[prev in list] [next in list] [prev in thread] [next in thread] 

List:       odtug-sqlplus-l
Subject:    RE: Modelling Facebook
From:       "Michelle, Suzanne" <Suzanne.Michelle () nyct ! com>
Date:       2011-03-08 15:19:40
Message-ID: 0B9952A150DC7F4D9341F981BDD9B41F3D5305 () NYCTEXVS06 ! transit ! nyct ! com
[Download RAW message or body]

One-room schoolhouse (pre-1900s) = self-contained classroom (mid-70's?) ...
(don't know what it's called now ...)

Suzanne (@ 2Bwy A13.32)
desk: 646-252-8663, cell: 347-907-1125

-----Original Message-----
From: odtug-sqlplus-l-bounce@fatcity.com [mailto:odtug-sqlplus-l-bounce@fatcity.com] \
                On Behalf Of John Flack
Sent: Tue, Mar 08, 2011 09:27
To: ODTUG-SQLPLUS-L@fatcity.com
Subject: RE: Modelling Facebook

Ohhhh, now I get it!  Not a hierarchical database, but a network database like \
developed by CODASYL and implemented in systems like Honeywell's IDS/2, and \
Cullinet's IDMS (now owned by CA).  From what I've heard, I'm awfully glad I've never \
used one of these.  System 2000 may have lacked flexibility, but at least it was \
sane.

Funny how people with no knowledge of history, keep re-inventing things, give them a \
new name and think they've invented something wonderful and new.

-----Original Message-----
From: odtug-sqlplus-l-bounce@fatcity.com [mailto:odtug-sqlplus-l-bounce@fatcity.com] \
                On Behalf Of David Wendelken
Sent: Monday, March 07, 2011 7:07 PM
To: ODTUG-SQLPLUS-L@fatcity.com
Subject: RE: Modelling Facebook

If you like the kind of thinking that brought us the XSLT syntax, you'll
love this approach.

Otherwise, like any sane person, run like hell away from it.


-----Original Message-----
From: odtug-sqlplus-l-bounce@fatcity.com
[mailto:odtug-sqlplus-l-bounce@fatcity.com] On Behalf Of Kent Graziano
Sent: Monday, March 07, 2011 2:44 PM
To: ODTUG-SQLPLUS-L@fatcity.com
Subject: Re: Modelling Facebook

This is pretty much how we would do it using the data vault modeling technique for a \
data warehouse (see www.danlisntedt.com or look for a presentation on DV in the ODTUG \
archives online).  
People = a hub
People Relationships = a link (i.e., m:m associative table)
 
Names, address, etc would all be in a Satellite off of the People Hub, the \
relationship "name" would be in a satellite table off the relation.  
I asked Dan about this issue (since I know absolutely NOTHING about graph databases). \
He said "the Data Vault model IS A GRAPH REPRESENTATION in a RELATIONAL DATABASE." \
(Hope that makes sense to some of you).  
Anyway take a look at the Wikipedia entry on graph databases \
(http://en.wikipedia.org/wiki/Graph_database):  
"A graph database is a kind of NoSQL database that uses graph structures with nodes, \
edges, and properties to represent and store information. General graph databases \
that can store any graph are distinct from specialized graph databases such as \
triplestores and network databases.  
Graph databases are based on graph theory. Graph databases employ nodes, properties, \
and edges. Nodes are very similar in nature to the objects that object-oriented \
programmers will be familiar with.  
Nodes represent entities such as people, businesses, accounts, or any other item you \
might want to keep track of.  
Properties are pertinent information that relate to nodes. For instance, if \
"Wikipedia" were one of our nodes, we might have it tied to properties such as \
"website", "reference material", or "word that starts with the letter 'w',"  \
depending on which aspects of "Wikipedia" are pertinent to our particular database.  
Edges are the lines that connect nodes to nodes or nodes to properties and they \
represent the relationship between the two. Most of the important information is \
really stored in the edges. Meaningful patterns emerge when one examines the \
connections and interconnections of nodes, properties, and edges.[1]"  
Notice the mention of triplestores? Those are implemented using semantic \
technologies, RDF, and OWL for which Oracle does have a solution in the RDBMS!  \
(David Hay presented on this at ODTUG a number of years back). The demonstrations I \
have seen (even recently) of this approach clearly can answer the question (and then \
some) quite elegantly. Here is an OOW 2006 paper on the  10g version: 
http://www.oracle.com/technetwork/database/options/semantic-tech/oow2006-semantics-061128-130119.pdf
  
There is also a modeling technique called Anchor Modeling that is similar to Data \
Vault that also provides an elegant way to model this data effectively in a \
relational database.  
Kent Graziano
Senior BI/DW Consultant
TrueBridge Resources at
MD Anderson Canter Center, DMAS
PPlease consider the environment before printing this email.


 


[prev in list] [next in list] [prev in thread] [next in thread]