Skip to main content
Fachgebiete
Automobil + Motoren Bauwesen + Immobilien Business IT + Informatik Elektrotechnik + Elektronik Energie + Nachhaltigkeit Finance + Banking Management + Führung Marketing + Vertrieb Maschinenbau + Werkstoffe Versicherung + Risiko
DE
EN
Bücher
Zeitschriften
Themenseiten
Organisationspsychologie Projektmanagement Marketing Smart Manufacturing
Weitere Formate
Podcasts Webinare Technik Webinare Wirtschaft Kongresse Veranstaltungskalender Awards
Jetzt Einzelzugang starten
Zugang für Unternehmen
Referenzkunden
SLX-Digitalkonferenz: Zukunftswerkstatt 2024

Mehr Umsatz mit Top-Kontakten

Am 7. Mai erfahren Sie, wie Vertriebsteams Leadgenerierung, Leadnurturing und Leadstrategien effizient steuern können.
Erweiterte Suche
Anmelden
nach oben
Erschienen in:
Introduction Kapitel lesen Erstes Kapitel lesen

2018 | OriginalPaper | Buchkapitel

6. Logos Model: Design and Performance

verfasst von : Bernard Scott

Erschienen in: Translation, Brains and the Computer

Verlag: Springer International Publishing

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

In this Chapter we describe Logos Model in terms of five fundamental decisions that determined its design: (i) how linguistic knowledge is to be represented; (ii) how that linguistic knowledge is to be stored; (iii) how stored linguistic knowledge is to be applied to the input stream; (iv) how target sentences are to be generated from source analysis; (v) how complexity effects are to be dealt with. We illustrate the implementation of these design factors with translations in a variety of linguistic contexts. Translations by Logos Model are in French and/or German. Translations are also provided by other MT systems for comparative purposes. Finally, we give an example of a translation that exemplifies capabilities that are unique to an MT system like Logos Model, and we give an account of how this particular translation is accomplished.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Jetzt informieren
Vorheriges Kapitel Syntax and Semantics: Dichotomy Versus Integration
Nächstes Kapitel Some Limits on Translation Quality
Anhänge
Nur mit Berechtigung zugänglich
Fußnoten
1
The example of a plural noun in English taking a singular noun in German may seem inconsequential, but it illustrates the potential for confusion that may arise in source and target for terms that look alike but have significant differences (faux amis). For example, the French verb ignorer looks very much like the English verb ignore, but its French meaning is entirely different.
 
2
In the following sections, we compare Logos Model translations with output from various other MT systems. See Postscript 6-B of this chapter for a collection of these comparisons.
 
3
Historically, some otherwise effective systems have been known to founder largely on these grounds (e.g., TAUM Aviation (Juola 1989)).
 
4
As stated in earlier chapters, it is misleading to speak of Logos Model as a rule-based translation system (RBMT). Logos Model is data-driven and pattern-based. What especially differentiates Logos Model from rule-based models concerns the manner in which matching is accomplished. We remind the reader that in Logos Model, pattern-rules are ordered purely on their SAL pattern, and matching is effected by the input stream functioning like a search argument against stored pattern-rules, not unlike in dictionary lookup. No metarule, discrimination network, or decision logic of any kind is involved in effecting these matches.
 
5
Logos Model software is now written in C.
 
6
See Part II Postscript for discussion about the use of numbers and its advantages. Also notice the numbers in Fig. 6.1.
 
7
Microsoft’s Bing NMT Translator and Google GNMT Translate both translated (4) correctly with respect to parts of speech, but Google Translate failed to recognize the relative clause, and Bing mistranslated costs are building up in the German as bauen die Kosten auf.
 
8
See Postscript 6-A for this Chapter for further illustration of semantic resolutions by SEMTAB. Note that SEMTAB pattern-rules entail deep-structure patterns, which means that a single, deep structure SEMTAB pattern-rule will handle a variety of different surface structures for a given verbal concept, such as process noun, verbal adjective, as well of course as the verb itself, in all its conjugations.
 
9
Subjectless passives in German require that the main clause subject not be the agent of the passive verb, and that the clause’s passive verb construction entail the auxiliary werden. Both conditions were tested for and satisfied by the target action component responsible for the translation in (39)(ii).
 
10
Deficiencies in source analysis chiefly concerned shortcomings in ambiguity resolution, particularly ambiguities having to do with part of speech. For data-driven SMT on the other hand, target generation poses the chief difficulty, particularly with respect to proper target word order and morphological agreement among chunks of text. The reason for this is manifold, but the deepest reason is that SMT is phrase-based and not sentence-based.
 
11
Note translation of (39) by the University of Montreal LISA Lab’s neural MT system, which is basically structure-preserving: D’autres moyens de stockage, comme les cartes magnétiques et la bande magnétique , peuvent également être accessibles au moyen de dispositifs facultatifs et de leur entrée dans le système. The LISA system however interestingly transforms can be accessed to can be accessible in the French. The complete mistranslation here of the second clause of (39) is underlined.
 
12
See Fig. 4.​9 in Chap. 4 for a graphic picture of Semantic Table interaction in Logos Model.
 
Literatur
Somers HL (1992/3) Current research in machine translation. Mach Trans 7:231–246CrossRef
Metadaten
Titel
Logos Model: Design and Performance
verfasst von
Bernard Scott
Copyright-Jahr
2018
Buch
Translation, Brains and the Computer

Print ISBN: 978-3-319-76628-7

Electronic ISBN: 978-3-319-76629-4

Copyright-Jahr: 2018

DOI
https://doi.org/10.1007/978-3-319-76629-4_6