The aim of this paper is to conduct a comparative study between liberal and prescriptive cultures (in this case Arab and British audiences) to extract the views of Web designers i...
Entity detection and tracking is a relatively new addition to the repertoire of natural language tasks. In this paper, we present a statistical language-independent framework for ...
Radu Florian, Hany Hassan, Abraham Ittycheriah, Ho...
We introduce factored language models (FLMs) and generalized parallel backoff (GPB). An FLM represents words as bundles of features (e.g., morphological classes, stems, data-drive...
We propose a comprehensive theory of codemixed discourse, encompassing equivalencepoint and insertional code-switching, palindromic constructions and lexical borrowing. The starti...
We present Promodes, an algorithm for unsupervised word decomposition, which is based on a probabilistic generative model. The model considers segment boundaries as hidden variable...