Re: finding a thesis topic

How about data access to and homogenous views over heterogeneous data sources, whether relational, hierarchical, network, object-oriented, etc? It seems to me that the greatest problem, at least for larger organizations, is the ability to integrate and access information with some sense of location (and system) independence.

I see middleware solutions for sloshing data through api's and complex interchange mechanisms these days, but the alternative of having integration happen at the data management layer seems to be getting short shrift (unless you think of XML in this regard). Are there new developments with federated database and mediators, multi-database, and true distributed database (true peer-to-peer collaboration) frameworks?

I have also once read that the ODMG ODL and OML definitions were being incorporated to some small extent into the SQL3 standard? Was this a mis-statement, or have there been efforts to extend and merge the two standards? It could be that I read material that was published before the big ORDBMS implementation push, which might have precluded the incorporation of pure object-oriented data definition and manipulation extensions into SQL-1999. These might be interesting topics also.

Dan

"karl wettin" <wettin_at_users.sourceforge.net> wrote in message news:20030302011131.08386c77.wettin_at_users.sourceforge.net...
> On 1 Mar 2003 15:21:55 -0800
> amalkaau_at_yahoo.com (Amal) wrote:
>
> > Will, am a master student and I have 2 start thinking about my thesis
> > topic, but the problem am really confused, really really confused.!! I
> > don't know what kind of topics are interesting to start with !!
> > My interest field is database ..but still there are a lot of topics 2
> > talk about under that field !!! so what am asking about , is helping
> > me 2 find a topic that r interesting and not only that but also a
> > topic which I will find a lot of resources talked about it. So I can
> > make extension to their work..:)
>
> The topic "databases" is a huge field.
>
> I for one would love a paper that descibed what database to use for what
> design. When do I use a herichial- instead of a relation- or objection
> oriented database, and why? Is there an algorithm I could apply my class
> diagram to find out what DBMS(s?) I should use? What do I win and loose by
> combining multiple datasources? What can I win by replacing my RDBMS with
> an OODBMS and navigate to any data? Can I navigate to any data?
>
> There is about ten thousand thesises I can think of right now. I think you
> need to narrow down what it is you like with databases. Unless you want to
> write a paper on how, when and why, you need to focus on heirichial-,
> relational- or object oriented databases. (Some people might argue.)
>
> > By the way ,,,,,I wanna ask u about "XML document" and what u think of
> > topics related to it.
>
> XML is, if possible, a larger topic than "databases". It's like writing a
> paper on "ASCII".
>
> You really need to be more specific.
>
>
> karl
Received on Fri Mar 07 2003 - 01:01:52 CET