Evaluation of a simple and effective music information retrieval method

ABSTRACT

We developed, and then evaluated, a music information retrieval (MIR) system based upon the intervals found within the melodies of a collection of 9354 folksongs. The songs were converted to an interval-only representation of monophonic melodies and then fragmented t into length-n subsections called n-grams. The length of these n-grams and the degree to which we precisely represent the intervals are variables analyzed in this paper. We constructed a collection of “musical word” databases using the text-based, SMART information retrieval system. A group of simulated queries, some of which contained simulated errors, was run against these databases. The results were evaluated using the normalized precision and normalized recall measures. Our concept of “musical words” shows great merit thus implying that useful MIR systems can be constructed simply and efficiently using pre-existing text-based information retrieval software. Second, this study is a formal and comprehensive evaluation of a MIR system using rigorous statistical analyses to determine retrieval effectiveness.

REFERENCES

• 1: Barlow, Harold, and Sam Morgenstern. A dictionary of musical themes. London: Ernest Benn, 1949.
• 2: Brook, Barry S. Thematic catalogue. In The new Grove dictionary of music and musicians, ed. Stanley Sadie. London: Macmillan Publishers, 1980.
• 3: Brook, Barry S., and Murray J. Gould. Notating music with ordinary typewriter characters (A Plaine and Easie code for Musicke). Fontes Artis Musicae 11: 142, 1964.
• 4: Dowling, W. Jay. Scale and contour: Two components of a theory of memory for melodies. Psychological Review 85 (4): 341-354, 1978.
• 5: Downie, J. Stephen. The MusiFind Music Information Retrieval Project, Phase III: Evaluation of indexing options. In Connectedness: Information, systems, people, organizations: Proceedings of the 23rd annual conference of the Canadian Association for Information Science, 7-10 June 1995, Edmonton, Alberta, 135-146. Toronto: Canadian Association for Information Science, 1995.
• 6: Downie, J. Stephen. Representing melodies as collections of"musical words": Networks Poster presented at ALISE '99, 26-29January 1999, Philadelphia, PA., 1999.
• 7: J. Stephen Downie, Music retrieval as text retrieval (poster abstract): simple yet effective, Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval, p.297-298, August 15-19, 1999, Berkeley, California, United States [doi>10.1145/312624.312727]
• 8: Downie, J. Stephen. Evaluating a simple approach to music information retrieval : conceiving melodic n-grams as text. London, Ont. : Faculty of Graduate Studies, University of Western Ontario, 1999. {dissertation}
• 9: Duggan, Mary K.. Electronic information and applications in musicology and music theory. Library Trends 40 (4): 756-780, 1992.
• 10: Donna Harman, Ranking algorithms, Information retrieval: data structures and algorithms, Prentice-Hall, Inc., Upper Saddle River, NJ, 1992
• 11: Hewlett, Walter B., and Eleanor Selfridge-Field, eds. Computing in musicology. Vol. 11, Melodic similarity: Concepts, procedures, and applications. Menlo Park: Center for Computer Assisted Research in the Humanities, 1998.
• 12: BKeller, Kate van Winkle, and Carolyn Rabson. National tune index, 18th century secular music. New York: University Music Edition, 1980.
• 13: Mark T. Kinnucan , Michael J. Nelson , Bryce L. Allen, Statistical methods in information science research, Annual review of information science and technology, vol. 22, Elsevier Science Inc., New York, NY, 1987
• 14: McLane, Alexander. Music as information. Annual Review of lnformation Science and Technology 31: 225- 262, 1996.
• 15: Rodger J. McNab , Lloyd A. Smith , Ian H. Witten , Clare L. Henderson , Sally Jo Cunningham, Towards the digital music library: tune retrieval from acoustic input, Proceedings of the first ACM international conference on Digital libraries, p.11-18, March 20-23, 1996, Bethesda, Maryland, United States [doi>10.1145/226931.226934]
• 16: McNab, Rodger J., Lloyd A. Smith, David Bainbridge, and Ian H. Witten. The New Zealand Digital Library MELody inDEX. D-Lib Magazine (May), 1997. Available at: http://www.dlib.org/dlib/may97/meldex/O5witten.html
• 17: Parsons, Denys. The directory of tunes and musical themes. New York: Spencer Brown, 1975.
• 18: Prechelt, Lutz and Rainer Typke. An interface for melody input. Unpublished manuscript, 1998. See also: http ://wwwipd.ira.uka.de/tuneserver
• 19: Randel, Don Michael, ed. The new Harvard dictionary of music. Cambridge, MA: Belknap Press, 1986.
• 20: RISM. Repertoire international des sources musicales: International inventory of musical sources. Series .4/11, Music manuscripts after 1600. CD-ROM database. Munich: K. G. Saur Verlag, 1997.
• 21: Gerard Salton , Michael J. McGill, Introduction to Modern Information Retrieval, McGraw-Hill, Inc., New York, NY, 1986
• 22: Setfridge-Field, Eleanor. 1994. The MuseData universe: A system of musical information. Computing in Musicology 9:11-30.
• 23: Selfridge-Field, Eleanor. Conceptual and representational issues in melodic comparison. Computing in Musicology 11: 3-64, 1998.
• 24: Jean Tague-Sutcliffe, The pragmatics of information retrieval experimentation, revisited, Information Processing and Management: an International Journal, v.28 n.4, p.467-490, July-Aug. 1992 [doi>10.1016/0306-4573(92)90005-K]
• 25: Tonta, Yasar. Analysis of search failures in document retrieval systems: A review. Public-Access Computer Systems Review 3 (2): 4-53, 1992.
• 26: Uitenbogerd, Alexandra, and Justin Zobel. Matching algorithms for large music databases. A technical report. Melbourne, Australia: Department of Computer Science, RMIT University, 1999.

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.4 Systems and Software
          Subjects: Performance evaluation (efficiency and effectiveness)

Additional Classification:
  H. Information Systems
  H.2 DATABASE MANAGEMENT
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.1 Content Analysis and Indexing
          Subjects: Indexing methods

  J. Computer Applications
  J.5 ARTS AND HUMANITIES
      Subjects: Performing arts (e.g., dance, music)

General Terms:
Algorithms, Design, Performance, Reliability

Keywords:
efficient search over non-textual information, results analysis and presentation for MMIR