D
Douwe Kiela
Researcher at Facebook
Publications - 153
Citations - 11846
Douwe Kiela is an academic researcher from Facebook. The author has contributed to research in topics: Natural language & Computer science. The author has an hindex of 45, co-authored 152 publications receiving 8464 citations. Previous affiliations of Douwe Kiela include University of Cambridge & Katholieke Universiteit Leuven.
Papers
More filters
I love your chain mail! Making knights smile in a fantasy game world
Shrimai Prabhumoye,Margaret Li,Jack Urbanek,Emily Dinan,Douwe Kiela,Jason Weston,Arthur Szlam +6 more
TL;DR: A goal-oriented model with reinforcement learning against an imitation-learned ``chit-chat'' model with two approaches is trained and it is shown that both models outperform an inverse model baseline and can converse naturally with their dialogue partner in order to achieve goals.
Posted Content
Learning Visually Grounded Sentence Representations
TL;DR: The authors introduce a variety of models, trained on a supervised image captioning corpus to predict the image features for a given caption, to perform sentence representation grounding, and train a grounded sentence encoder that achieves good performance on COCO caption and image retrieval.
Posted Content
Improving Question Answering Model Robustness with Synthetic Adversarial Data Generation
TL;DR: In this paper, the authors investigate several answer selection, question generation, and filtering methods that form a synthetic adversarial data generation pipeline that takes human-generated adversarial samples and unannotated text to create synthetic question-answer pairs.
Concreteness and Corpora: A Theoretical and Practical Study
TL;DR: The extent to which concreteness is reflected in the distributional patterns in corpora is explored and the quality of the representations of abstract words in LDA models can be improved by supplementing the training data with information on the physical properties of concrete concepts.
Proceedings Article
Robust Text Classification for Sparsely Labelled Data Using Multi-level Embeddings
TL;DR: This work presents a novel approach for learning embedded features that jointly learns embeddings at different levels of granularity (word, sentence and document) along with the class labels to help classify sparsely labelled data.