Fabrizio Gotti

EBMT by tree-phrasing

This article presents an attempt to build a repository storing associations between simple syntactic dependency treelets in a source language and their corresponding phrases in a target language. We assess the usefulness of this resource in two different settings. First, we show that it improves upon a standard subsentential translation memory. Second, we observe improvements in translation quality when a standard statistical phrase-based translation engine is augmented with the ability to exploit such a repository.

Enhancing the Bilingual Concordancer TransSearch with Word-Level Alignment

Despite the impressive amount of recent studies devoted to improving the state of the art of Machine Translation (MT), Computer Assisted Translation (CAT) tools remain the preferred solution of human translators when publication quality is of concern. In this paper, we present our perspectives on improving the commercial bilingual concordancer TransSearch , a Web-based service whose core technology mainly relies on sentence-level alignment. We report on experiments which show that it can greatly benefit from statistical word-level alignment.

Phrase-Based SMT with Shallow Tree-Phrases

In this article, we present a translation system which builds translations by gluing together Tree-Phrases, i.e. associations between simple syntactic dependency treelets in a source language and their corresponding phrases in a target language. The Tree-Phrases we use in this study are syntactically informed and present the advantage of gathering source and target material whose words do not have to be adjacent. We show that the phrase-based translation engine we implemented benefits from Tree-Phrases.

Mood at work: Ramses versus Pharaoh

A portable and collapsible bed includes a generally rectangular tubular frame and an inflatable mattress with the inflatable mattress having a connecting means that has loose flaps extending towards the periphery of the hollow main flexible body with the flaps having connectors on the ends thereof which can be connected to the tubes of the tubular frame. The tubular frame is formed in a plurality of sections that are easily separable and can be collapsed to a small fraction of the overall size of the unit while the air mattress, in a deflated condition, remains supported on the frame. The unit also includes an inflating unit that can be converted to a deflating unit without any extraneous tools.

NUKTI: English-Inuktitut Word Alignment System Description

Machine Translation (MT) as well as other bilingual applications strongly rely on word alignment. Efficient alignment techniques have been proposed but are mainly evaluated on pairs of languages where the notion of word is mostly clear. We concentrated our effort on the English-Inuktitut word alignment shared task and report on two approaches we implemented and a combination of both.

Harnessing Open Information Extraction for Entity Classification in a French Corpus

We describe a recall-oriented open information extraction system designed to extract knowledge from French corpora. We put it to the test by showing that general domain information triples extracted from French Wikipedia can be used for deriving new knowledge from domain-specific documents unrelated to Wikipedia. Specifically, we can label entity instances extracted in one corpus with the entity types identified in the other, with little supervision. We believe that the present study is the first one that focusses on such a cross-domain, recall-oriented approach in open information extraction.

RALI: SMT Shared Task System Description

Thanks to the profusion of freely available tools, it recently became fairly easy to built a statistical machine translation (SMT) engine given a bitext. The expectations we can have on the quality of such a system may however greatly vary from one pair of languages to another. We report on our experiments in building phrase-based translation engines for the four pairs of languages we had to consider for the SMT shared-task.

Designing a machine translation system for Canadian weather warnings: A case study

In this paper we describe the many steps involved in building a production quality Machine Translation system for translating weather warnings between French and English. Although in principle this task may seem straightforward, the details, especially corpus preparation and final text presentation, involve many difficult aspects that are often glossed over in the literature. On top of the classic Statistical Machine Translation evaluation metric results, four manual evaluations have been performed to assess and improve translation quality. We also show the usefulness of the integration of out-of-domain information sources in a Statistical Machine Translation system to produce high quality translated text.

From French Wikipedia to Erudit: A test case for cross-domain open information extraction

In this paper, we describe an open information extraction pipeline based on ReVerb for extracting knowledge from French text. We put it to the test by using the information triples extracted to build an entity classifier, ie, a system able to label a given instance with its type (for instance, Michel Foucault is a philosopher). The classifier requires little supervision. One novel aspect of this study is that we show how general domain information triples (extracted from French Wikipedia) can be used for deriving new knowledge from domain‐specific documents unrelated to Wikipedia, in our case scholarly articles focusing on the humanities. We believe that the present study is the first that focuses on such a cross‐domain, recall‐oriented approach in open information extraction. While our systems performance shows room for improvement, manual assessments show that the task is quite hard, even for a human, in part because of the cross‐domain aspect of the problem we tackle.