Thank you so much for the wonderful discussion! I just put another star to this message. I agree that it's critical to have accurate and specific place name info in a record for various reasons.
As far as I am concerned, that is the value-added library metadata services that a cataloger (a.k.a. informatics specialist) can provide. Obviously, the actual cataloging work can be assisted with some visualization tools for potential match(es) to the place name that she/he is going to assign to the publication. To save cataloger's time for researching the "actual name of the larger jurisdiction or jurisdictions (state, province, etc., and/or country)", I am just wondering why we can't inject some inference rules to the cataloging tools in the context of cataloging the publication. The tools may be augmented with: 1. controlled vocabularies, e.g. place name from Getty TGN, etc.; 2. publisher name from LC authority file, 3. top query expansion log files from Google Universal Search, and 4. additional tools from Google Map, Google Earth, etc. for larger jurisdiction name(s) verification at the time of record creation, etc. For example, an older publication from Cambridge University Press, in MARC 260 field is assigned as the following: *260* *##**$a*Cambridge [Cambridgeshire] :*$b* University of Cambridge Press,*$c*1980. How can we get this output? The cataloger is the gate keeper and domain expert on data quality for every field in a cataloging record. The visualized tool injected with inference rules for location name resolution will assist his/her decision making in the choice of the larger jurisdiction name(s) like the following: 1. Acquire data usage pattern(s) of the place name and publisher in the publication, e.g. Cambridge is used* in conjunction with** *University of Cambridge Press; 2. Lookup place name in place name thesaurus, e.g. TGN, and publisher name in name authority files, e.g. LC NAF, etc.; 3. Retrieve and suggest possible choice(s) it finds a match or more as the following: *S1: From Data Source TGN* * ID*:7010874 ... Cambridge (preferred, C, V) ... *Coordinates*: Lat: 52 12 00 N degrees minutes Long: 000 07 00 E degrees minutes Lat: 52.2000 *decimal degrees* Long: 0.1167 *decimal degrees* Hierarchical Position: Hierarchy of World (facet) World (facet) Hierarchy of Europe (continent) .... Europe (continent) (P) Hierarchy of United Kingdom (nation) ........ United Kingdom (nation) (P) Hierarchy of England (country) ............ England (country) (P) Hierarchy of Cambridgeshire (county) ................ Cambridgeshire (county) (P) .................... Cambridge (inhabited place) (P) *S2: From Data Source LCSH* ... 010 __* |a *n 50059133 110 2_* |a *Cambridge University Press 410 2_* |a *University of Cambridge.* |b *Press ... *S3: From Google Map or Google Earth API Plugin * ... *Fly to*: 52.2000, 0.1167 *AND* Cambridge University Press to obtain the potential address as: Cambridge University Press, Shaftesbury Road, Cambridge CB2 8BS, United Kingdom ... OR if possible *S4: From Google Top Query Expansion Log Files* ... Cambridge University Press, Cambridge. United Kingdom ... *Inference rules* can be made as the following using plain English: Since *S2*, *S3*, or *S4* have associated Cambridge and Cambridge University Press with United Kingdom, AND *S2* indicates that the larger jurisdiction name of "Cambridge" is "Cambridgeshire," therefore, the possible choice for 260$a is *Cambridge**[Cambridgeshire]. *To get detailed data flow diagram for controlled vocabulary movement at 5000 feet above the ground, please feel free to check* *slide # 26 of 32 "DFD - Data Flow Diagram" available from Vision of Library Technical Services (2009) * http://www.slideshare.net/elephantsmith/vision-of-libraries-technical-services . *In summary, for additional data application services that we would like to provide to our community, we have to have some good data of high quality to start with. External data sources and tools will assist catalogers with decision making in the choice of form when constructing a heading for a place name and its larger jurisdiction name. This is just a high level description of how to get there. Sincerely yours, Amanda Xu On Wed, Apr 27, 2011 at 6:42 AM, <hec...@dml.vic.edu.au> wrote: > Quoting "Danskin, Alan" <alan.dans...@bl.uk>: > > It is not clear what >> benefit you perceive is derived from the addition of information about >> the larger jurisdiction. >> > > The benefit is to inform the catalogue user where the document was issued. > > There are many, many places which may appear in this element of a resource > description, but which share a name with other places far distant. One of > the FRBR things that seldom reaches our consciousness is "context" -- the > set of conditions in which a work or an expression was created, or a > manifestation published. > > Context also applies when a user searches the catalogue. In my own > environment (Australia), "Melbourne" as unqualified place is inevitably > taken to denote the capital of the state of Victoria. In a document > description, it might well be the homonymous place in Florida, or in England > in Humberside or in Derbyshire -- no doubt there are others as well. > > Elaine Svenonius, in expounding the principle of representation ("to > reflect the way bibliographic entities represent themselves")* states the > need for truth in transcription to support accuracy; she also says, "A > description is inaccurate if it in any way misrepresents an entity, making > it seem what it is not." > > No description can be called accurate if the omission of information > misleads a proportion of the users of the catalogue where it appears. A > great many users outside Ontario who read "London" will inevitably suppose > it to denote the capital of England -- the bibliographical universe is > indeed universal. > > No single principle can be carried to the utmost in implementation without > producing an absurd result: there always have to be checks and balances. One > strand of check and balance is the normal expectations of the user of the > catalogue -- a factor modified by environment, but one of which we can make > an easy guess in the case of "London". London, England, is not the same > place as London, Ontario (nor London, Kentucky; London, Kiribati; nor a > number of other places). > > Accurate knowledge of the place of publication is often one of the criteria > for selecting the resource which best meets the user's requirements; the > more so as selection is generally made initially from a brief record > display, not the full set of data. > > To deprive the user of the necessary identifying information presented in > conjunction with the primary place name is doing the user a disservice -- > and the highest principle, as Svenonius (p. 68-70, following Ranganathan and > others) reminds us, is the principle of user convenience: "Decisions taken > in the making of descriptions should be made with the user in mind." (p. 68) > > How does refusal to specify the jurisdiction which contains the place named > as the place of publication, and necessary to enable the user to identify it > properly, do anything but offer an obstacle to the catalogue user? Is the > principle of representation really so absolutely inviolable that > interpolation (clearly marked as such by square brackets) of necessary > information into a descriptive element that is not complete, and which is a > minor element in forming a citatioin for a document, really transgresses it? > > I rate the principle of user convenience higher, and judge that bracketed > information, if useful, should be supplied. > > *_The Intellectual Foundation of Information Organization_. Cambridge, > Mass. : MIT Press, 2000. (p. 71) > > Hal Cain > Melbourne, Australia > hec...@dml.vic.edu.au > > ---------------------------------------------------------------- > This message was sent using IMP, the Internet Messaging Program. > >
