Most of the information on this project is in German only. Sorry!
But some information on this project you can get from some emails I wrote in English to explain this project to some persons interested in it. I'll try to assemble the various bits under the following headers:
B.) Situation and problems: ----------------------- We have in various places various interneto- graphies, clearinghouses and the like for various subjects. Entries not rarely are multiplied accross these indexes. Part of these indexes are for electronic material only and part of them are for printed things as well. "Normal cataloguing" is found by the cataloguers as not suited for at least part of the electronic material. Part of the material catalogued is pertinent for more than one subject, and has to be indexed in more than one way more or less specicific for certain disciplines only. Cataloguers suffer from extra work due to lack of cooperation when indexing "non standard" materials. Customers and librarians complain that almost anything they need is indexed somewhere in the various catalogues, internetographies and the like, but that it becomes less and less likely that they will be able to find out where and how information about the item they seek might be retrieved. The "big" search engines are not enough help due to lack of really specific indexing and "up-to- date-ness". Maintaining the structures of the internetographies becomes more and more clumbersome. Meta-data contained in the electronic documents is scarce, and often not sufficient to retrieve the items searched. Inexing by authors tehmselves often is no viable solution.
We have various subject guides on various subjects. Users have complained, that they have developped in a way that makes them difficult to handle, that it takes too long to find out what is listes in which of these internetographies, etc.. Besides quite a number of resources have been indexed by several of our librarians and other specialists in several guides from several points of view. Consent has been reached, that any sort of "universal classification" won't be acceptable for specialist customers' needs (we have some 80 institutes doing basic research in more than 80 [in most cases rather "esoteric"] fields in STM and the humanities).
Our site has got several hierarchical entry-points and "internetographies" and other stuff on various subjects. Worked fine ... . At the start at least. But now, as the number of 1000 entries has been passed long ago, user-feedback says "there's almost anything that we are interested in indexed at your site, but it takes soooo much time to find it"; and my own experience says, user-feedback is right. And new entries to our indices are inserted every week ... . No, the situation and user-friendlyness does *not* improve ... . Just adding a "go for keywords you might imagine and hope that the author of the text or the indexer might have shared your imagination"-search- engine would not help (at least not on the long run).
A.) About our institution: ---------------------- Max-Planck-Society (non profit, mostly sponsored by public money) has some 80 institutes doing basic research in various science and humanities disciplines. This research is supported by some 80 libraries and other units for information gathering and providing most of which are part of the institutes.
Project managers are Diann Rusch-Feja, Peter Scherber, Ernst von Biron, and Heinrich C. Kuhn :
C.) Project steering comitee: ------------------------- 2 librarian+active research+electronic dataprocessing persons 1 electronic data processing+documentation+ active research person 1 electronic data processing person
D.) Solution: --------- Combination of intellectual and automatic indexing according to a vary detailed and very flexible set of categories. Use of a relational database for indexing, retrie- val and automatic generation of several types of documents. WWW-based interface both for indexing and retrieval. Cooperative indexing. Indexing can (but does not have to) make use of several indexing schemes at the same time.
> 1. When did the project start? Late autumn 1996. > 2. How many records have you done? 0 in the "solution database". Several thousands in clearinghouses, internetographies etc. that are to be merged into the "solution database". > 3. Are you using any specific software? RDBMS plus some things that we'll have to program or will have to have programmed. > 4. Does the project concentrate on one type of materials only? It will start with various types of electronic material, but th structure is such, that it can incorporate library catalogues and "classical" biblographic database entries as well. > 5. Are there any written descriptions or documentation, on the web > or in print? There is a description of the set of categories used for indexing. Descriptions are - alas - in German only. I can send you a copy if you are interested.
So we decided to develop a framework for so- phisticated "cataloguing" of internet ressources that is open for general and specialized indexing of all types, that permits "cooperative indexing" and the use of atomatically generated, author-generated and indexer-generated information to describe one and the same ressource. It permits as well to build a common database both for electronic and printed ressources. And, of course, it permits to include full texts. The final solution is intended to replace our present internetographies by documents created "on the fly".
So we decided to go for a combination of several things: - transforming most of our stuff from entries in more or less flat files to entries in a database - hierarchies built on the fly according to search-entries by users - a search-engine with quite a number of interfaces, hand-tailored to the various needs of various groups of users - the possibility to combine searches in hierarchies with searches via keywords and classification codes - and some add-ons ... . We hope to have the new design up and working sometime this summer ... .