Browse Prior Art Database

Method of Showing Word Equivalents in a Readable Format in Machine Translation

IP.com Disclosure Number: IPCOM000101634D
Original Publication Date: 1990-Aug-01
Included in the Prior Art Database: 2005-Mar-16
Document File: 5 page(s) / 151K

Publishing Venue

IBM

Related People

Nomiyama, H: AUTHOR

Abstract

Disclosed is a mechanism to show word equivalents in a readable format in machine translation by adding formatting information automatically.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 53% of the total text.

Method of Showing Word Equivalents in a Readable Format in Machine Translation

       Disclosed is a mechanism to show word equivalents in a
readable format in machine translation by adding formatting
information automatically.

      The correspondences of translated words are not always word-to-
word relations but may be word-to-structured-segment, such as
word-to- compound-word, word-to-phrase, or word-to-sentence,
relations. Words with such relations may occur several times in the
same document. Translating them in the same way, each time they
occur, may degrade the readability of the document.

      In such cases, various methods of improving readability, such
as acronyms, parentheses, and footnotes, are commonly used.

      Our system takes documents with formatting information, such as
SCRIPT, as input, and outputs documents with other formatting
information added, utilizing such formatting techniques.
      An overview of the system is shown in Fig. 1.
      The components are described below.
1. Input Document and Output Document
      An input document is a document in the source language
      (SL) that includes formatting information. An output
      document is a document in the target language (TL). A
      formatter needs an output document, a header file that
      involves contracted word equivalents and their
      formatting information, and a procedure file that
      involves definitions of procedures for added formatting
      information.
2. Format Preservation
      This process preserves the formatting information of an
      input document and transfers it to an output document
      unchanged (1,2).
3. Translation

      This process translates a document in SL into a
document in TL.
4. Bilingual Dictionary
      Correspondences between words in SL and word
      equivalents in TL are described in a bilingual
      dictionary. Our method needs additional information on
      contraction levels and contracted forms. Contraction
      levels are shown in Table 1.
5. Dictionary Lookup
      At first this process returns to the translation
      process a content that is obtained from a bilingual
      dictionary. Since the content may contain several
      candidates, it should receive notification from the
      translation process of which word was selected. It then
      decides whether a contracted form should be adopted by
      the evaluation function.
      If it i...