Music similarity: improvements of edit-based algorithms by considering music theory

ABSTRACT

Estimating the symbolic music similarity is one of the major open problems in the music information retrieval research domain. Existing systems consider sequences of notes characterized by pitches and durations. Similarity estimation is mainly based on variations of pitches and durations and does not consider any other musical elements. However, musical elements such as tonality or rhythm are particularly important in the perception of music. In this paper we propose to investigate some algorithmic improvements that allow edit-based systems to take into account important musical elements: tonality, passing notes, strong and weak beats. These elements are illustrated with a few monophonic musical examples which lead to important errors in usual systems. First experiments with these examples show that the improvements induced are significant. Furthermore, experimental results obtained with the MIREX 2005 database are very good. All the results are thus very promising since they confirm that considering musical information improves the accuracy of music retrieval systems.

REFERENCES

• 1: E. Chew. Towards a Mathematical Model of Tonality. PhD thesis, MIT Cambridge, MA, 2000.
• 2: Shyamala Doraisamy , Stefan Rüger, Robust Polyphonic Music Retrieval with N-grams, Journal of Intelligent Information Systems, v.21 n.1, p.53-70, July 2003 [doi>10.1023/A:1023553801115]
• 3: J. S. Downie, K. West, A. F. Ehmann, and E. Vincent. The 2005 Music Information retrieval Evaluation Exchange (MIREX 2005): Preliminary Overview. In ISMIR, pages 320--323, 2005.
• 4: P. Ferraro and P. Hanna. Optimizations of Local Edition for Evaluating Similarity Between Monophonic Musical Sequences. In Proceedings of the International Conference on Information Retrieval -- Large-Scale Semantic Access to Content (RIAO), Pittsburgh, PA, USA, 2007.
• 5: M. Grachten, J. L. Arcos, and R. L. de Mántaras. Melodic Similarity: Looking for a Good Abstraction Level. In Proceedings of the International Conference on Music Information Retrieval (ISMIR), 2004.
• 6: Dan Gusfield, Algorithms on strings, trees, and sequences: computer science and computational biology, Cambridge University Press, New York, NY, 1997
• 7: M. Mongeau and D. Sankoff. Comparison of Musical Sequences. Computers and the Humanities, 24(3):161--175, 1990.
• 8: S. Needleman and C. Wunsch. A General Method Applicable to the Search for Similarities in the Amino Acid Sequences of Two Proteins. Journal of Molecular Biology, 48:443--453, 1970.
• 9: David Rizo , José M. Iñesta , Pedro J. Ponce de León, Tree model of symbolic music for tonality guessing, Proceedings of the 24th IASTED international conference on Artificial intelligence and applications, p.299-304, February 13-16, 2006, Innsbruck, Austria
• 10: D. Sankoff and J. B. Kruskal, editors. Time Wraps, Strings Edits, and Macromolecules: the Theory and Practice of Sequence Comparison. Addison-Wesley Publishing Company Inc, University of Montreal, Quebec, Canada, 1983.
• 11: M. Simon. Cours complet théorique et pratique des principes de la musique. ed. Jacquot, 1947. in french.
• 12: David Temperley, A Bayesian Approach to Key-Finding, Proceedings of the Second International Conference on Music and Artificial Intelligence, p.195-206, September 12-14, 2002
• 13: R. Typke, M. den Hoed, J. de Nooijer, F. Wiering, and R. C. Veltkamp. A Ground Truth For Half A Million Musical Incipits. Journal of Digital Information Management, 3(1):34--39, 2005.
• 14: Rainer Typke , Remco C. Veltkamp , Frans Wiering, Searching notated polyphonic music using transportation distances, Proceedings of the 12th annual ACM international conference on Multimedia, October 10-16, 2004, New York, NY, USA [doi>10.1145/1027527.1027551]
• 15: R. Typke, R. C. Veltkamp, and F. Wiering. A Measure for Evaluating Retrieval Techniques Based on Partially Ordered Ground Truth Lists. In Proceedings of the International Conference on Multimedia and Expo (ICME), pages 128--135, Toronto, Canada, 2006.
• 16: E. Ukkonen, K. Lemström, and V. Mäkinen. Geometric Algorithms for Transposition Invariant Content-Based Music Retrieval. In Proceedings of the International Conference on Music Information Retrieval (ISMIR), pages 193--199, Baltimore, USA, 2003.
• 17: Robert A. Wagner , Michael J. Fischer, The String-to-String Correction Problem, Journal of the ACM (JACM), v.21 n.1, p.168-173, Jan. 1974 [doi>10.1145/321796.321811]

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.3 Information Search and Retrieval

General Terms:
Algorithms

Keywords:
edit-based systems, information retrieval, music similarity, music theory