Discourse & Dialogue
Towards Robust Multimodal Sentiment Analysis with Incomplete Data
Recognizing that the language modality typically contains dense sentiment information, we consider it as the dominant modality and present an innovative Language-dominated Noise-resistant Learning Network (LNLN) to achieve robust MSA.
Approximating Interactive Human Evaluation with Self-Play for Open-Domain Dialog Systems
Building an open-domain conversational agent is a challenging problem.
A Reduction for Efficient LDA Topic Reconstruction
Matteo Almanza, Flavio Chierichetti, Alessandro Panconesi, Andrea Vattani
Neural Information Processing Systems http://nips.cc/
d921c3c762b1522c475ac8fc0811bb0f-AuthorFeedback.pdf
We wish to thank all of the reviewers for their time and thorough reading of our paper! We appreciate the reviewer's suggestions regarding clarity. We have added the suggested summary sentence "the key We started with binary sentiment classification, but are actively working on more tasks. RNN hidden states onto the top two PCs for two different input sequences that differ only by two tokens (replacing ' The trajectories start out the same as the initial tokens are identical. We have added a footnote noting this in the main text.
Multilingual Anchoring: Interactive Topic Modeling and Alignment Across Languages
Michelle Yuan, Benjamin Van Durme, Jordan L. Ying
Neural Information Processing Systems http://nips.cc/
Precision-Recall Balanced Topic Modelling
Neural Information Processing Systems http://nips.cc/
