[BangPypers] How to compare the relevancy between news headlines?
Anand Balachandran Pillai
abpillai at gmail.com
Tue Jun 14 14:14:59 CEST 2011
On Tue, Jun 14, 2011 at 5:16 PM, Gopalakrishnan Subramani <
gopalakrishnan.subramani at gmail.com> wrote:
> Dear All,
> Thanks for very informative answers. I would follow NLT & also try out
> Patra's advice as well. Post you my updates.
You can also do this by utilizing the entity extraction/categorization
online Linked-data/clustering providers. One such service is OpenCalais
provided by Thomson-Reuters.
I had played around with Calais one year back and found it very useful
to learn about Linked-data and entity extraction.
In fact, there is an already existing news clustering service written
on top of OpenCalais.
Klezio.com was also a very good example of this, but now it seems
to be down.
In fact I cooked up an application to pull out the cost of mobile
phones using Opencalais and demoed it in a BangPypers meeting
some time back, here is an archive of it.
> On Tue, Jun 14, 2011 at 4:55 PM, Gora Mohanty <gora at mimirtech.com> wrote:
> > On Tue, Jun 14, 2011 at 1:28 PM, Gopalakrishnan Subramani
> > <gopalakrishnan.subramani at gmail.com> wrote:
> > > Thanks for suggestion. Can you give me a specific NLT toolset/approach
> > with
> > > example if you have experience already?
> > Try clustering with Python NLTK ( http://www.nltk.org/ ).
> > NLTK is well documented, and what you are looking for
> > will probably be available from nltk.cluster. The site has
> > a free, downloadable book , and extensive documentation.
> > See, e.g.,
> > If you are willing to go away from Python, there are various
> > options, such as:
> > * Carrot 2: http://search.carrot2.org/stable/search
> > * Mahout: http://lucene.apache.org/mahout/
> > Regards,
> > Gora
> > _______________________________________________
> > BangPypers mailing list
> > BangPypers at python.org
> > http://mail.python.org/mailman/listinfo/bangpypers
> BangPypers mailing list
> BangPypers at python.org
More information about the BangPypers