AAAI Publications, Workshops at the Thirty-Second AAAI Conference on Artificial Intelligence

Font Size: 
Online Goal Recognition as Reasoning over Landmarks
Mor Vered, Ramon Fraga Pereira, Mauricio Cecilio Magnaguagno, Felipe Meneguzzi, Gal A. Kaminka

Last modified: 2018-06-20

Abstract


Online goal recognition is the problem of recognizing the goal of an agent based on an incomplete sequence of observations with as few observations as possible. Recognizing goals with minimal domain knowledge as an agent executes its plan requires efficient algorithms to sift through a large space of hypotheses. We develop an online approach to recognize goals in both continuous and discrete domains using a combination of goal mirroring and a generalized notion of landmarks adapted from the planning literature. Extensive experiments demonstrate the approach is more efficient and substantially more accurate than the state-of-the-art.

Keywords


plan recognition;mirroring;landmarks;planning

Full Text: PDF