On 11/06/2013 16:38, Anderson, William wrote:
It would seem the first logical step would be “what is the job”, the tools will emerge from that. It occurs to me that we do need something like FRBR user tasks, if not “the” FRBR user tasks. Does the word “tasks” start one step to far along the process than what is needed. As James stated below, returning to brass tacks, “what do people want from information”, than moving on the “tasks” to provide it. The DSM IV occurred to me as a model. One wonders if we could have a manual of general “information diagnoses” (IDM?) with evolving “treatments”, i.e. need à task. The framework around which to build tools. (To head it off, no I don’t mean anything as absurd as knowledge base to answer every question. Mental illnesses, pardon the analogy, each are very individualized whatever diagnosis is chosen from DSM, ditto any “information illness” for a librarian to doctor.). It would be an interesting idea for a committee (I almost wish we’d done this before RDA).
I like the idea of basing it on the DSM. One idea I have had is if you could get a massive database of questions from reference librarians (the first stop toward discovering what the public really wants. Does something like this exist already?) along with the answers that the librarians deemed satisfactory, you should be able to encode those answers as queries, e.g. for this kind of question from this kind of searcher, look in these databases, look at these tools, and use these methods. The expert would fill in the necessary information. These are the types of things that can definitely be automated at least up to a point, but the final product would necessarily be highly complex to use. Bingo! You have the beginnings of a tool that definitely works, giving people what they want (which would have already been proven since it came from real-life practice from reference librarians) and you would need an expert librarian to use it. The key is to not dumb it down. The expert queries themselves should be cataloged for later retrieval for others, possibly using the DSM as a basic model for arrangement, once again for use by expert librarians. And of course, the queries themselves would be going through constant revision as librarians learned more about what people want, and what information is available out there.
Again, something like this could follow a DSM pattern fairly well, I think. (By the way, here are the codes http://en.wikipedia.org/wiki/DSM-IV_codes and yes, they can be found alphabetically too!) In the place of the DSM arrangement, there could be sections by topics (art, architecture, literature etc.), by formats (DVDs, streaming, paper, parchment), by creators, by time, by types of researchers (children, high school students, interested adults, journalists, researchers) etc. and these queries could be combined or separated as needed. Each query could have lots of technical information explaining more clearly what it does and doesn’t do, problems with it, etc. etc. etc. I am sure the system for something like this could be built relatively easily and something similar probably already exists. The real problem would be getting the questions and decent answers, along with general buy in.
But it is obviously more important to figure out what parts go with the work, the expression, the manifestation, and the item! Doing this will help people find things in our catalogs even though the structure will apparently not be followed in Bibframe, and nothing has to be demonstrated. 🙂