Comments on: Idea #15: Beltway Analytics

By: regles tournoi poker

regles tournoi poker — Thu, 12 Jun 2008 13:54:54 +0000

cricket free ringtones…

Similarly slots online poker stars net casino gratuites sans depot ganar dinero real online cingular free real ringtones…

By: Michael Krenz

Michael Krenz — Thu, 13 Jul 2006 09:29:56 +0000

Hey, this reminds me of a site with a slightly similar approach in the UK (maybe the neural networks stuff is not as sophisticated as in your idea, but still): http://www.theyworkforyou.com/. There's a blog post on sitepoint.com with some background info and furter links. "... a web app that allows UK citizens to view their local politicians performance as well as providing an interface for searching debates that have taken place in the House of Commons (the UK’s main political “forum”). It works by re-packaging information from other online sources, such as Hansard (the transcript of all debates in Parliament) into a form real people might actually be interested in."

By: saffron

saffron — Fri, 07 Jul 2006 22:18:04 +0000

How about Underbelt Analytics?
Simply tracking speeches of polititians in time ( bankers, CEOs etc. ) about never ending promises we hear everyday, that just do not turn out the way they were promised to.

Example please?:
Gerhard Schröder’s view of unemployment in Germany was in 1999 ( or so :-) during his campaign ( “If I do not cut the uneployment I do not deserve to be reelected…” ), compared to the sobering reality of his achievements and later campaigns for the same things – just like the Spiegel cover earlier this year…

Showing what strong words were used when… …just reflecting public opinion of the moment.

Showing that most of the politics is just averaging opinions and not really having an opinion about anything.

Showing the determination before the big day vs. wishy-washy excuses few years ( days? ) later…

Showing who is copying who’s speaches…

Showing who is consistent and who not over larger period of time than a discussion on TV…

Showing who is falling back on proven phrases…

Showing who has read too many books…

Maybe you could negotiate to buy the http://www.weathercock.com domain… :-)

Collecting from respected sources, newspapers, public corpora, so that everybody can verify that between the words of 1990 and 2000 are galaxies in distance. Transforming speech to text and having TV card tuned into all major news networks following everything -they- say.
I guess your work on masivelly parallel processing comes in handy here!

Pros:
- you can always sell to the other party
- you can sell exclusive rights just for a certain campaign
- you can reaaaaally personalize this! the victim will love you! ( pity billg is leaving us soon, we might not be able to collect all the material retrospectively )
- you are not doing anything illegal by repeating their own words in seemingly random order in time…. :-)
- storage and bandwidth is getting cheaper day by day ( there will be a lot to collect )
- useful for any nation/language
- you can follow any given topic
- worst case you do not sell anything, you will be bought to shut up
- you will be helping to make the world a better place

Cons:
( in cons order ASC )
- you might tend to read abc of democracy periodically searching for hidden meanings you might have missed when in college
- you might never take part in elections again
- you might get even more depressive
- you might get shot

If I had your skills and knowledge I would have started yeaterday. Unfortunately I am quite tired when I get home in the evening since I have to work on a CRUD app most of the day…
But… …if you get this going, I promise to do the usability testing for you! If I survive that, I will apply for the position of the sales lead.

Good luck!

By: Evgeny Shadchnev

Evgeny Shadchnev — Fri, 07 Jul 2006 14:54:26 +0000

The idea rocks. It’s really good but very hard to implement, I believe. I agree with anon, you must be extremely good to make a public beta in half a year. However, chances are this site will be popular.
By the way, you forgot one more advantage. It will be relatively easy to make a similar site for other countries, which means that you’ll be able to multiply the number of visitors several times at almost no cost.

By: anon

anon — Fri, 07 Jul 2006 13:05:22 +0000

Great idea, but it does sound like a tough cookie. Parsing text and trying to get the computer to make intelligent sense out of it is no easy feat! You must be _really_ good to expect v1.0 to be done in 6 months! :)

Assuming you can pull it off, how do you plan to rise above the copycats? If you are thinking “this is so difficult that not very many people cant do it”, think again. I agree it is perhaps more difficult than copying a blogsearch website, but still… If your only source of income is pageviews and ads, I think your idea analysis should address the copycat issue as well.

Good luck.

By: Dilip

Dilip — Fri, 07 Jul 2006 12:16:07 +0000

Apart from just political stuff,
If you could come with a kind of frame work, where you just point this framework to data source and some mapping information (to get some sense out of data).

then this analytics could be applied to any thing. But the idea is wonderful, it will certainly empower citzens.

By: Phil

Phil — Fri, 07 Jul 2006 10:27:45 +0000

Great idea! Back in the day there used to be Yahoo! Politics, where you could find every congressperson and see every bill they voted for and which direction, what their salary was, hometown and all kinds of other statistics in a very detailed way like their Finance site. Congress’s own site is terrible for finding this information. The Washington Post is the only site that does anything vaguely similar these days.

By: Nico

Nico — Fri, 07 Jul 2006 08:40:28 +0000

I like this idea a lot.