Rationalizing Neural Predictions

EMNLP, pp. 107-117, 2016.

Cited by: 329|Bibtex|Views413|Links
EI

Abstract:

Prediction without justification has limited applicability. As a remedy, we learn to extract pieces of input text as justifications -- rationales -- that are tailored to be short and coherent, yet sufficient for making the same prediction. Our approach combines two modular components, generator and encoder, which are trained to operate ...More

Code:

Data:

Your rating :
0

 

Tags
Comments