Latent Dirichlet allocation is a fully generative statistical language model that has been proven to be successful in capturing both the content and the topics of a corpus of docum...
This paper presents a probabilistic model for sense disambiguation which chooses the best sense based on the conditional probability of sense paraphrases given a context. We use a...
In hierarchical phrase-based SMT systems, statistical models are integrated to guide the hierarchical rule selection for better translation performance. Previous work mainly focus...
Lei Cui, Dongdong Zhang, Mu Li, Ming Zhou, Tiejun ...
Abstract. We compare two image bases with respect to their capabilities for image modeling and steganalysis. The first basis consists of wavelets, the second is a Laplacian pyramid...
Valentin Schwamberger, Pham Hai Dang Le, Bernhard ...
In this paper, we present an unsupervised hybrid model which combines statistical, lexical, linguistic, contextual, and temporal features in a generic EMbased framework to harvest...