NINES-TEI Workshop -- Day 3
Notes from Day 3 of NINES-TEI Conference:
_Customization the TEI schema: options_
--Select modules
--Delete unnecessary elements
--Add new elements or attributes
--Change element or attribute names
--Constrain attribute values (constrain data early and often. Tighter you make the schema, the fewer errors you'll have. But, don't contrain too much).
--Constrain structure
--Manipulate functional groupings of elements
--Produce an internationalized version of the TEI
_Contextual Information_
Information we know that is relevant to an understanding of the text:
--The identity of things named in the text: people, places, books, etc.
--Information about things named in the text: birthdates, geographical locations, date published, etc.
--Interpretive information: themes, keywords
--Normalization of measurements, dates, etc.
_Contextual Information in the TEI_
The TEI provides several different structures for encoding contextual information:
--’Ographies: prosopography (personography), gazetteers(placeography),orgography, bibliography
--keywords applied to the text as a whole
--thematic or interpretive information applied to specific places in the text
--attributes for supplying normalized values
_Personography_
--Like a local name authority file
--Can be simple or very detailed
--Can be kept in your encoded file or externally
--Includes specific elements for the most common data
--Also includes general elements for the unforeseen
_Placeography (Gazetteer)_
Key points:
--Very similar to personography...but for places!
--Can be linked to maps via geographic information data
* use XLST for organization, CSS for display, database (e.g. Oracle, Lucene) for search capabilities. Do in-house, NOT via outsourcing
_Successful Digital Projects_
--Well-planned workflow (planning before execution, communication mechanisms, etc.)
--Phased approach (show progress early on)
--Successful staff plan (needs versus practicality)
--Realistic technical implementation plan (identify tools at the cale you need. Oracle database?)
--Funding...
_Workflow Issues_
Source --> Transcription --> Corrected Transcription --> HTML output
information gain should go to "corrected transcripton." "Information loss" may occur at HTML output.
_Workflows that Make Sense_
1. Craft approach
detail initial capture by hand (i.e. you encode it) --> review and error correction --> interface created that allows you to do various things
2. Expertise in the Craft Approach
detail initial capture by hand (i.e. you encode it) --> review and error correction --> scholarly and technical expertise interacts with review-and-error correction --> display
3. Phased Approach
simple initial capture (not detailed capture; could be automated via OCR capture) --> simple error correction (e.g. spell-checking) and standard publication tools that do simple searches (e.g. search on author) --> simple reading and search interface output --> then do more advanced information (i.e. add to the XML markup) --> new output
Occurs often in large projects
_Project Life Cycle: Starting Out_
1. initial idea. Essentially volunteer
2. Seed funding (e.g. NEH DHSG)
3. Implementation that works, with a real audience. Requires serious funding (NEH, Mellon, Gettey, NSF,...)
4. Discover flaws and re-do; get funding because it's still an interesting project
5. wrap up and archive; institutional funding via institutional repository
-OR- sustainability model for on-going project ad infinitum --> ongoing redevelopment/new prototypes
Comments