Recent advances in the field of natural language processing were achieved with deep learning models. This led to a wide range of new research questions concerning the stability of such large-scale systems and their applicability beyond well-studied tasks and datasets, such as information extraction in non-standard domains and languages, in particular, in low-resource environments. In this work, we address these challenges and make important contributions across fields such as representation learning and transfer learning by proposing novel model architectures and training strategies to overcome existing limitations, including a lack of training resources, domain mismatches and language barriers. In particular, we propose solutions to close ...
© Dr Long DuongNatural language processing (NLP) aims, broadly speaking, to teach computers to under...
The current generation of neural network-based natural language processing models excels at learning...
Large-scale annotated datasets are an indispensable ingredient of modern Natural Language Processing...
Recent advances in the field of natural language processing were achieved with deep learning models....
Deep learning has achieved state-of-the-art performance on a wide range of tasks, including natural ...
This thesis explores information extraction (IE) in \textit{low-resource} conditions, in which the q...
Despite much success, the effectiveness of deep learning models largely relies on the availability o...
Stochastic natural language generation systems that are trained from labelled datasets are often dom...
[Abstract] The recent trend toward the application of deep structured techniques has revealed the l...
Recent approaches based on end-to-end deep neural networks have revolutionised Natural Language Proc...
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Electrical Engineering and Comp...
Information extraction (IE) plays a significant role in automating the knowledge acquisition process...
Supervised deep learning-based approaches have been applied to task-oriented dialog and have proven ...
In Machine Learning, a good model is one that generalizes from training data and makes accurate pred...
This paper investigates very low resource language model pretraining, when less than 100 thousand se...
© Dr Long DuongNatural language processing (NLP) aims, broadly speaking, to teach computers to under...
The current generation of neural network-based natural language processing models excels at learning...
Large-scale annotated datasets are an indispensable ingredient of modern Natural Language Processing...
Recent advances in the field of natural language processing were achieved with deep learning models....
Deep learning has achieved state-of-the-art performance on a wide range of tasks, including natural ...
This thesis explores information extraction (IE) in \textit{low-resource} conditions, in which the q...
Despite much success, the effectiveness of deep learning models largely relies on the availability o...
Stochastic natural language generation systems that are trained from labelled datasets are often dom...
[Abstract] The recent trend toward the application of deep structured techniques has revealed the l...
Recent approaches based on end-to-end deep neural networks have revolutionised Natural Language Proc...
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Electrical Engineering and Comp...
Information extraction (IE) plays a significant role in automating the knowledge acquisition process...
Supervised deep learning-based approaches have been applied to task-oriented dialog and have proven ...
In Machine Learning, a good model is one that generalizes from training data and makes accurate pred...
This paper investigates very low resource language model pretraining, when less than 100 thousand se...
© Dr Long DuongNatural language processing (NLP) aims, broadly speaking, to teach computers to under...
The current generation of neural network-based natural language processing models excels at learning...
Large-scale annotated datasets are an indispensable ingredient of modern Natural Language Processing...