Distributed Submodular Maximization for Large Vocabulary Continuous Speech Recognition | Publicación