Comments on: Graph database http://visualizationtools.net/default/graph-database/ Think. Visualize. Understand. Wed, 28 Dec 2011 13:39:14 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Ido Ran http://visualizationtools.net/default/graph-database/#comment-341 Ido Ran Thu, 24 Jun 2010 05:54:17 +0000 http://visualizationtools.net/default/?p=4159#comment-341 Hi, I am working for about a year now on a project that also has graph (network) in it's core and I also have to deal with persist the data to a database (RDBMS). As you said, OO DB like DB4O is not production-ready, but I did look into "document-oriented database" like CouchDB and MongoDB but I end up using regular RDBMS as back-end database and J2EE application as the front-end of the database. This give me the flexibility to "talk" to the client application in more "graph language" while still persist the data into tables at the end - but the client application is clean from any table knowledge. There is no chance connecting this database to a reporting application because the reporting application was "table oriented" and not "graph oriented" so it will be able to create very limited reports on it. If you give a molecular scientist a hamer - who far can it reach? Thank you for the post and I'm looking forward for more. Ido Hi, I am working for about a year now on a project that also has graph (network) in it’s core and I also have to deal with persist the data to a database (RDBMS).
As you said, OO DB like DB4O is not production-ready, but I did look into “document-oriented database” like CouchDB and MongoDB but I end up using regular RDBMS as back-end database and J2EE application as the front-end of the database. This give me the flexibility to “talk” to the client application in more “graph language” while still persist the data into tables at the end – but the client application is clean from any table knowledge.

There is no chance connecting this database to a reporting application because the reporting application was “table oriented” and not “graph oriented” so it will be able to create very limited reports on it.
If you give a molecular scientist a hamer – who far can it reach?

Thank you for the post and I’m looking forward for more.
Ido

]]> By: Andriy Khoruzhenko http://visualizationtools.net/default/graph-database/#comment-340 Andriy Khoruzhenko Mon, 07 Jun 2010 09:40:15 +0000 http://visualizationtools.net/default/?p=4159#comment-340 I recommend to use combination of two approaches. For data storing use RDMS with constrains and data integration and use XML data for visualization stuff. Both Oracle and SQL Server allows manipulate with XML documents stored in blobs and load partially. Such combination I've use in OCR management tool and get very good results and performance. I recommend to use combination of two approaches. For data storing use RDMS with constrains and data integration and use XML data for visualization stuff.

Both Oracle and SQL Server allows manipulate with XML documents stored in blobs and load partially.

Such combination I’ve use in OCR management tool and get very good results and performance.

]]>