Published on

Riley, Jenn. “Variations2.” Women in Computing Technical Hour, February 15, 2005.

Published in: Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • The Variations2 Data Model includes the following entities and relationships:
    The Work represents the abstract concept of a musical composition or set of compositions.
    It is manifested in the Instantiation, which represents a manifestation of a work as a recorded performance or a score.
    The physical level is represented by the Container, which is the item or set of item(s) on which one or more instantiations of works can be found, e.g. a CD or published score.
    This is actually the level which is most typically represented in the traditional USMARC catalogs.
    Finally, the Media Object represents a piece of digital media content, such as a sound file or score image, and Contributors represent people or groups that contribute to a work, instantiation, or container.
  • Here is an example of how real recordings and scores might fit into the Variations2 Data Model.
    A CD titled “Mozart, Piano Works” would represent the container level in the Data Model.
    The recordings of two different pieces on this CD—Sonata K.279 and Fantasia K.397—would become instantiations, each having its own properties (recording date, place, etc.).
    The performers of these recordings would receive appropriate Contributor records with their own attributes (dates of birth, etc.). This is especially valuable because unlike in the USMARC system, Variations2 Data Model accommodates very clear and precise linking of performers and works.
    Each of the instantiations would be linked to an abstract work record, with its own “global” work properties. The work record, in its turn, is linked to its contributor, the composer Wolfgang Amadeus Mozart.
    A score would be accommodated in a similar way. For example, a score of the Piano Fantasia K. 397 would be linked to its instantiation, which would be related to the appropriate work and its contributor.
  • Variations2

    1. 1. Variations2 Women in Computing Technical Hour Jenn Riley Metadata Librarian Indiana University Digital Library Program
    2. 2. IU Digital Library Program (1)  A collaborative organizational unit – campuswide and systemwide – created to:  Provide financial support and human resources to support for existing digital library initiatives;  Provide infrastructure, financial support, and expertise to develop new digital initiatives across the campuses of Indiana University.  provide leadership in the development of digital libraries locally, nationally, and internationally.
    3. 3. IU Digital Library Program (2)  Joint project of IU Libraries and UITS  Research support from SLIS and Informatics  Created in 1997  12 FTE permanent staff  6 Libraries  4 UITS  2 jointly funded  10 FTE additional grant-funded staff
    4. 4. The Variations2 Project  $3M grant from the National Science Foundation through the Digital Libraries Initiative—Phase 2 program  Builds on the high-profile Variations project  ~7000 recordings  ~200 scores  Funding through UITS to the Digital Library Program to support Music Library and School of Music activities  Will end in September 2005; additional funding applied for
    5. 5. Goals of Variations2  An integrated multimedia library system to provide navigation, search, and retrieval functions for a large and diverse information space  A software framework to make digital music objects accessible to music instructors and application developers, using a component- based programming architecture.
    6. 6. Formats currently covered  Digital audio (from CD, LP, open reel tape, etc.)  Scanned score pages  Encoded scores  What’s missing?  Liner notes  Video  Record jackets  Concert programs
    7. 7. Variations2 staffing  5 FTE grant-funded staff  7 GAs  Student employees  Involvement from at least 8 other Libraries, School of Music and UITS staff and faculty  Other faculty as project investigators
    8. 8. Research areas  Copyright  Metadata  Instruction  System Design  Usability
    9. 9. Research areas  Copyright  Metadata  Instruction  System Design  Usability
    10. 10. Some staffing breakdowns  Development team  5 grant-funded staff (male)  1 GA (male)  1 manager (male)  Metadata team  3 librarians (female)  1-2 GAs (female)  Usability team  Headed by development team member (male)  1 GA (female)
    11. 11. Variations2 demo Schubert pilot
    12. 12. Technical environment  Java client application / RMI servers  Windows and Mac OS X client platforms  Unix server  QuickTime for Java, Darwin Streaming Server  IBM DB2  AT&T DjVu image compression
    13. 13. Variations2 Data Model  Uses entity-relationship analysis to identify key concepts, properties, and relationships of musical objects  Identifies, separates, and relates logical and physical layers of musical works and their physical manifestations  Similar to FRBR from IFLA, but designed specifically for music  In addition to descriptive metadata, also includes structural and technical metadata
    14. 14. is represented by MEDIA OBJECT represents a piece of digital media content (e.g., sound file, score image) is enclosed in CONTAINER represents the physical item or set of items on which one or more instantiations of works can be found (e.g., CD, score) is manifested in INSTANTIATION represents a manifestation of a work as a recorded performance or a score WORK represents the abstract concept of a musical composition or set of compositions Data Model: Entities is created by CONTRIBUTOR represents people or groups that contribute to a work, instantiation, or container
    15. 15. Broder, editor Prepared from autographs in 1960 Mozart, composer Fantasia K.397Sonata K. 279 Horowitz, pianist Uchida, pianist Sonata K. 279 recorded in 1965, Carnegie Hall Fantasia K.397 recorded in 1991, Tokyo, Suntory Hall Data Model: Example CONTAINERS INSTANTIATIONS WORKS CONTRIBUTORS CD Mozart, Piano Works Score Mozart, Piano Fantasia K.397
    16. 16. Data Model: Benefits  Increases comprehensiveness and precision of search results  Provides linkage of works in multiple formats on various levels  Allows for navigation within the work and between its different instantiations  Provides appropriate and complete descriptive, administrative, and structural metadata for each entity  Provides for Variations2 as a research system in addition to a discovery system
    17. 17. Usability research  To “study how students, faculty, and library patrons use digital music resources—and learn how technology can be used or adapted to help this process.”  Users involved continuously at all stages of project development  General evaluation of entire system and individual studies of specific aspects
    18. 18. Methodologies employed  Contextual inquiry: observations of Variations and Variations2 actual usage  Lab-based usability testing of prototypes and finished releases  Questionnaire studies during pilot use of Variations2 by classes  Interviews of instructors after using Variations2  Interviews with instructors who do NOT use Variations2  Log file analysis of Variations2 usage, both during pilot projects and during normal use
    19. 19. General usability evaluations  Searching  Audio Player  Score Viewer  Printing  Bookmarking  Timeline Tool  Paired vs. Individual User  Satisfaction Ratings
    20. 20. Usability evaluation of Variations  Aid in user requirements and task analysis for Variations2  Measure a baseline for comparison of satisfaction ratings between the original Variations system and Variations2  Compare satisfaction measures in a usability lab testing situation vs. real contextual usage.
    21. 21. Usability evaluation of searching  Evaluate the usability of the search interface  Evaluate the functionality of the search results  Investigate use of diacritics and multiple spellings in relation to searching
    22. 22. Usability evaluation of installation  Usability of new installer program  Effectiveness of installation help text
    23. 23. Usability evaluation of MMTT  Evaluate student learning potential in light of a variety of lesson content and presentation formats (i.e. question difficulty, appropriateness of musical examples, ease of completing harmonic analysis, melodic dictation, etc.)  Gauge student interest in using similar computer-based applications to complete music theory exercises  Assess the usability of the interfaces in terms of navigation, content layout and design
    24. 24. Plans for Variations3  Extend Variations2 into an open-source tool usable by music libraries of various types  Improve sustainability of the Variations2 metadata model  Other grants will cover:  OMR  Instructional use  Development of more MMTT modules
    25. 25. Further reading  Mark Notess, Jenn Riley, and Harriette Hemmasi. From Abstract to Virtual Entities: Implementation of Work-B . ECDL 2004. Bath, UK, September 2004.  Harriette Hemmasi. Why Not MARC? International Conference on Music Information Retrieval, Paris, France, October 13-17, 2002.  Mark Notess. Three Looks at Users: A Comparison of Methods for Stud . Toward a User-Centered Approach to Digital Libraries, Espoo, Finland, September 8-9, 2003.
    26. 26. Acknowledgments  For “borrowed” content  Jon Dunn  Mark Notess  Natalia Minibayeva  The rest of the Variations2 staff  WIC!
    27. 27. Required disclaimer This material is based upon work supported by the National Science Foundation under Grant No. 9909068. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
    28. 28. Questions?  For more information:  Variations2 web site   These presentation slides: <>