NLIZE: A Perturbation-Driven Visual Interrogation Tool for Analyzing and Interpreting Natural Language Inference Models.
Shusen LiuZhimin LiTao LiVivek SrikumarValerio PascucciPeer-Timo BremerPublished in: IEEE Trans. Vis. Comput. Graph. (2019)
Keyphrases
- natural language
- learning algorithm
- decision trees
- object recognition
- statistical inference
- random fields
- complex systems
- statistical models
- visual tasks
- data sets
- visual representations
- structured prediction
- visual perception
- computational model
- information extraction
- probabilistic model
- low level
- hidden markov models
- bayesian networks