SHEF-MIME: Word-level Quality Estimation Using Imitation Learning
By Daniel Beck, Andreas Vlachos, Gustavo Paetzold, Lucia Specia. 2016
4 citations recorded by the source corpus.
We describe University of Sheffield’s sub-mission to the word-level Quality Estimation shared task. Our system is based on imitation learning, an approach to structured prediction which relies on a classifier trained on data generated appropriately to ameliorate error propagation. Compared to other structure prediction approaches such as conditional random fields, it allows the use of arbitrary information from previous tag predictions and the use of non-decomposable loss functions over the structure. We explore these two aspects in our submission while using the baseline features provided by the shared task organ-isers. Our system outperformed the conditional random field baseline while using the same feature set.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Varvara Logacheva
- Mrinmaya Sachan
- Eric P. Xing
- Frédéric Blain
- Lucia Specia
- Mohit Bansal
- Jianfeng Gao
- Graham Neubig
In Computer Vision and Pattern Recognition
- Mrinmaya Sachan
- Eric P. Xing
- Varvara Logacheva
- Lucia Specia
- Frédéric Blain
- Mohit Bansal
- Luc Van Gool
- Ming-Hsuan Yang
Citations in the corpus
- Learning to Solve Geometry Problems from Natural Language Demonstrations in Textbooks (2017, methodology): “Imitation learning is increasingly being used in NLP (Vlachos and Clark, 2014; Berant and Liang, 2015; Augenstein et al., 2015; Beck et al., 2016; Goodman et al., 2016a,b).”
- USFD’s Phrase-level Quality Estimation Systems (2016, methodology): “The word-level prediction features are produced by the SHEF-MIME QE sys-tem (Beck et al., 2016), which uses imitation learning to predict translation quality at the word level.”