MindMiner: A Mixed-Initiative Interface for Interactive Distance Metric Learning - Human-Computer Interaction – INTERACT 2015 Access content directly
Conference Papers Year : 2015

MindMiner: A Mixed-Initiative Interface for Interactive Distance Metric Learning

Xiangmin Fan
  • Function : Author
  • PersonId : 1018463
Youming Liu
  • Function : Author
  • PersonId : 1018464
Nan Cao
  • Function : Author
  • PersonId : 1018465
Jason Hong
  • Function : Author
  • PersonId : 1018466
Jingtao Wang
  • Function : Author
  • PersonId : 1018467

Abstract

We present MindMiner, a mixed-initiative interface for capturing subjective similarity measurements via a combination of new interaction techniques and machine learning algorithms. MindMiner collects qualitative, hard to express similarity measurements from users via active polling with uncertainty and example based visual constraint creation. MindMiner also formulates human prior knowledge into a set of inequalities and learns a quantitative similarity distance metric via convex optimization. In a 12-subject peer-review understanding task, we found MindMiner was easy to learn and use, and could capture users’ implicit knowledge about writing performance and cluster target entities into groups that match subjects’ mental models. We also found that MindMiner’s constraint suggestions and uncertainty polling functions could improve both efficiency and the quality of clustering.
Fichier principal
Vignette du fichier
346942_1_En_47_Chapter.pdf (852.42 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01599869 , version 1 (02-10-2017)

Licence

Attribution

Identifiers

Cite

Xiangmin Fan, Youming Liu, Nan Cao, Jason Hong, Jingtao Wang. MindMiner: A Mixed-Initiative Interface for Interactive Distance Metric Learning. 15th Human-Computer Interaction (INTERACT), Sep 2015, Bamberg, Germany. pp.611-628, ⟨10.1007/978-3-319-22668-2_47⟩. ⟨hal-01599869⟩
79 View
140 Download

Altmetric

Share

Gmail Facebook X LinkedIn More