Browse Prior Art Database

Interactive Author-Assistance Tool

IP.com Disclosure Number: IPCOM000060754D
Original Publication Date: 1986-May-01
Included in the Prior Art Database: 2005-Mar-09
Document File: 3 page(s) / 51K

Publishing Venue

IBM

Related People

Cohen, PS: AUTHOR [+6]

Abstract

An interactive author-assistance tool (IAAT) or program provides an online system allowing authors of documents, which are written in one natural language, such as English, to make such documents more translatable into other natural languages. The translation can be done by computer-based systems intended to perform mechanical translation (MT) or machine-aided translation (MAT). The program has three semi-automatic modules implemented interactively on a computer to perform the functions generally shown in Fig. 1. The text or input document 10 is first analyzed by a spelling checker 11 to detect spelling and typographical errors which can then be corrected by the author. Next, the text is subjected to a readability analysis 12 which generally detects and warns the author about overly long sentences and undefined acronyms.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 3

Interactive Author-Assistance Tool

An interactive author-assistance tool (IAAT) or program provides an online system allowing authors of documents, which are written in one natural language, such as English, to make such documents more translatable into other natural languages. The translation can be done by computer-based systems intended to perform mechanical translation (MT) or machine-aided translation (MAT). The program has three semi-automatic modules implemented interactively on a computer to perform the functions generally shown in Fig. 1. The text or input document 10 is first analyzed by a spelling checker 11 to detect spelling and typographical errors which can then be corrected by the author. Next, the text is subjected to a readability analysis 12 which generally detects and warns the author about overly long sentences and undefined acronyms. The author is required to shorten the sentences and define each acronym at its first occurrence. Such shortened sentences are generally grammatically simpler and less ambiguous so that fewer errors are produced in the eventual translation. The readability analysis produces a report allowing the author to determine whether readability criteria had been met. If not, the author can edit the text at 14 and then input it back through steps 11 and 12 until the criteria are met. The readability criteria can be measured by the Flesch Index, the Fog Index, reading grade levels, Kincaid Index, and the British "Reading Age". When the readability criteria are met, the text is then processed by a linguistic analysis module 15 the detailed functions of which are shown in Fig. 2. After such linguistic analysis, a final spelling check is performed at 16 and the output of that step is a translatable document 17 that can be translated by MT or MAT systems. During linguistic analysis, the text from step 13 is first subjected to a sentence separator 20 that divides the text into individual sentences. The first sentence at step 21 is then passed to and analyzed by dictionary look-up and grammatical parse step 23. The parse determines whether the sentence is over...