Learning Relational Grammars from Sequences of Actions

  • Blanca Vargas-Govea
  • Eduardo F. Morales
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5856)


Many tasks can be described by sequences of actions that normally exhibit some form of structure and that can be represented by a grammar. This paper introduces FOSeq, an algorithm that learns grammars from sequences of actions. The sequences are given as low-level traces of readings from sensors that are transformed into a relational representation. Given a transformed sequence, FOSeq identifies frequent sub-sequences of n-items, or n-grams, to generate new grammar rules until no more frequent n-grams can be found. From m sequences of the same task, FOSeq generates m grammars and performs a generalization process over the best grammar to cover most of the sequences. The grammars induced by FOSeq can be used to perform a particular task and to classify new sequences. FOSeq was tested on robot navigation tasks and on gesture recognition with competitive performance against other approaches based on Hidden Markov Models.


Hide Markov Model Mobile Robot Gesture Recognition Generalization Process Inductive Logic Programming 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Adriaans, P.W., Trautwein, M., Vervoort, M.: Towards high speed grammar induction on large text corpora. In: Jeffery, K., Hlaváč, V., Wiedermann, J. (eds.) SOFSEM 2000. LNCS, vol. 1963, pp. 173–186. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  2. 2.
    van Zaanen, M.V.: Abl: alignment-based learning. In: Proc. 17th Conf. on Computational linguistics, Association for Computational Linguistics, pp. 961–967 (2000)Google Scholar
  3. 3.
    Bernard, M., de la Higuera, C.: Gift: Grammatical inference for terms. In: International Conference on Inductive Logic Programming (1999)Google Scholar
  4. 4.
    Nevill-Manning, C., Witten, I.: Identifying hierarchical structure in sequences: A linear-time algorithm. Journal of Artificial Intelligence Research 7, 67–82 (1997)zbMATHGoogle Scholar
  5. 5.
    Amit, R., Mataric, M.J.: Learning movement sequences from demonstration. In: ICDL 2002: Proc. 2nd International Conf. on Development and Learning, Cambridge, MA, pp. 12–15 (2002)Google Scholar
  6. 6.
    Avilés-Arriaga, H., Sucar, L., Mendoza, C.: Visual recognition of similar gestures. In: 18 International Conference on Pattern Recognition ICPR 2006, vol. 1, pp. 1100–1103 (2006)Google Scholar
  7. 7.
    Agrawal, R., Srikant, R.: Fast algorithms for mining association rules. In: Bocca, J.B., Jarke, M., Zaniolo, C. (eds.) Proc. 20th Int. Conf. Very Large Data Bases, VLDB, pp. 487–499. Morgan Kaufmann, San Francisco (1994)Google Scholar
  8. 8.
    Plotkin, G.: A note on inductive generalization. Machine Intelligence 5, 153–163Google Scholar
  9. 9.
    Benson, S., Nilsson, N.J.: Reacting, planning, and learning in an autonomous agent. Machine Intelligence 14, 29–62 (1995)Google Scholar
  10. 10.
    Vargas, B., Morales, E.F.: Learning navigation teleo-reactive programs using behavioural cloning. In: IEEE International Conference on Mechatronics (2009)Google Scholar
  11. 11.
    Vaughan, R.T., Gerkey, B.P., Howard, A.: On device abstractions for portable, reusable robot code. In: IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 2421–2427 (2003)Google Scholar
  12. 12.
    Avilés, H.H., Sucar, L.E., Morales, E.F., Vargas, B.A., Sánchez, J., Corona, E.: Markovito: A flexible and general service robot, January 2009. Studies in Computational Intelligence, vol. 177, pp. 401–423. Springer, Heidelberg (2009)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Blanca Vargas-Govea
    • 1
  • Eduardo F. Morales
    • 1
  1. 1.Computer Science DepartmentNational Institute of Astrophysics Optics and ElectronicsTonantzintlaMéxico

Personalised recommendations