Comments on: PhD proposal on distributed graph data bases

By: Aurelius Titan graph enables realtime querying with 2400 concurrent users on a distributed graph database!

Wed, 21 Aug 2013 10:43:08 +0000

[…] a blog by a Webscience PHD student PhD Thesis […]

By: Aurelius Titan graph enables realtime querying with 2400 concurrent users on a distributed graph database!

Wed, 21 Aug 2013 10:43:08 +0000

[…] a blog by a Webscience PHD student PhD Thesis […]

By: Gerd

Gerd — Sat, 31 Mar 2012 13:59:35 +0000

Hi René,
your proposal is quite challenging and it is
probably to much for one thesis.
In general, it is no problem if you target in your proposal a rather
broad or general topic like distributed graph databases,
and you will narrow your topic while working on your thesis.
Details:
Introduction:
I also like the article of Jennifer Widom, especially for the introduction it is a valuable guideline.
Thus, I would recommend to consider this structure
(five points of an intro) also in your introduction.
(Currently, is is only a use case description.)
Related Work:
You cover a large field of related work, which is important for a
PhD thesis (and also for your research).
Section 3 (Formalization)
I consider this as an important part of a proposal since this should
motivate your work, i.e., what is the problem?,
what are the limitations of current approaches/state of the art?
What is the gap you want to close with your work?
As you mentioned, this is hard, but quite important and I have the
feeling this should be further elaborated.
(This is probably quite difficult at the moment since your topic is quite broad.)
Sections 4, 5 and 6 (block: graph algebra, CRUD, hybrid DS)
These three sections could be a thesis on their own.
Regarding the content of these paragraphs,
I agree with your problem statements and your research plan.
(Personally, I like this (sub-) topic a lot.)
I was wondering whether the structure could be like this:
(1) analysis of basic CRUD operations
(2) hybrid data structure (based on results of the analysis)
(3) algebra and query language for a hybrid structure?
Sections 7,8 and 9 (block: ACID)
Section 7 reads like “making experiments and then recognize
research questions”.
I would suggest to be more precise and systematic, and formalize the
particular problems: (1) ACID in general, (2) ACID in distributed systems,
(3) ACID in graphs (graph data bases).
Afterwards, the experiments are more focused and it might
be easier to come up with the best data structure (in Sect. 8).
Regards,
Gerd

By: Gerd

Gerd — Sat, 31 Mar 2012 13:59:35 +0000

By: Daniel

Daniel — Wed, 28 Mar 2012 16:32:59 +0000

Hi René, we talked about JGraLab this morning and you asked me to post the link here: https://github.com/jgralab Some more informations can be found on the official homepage: I'm sure, it is interesting for everybody working with graphs. Regards, Daniel

By: Daniel

Daniel — Wed, 28 Mar 2012 16:32:59 +0000

By: PhD proposal on distributed graph data bases « Another Word For It

PhD proposal on distributed graph data bases « Another Word For It — Wed, 28 Mar 2012 00:16:57 +0000

[…] PhD proposal on distributed graph data bases by René Pickhardt. […]

By: PhD proposal on distributed graph data bases « Another Word For It

PhD proposal on distributed graph data bases « Another Word For It — Wed, 28 Mar 2012 00:16:57 +0000

[…] PhD proposal on distributed graph data bases by René Pickhardt. […]

By: Max De Marzi

Max De Marzi — Tue, 27 Mar 2012 17:10:23 +0000

René,
Sounds awesome, but super ambitious. If you choose to build a Gremlin compatible system, it removes a bunch of work, makes it easier to benchmark against others, and lets you focus on the underlying system. It would also make it easier to gain adoption if you were to turn this into a real product at some point.
Regards,
Max

By: Max De Marzi

Max De Marzi — Tue, 27 Mar 2012 17:10:23 +0000