number of pages13
Volume222
Publication statusPublished - 1 Aug 2010

Fingerprint

Trajectories

Keywords

  • goal recognition
  • goal recognition system
  • computer systems design

Cite this

Pattison, D., Long, D., & Agnotes, T. (Ed.) (2010). Domain independent goal recognition. In STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium (Vol. 222, pp. 238 - 250)
Pattison, D. ; Long, D. ; Agnotes, T. (Editor). / Domain independent goal recognition. STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium. Vol. 222 2010. pp. 238 - 250
@inproceedings{2f75de542dc94f1786295a82b5cd1f40,
title = "Domain independent goal recognition",
abstract = "Goal recognition is generally considered to follow plan recognition. The plan recognition problem is typically defined to be that of identifying which plan in a given library of plans is being executed, given a sequence of observed actions. Once a plan has been identified, the goal of the plan can be assumed to follow. In this work, we address the problem of goal recognition directly, without assuming a plan library. Instead, we start with a domain description, just as is used for plan construction, and a sequence of action observations. The task, then, is to identify which possible goal state is the ultimate destination of the trajectory being observed. We present a formalisation of the problem and motivate its interest, before describing some simplifying assumptions we have made to arrive at a first implementation of a goal recognition system, AUTOGRAPH. We discuss the techniques employed in AUTOGRAPH to arrive at a tractable approximation of the goal recognition problem and show results for the system we have implemented.",
keywords = "goal recognition, goal recognition system , computer systems design",
author = "D. Pattison and D. Long and T. Agnotes",
year = "2010",
month = "8",
day = "1",
language = "English",
volume = "222",
pages = "238 -- 250",
booktitle = "STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium",

}

Pattison, D, Long, D & Agnotes, T (ed.) 2010, Domain independent goal recognition. in STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium. vol. 222, pp. 238 - 250.

Domain independent goal recognition. / Pattison, D.; Long, D.; Agnotes, T. (Editor).

STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium. Vol. 222 2010. p. 238 - 250.

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

TY - GEN

T1 - Domain independent goal recognition

AU - Pattison, D.

AU - Long, D.

A2 - Agnotes, T.

PY - 2010/8/1

Y1 - 2010/8/1

N2 - Goal recognition is generally considered to follow plan recognition. The plan recognition problem is typically defined to be that of identifying which plan in a given library of plans is being executed, given a sequence of observed actions. Once a plan has been identified, the goal of the plan can be assumed to follow. In this work, we address the problem of goal recognition directly, without assuming a plan library. Instead, we start with a domain description, just as is used for plan construction, and a sequence of action observations. The task, then, is to identify which possible goal state is the ultimate destination of the trajectory being observed. We present a formalisation of the problem and motivate its interest, before describing some simplifying assumptions we have made to arrive at a first implementation of a goal recognition system, AUTOGRAPH. We discuss the techniques employed in AUTOGRAPH to arrive at a tractable approximation of the goal recognition problem and show results for the system we have implemented.

AB - Goal recognition is generally considered to follow plan recognition. The plan recognition problem is typically defined to be that of identifying which plan in a given library of plans is being executed, given a sequence of observed actions. Once a plan has been identified, the goal of the plan can be assumed to follow. In this work, we address the problem of goal recognition directly, without assuming a plan library. Instead, we start with a domain description, just as is used for plan construction, and a sequence of action observations. The task, then, is to identify which possible goal state is the ultimate destination of the trajectory being observed. We present a formalisation of the problem and motivate its interest, before describing some simplifying assumptions we have made to arrive at a first implementation of a goal recognition system, AUTOGRAPH. We discuss the techniques employed in AUTOGRAPH to arrive at a tractable approximation of the goal recognition problem and show results for the system we have implemented.

KW - goal recognition

KW - goal recognition system

KW - computer systems design

UR - http://personal.cis.strath.ac.uk/~pattison/pub/domain_independent_goal_recognition.pdf

M3 - Conference contribution book

VL - 222

SP - 238

EP - 250

BT - STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium

ER -

Pattison D, Long D, Agnotes T, (ed.). Domain independent goal recognition. In STAIRS 2010: Proceedings of the Fifth Starting AI Researchers' Symposium. Vol. 222. 2010. p. 238 - 250