TY - JOUR
T1 - Temporally extended goal recognition in fully observable non-deterministic domain models
AU - Pereira, Ramon Fraga
AU - Fuggitti, Francesco
AU - Meneguzzi, Felipe
AU - De Giacomo, Giuseppe
N1 - This work has been partially supported by the ERC-ADGWhiteMech (No. 834228), the EU ICT-48 2020 project TAILOR (No. 952215), the PRIN project RIPER (No. 20203FFYLK),and the PNRR MUR project FAIR (No. PE0000013).
PY - 2024/1
Y1 - 2024/1
N2 - Goal Recognition is the task of discerning the intended goal that an agent aims to achieve, given a set of goal hypotheses, a domain model, and a sequence of observations (i.e., a sample of the plan executed in the environment). Existing approaches assume that goal hypotheses comprise a single conjunctive formula over a single final state and that the environment dynamics are deterministic, preventing the recognition of temporally extended goals in more complex settings. In this paper, we expand goal recognition to temporally extended goals in Fully Observable Non-Deterministic (fond) planning domain models, focusing on goals on finite traces expressed in Linear Temporal Logic (ltl$$_f$$) and Pure-Past Linear Temporal Logic (ppltl). We develop the first approach capable of recognizing goals in such settings and evaluate it using different ltl$$_f$$and ppltl goals over six fond planning domain models. Empirical results show that our approach is accurate in recognizing temporally extended goals in different recognition settings.
AB - Goal Recognition is the task of discerning the intended goal that an agent aims to achieve, given a set of goal hypotheses, a domain model, and a sequence of observations (i.e., a sample of the plan executed in the environment). Existing approaches assume that goal hypotheses comprise a single conjunctive formula over a single final state and that the environment dynamics are deterministic, preventing the recognition of temporally extended goals in more complex settings. In this paper, we expand goal recognition to temporally extended goals in Fully Observable Non-Deterministic (fond) planning domain models, focusing on goals on finite traces expressed in Linear Temporal Logic (ltl$$_f$$) and Pure-Past Linear Temporal Logic (ppltl). We develop the first approach capable of recognizing goals in such settings and evaluate it using different ltl$$_f$$and ppltl goals over six fond planning domain models. Empirical results show that our approach is accurate in recognizing temporally extended goals in different recognition settings.
U2 - 10.1007/s10489-023-05087-1
DO - 10.1007/s10489-023-05087-1
M3 - Article
SN - 1573-7497
VL - 54
SP - 470
EP - 489
JO - Applied Intelligence
JF - Applied Intelligence
ER -