Talk:Wikicite/Wikicat

Proposed name-change

Suggest immediate name-change (before page becomes fully established), as Wikicat has no real connotation of bibliographic catalog, or anything else bibliographic for that matter, and the name applies equally well to any substantial, organized, list of items, for any subject at all, across the entire Wikiverse. It should really be reserved for a generic wiki-function. (Perhaps a machine-readable list of Wiki subjects?)

Alternatives:

WikiBibCat is probably too long, Wikibic, a little short, and WikiBib is taken.

But WikiBibs is free [Update: Now reserved for Wikicat], and the best choice of all, anyway!

4.232.231.116 06:17, 16 April 2006 (UTC)

Sorry for the belated response- life sometimes gets in the way :)I . In any case I chose Wikicat as it is analogous to Worldcat, the world-wide bibliographic catalog. I'm not set on any particular name at this point, though I'd prefer something a litte more mellifluous than WikiBibs :)

Jleybov 03:41, 28 April 2006 (UTC)

I think the name wikicat is OK. --LA2 00:02, 30 July 2006 (UTC)

Legal issues

The current project description presumes it is legal to store and republish information that one gets over a Z39.50 interface. Is this so? And is it always so? What are the limitations and where are these documented? --LA2 00:02, 30 July 2006 (UTC)

This will definitely have to be investigated. I am fairly confident that government/public sources such as the Library of Congress would not have legal encumberances on using/storing data. Here is what a commercial maker of Z39.50 software says:

We do know that if you are in the U.S.A. you are permitted, even encouraged, to copy records from the Library of Congress for your own use.[1]

Most catalogs with restrictions (e.g. OCLC) do not offer free access anyway. Yet there could be the problem that records originally cataloged by a group that imposes restrictions (OCLC) might end up in the the database of a free institution (New York Public Library)...

I will start looking around, but it would be nice if the Foundation were to officially take up this task so that if some sort of institutional agreement were needed to get access, someone with actual authority could take action.

Jleybov 15:21, 31 July 2006 (UTC)

Even if WMF could get an agreement with OCLC to use their data, that would not be in line with the WMF/GFDL mission to accumulate and disseminate free information. An important function of the Wikicat project could be to improve (wash and extend) free bibliographic data (from LoC etc.) and feed this back into the library world. One risk is that our collection is "contaminated" with non-free bibliographic records, perhaps by accident by less knowing persons who happen to have access to OCLC data, similar to naive contributors who copy text from copyrighted encyclopedias into Wikipedia. Wikicat would have to draw a clear line between free and non-free bibliographic data. --LA2 23:45, 31 July 2006 (UTC)

Here are OCLC's terms, which are actually quite reasonable:

In addition to transfers to libraries, each member and nonmember library may transfer records of its own holdings without restriction, to (a) member networks, (b) state and multi-state library agencies and (c) all other noncommercial firms.

http://www.oclc.org/support/documentation/worldcat/records/guidelines/

And though IANAL, I remember reading that according to U.S. copyright law, while a database of public/free information may be copyrighted (e.g. a phone book), any particular data within it may not be and so in practice I think it will be almost impossible to prove copyright violations against us as almost all the data in the catalog will be derivable from the item itself. Still, policies should be in place for this sort of contingency just in case.

Jleybov 19:32, 1 August 2006 (UTC)

Sorry, I think wikicat is going nowhere as long as the legal issue is unresolved. I think wikicat's internal database of copied bibliographic records must be published in full, worldwide, as public domain or under the GFDL, and any restriction to "non-commercial use" or "personal use for U.S. citizens" is going to block this. The OCLC conditions that you cited are also restricted to bibliographic records representing the library's own holdings, and wikicat needs to deal with bibliographic records for many books that we don't hold. It is no surprise that we cannot use data from OCLC, just like Wikipedia cannot copy text from Britannica. The question is if we can use data from the Library of Congress. --LA2 21:08, 13 August 2006 (UTC)

The stipulation about the library's own holdings is only put there, I think, to prevent one library giving away access to the entirety of OCLC's data; plus in practice the query results for a partiuclar OPAC are going to be limited to that library or library system's holdings anyway.

Regardless, I think this will all be a moot point in the end. Wikicat will be seeded by external records, true, but through a process of extracting, transforming, and later extending the data contained within them. The actual records will not be kept, nor will they be divulged should Wikicat itself become an OPAC in the future. Again, you cannot copyright publicly available data, and that is the most Wikicat will retain of any given bibliographic record.

I'll still had a legal issues section for each entry in the OPAC target list, though. -- 03:35, 16 August 2006 (UTC)

Here is confirmation, from the source, that Library of Congress bibliographic records indeed are free of legal encumberances, both for copying and redistribution: http://lists.indexdata.dk/pipermail/yazlist/2006-December/001789.html

Jleybov 17:26, 6 December 2006 (UTC)

Relationship to other projects

On the foundation-l, wikicat was described as similar to IMDb but for more than films and actors. This would include MusicBrainz, LibraryThing and CiteULike. The current project description doesn't include any reference to these projects. Some (MB, CUL) are similar in to Wikimedia's openness and nonprofit status, others (LT) are more commercial. Should we import their data? Should we cooperate? What relationship should wikicat have to those other projects? --LA2 00:02, 30 July 2006 (UTC)

I haven't heard of these other sites, but there definitely could be the potential for collaboration. One problem I can immediately see, though, is that we probably could not enforce any sort of data quality standards on these other sites. Thus Wikicat would have data coming in from professional/institutional catalogs via Z39.50, data manually entered by Wikcat users/editors, and data coming in from external sites like MusicBrainz. This obviously multiplies the potential for data becoming inconsistent/incoherent.

Jleybov 15:21, 31 July 2006 (UTC)

Developing an annotated bibliography on Wikiversity

Hi, for a research projekct on Wikiversity I started developing a small annotated bibliography (v:Topic:Annotated Bibliography of African Philosophy. I am using pages in Topic namespace as "library cards" and I can group them into different categories. That is better than a simple list, but it is still an ad hoc approach and cannot be the final solution, especially if the number of references is getting larger. Annother user hinted me at an interesting solution on another Wiki. He wrote to me:

Hi Nannus - I looked at v:Topic:Annotated Bibliography of African Philosophy and I saw your desire to create a "ResourceDescriptor-Namespace". You might be interested to see the Science Online wiki, which has implemented a reference namespace (example) - is this along the lines of what you were thinking of? Cheers. Cormaggio ^beep 01:38, 2 March 2007 (UTC)

In my reply to him I have described my vision of such a system. I would like to know if the functionality I have in mind is covered by the current Wikicat project. Maybe some of my ideas are useful here. Here is what I have answered (Retrieved from "http://en.wikiversity.org/wiki/User_talk:Cormaggio")

Hi, thanks for the hint. Yes, that goes in the direction I am imagining. I like the idea that articles must not have direct external references but must go through an intermediate level of references. This seems to make life harder on the first look but would actually be very usefull. You can see on that site for each reference which articles are using it. That is useful. And you could add abstracts, annotations and reviews and discuss a sources credibility etc. on the article about the source instead of in your article or its talk page. I have also seen that there is a Meta project called Wikicat, but I have not yet had the time to really look at what they are doing there and if that project in the present form covers what I am thinking about. I think I should go there and ask them. What I imagine is a site that would be outside Wikipedia, Wikiversity etc. because it provieds references for all. It would also be multi language. The "reference cards" should have a structured part with the bibliographic information plus a free from part that is multilanguage. If such an object is created, either from an import from a library server or by an author or reader entering it, it would immediately be visible for all languages. You could then add new text parts in different languages. For example, initially the text could be in english, but somebody else could then add a french or japanese text part on a parallel page (maybe users could define a sequence of language preferences). Annotations, abstracts etc. for a reference can so be centralized instead of being dispersed over different articles using a reference. Discussions about the validity or quality of literature items could be moved out of single articles in Wikipedia etc. to the talk pages of such objects, making them visible from any article using that reference. References could be categorized by lots of overlapping categories, making traditional hierachical bibliographical signature systems obsolete (although these traditional classifications would still be included). Category-like information should be in categories, not in fields on the cards. The cards could include links to other cards with a "cites" meaning (and the software could offer the opposite "cited by"). It should be possible to annotate (fre form or partially formalized) such links in turn so that information like "a confirms b", "a rejects/contradicts b" could optionally be captured on the cards. This would especially be useful in the humanities/social sciences. There should be the possibility to group different editions, translations etc. into group entities (my first look at the Wikicat page indicates that they have included such features already). The data model could include pages on authors and institutions. The system could include an interface to a distributed system of library servers through which one could find the next library (archive, museum...) containing the referenced resource. Library servers could include interfaces to such a system. From the user's pint of view, this would look like one server (demanding project, but seems feasable to me). I am not sure to what extent these ideas are covered by the present wikicat project. In any cas there should be free form text space (i.e. if you overformalize, it becomes less usefull) and discussion pages attachable to each "card". Sorry for getting so lenghty. I think I will just carry this over to the Wikicat project's talk page.Nannus 18:09, 3 March 2007 (UTC)

Some suggestions and Another Wiki idea utilising the catalogue and expanding it's use

Below I have pasted a post I made at the 'Wikibibliography discussion' page ([[2]]). It outlines some solutions to problems raised and then puts forward a proposal that I was thinking about before I was aware of these cataloging projects. In light of this project looking as though it might 'swallow' wikibiliography, and also because Wikicat seems to be much more in line with the thinking I had about the Catalogue half of my idea (with the well thought out 'work, expression, manifestation and item proposal), I thought that WIM would work much better as an extension of this project. What follows is the post I made on the Wikibibliography discussion.

I have come up with a similar but more expansive idea than Wikibibliography that would use a Wikibibliography to give a rough history of ideas - Wiki Ideas Map (WIM). I've outlined this proposal roughly and have pasted this below, but first wanted to address the problems proposed by 'Meta' above.

1) The first problem as i understand it is scale, and it is quite true that in some libraries in the world breath in and out more books in a year than one person could read in a lifetime. (I paraphrase 'Library: A History'). However, this it is not essential, at least initially, to have the entirety of every book written catalogued. Like Wikipedia, which does not have an entry on every single topic thought of or even named within itself, it does give the user the vast bulk of things they are likely to look up. One of the joys of a Wiki is that it expands in the directions users push it in.

Neither do I think it would be impossible to get a large amount of entries put in. Unlike Wikipedia a lot of the work could be done by those already doing it. I'm thinking of libraries and universities who catalogue all their new books each year. I think they would be happy to simply have their catalogue information spliced together in a Wiki format, and/or, put the new items straight on to Wiki. If the format could be set up correctly this could leave them simply performing the same task with the information simultaneously entering the Wiki and at the same time their own library catalogue. If the Wiki became expansive i think it would become of great use to the library community.

There are also companies like 'Amazon' that catalogue their books. Whilst I'm not generally for advertising I think this may be a case where it could be utilised well. One of the ways the Amazon advertises is to have a 'buy this book at amazon' link on a book entry, such as in library catalogues. They could be collaborators as well as financers.

2) The second problem I understand to be verification. I could understand this would be a problem if we only had 'any tom, dick or harry' entering information in. One solution to this would simply to have verification system. I think I’m correct in saying Wikipedia has something of the sort. This would be a little different though. This system would work on a basis of 'user trust' and 'multiple verification'.

By 'user trust' each piece of information could contain a tag that told you who entered the information. Some people would abuse this Wiki, but in a short space of time genuine trustworthy editors would become know by name.

By 'multiple verification' i anticipate a check system. So any editor could have a book that is on the Wiki already and simply check the information. This check would operate as simply a list of user names by the entry rather than one.

!However! I do not see this as at all being a 'tom, dick or harry' situation. It would be in the financial interests of printers, publishers, (internet) book sellers, authors, even libraries and universities, to have information about (their) books up on the Wiki. Then 'user trust' and 'multiple verification' become heavily bolstered by verifiable trusted users, that is users whose identity has been verified by the relevant companies and institutions as being who they say they are.

Ok, well I’ll now get onto my proposal which I previously emailed to the Wikimedia email address before getting sent here. I called it the Wiki Ideas Map, WIM, and at it core it is a Wikibibliography/biography/other covering people as well as art and books.

The Idea: 'Wiki Ideas Map'

Initial Impetus: The people that the Wiki Ideas Map (WIM) is formulated upon helping are those for whom following references back through numerous books and other works is necessary. I'm thinking of those doing research on a particular area of history, an author or authors, or an idea; students and their professors in the main, but also journalists and writers as other examples of those wishing to get to grips with the span of an topic.

If such researchers could follow the references back without having to locate and find each individual text but instead have the 'map' of references on the web then naturally following the references would be much easier. This map is the WIM, which would be roughly a combination of a detailed (see below for the key details) universal library catalogue and a historical spider diagram, the former supplying the database for the latter.

How It Would Work In Brief: The two interfaces of the WIM are the 'Catalogue' and the 'Map'.

The Catalogue would contain much the same information as a library catalogue, hence the name. It would be the working area of the Wiki, into which information was added and editing done, and it would also be research tool in itself. It would though contain more information than a library catalogue, including, i envisage, publishers and printers, and it would be divided into entries on individual books, articles, authors etc, much like a wikipedia or wikidictionary entry. The key additional piece of information would be the list of references. These would be much like the references you see at the end of an essay or dissertation or non-fiction book. They would include references of this work (such as a book) to other works, references of other works to this work and references of those claiming a link from or to this work.

Note though, that whilst I'm referring to books and articles as the entries in the WIM, the WIM catalogue would potentially contain entries on every noun that was concrete or imagined to be concrete as long as they have been referably referenced (or could be), such as works of art, publishing houses, authors and artists and other individuals, etc.

For example, lets say someone was doing a typical research assignment for university on Virginia Woolf. By looking at Virginia Woolf they would find a list of her works, such as Mrs. Dalloway. They open the WIM 'Mrs Dalloway' entry and there will be a link to the various publishers of Mrs. Dalloway, the various printers, the various editions (which would form part of the same WIM entry) and all references made by or to Mrs. Dalloway, whether directly and specifically made by the text itself or whether despecified by a third party. Our researcher could then follow one of those references to the next entry, and could potentially follow a line of references back.

The Catalogue would form the database for the Map.

The Map would be the initial interface for users. If you like, the Catalogue allowed the user to look at the details, plunging the depths like a diver. The Map is what allows the process to really speed up, allowing the diver to skim the surface of the water on speed boat.

The Map would essentially be a spider diagram, with each point being an entry from the Catalogue. Each line between the points would be an indication that a reference had been made from one entry to another. Other useful details would include a numerical value, indicating the number of references, the ability to click on a line to bring up the list of references, the ability to turn on and off type of referencing, and the ability to condense entries into ‘topics’.

Let’s return to our previous research example of Virginia Woolf. Using the Map, with our initial ‘point’ of the Catalogue entry ‘Virginia Woolf’ we select the option to see all works by Virginia Woolf. Our researcher could then select to see all works referenced by Mrs. Dalloway. The Map would now show Virginia Woolf point with lines to the individual works of Virginia Woolf, one of which, Mrs. Dalloway, has lines linking to the works referenced by it. In this particular case there are no lines from the work, Mrs Dalloway, so our researcher opts to add to the Map all third party references, producing the new points and lines.

Our researcher might now want to look more generally at those links made to Virginia Woolf and here works in general. In this case they made ‘contract’ each of those separate points, Virginia Woolf and the individual works, into one point ‘Virginia Woolf and her works’. The Map will now treat these entries as one point and display them as such.

Uses: I hope the initial uses are fairly apparent, that is the ability to wade through reams of previous references, and thereby skim the previous research on a topic to get one’s bearing more fully on a topic, preventing repeating research and enabling quicker, better selection of books. However I envisage further uses for this WIM. For example, as the WIM grew in size it could highlight gaps in research, assumptions that may prove to be false, debates, and theories that have slipped and been largely forgotten.

Problems: There will of course be technical problems, most of which I believe are solvable and indeed in my head are solved. I’ll mention the main one here: How do we decide whether or not something is referenced? The solution is simple: a reference must be overt. If a reference is not overt but covert then it simply will not be included. This is where third party references come in. I will demonstrate with an example:

I write a piece of text that says: ‘I am Cain’. I add a reference saying ‘A.N.Author, Book of Cain pp19-20, 1998. This is an overt reference and can be included in the Catalogue. A third party though may think that I am also referencing the Tome of Cain. There is no reference to this so then it is not able to be included in the Catalogue. However our third party states his belief on a radio show. This quote from the radio show becomes our third party reference.

Starting Point: Firstly I someone is needed who can write the WIM, I don’t know anywhere near enough computer languages to know where to start. However, I think I could 'guide' such a person.

I propose a tester using the Victorian era writers of Great Britain. They are numerous enough to provide a good test, with I think the perfect combination of clichéness and references to outside the era.

Please let me know what you all think and if anyone would like to go ahead with the idea and if Wikimedia would like to support the test.

I hope to see about adding these suggestion to the Wikicat idea, but right now I don't know how much time I can devote to this.

--Gris 10:24, 9 May 2007 (UTC)

Manifestations and Items - A few slight redefinitions needed?

On the 'Design' specification: 'Entity': 'Item', the definition of 'Item' ended:

"Since an Item can consist of several distinct physical objects, its unity is pragmatically determined, typically according to how it is physically packaged as well as its release history. A box set of CD's is obviously one Item, but so are two separately bound volumes with no common sleeve/box that were issued and sold together."[[3]]

I think that this definition needs changing so that 'two separately bound volumes that were issued and sold together constitute two separate Items. The history of the volumes do not constitute them being one Item.' I would also add the idea of a 'Item History' wherein this event of the two items being issued and sold together could be noted in each Item's History. It might also be useful for the Design: Entity: Manifestation to have a History.

If however those two volumes were bound together then they would constitute one Item.

Alternatively, two volumes issued together could constitute both one Item and two Items, all three of these having a note in their History.The segment above should then read 'two separately bound volumes that were issued and sold together constitute two separate Items and one Item, constituting three Items in all.' We might sub-define this third item as a 'coincident-Item'. This would be distinct from the 'single-Item', our base unit, and the 'composite-Item'. A 'composite-Item' is not two Manifestations bound together, even if those two Manifestations have been issued as separate items. A 'composite-Item' is two volumes still separate but issued and sold together within one uniting physical entity.

To clarify with an example then: If we were to take Neil Gaiman's Sandman comics, they have been issued as individual comics. Individual Comics 1-8 issued and sold separately would constitute eight different (single-)Items. Individual Comics 1-8 issued and sold together would constitute nine items, the eight individual comics (eight single-Items) and the eight comics together (one coincident-Item). If these eight individual comics where placed in a uniting card sheath then this would constitute ten Items, the eight individual comics (eight single-Items), the card sheath (one single-Item) and the eight comics together with the card sheath (one composite-Item). If the eight individual comics were bound together into one Graphic Novel then this would constitute one (single-)Item.

It may not be necessary to capture all of these Items, but I think the option should be there.

--Gris 11:43, 9 May 2007 (UTC)

Page look

How would look one page on Wikicat? Ididn't read it enywhere. --StjepanG 12:49, 31 December 2009 (UTC)