This paper describes the application of discriminative reranking techniques to the problem of machine translation. For each sentence in the source language, we obtain from a basel...
Named Entity (NE) extraction is an important subtask of document processing such as information extraction and question answering. A typical method used for NE extraction of Japan...
In this paper, the automatic labeling of semantic roles in a sentence is considered as a chunking task. We define a semantic chunk as the sequence of words that fills a semantic...
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...
Leximancer is a software system for performing conceptual analysis of text data in a largely language independent manner. The system is modelled on Content Analysis and provides u...