As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites
This old website and all of its content will stay on as archive – http://archive.ifla.org
and
The timely publication of the results of the IFLA committee on 'Functional Requirements for Bibliographic Records' [2], meant that the workshop could focus on this as a de facto model and bring to bear some of the requirements and methods of Object Orientation. The varied experiences and long-term theoretical and practical interests of the members meant that the succession of examples used to examine the model was varied and complex.
The four levels are: Work, Expression, Manifestation, Item.
Each is an object derived of the preceding one and inherits properties from it. The most obvious example is (starting from the bottom) the Item. It is common practice that two Items (or copies) of a book, for instance, share all their bibliographic properties and differ only in their physical properties (Barcode number, placement, physical state, etc.).
The Manifestation and Expression levels formalise in a theoretical model the pragmatic decisions made in the MARC Format Integration changes, where data elements peculiar to a particular physical Manifestation became repeatable and were grouped together.
The Work level is an extension 'upwards' to allow the creation of a completely intellectual construct which is devoid from any processing or physical properties so that the unique 'absolute content' of bibliographic objects may be described.
The model is described as 'bibliographic', but it applies equally well to other objects, such as the authority records, which we model in the whole of 'bibliographic space'.
This cleaner model is easier to maintain, has increased flexibility of representation, may be better searched and may be more accurately processed according to rule based processes. It is very much in line with the results of the analysis of the digital resources, integration of which with the "traditional" material is very important these days.
Because of the existence of more bibliographic levels there will be more 'clustering' of individual real world entities (e.g. books and videos and computer games) under a single Work level object. This allows for 'intuitive relationships' (relations such as 'the movie of the book', or 'videos on the same subject as this song', etc.) searching to be applied. This opens a whole new area of searching possibilities.
The increased conceptual clarity of the 4-level model which includes the traditional authority files will enhance the navigation capabilities of OPACs. It becomes possible to navigate between objects at different levels as a meaningful search process. The authority files (Authors, etc. and Subjects) can be included in this navigation process as objects in their own right as they have the potential for significant information content at the different levels.
Simplified catalogue displays through 'information hiding', i.e. the use of the levels of abstraction in order to show the user - at will - only the Works (or Expressions) instead of a multitude of Manifestations during browsing of catalogues. That is, the forming of an hierarchic catalogue, which allow shorter, more compact, lists for friendlier, faster browsing from broad searches, providing an overall quicker path from start to the final desired object.
The model allows the natural 'media integration', i.e. the merging of the library catalogue with museum and archive catalogues, offering the user increased informational capabilities. The extension of the conceptual model to these related disciplines is almost automatic. That is not to say that the mapping of data elements becomes much easier, just that the framework is better and there is more chance of a unified approach from the users point of view.
The catalogue may be transformed in a language/culture/region dependent manner. By separation of the cataloguing elements in the object's language from the elements in the cataloguer's language and tagging the elements by the language/culture/region identifiers, it becomes possible to automatically deliver the culturally correct version of the object to the end user.
As discussed above for the end users, the improved definition of the objects allows for better chances of 'translation' and an increase in the possibilities of sharing data elements from 'foreign' sources.
The improvement in structure allows that direct control of authorities will be easier. Single Work level changes will be propagated downwards, objects will be simpler and more accurately fit the concept in the cataloguer's mind. Besides, the "cataloguing" of Works could be mainly the job of the national bibliographic or cataloguing centres, thus emphasising the "division of labour" within the library world, and also an efficient way towards UBC.
Direct interchange of records would become simpler and quicker. The records would be smaller and the mapping from one scheme of representation (e.g. MARC format) to another would be simpler.
Recording copyright information at Work and Expression levels will offer copyright management functionality to the library catalogue. Including comprehensive rights and permissions now becomes a technical possibility with the increased levels of abstraction. Updating and maintenance is easier. Also the 'Books in Print' catalogues could be more easily integrated with the union catalogues.
From the group's investigations it became clear that the theoretical framework of the new bibliographic model (dubbed 4LBM) needs to be enhanced and extended to 'connect' to the real world of library automation systems. Those systems (either existing or planned) which attempt any form of bibliographic model are rare. This reflects the overwhelming use of the MARC record as the sole database modelling foundation, despite its acknowledged inefficiencies. Thus most current systems are solely 'single record' based despite their use of database management systems.
Of the few systems which have ventured a different bibliographic model base, most have a scheme which is approximately that proposed by the 4LBM. These prove, in isolation, that the model is practical and even show some of its benefits.
To utilise the 4LBM would require changes to the database interface to the majority of existing systems. One of the projects the authors are proposing is just such an investigation of the minimal (and maximal) work needed to utilise the model and the (minimal) gains to be achieved.
With the more accurate definition of data elements in modelling terms and the whole issue of 'translation' between systems becomes an area for renewed study.
The results of this workshop are being deliberated by its members and other interested members of ELAG through a listserv and a web site. Interested groups and individuals are encouraged to subscribe and participate. The workshop members have also undertaken a search for researchers and projects attempting to use the model and are compiling a directory of these products, people and resources to help further co-ordinated development of the model. Further information can be found at
http://www.edulib.com/library/bibliographic/ or through an email to information@edulib.com