Group items tagged

"Alignment tools[edit] The Hunalign sentence aligner GIZA++ alignment tool An implementation of the Gale and Church sentence alignment algorithm An alignment tool by Jörg Tiedemann"

"Bitext word alignment or simply word alignment is the natural language processing task of identifying translation relationships among the words (or more rarely multiword units) in a bitext, resulting in a bipartite graph between the two sides of the bitext, with an arc between two words if and only if they are translations of one another. Word alignment is typically done after sentence alignment has already identified pairs of sentences that are translations of one another."

hunalign aligns bilingual text on the sentence level. Its input is tokenized and sentence-segmented text in two languages. In the simplest case, its output is a sequence of bilingual sentence pairs (bisentences). In the presence of a dictionary, hunalign uses it, combining this information with Gale-Church sentence-length information. Like most sentence aligners, hunalign does not deal with changes of sentence order: it is unable to come up with crossing alignments, i.e., segments A and B in one language corresponding to segments B' A' in the other language.

Cfr. LF Aligner

Most commercial translation memory applications now include an aligner as part of a suite of tools. Aligners are used to produce translation memory files (e.g. in the industry-standard TMX format) from legacy translations and their corresponding source files. The resulting file can then be used within a translation memory application, which provides easy access to the source/target segment pairs in the translation memory file. Although the advantage of aligning relevant legacy texts is obvious, a major drawback is that such tools generally align non-intelligently. Where the source and target texts have different numbers of segments (e.g. because at some point, the original translator merged two sentences to form a single sentence in the translation), the resulting translation memory file is misaligned. Aligners generally provide an interface through which such misalignments can be corrected manually. This can be a time-consuming process, however, and is usually worthwhile only when the legacy translation is known to be useful, such as when the original source text has been modified and a new translation of it is required.

Pàgina con ligas de acceso a herramientas de alineación de linux.

ABBYY Aligner 2.0 is a professional tool for aligning parallel texts and creating Translation Memory databases. This easy-to-use and convenient software accurately finds matching segments in parallel texts, provides capabilities to edit the alignment results and allows saving them into TMX files for further use in CAT-tools or into RTF files.

LF Aligner helps translators create translation memories from texts and their translations. It relies on Hunalign for automatic sentence pairing. Input: txt, doc, docx, rtf, pdf, html. Output: tab delimited txt, TMX and xls. With web features.

Con base en la última liga que añadí al grupo me parece que LF Aligner es una herramienta de Linux y por eso no puedo correrla correctamente en el sistema operativo de Windows. Cuando corre en Windows primero me pide que seleccione el tipo de formato que voy a utlizar en el proceso de alineaciòn; en este caso txt. Después pide el nùmero de lenguajes: 2. Luego pide la clave del texto fuente (en) y del meta (es). Finalmente pide que escoja los archivos para los textos fuente y meta respectivamente. El problema es que no sucede nada. De todas formas voy hacer pruebas en mi otra máquina.

LF Aligner se programa para los tres sistemas operativos: MacOS, Linux y Windows. Así que el problema no proviene de ahí. Lo bajé e intenté correrlo, pero se requiere tener dos archivos de texto o de MS Word con un texto en una lengua y su traducción a otra, muy parecidos en formato, para poder hacer una prueba. En todo caso, es absolutamente necesario leer y comprender el archivo LF_aligner_readme.txt que viene dentro del paquete, para poder utilizarlo correctamente. El autor presume de aprovechar un algoritmo diseñado por alguien más (al parecer, un húngaro) para hacer la alineación. Ese algoritmo está tomado de: http://mokk.bme.hu/resources/hunalign

With Youalign, you can quickly and easily create bitexts from your archived documents. A bitext contains a document and its translation aligned at the sentence level. generates TMX files that can be loaded into your translation memory.

Es una buena herramienta en línea. Pensé que era mejor opción que ABBYY Aligner en cuestión de precios pero ABBYY es 2,500 pesos más barato. Seguiré buscando otras opciones.

"tools for creating annotated parallel corpora including pre-trained tagging and parsing models for various languages, sentence alignment tools and word alignment tools. "

User-Friendly Translation Memory Software AnyMem is a user-friendly translation memory software. AnyMem is compact, stable, easy to learn and to use. AnyMem CAT engine fully supports Unicode and can work with any language pair without the need of external components, like .NET Framework or Java Virtual Machine, commonly required by other translation memory software. Features: Convenient translation memory management: translation memory is stored in a single MEM file, for more convenient exchange. Export translation memory to TMX: translation memory can be exported to TMX 1.1, TMX 1.4. Import translation memory from TMX or MS Word files: AnyMem 2.0 can import translation memory from a TMX 1.1, or TMX 1.4 file, or from a bilingual MS Word file. Flexible text segmentation: configure exceptions to text segmentation rules by adding abbreviations and ordinal follower words. Selective translation: translate only the text which must be translated by specifying styles of paragraphs, which are not to be translated. Document CAT analysis: AnyMem 2.0 can analyze documents to determine the number of matching segments (complete and fuzzy match), and provide accurate CAT text count in words, segments or percent. Terminology search: AnyMem 2.0 can search translation memory for matching segments which contain certain term. Flexible data display: configure the way AnyMem 2.0 displays translation memory data with the help of templates. Automatic window alignment: prepare AnyMem 2.0 and MS Word windows for work by correctly aligning them with a few clicks, and without the need to manually drag window borders MS Word integration: quickly access most frequently used AnyMem 2.0 commands and functions directly from MS Word with the help of AnyMem 2.0 toolbar. You can also quickly run AnyMem 2.0 directly from MS Word. Unicode support: AnyMem 2.0 CAT engine fully supports Unicode without the need to install additional library components (like.NE

Across translation memory - a database that stores previously translated segments allowing reuse of previously translated text passages, but also serving as a basis for assisting authors in the consistent creation of source texts, and for assuring the quality of both the source text and the translations. As a learning system, Across translation memory (TM) has to be filled initally either by importing existing translation memory data or by performing an "alignment" in which source and target text pairs are analyzed and matching sentences are added to the translation memory. After that, the database grows with each new translation. New sentences within translation jobs are identified automatically and added to the translation memory either directly or after a quality check. By saving every translated sentence in the database, new sentences that need to be translated can be compared with this database. The purpose is to reuse as many previously translated sentences as possible. This speeds up the actual translation, improves the consistency, and saves money. Across translation memory assists with your translations As a basis for automatic pre-translation of sentences that are 100% matches As a basis for translators in case of partial matches As a basis for concordance searches (i.e., when researching the context in which a word or an expression is used) As a basis for translation-oriented authoring, in which the source text author selects wording variations for which a translation is already available.

Esta página proporciona ligas para obtener herramientas de alineación. Sin embargo, la página que me llevó aquí es una llamada "Linux for Translators". Quizá estas herramientas no funcionen en el sistema operativo de Windows. Tengo que investigar más acerca de ésto.

"MyMemory is the world's largest Translation Memory. MyMemory is 100% free. It has been created collecting TMs from the European Union, United Nations and aligning the best domain specific multilingual websites. You can download memories for free in TMX format and use them with your CAT tool of choice. You can also contribute to make MyMemory even better with your memories, or editing entries. When no human translation is found for your sentences, MyMemory will help you with state-of-the-art Statistical Machine Translation technology."

Bruno, quizás este servicio ayude a perfeccionar la memoria de traducción que tienes construida.

Gracias Julio. Veremos si esta tecnología podría ayudarnos en el futuro.

MultiCorpora is the language technology experts providing innovative computer assisted translation systems & translation management system (TMS). Our solutions include translation memory, terminology management, project management and more.

Liga para descargar y probar Trados. El precio del producto está entre 10 mil y 15 mil pesos.

"Concordancier bilingue"

"Translation Table creates a two column table, where in the first column you get the source text segmented by sentence, each sentence in a cell, and in the second column you can put your translated sentences, each one in opposite cells to the source sentences. It is not a CAT (Computer Aided Translation) tool. For that, you may look to the Anaphraseus extension at http://anaphraseus.sourceforge.net/ or the OmegaT project at http://www.omegat.org/. "

Gratuito y portátil