Abstract
Neural discourse models proposed so far are very sophisticated and tuned specifically to certain label sets. These are effective, but unwieldy to deploy or repurpose for different label sets or languages. Here, we propose a robust neural classifier for non-explicit discourse relations for both English and Chinese in CoNLL 2016 Shared Task datasets. Our model only requires word vectors and simple feed-forward training procedure, which we have previously shown to work better than some of the more sophisticated neural architecture such as long-short term memory model. Our Chinese model outperforms feature-based model and performs competitively against other teams. Our model obtains the state-of-the-art results on the English blind test set, which is used as the main criteria in this competition.