Why Analytical Applications Fail(juiceanalytics.com)
juiceanalytics.com
Why Analytical Applications Fail
http://www.juiceanalytics.com/writing/why-analytical-applications-fail/
3 comments
This has been referenced before on News.YC; this is an interesting paper about the design and implementation of ITA's optimizer.
"Computational Complexity of Air Travel Planning" by Carl de Marcken (2003) http://www.demarcken.org/carl/papers/ITA-software-travel-com...
"Computational Complexity of Air Travel Planning" by Carl de Marcken (2003) http://www.demarcken.org/carl/papers/ITA-software-travel-com...
Cool!
Its a very detailed analysis of analytical sites. Examples shown serve different fields - travel and web traffic. And even usability of analytical sites is broken down by time!
That was a great read for. I am developing a set of news analysis tools and this post helps.
But I am still wondering what could be done in my case?
I have a totally varied set of tools. Some of them show graphs, some maps and some text data. I have separated the mashup into sub-tools. But now after reading this post, it makes me think about presenting the overview of all the types of results on the overview page like the Google Analytics dashboard. Then, I still have to think about users with slow internet connections, since the page load time might increase (and there might be time out errors from my servers resulting in half-loaded pages).
Its a very detailed analysis of analytical sites. Examples shown serve different fields - travel and web traffic. And even usability of analytical sites is broken down by time!
That was a great read for. I am developing a set of news analysis tools and this post helps.
But I am still wondering what could be done in my case?
I have a totally varied set of tools. Some of them show graphs, some maps and some text data. I have separated the mashup into sub-tools. But now after reading this post, it makes me think about presenting the overview of all the types of results on the overview page like the Google Analytics dashboard. Then, I still have to think about users with slow internet connections, since the page load time might increase (and there might be time out errors from my servers resulting in half-loaded pages).
I think the target user for sites like travelocity is probably fairly inflexible in their preferences. They want higher income, price insensitive business people who specifically want from point A to point B at time X. The sites and airlines don't really want to cater to College Kid Joe who's carefully weighing his options and playing around with them to save $150. Those people just don't drive revenue.
http://matrix.itasoftware.com/
It gives a color coded matrix of flights, and all the info is captured in one screen. I think orbitz and others use their optimizer as a backend.