Section: Highlights of the Year
Highlights of the Year
We have published two papers at NIPS  ,  , the leading conference in machine learning. The first paper presents novel results on large-scale learning with higher-order risk functionals, which has applications in link prediction, graph inference and metric learning (among others). The second paper proposes new gossip algorithms for decentralized estimation of pairwise statistics in networks.
We have published a paper at AAAI  , one of the top conferences in Artificial Intelligence. The contribution is a new structured model for learning anaphoricity detection and coreference resolution, which achieved the best score to date on the popular CoNLL benchmark with gold mentions.
We have published a paper at EMNLP  , a leading conference in Natural Language Processing. The work presents a detailed comparative framework for assessing the usefulness of popular unsupervised word representations for identifying so-called implicit discourse relations.