We provide a unified account of sentence-level and text-level anaphora within the framework of a dependency-based grammar model. Criteria for anaphora resolution within sentence boundaries rephrase major concepts from GB's binding theory, while those for text-level anaphora incorporate an adapted version of a Grosz-Sidner-style focus model.

This paper proposes subjLDA for sentence-level subjectivity detection by modifying the latent Dirichlet allocation (LDA) model through adding an additional layer to model sentence-level subjectivity labels. A variant, called joint-subjLDA, has also been described. The model inference and parameter estimation algorithms, and Gibbs sampling procedure are presented.

The computing cost of many NLP tasks increases faster than linearly with the length of the representation of a sentence. For parsing the representation is tokens, while for operations on syntax and semantics it will be more complex. In this paper we propose a new task of sentence chunking: splitting sentence representations into coherent substructures. Its aim is to make further processing of l...

This study investigates the neural substrates underlying the perception of two sentence-level prosodic phenomena in Mandarin Chinese: contrastive stress (initial vs. final emphasis position) and intonation (declarative vs. interrogative modality). In an fMRI experiment, Chinese and English listeners were asked to selectively attend to either stress or intonation in paired 3-word sentences, and ...

This paperintroduces a simple mixtare language model that attempts to capture long distance conslraints in a sentence or paragraph. The model is an m-component mixture of Irigram models. The models were constructed using a 5K vocabulary and trained using a 76 million word Wail Street Journal text corpus. Using the BU recognition system, experiments show a 7% improvement in recognition accuracy ...

Lipreading is the task of decoding text from the movement of a speaker’s mouth. Traditional approaches separated the problem into two stages: designing or learning visual features, and prediction. More recent deep lipreading approaches are end-to-end trainable (Wand et al., 2016; Chung & Zisserman, 2016a). However, existing work on models trained end-to-end perform only word classification, rat...

-Sentiment analysis addresses the computational of feeling, sentiments and subjectivity in content, has reached an impressive consideration in recent years. Rather than the customary coarse-grained sentiment analysis tasks, for example document-level sentiment analysis. Aspect-oriented opinion mining aims to identify product aspects (features of products) about which opinion has been expressed ...