musicSpace

About musicSpace

Efforts over the past decade to digitise scholarly musicological materials have revolutionised the academic research process.  However, online research in musicology remains constrained by the segregation of knowledge into isolated databases, and by archaic metadata (i.e. the information used to identify and classify materials in a collection).  Many musicological research questions are rendered unanswerable because of this lack of quality metadata, and due to poor communication between database systems. The musicSpace project has taken a dual approach to exploring solutions to this problem:

  1. Design back-end services to integrate (and where necessary enhance) the information stored on partner sites (i.e. key online musicology data providers).
  2. Provide a front-end interface to support rich exploratory online searches.

We have unified online partner data through the use of Semantic Web technologies, and in some cases enriched the data, allowing for more refined filtering and browsing of records via our search user interface.  Significantly, the musicSpace framework allows all available information to be used by musicologists for the purposes of querying datasets, thereby aiding the process of knowledge discovery and creation.

Our benchmarks have been twofold: to make tractable previously intractable queries, and to accelerate knowledge discovery. The technologies that we have developed have been adopted by, for example, the Francophone Music Criticism project (http://music.sas.ac.uk/fmc).

Tour musicSpace

Only a decade or so ago, researchers in musicology – in common with many humanities disciplines – had to criss-cross the globe to consult material held in libraries, archives and museums all over the world. These days, thanks to efforts over the past decade that continue today, research data is increasingly available digitally, from the comfort and convenience of one’s computer.

This should mean that data is now readily accessible and usable for research purposes, but this has not entirely been the case.  Research data has become ‘siloed’ (contained within separate databases which don’t communicate with one another) into countless online repositories according to, for example:

  • Media type (i.e. whether the data relates to text, image, audio or video objects)
  • Date of creation and/or publication
  • The subject area of the material
  • Language of the material
  • Copyright holder

Using current online data resources presents barriers at all stages of the research process.  For example, it is hard to speculatively browse around a subject area.  Multi-part queries, such as the following, are effectively intractable:

“(i) What recording of works by John Cage exist, (ii) which performers have recorded a particular work by Cage, and (iii) what else by Cage have they recorded?”

Prior to musicSpace, such a query process would be laborious; the musicologist would have to use the search interfaces of the British Library Sound Archive, Copac (the amalgamated UK research libraries catalogue) and Naxos.  Metadata from one source could not guide another source, while multi-part queries were required to be broken down and the results filtered and collated manually.   This process was exacerbated by insufficient granularity (richness and specificity) of the data and search options.

musicSpace, by contrast, uses a multi-level hierarchy based on metadata type.  It makes this hierarchy explicit to users, making both broad and narrow searching possible.

The musicSpace system presents data as columns of attributes, allowing users to make selections in the columns in order to filter down results.  Columns can be rearranged easily, allowing users to prioritise search criteria and update search results on the fly.  Search results can be generated instantly for printing or email distribution.

Since the completion of the musicSpace project in 2010, the mSpace technology developed for this project has successfully been implemented in current musicology platforms such as the Francophone Music Criticism project (http://music.sas.ac.uk/fmc).  This work has been taken further by the MusicNet project.

Partner Organisations:

  • British Library & British Library Sound Archive (http://www.bl.uk/)
  • Grove Music Online (http://www.bl.uk/)
  • Copac (http://copac.ac.uk/)
  • RISM UK & Ireland (http://www.rism.org.uk/)
  • Cecelia (http://www.cecilia-uk.org/)
  • Naxos Music Library  (http://www.naxosmusiclibrary.com/)

Funders

  • JISC
  • AHRC
  • EPSRC

Duration

  • 1st September 2007 to 1st September 2010