An LP-Based Approach for Goal Recognition as Planning
Authors: Luísa R. A. Santos, Felipe Meneguzzi, Ramon Fraga Pereira, André Grahl Pereira11939-11946
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate our approach empirically over a large data set, analyzing its components on how each can impact the quality of the solutions. In general, our approach is superior to previous methods in terms of agreement ratio, accuracy, and spread. |
| Researcher Affiliation | Academia | 1Federal University of Rio Grande do Sul, Brazil 2Pontifical Catholic University of Rio Grande do Sul, Brazil 3Sapienza University of Rome, Italy |
| Pseudocode | No | The paper describes methods and processes in narrative text and mathematical formulations but does not include any explicit pseudocode or algorithm blocks. |
| Open Source Code | Yes | Source-code and benchmark are available at: https://bit.ly/lp-goal-recognition |
| Open Datasets | Yes | We create a new benchmark by adapting the one introduced by Pereira, Oren, and Meneguzzi (2017) to use the agreement ratio evaluation metric from Ram ırez and Geffner (2009). |
| Dataset Splits | No | The paper describes how the benchmark domains and various data sets (optimal, sub-optimal, noisy) were generated with different observability levels, including the total number of goal recognition tasks (8,288). However, it does not specify explicit train/validation/test dataset splits for these tasks. |
| Hardware Specification | Yes | We ran all experiments with Ubuntu running over an Intel Core i7 930 CPU (2.80 GHz) with a memory limit of 1 GB, all methods solved each goal recognition task under a time limit of five seconds. |
| Software Dependencies | Yes | Our implementation uses Fast Downward version 19.06 (Helmert 2006), a Python prepossessing layer, and the CPLEX 12.10 LP solver. |
| Experiment Setup | Yes | We ran all experiments with Ubuntu running over an Intel Core i7 930 CPU (2.80 GHz) with a memory limit of 1 GB, all methods solved each goal recognition task under a time limit of five seconds. |