A Probabilistic Model of Plan Recognition

By Eugene Charniak, Robert P. Goldman. 1991

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Explanation in Artificial Intelligence: Insights from the Social Sciences (2017, background): “Using abductive reasoning to generate explanations has a long history in artificial intelligence [97], aiming to solve problems such as fault diagnosis [143], plan/intention recognition [24], and generalisation in learning [133].”
Programming by Demonstration Using Version Space Algebra (2003, unknown): “Incontrast,CharniakandGoldman(1991)presentaprobabilisticmethodfor decidingbetweenplanhypotheses,butdonotaddresstheproblemofdeterminingwhich plansareconsistentwiththedata.”
Focusing Construction and Selection of Abductive Hypotheses (1993, background): “Its approach to judging plausibility of assumptions is in the spirit of systems using probabilistic criteria (e.g., [Charniak and Goldman, 1991]), but the situations it attempts to explain are situations for which probability information is unlikely to be available.”
Techniques for Plan Recognition (2001, background): “Charniak and Goldman (1991, 1993) constructed the ¢rst Bayesian plan inference system.”
Abduction, experience, and goals: a model of everyday abductive explanation (1995, background): “…abduction almost universally bases this decision solely on criteria that in some sense measure the \plausibility" of explanations (Charniak, 1986; Charniak & Goldman, 1991; Charniak & Shomony, 1994; Hobbs et al., 1993; Josephson & Josephson, 1994; Kautz & Allen, 1986; Kass, 1990; Konolige, 1990;…”
Implicit Communication of Actionable Information in Human-AI teams (2019, background): “Abduction has been studied in AI [13, 15, 30, 36, 41, 50, 58, 63, 65] and employed to solve a variety of real-world problems [36, 57, 66, 81, 82] and to perform plan recognition [5, 21, 85].”
Action understanding as inverse planning. (2009, background): “However, these approaches assume that the distribution over actions, conditioned on goals, is either available a priori (Bui et al., 2002; Charniak & Goldman, 1991), or must be estimated from a large data-set of observed actions (Liao et al., 2004).”