Descriptive Metadata for
Audio-Oriented Digital Collections


Conclusion

What do the findings tell us about best practices for audio oriented digital libraries and online collections? This study examined metadata records, but did not evaluate whether that metadata was successful in supporting the desired functionality. Given that there was such a wide array of implementations, it is not possible to give a single model to emulate. However, some specific recommendations for audio-oriented metadata have emerged from this study.

When developing metadata requirements for a pre-existing collection of content, in addition to evaluating user needs and modes of expected, the characteristics and consistency of the content should be assessed to determine the degree to which it is more or less homogeneous, as well as the extent of any pre-existing data that should be accommodated. Furthermore, in general, metadata schemas for audio-oriented content should accommodate:

  • multiple contributor roles
  • data about time and place of recording
  • identification and issue numbers
  • explanatory or descriptive notes

To some extent, these findings confirm basic understandings about sound recordings, previously identified in earlier cataloguing rules, such as the ARSC and IASA rules. The challenge of incorporating these recommendations into practice will be in striking a satisfactory balance between the conflicting imperatives of specificity and flexibility: specificity as required by the nature of the content and flexibility in achieving interoperability. Working within existing standards is recommended.

Future Research

This study was exploratory in nature, and further research in the area of audio-oriented online collections would be welcome. Further work with the group of archives and collections in this study could conduct further sampling to improve the reliability of the results, or take a closer look at the metadata schemas in use. The pool of collections studied could be broadened to include additional collections subsequently identified by the author, such as the Densho digital archive and the Maria Rogers Oral History Program digital archive. As mentioned earlier, the measurement taken in this study was the metadata elements in use in the online collections, but the fact that particular elements are in use does not allow any conclusions to be drawn about their effectiveness in meeting the actual needs of users. With audio recordings in particular, there is a tendency, in some cases, to want to create highly detailed and descriptive records, without any sense of whether the added detail is meeting the needs of the users. The analysis of user tagging and descriptions, as noted in the Freesound project, would provide one way to evaluate user perspectives on metadata requirements.

Acknowledgements

The author wishes to thank Dr. Ali Shiri for his guidance and support over the course of this research project. The author also wishes to thank Dr. Dinesh Rathi for his assistance in adaptating the research paper for the web.

This website was created by John Huck in March, 2010
School of Library and Information Studies
University of Alberta