home contents changes issues options help subscribe edit

Conversation with Associate Director/Historian Dr. Joe Mosnier, Southern Oral History Program, University of North Carolina at Chapel Hill

Focus of conversation: Controlled vocabulary for oral history material

Approximately 500 oral history interviews completed and documented already

Indexing

Each interview indexed using basic fields (name of interviewee, topic, time period, subject(s)). Selected excerpts are also indexed (if deemed significant by project team). In the future want to make full text searchable (will use some kind of Google-like “intelligent” search software)

How did they create their subject vocabulary?

We created vocab drawing from 1) existing Library of Congress subheadings (found in “red books” in library ready reference and on the web), 2) indexes of scholarly monographs. LC subheadings: too general, but they started with this and elaborated on; combined with monograph index headings. “Just like if you wrote monograph, index is author created/driven.” [JI: I would think they also drew from scholarly journal literature]?

Active scholars determined best indexing/subject headings/taxonomy, what they and colleagues would fine useful.

So indexing is time-bound, reflects current intellectual conceptualizations. How strong will utility be in the future? Have to make that make best judgment for this point in time.

Tools to make sense of interviews and tools to access interviews: vocab/taxonomy/finding aid/cross-reference/map

Search features

Boolean search feature (and, or, not). Abstract is searchable.

Project Software

Use open source software to run the database (have technical team, based in library, who manage this aspect).

Formats

Digital recordings, 16 bit 14.4 khz (wav files), down to mp3, recording to mono. Audio files on server. Mp3 delivered online. Open source software to serve them.

Todd Cooper, project manager, library UNC-CH. cojere@email.unc.edu He can talk more from the tech side.

Project Goal: Find 500 interviews on the web, coded, commented upon. Themes, coding scheme. Multiple strategies for searching. Advanced search feature.

Costs

$600 per interview before processing (3-4 hrs. for entering interview into database), dbase automated

$100 for processing is definitely worth it, to make easier to use and more accessible. To increase use and awareness of materials. Rather than just sitting on the shelf.

$600 + $100 = $700 per interview/ total cost they calculate.




subject:
  ( 4 subscribers )