Abstract
In this paper, we propose a domain learning process build on a machine learning-based process that, starting from plan traces with (partially known) intermediate states, returns a planning domain with numeric predicates, and expressive logical/arithmetic relations between domain predicates written in the planning domain definition language (PDDL). The novelty of our approach is that it can discover relations with little information about the ontology of the target domain to be learned. This is achieved by applying a selection of preprocessing, regression, and classification techniques to infer information from the input plan traces. These techniques are used to prepare the planning data, discover relational/numeric expressions, or extract the preconditions and effects of the domain’s actions. Our solution was evaluated using several metrics from the literature, taking as experimental data plan traces obtained from several domains from the International Planning Competition. The experiments demonstrate that our proposal—even with high levels of incompleteness—correctly learns a wide variety of domains discovering relational/arithmetic expressions, showing F-Score values above 0.85 and obtaining valid domains in most of the experiments.
Similar content being viewed by others
Availability of data and material
Data used in experimentation are available at http://www.icaps-conference.org/index.php/Main/Competitions
Code Availability
Code presented in this work is available in https://github.com/Leontes/PlanMiner
References
Aineto D, Jiménez S, Onaindia E (2018) Learning strips action models with classical planning. In: International conference on automated planning and scheduling, ICAPS-18
Aineto D, Celorrio SJ, Onaindia E (2019) Learning action models with minimal observability. Artif Intell 275:104–137. https://doi.org/10.1016/j.artint.2019.05.003. http://www.sciencedirect.com/science/article/pii/S0004370218304259
Cabot J, Gogolla M (2012) Object constraint language (ocl): a definitive guide. In: International school on formal methods for the design of computer, communication and software systems. Springer, pp 58–90
Chatterjee S, Hadi AS (2015) Regression analysis by example, Wiley, New York
Cook SA (1971) The complexity of theorem-proving procedures. In: Proceedings of the third annual ACM symposium on theory of computing, pp 151–158
Cresswell SN, McCluskey TL, West MM (2013) Acquiring planning domain models using locm. Knowl Eng Rev 28(2):195–213. https://doi.org/10.1017/S0269888912000422
Fernandez A, Herrera F, Cordon O, Jose del Jesus M, Marcelloni F (2019) Evolutionary fuzzy systems for explainable artificial intelligence: why, when, what for, and where to? IEEE Comput Intell Mag 14 (1):69–81
Fox M, Long D (2002) Pddl+: modeling continuous time dependent effects. In: Proceedings of the 3rd international NASA workshop on planning and scheduling for space, vol 4, p 34
García D, Gámez JC, González A, Pérez R (2015) An interpretability improvement for fuzzy rule bases obtained by the iterative rule learning approach. Int J Approx Reason 67(C):37–58. https://doi.org/10.1016/j.ijar.2015.09.001
Ghallab M, Howe A, Knoblock C, Mcdermott D, Ram A, Veloso M, Weld D, Wilkins D (1998) PDDL—the planning domain definition language. http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.37.212
Gregory P, Lindsay A (2016) Domain model acquisition in domains with action costs. In: ICAPS, pp 149–157
Hansen P, Jaumard B (1990) Algorithms for the maximum satisfiability problem. Computing 44(4):279–303
Hart PE, Nilsson NJ, Raphael B (1968) A formal basis for the heuristic determination of minimum cost paths. IEEE Trans Syst Sci Cybern 4(2):100–107
Hoffmann J (2003) The metric-FF planning system: translating “ignoring delete lists” to numeric state variables. Artif Intell Res 20:291–341
Howey R, Long D (2003) Val’s progress: the automatic validation tool for pddl2.1 used in the international planning competition. In: Proceedings of the ICAPS 2003 workshop on the competition: impact, organization, evaluation, benchmarks, Trento, pp 28–37
Jilani R, Crampton A, Kitchin DE, Vallati M (2014) Automated knowledge engineering tools in planning: state-of-the-art and future challenges. Knowledge Engineering for Planning and Scheduling
Jiménez S, Rosa TDL, Fernández S, Fernández F, Borrajo D (2012) A review of machine learning for automated planning. Knowl Eng Rev 27(4):433–467
Kotsiantis SB, Zaharakis I, Pintelas P (2007) Supervised machine learning: a review of classification techniques. Emerg Artif Intell Appl Comput Eng 160:3–24
Koza JR (1994) Genetic programming as a means for programming computers by natural selection. Stat Comput 4(2):87–112. https://doi.org/10.1007/BF00175355
Lanchas J, Jiménez S, Fernández F, Borrajo D (2007) Learning action durations from executions. In: Proceedings of the ICAPS. Citeseer
Long D, Fox M (2003) The 3rd international planning competition: results and analysis. Artif Intell Res 20:1–59
Matloff NS (2009) From algorithms to Z-Scores: probabilistic and statistical modeling in computer science. University Press of Florida, Gainesville
McCluskey TL, Richardson NE, Simpson RM (2002) An interactive method for inducing operator descriptions. In: AIPS, pp 121–130
McCluskey TL, Cresswell S, Richardson NE, West MM (2009) Automated acquisition of action knowledge. In: International conference on agents and artificial intelligence (ICAART)
Mitchell TM (1997) Machine learning, 1 edn. McGraw-Hill Inc., New York
Nau D, Ghallab M, Traverso P (2004) Automated planning: theory & practice. Morgan Kaufmann Publishers Inc., San Francisco
Quinlan JR (1990) Learning logical definitions from relations. Mach Learn 5(3):239–266
Russell SJ, Norvig P (2016) Artificial intelligence: a modern approach Malaysia. Pearson Education Limited
Suárez-Hernández A, Segovia-Aguas J, Torras C, Alenyà G (2020) Strips action discovery. arXiv preprint arXiv:2001.11457
Var I (1998) Multivariate data analysis. Vectors 8(2):125–136
Worm T, Chiu K (2013) Prioritized grammar enumeration: symbolic regression by dynamic programming. In: Proceedings of the 15th annual conference on genetic and evolutionary computation. ACM, pp 1021–1028
Zhuo HH, Yang Q, Hu DH, Li L (2010) Learning complex action models with quantifiers and logical implications. Artif Intell 174(18):1540–1569. https://doi.org/10.1016/j.artint.2010.09.007. http://www.sciencedirect.com/science/article/pii/S0004370210001566
Yang Q, Wu K, Jiang Y (2007) Learning action models from plan examples using weighted MAX-SAT. Artif Intell J 107– 143
Zhuo HH, Nguyen T, Kambhampati S (2013) Refining incomplete planning domain models through plan traces. In: Twenty-third international joint conference on artificial intelligence
Acknowledgements
This research is being developed and partially funded by the Spanish MINECO R&D Project TIN2015-71618-R and RTI2018-098460-B-I00
Funding
This research is being developed and partially funded by the Spanish MINECO R&D Project TIN2015-71618-R and RTI2018-098460-B-I00
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of Interests
Authors state that there is no conflict of interest
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Segura-Muros, J.Á., Pérez, R. & Fernández-Olivares, J. Discovering relational and numerical expressions from plan traces for learning action models. Appl Intell 51, 7973–7989 (2021). https://doi.org/10.1007/s10489-021-02232-6
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-021-02232-6