Filtering content
using click data
Data: 1 Jan 2013 - 25 Feb 2015 clicking within the TRIP medical database (799,480 clicks between 296,697 articles)
General discussion of approach at http://www.pontneo.com/trip/self-organising-data.html
Click-through filter described at http://www.slideshare.net/pontneo/click-through-filter
Initial results analysis at http://www.slideshare.net/pontneo/click-through-filterprototyperesultsv2
Click-through filter tests: Search Results
Results reordered and broadened by click traffic, resulting in time-varying, adaptive returns. Here, using a simple search on titles. Change settings to control output, noise and response over time.

More detail here >>
Recommendations
Personalised, time-dependent content based on a user's history and the material related to it by click traffic.
Popular Items
The most visited items over a given period and at a given time for different components of click traffic.

Settings:

Influence100 %
Sensitivity86 %
Detail58 %
Date2015-02-25
Traffic

Search for:

= reordered text match
= click-related non-text match

Output:

@ CTR dropoff (Google uses 1.55)
Improvement: 0 %
Click Density: 0 %
Data Period: 0 days
Traffic Density: 0 % per day



Enter a search query to continue ...