Broad coverage, high quality parsers are available for only a handful of languages. A prerequisite for developing broad coverage parsers for more languages is the annotation of text with the desired linguistic representations (also known as "treebanking"). However, syntactic annotation is a labor intensive and time-consuming process, and it is difficult to find linguistically annotated text in sufficient quantities. In this article, we explore using parallel text to help solving the problem of creating syntactic annotation in more languages. The central idea is to annotate the English side of a parallel corpus, project the analysis to the second language, and then train a stochastic analyzer on the resulting noisy annotations. We discuss our background assumptions, describe an initial study on the "projectability" of syntactic relations, and then present two experiments in which stochastic parsers are developed with minimal human intervention via projection from English.

No abstract

Scanning books, magazines, and newspapers has become a widespread activity because people believe that much of the worlds information still resides off-line. In general after works are scanned they are indexed for search and processed to add links. This paper describes a new approach to automatically add links by mining popularly quoted passages. Our technique connects elements that are semantically rich, so strong relations are made. Moreover, link targets point within a work, facilitating navigation. This paper makes three contributions. We describe a scalable algorithm for mining repeated word sequences from extremely large text corpora. Second, we present techniques that filter and rank the repeated sequences for quotations. Third, we present a new user interface for navigating across and within works in the collection using quotation links. Our system has been run on a digital library of over 1 million books and has been used by thousands of people.

In this paper, we take a pattern recognition approach to correcting errors in text generated from printed documents using optical character recognition (OCR). We apply a very general, theoretically optimal model to the problem of OCR word correction, introduce practical methods for parameter estimation, and evaluate performance on real data. General TermsOCR error modeling and correction

In this paper, we introduce a generative probabilistic optical character recognition (OCR) model that describes an end-to-end process in the noisy channel framework, progressing from generation of true text through its transformation into the noisy output of an OCR system. The model is designed for use in error correction, with a focus on post-processing the output of black-box OCR systems in order to make it more useful for NLP tasks. We present an implementation of the model based on finitestate models, demonstrate the model's ability to significantly reduce character and word error rate, and provide evaluation results involving automatic extraction of translation lexicons from printed text.