Proactive Information Retrieval by User Modeling from Eye Tracking thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Proactive Information Retrieval by User Modeling from Eye Tracking

Published on Feb 25, 20075359 Views

Related categories

Chapter list

Proactive Information Retrieval by User Modeling from Eye Tracking00:00
People can infer a lot from gaze00:14
Proactive user interface00:45
Gaze direction and target can be measured pt 101:25
Gaze direction and target can be measured pt 202:17
Pilot study: Inferring relevance from eye movements03:19
Experimental setup03:35
Pascal NoE Challenge04:36
There are other sources of implicit feedback as well pt 105:19
There are other sources of implicit feedback as well pt 206:06
Case study: Infer the relevance of titles of scientific articles06:40
Setting06:57
Eye movements07:27
Feature extraction07:57
Predicting relevance with Discriminative Hidden Markov Models08:38
Performance measures09:25
Results pt 109:53
Collaborative Filtering = Relevance out of others’ interests pt 110:28
Collaborative Filtering = Relevance out of others’ interests pt 210:32
Experimental setup pt 111:33
Experimental setup pt 211:48
User Rating Profile (URP) Model pt 111:56
User Rating Profile (URP) Model pt 212:18
Results pt 212:21
Combining collaborative filtering and eye movements13:10
Combining predictions13:14
Dirichlet Mixture Model pt 113:43
Dirichlet Mixture Model pt 214:30
Results pt 314:32
Eyeogle15:20
Conclusions16:25
The research consortium17:32