[Edu-sig] back on topic: apropos Go
kirby.urner at gmail.com
Sun Feb 13 00:41:53 CET 2011
Thx for this thread.
One of my colleagues here in Portland taught himself geek skills from
scratch and soon had his own buzzbot in the cloud, all written in Python
and backending into MySQL. Beautifulsoup.
Screen scraping, like skateboarding, is not a crime.
I've suggested on this list that we co-develop such a buzzbot as one of
the archetypal "learning projects". This involves registering as a dev
with a search engine company (one or more) and accessing its API.
There's some obvious division of labor in these kinds of apps.
You need your web crawlers tagging sites based on preliminary
tests, followed by data harvesters, various filters and cleaners.
Then you'll want your scoring / grading algorithms (another kind
of filtering), that applies more relevant tests, saves the rankings.
Having all these facilities in open source Python, in however many
versions, will continue to attract curious learners and clients for
their services. Patrick is getting pinged by Hollywood deal pitchers
looking for metrics in a mostly intuitive business.
On Sat, Feb 12, 2011 at 2:15 PM, Mark Engelberg <mark.engelberg at gmail.com>wrote:
> Getting back on topic, there's a good Python book about this sort of "data
> mining" approach to AI with lots of great projects that are perfect for
> advanced students:
> Edu-sig mailing list
> Edu-sig at python.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Edu-sig