Fréderic Godin
- ORCID iD
-
0000-0003-2465-8379
Show
Sort by
-
- Conference Paper
- C1
- open access
Zero-shot cross-lingual sentiment classification under distribution shift : an exploratory study
-
- Conference Paper
- C1
- open access
IDAS : intent discovery with abstractive summarization
-
Robustifying sentiment classification by maximally exploiting few counterfactuals
-
Learning when not to answer : a ternary reward structure for reinforcement learning based question answering
-
Verbetering en interpretatie van neurale netwerken voor voorspellingsopdrachten op woordniveau in natuurlijketaalverwerking
(2019) -
- Conference Paper
- C1
- open access
Predefined sparseness in recurrent sequence model
-
SpliceRover : interpretable convolutional neural networks for improved splice site prediction
-
- Conference Paper
- C1
- open access
Explaining character-aware neural networks for word-level prediction : do they discover linguistic rules?
(2018) p.3275-3284 -
Dual Rectified Linear Units (DReLUs) : a replacement for tanh activation functions in quasi-recurrent neural networks
-
- Conference Paper
- C1
- open access
Improving language modeling using densely connected recurrent neural networks
(2017) p.186-190 -
A distance-based approach for semantic dissimilarity in knowledge graphs
-
Normalized semantic web distance
-
Multimedia lab @ ACL W-NUT NER sharedtask: named entity recognition for Twitter microposts using distributed word representations
-
Character-level modelling of noisy microposts
-
How can Twitter data and viewing rates enrich one another? A conceptual and empirical comparison between viewing rates and Twitter data for TV audience research