In this article we present a syntax-based translation system, called TABL (Translation using Alignment-Based Learning). It translates natural language sentences by mapping grammar rules (which are induced by the Alignment-Based Learning grammatical inference framework) of the source language to those of the target language. By parsing a sentence in the source language, the grammar rules in the ...

In this paper we present our winning system in the WMT16 Shared Task on CrossLingual Pronoun Prediction, where the objective is to predict a missing target language pronoun based on the target and source sentences. Our system is a deep recurrent neural network, which reads both the source language and target language context with a softmax layer making the final prediction. Our system achieves ...

This paper proposes an unsupervised word segmentation algorithm that identifies word boundaries in continuous source language text in order to improve the translation quality of statistical machine translation (SMT) approaches. The method can be applied to any language pair where the source language is unsegmented and the target language segmentation is known. First, an iterative bootstrap meth...

This paper explores whether an individual’s news source can explain their attitudes on immigration. We focus on the Spanish-speaking population in the U.S., since they have the option of accessing their news in English, Spanish or in both languages. Our audience influence hypothesis predicts that Spanish-language news will cover immigration in a more positive and informative manner than will En...

This paper presents a weak defense of rationalist approaches to Machine Translation (MT) based on the following line of reasoning. There is, independent of non-linguistic context, a one to many relationship between a source language text and the possible high quality target language translations of the text. This is due to the fact that there is a one to many relationship between a source langu...

T-Clipse is an Integrated Development Environment (IDE) using Eclipse developed for TwoLevel Grammar (TLG), a high-level formal specification language. In our research, given a programming language, TLG is used to specify the syntax and semantics of the language to generate an interpreter for that language. This IDE provides a user-friendly environment for editing and navigation of TLG source c...

We present a method we used for the quality estimation shared task of WMT 2012 involving IBM1 and language model scores calculated on morphemes and POS tags. The IBM1 scores calculated on morphemes and POS-4grams of the source sentence and obtained translation output are shown to be competitive with the classic evaluation metrics for ranking of translation systems. Since these scores do not req...

Model-driven engineering (MDE) can be used both to create new software systems and to evolve or modernize existing software systems. In model-driven software evolution, models are extracted from the system and then MDE techniques are applied to make it evolve. Most evolution scenarios involve dealing with existing source code written in some programming languages. Therefore, a bridge from gramm...

This paper describes a novel approach to handling translation divergences in a Generation-Heavy Hybrid Machine Translation (GHMT) system. The translation divergence problem is usually reserved for Transfer and Interlingual MT because it requires a large combination of complex lexical and structural mappings. A major requirement of these approaches is the accessibility of large amounts of explic...