There are many concepts for describing bibliographic data (eg marc21, frbr, spar, etc.). However, there is no general agreement about their usability in the field of bibliographic databases. Which model do you think is the best? Please justify your opinion.
In my honest opinion since there are so many standards, there will be no simple and straight answer to this question. Each standard focuses on different aspects and details, and therefore each has different drawbacks. However, since you have asked about bibliographic databases I would recommend more holistic approach, in which you will specify the upper set of features each of standards offer, define transformation protocols and simply speaking bring all the standards to one common ground. After that you will be able to conduct any kind of operation on the data that you like.
The problem with bibliographic models is that there in no one-size-fits-all model for bibliographic information. The most well-known is Functional Requirements for Bibliographic Records (FRBR) proposed by the International Federation of Library Association (IFLA), for describing documents and their evolution.
Dear Krystian,
thank you for response.
However your suggestion is exactly what I wanted to avoid :-) I have no doubt that the development of such a system is possible and even would be an art of engineering, but I doubt the quality of data collected this way. A large number of optionally filled data is not conducive to the automation of their processing. What's more, authors of FRBR warn that this is just a general concept that should be refined for specific applications.
Dear Athina,
thank you for the answer. There is also PRESSoo - specialization of object oriented FRBR for serials.
Dear Anna,
as for computer researcher that specialize in databases and knowledge bases I have to admit that FRBR was a state of the art in the time it was proposed. At that time the possibility to implement ER based concept into working database was in fact limited. However, there are new possibilities now and this concept can and should, in my honest opinion, be implemented. The concepts introduced by FRBR can be expressed by e.g. OWL, AODB or one of schema less databases.
On the other hand there is a risk of eruption of large number of attributes that would have to be filled or could be filled. This can be avoided by appropriate approach towards mapping of existing systems into proposed solutions. You can also assign specific classes to attributes to give them priorities in processing. Options are limitless and if needed I would be happy to elaborate :)
Dear Krystian,
I agree all you've wrote about capabilities and functionalities of modern database technologies. But these are only for implementation of concepts that formerly should have been developed and widely agreed. But unfortunately this has not happened yet. It's a bit like the behavior expressed in the attached drawing [source: https://xkcd.com], do not you think?
Dear Anna,
it is hard not to agree with that :)
However, the question is if you want to develop new standard or simply find solution that would cover your needs? In my previous answer I have addressed FRBR. As for me it was designed in a way that with proper implementation would easily win with others.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs