Â鶹¹ÙÍøÊ×Ò³Èë¿Ú News Algorithmic Sorter
Post categories: prototypes
| 13:18 UK time, Tuesday, 24 June 2008
Â鶹¹ÙÍøÊ×Ò³Èë¿Ú News Algorithmic Sorter an attempt to try and work out what the British public are finding important. The main Â鶹¹ÙÍøÊ×Ò³Èë¿Ú News website offers a glimpse of what’s popular, but as with all things that’s limited to the audience of the Â鶹¹ÙÍøÊ×Ò³Èë¿Ú. While this gives a somewhat true to form view of what people are interested in, I wanted to expand it, and thus came up with the BNAS site. The main site isn’t that impressive as the main focus was on the backend.
The script uses several external APIs from blogging communities, search engines and social networking sites to work out what people are talking about. Then using the last 50 Data from the Â鶹¹ÙÍøÊ×Ò³Èë¿Ú it compiles a list in order of the interest the British public has on the subject. At the time of writing this Lock keepers' cottage sale halted is the least interesting story, and Oil price up despite Saudi pledge the most. It’s not 100%, and sometimes odd results show. The system works like a automated Digg.com.
There is no human interaction or traffic statistics used. Ideally the system could be expanded into a full blown news site taking stories from a wide range of sites, but the lack of open APIs from other news sources inhibit that. I’m working on improving the visual data that the site could provide, and the first bit of that is the Balance Chart.
Using the same data from the Algo sorted it works out if the public are more interested in a select few events, or generally more spread out. For example if there was a major terrorist attack the pie chart would show a few very large segments, but on a normal day it tends to be pretty split up.
Comments