Menu

#245 provide suggested values for rs@type

GREEN
closed-accepted
5
2011-12-08
2010-09-11
No

The element definition for <rs> includes @type but does not give any suggested values (cf. the definition of <fw>). A few values are given in examplesa in the Guidelines, but sometimes seemingly simlilar concepts are expressed with different values (such as type="org" and type="organization"). For ease of interchange of texts, it would be good to recommend certain values, giving explanations of what these are intended for (as in the definition of <fw>).

For background, see http://listserv.brown.edu/archives/cgi-bin/wa?A2=ind1009&L=TEI-L&T=0&F=&S=&P=176 .

Discussion

  • Laurent Romary

    Laurent Romary - 2010-09-11

    I would suggest to keep in synch with existing named entity classifications, such as the widely referenced BBN classification used in evaluation conferences such as MUC.
    See http://www.ldc.upenn.edu/Catalog/docs/LDC2005T33/BBN-Types-Subtypes.html

     
  • Lou Burnard

    Lou Burnard - 2010-09-13

    Proposal is to supply some suggested values from the source proposed below, and to reference them, noting however that other taxonomies may be appropriate in other application areas.

     
  • Lou Burnard

    Lou Burnard - 2010-09-13

    Proposal is to supply some suggested values from the source proposed below, and to reference them, noting however that other taxonomies may be appropriate in other application areas.

     
  • Lou Burnard

    Lou Burnard - 2010-09-13
    • milestone: --> GREEN
     
  • Lou Burnard

    Lou Burnard - 2010-10-26
    • status: open --> open-accepted
     
  • Sebastian Rahtz

    Sebastian Rahtz - 2011-11-04

    I would vote against this. Let sleeping dogs lie. If we want interchangeable values for @type, it applies far wider than <rs>, and this seems like tinkering at the edges. The stuff at http://www.ldc.upenn.edu/Catalog/docs/LDC2005T33/BBN-Types-Subtypes.html looks much too specific to one domain to me.

     
  • James Cummings

    James Cummings - 2011-11-05

    I would agree that we can provide some suggested values, but we should probably be look at rationalising our recommendations of suggested values for @type across the guidelines as a whole.

     
  • Kevin Hawkins

    Kevin Hawkins - 2011-11-07

    Lou has changed the instance of <rs type="organization"> in P5 to <rs type="org"> per the generally preferred way to handle syntactic sugar (using existing element names). Will embellish the definition of rs@type to recommend use of http://www.ldc.upenn.edu/Catalog/docs/LDC2005T33/BBN-Types-Subtypes.html . In addition, add a note to att.typed:

    1) giving our approach to syntactic sugar (use the name of the sugar)
    2) use a taxonomy in your field if it's appropriate
    3) not all elements with @type use att.typed

    Kevin will implement.

     
  • James Cummings

    James Cummings - 2011-11-07
    • assigned_to: nobody --> kshawkin
     
  • Kevin Hawkins

    Kevin Hawkins - 2011-11-09
    • status: open-accepted --> pending-accepted
     
  • Kevin Hawkins

    Kevin Hawkins - 2011-11-27

    At revision 9830, modified the definition for <rs> and did (2) and (3). However, I'm not sure how to explain (1). I wanted to say:

    "Alternatively, use the attribute to create syntactic sugar for another TEI element. In this case, the value of the attribute correspond to the TEI element name. For example, ____"

    For the example, I thought of using name@type. But for a personal name, do you do <name type="person"> (corresponding to <person>), <name type="persName"> (corresponding to <persName>), or <name type="pers"> (corresponding to the "pers" in "persName")?

     
  • Kevin Hawkins

    Kevin Hawkins - 2011-12-01

    After discussion on tei-council, I see that I was misusing "syntactic sugar". After some investigation, I see that while @type often achieves syntactic salt for an element not available in the modules chosen for a particular customization (as is often the case with <name> or <rs>), I can't find any members of att.typed for which this is typically the case. That is, att.typed includes elements that are themselves already sugar or which don't seem to be sugar or salt of anything.

    So I'm now inclined to skip trying to say anything about syntactic sugar for att.typed.

     
  • Kevin Hawkins

    Kevin Hawkins - 2011-12-01
    • status: pending-accepted --> open-accepted
     
  • Lou Burnard

    Lou Burnard - 2011-12-01

    I agree with Kevin that there is no need to attempt to explain usage of syntactic sugar (or salt if you prefer that rather misleading metaphor), which was not in fact part of the original ticket. So this ticket might as well be closed.

     
  • Lou Burnard

    Lou Burnard - 2011-12-08
    • status: open-accepted --> closed-accepted