gleamed from SOCNET:
Date: Mon, 6 Oct 2008 10:57:39 +0200
From: Vladimir Batagelj <vladimir.batagelj@FMF.UNI-LJ.SI>
Subject: Re: Amazon Books
***** To join INSNA, visit http://www.insna.org *****
<<<-------- aparna basu-------->>>
> > I searched for the keyword 'Magnesium' on the Amazon site and found more
> > than 50,000 titles, of which only 20 appear on the first page. I would
> > like to draw a network from this data and would appreciate some help from
> > anyone on this list who may have done this before.
To draw a network you have first to define it. Assume that the
50,000 titles are vertices, what are the links?
Some years ago I wrote a program in Python to get the network from
Amazon based on "Customers who bought" ...
http://vlado.fmf.uni-lj.si/vlado/podstat/AO/Amazon.zip
http://vlado.fmf.uni-lj.si/pub/networks/data/econ/amazon/amazon.htm
Since the program was interrupted several times I built in also
the 'restart from' option.
For small networks (some houndreds of vertices) I wrote, as an
'how to' illustration for students, a similar program in R.
The structure of Amazon's pages changed - some adaptations in both
programs are needed.
See also
http://mrvar.fdv.uni-lj.si/pub/mz/mz4.1/kejzar.pdf
Vlado
-- Vladimir Batagelj, University of Ljubljana, FMF, Department of Mathematics Jadranska 19, 1000 Ljubljana, Slovenia http://vlado.fmf.uni-lj.si