Computer-aided Translation Research Articles

Unit Completion for a Computer-aided Translation Typing System

This work is in the context of, a system that watches over the users as they type a translation and repeatedly suggests completions for the text already entered. The users may either accept, modify, or ignore these suggestions. We describe the design, implementation, and performance of a prototype which suggests completions of units of texts that are longer than one word.

CAT Tools in an Academic Environment

AbstractMany universities have now incorporated commercially available translators ' workbench-style systems into their translator-training programmes. But, when it comes to computer-aided translation (CAT), the university's role need not be confined to teaching students how to operate some third party's system; rather new CAT tools open up whole new areas of research. For example, experience of Trados'sTranslator's Workbenchsuggests that workbench features such as automatic terminology recognition and translation memories stand to bring about fundamental changes in the way terminology is recorded and texts are authored. State-of-the-art CAT tools can also make a contribution to Descriptive Translation Studies and translation pedagogy.

Interactive translation of conversational speech

As communication becomes increasingly automated and transnational, the need for rapid, computer-aided speech translation grows. The Janus-II system uses paraphrasing and interactive error correction to boost performance. Janus-II operates on spontaneous conversational human dialogue in limited domains with vocabularies of 3,000 or more words. Current experiments involve 10,000 to 40,000 word vocabularies. It now accepts English, German, Japanese, Spanish, and Korean input, which it translates into any other of these languages. Beyond translating syntactically well-formed speech or carefully structured human-to-machine speech utterances, Janus-II research has focused on the more difficult task of translating spontaneous conversational speech between humans. This naturally requires a suitable database and task domain.

Automatic derivation of the governing equations that describe a transient electrochemical experiment, given a reaction mechanism of arbitrary complexity. Part 2. Governing equations in one-dimensional geometry

A systematic procedure for deriving the governing equations that describe a transient electrochemical experiment, given a reaction mechanism of arbitrary complexity under conditions of controlled-potential or controlled-current transients methods, is outlined. The procedure is based on the analysis and transformations of the stoichiometric matrix and other data, and applies to the class of reaction networks involving electrochemical, heterogeneous non-electrochemical and homogeneous reactions between bulk species (subject to semi-infinite diffusion and/or convection in one-dimensional spatial geometry) and interfacial species (located at electrodes). Equilibrium, non-equilibrium reversible and irreversible reactions are allowed, as well as the presence of species with invariant concentrations. The corresponding governing equations are proven to take the form of coupled sets of partial differential equations with initial and boundary conditions (for the concentrations of bulk species), and differential-algebraic equations (for the concentrations of interfacial species). The procedure is a part of the algorithm of an automatic, computer-aided translation of electrochemical reaction mechanisms into corresponding texts of the governing equations.

Automatic derivation of the governing equations that describe a transient electrochemical experiment, given a reaction mechanism of arbitrary complexity. Part 1. Problem parameters and initial conditions

We describe a systematic procedure for determining a self-consistent set of thermodynamic and kinetic parameters and initial conditions that are needed for the simulation of transient electrochemical experiments, given a reaction mechanism of arbitrary complexity. This is an extension of the algorithm recently suggested by Luo et al. (J. Electroanal. Chem., 368 (1994) 109) to the larger class of reaction networks involving elementary or non-elementary electrochemical, heterogeneous non-electrochemical and homogeneous reactions of unrestricted molecularities, subject to the power rate law with reaction orders that can be different from stoichiometric coefficients. Equilibrium, non-equilibrium reversible and irreversible reactions between bulk species (distributed in the electrolyte) and interfacial species (located at electrodes) are allowed, as well as the presence of species with invariant concentrations. Tests enabling verification of the correctness of the symbolically written reaction mechanisms are described. The procedure is part of an algorithm for automatic, computer-aided translation of electrochemical reaction mechanisms into corresponding texts of mathematical governing equations.

Closed Captioning in Real Time

The line-21 captioning system was developed in the 1970s as a means of delivering hidden subtitles (closed captions) to hearing-impaired television viewers. A more recent development is the ability to generate captions in “real time” during live broadcasts, using experienced court reporters and a computer-aided translation system that converts shorthand code into captions almost instantaneously. More than 70 hours of nationally broadcast programming are captioned weekly with this technology. This article describes the creation and broadcast of captions in real time, including equipment, staffing, and procedures.

Other Than Counting Words: A Linguistic Approach to Content Analysis

A linguistic technique for the content analysis of texts and transcripts is described and illustrated. The technique produces a quantitative description of texts that represents both the interrelations among words and their classification into meaning categories. The face validity of the description is demonstrated with a computer-aided translation of coded data back into text. The advantages of linguistic content analysis over qualitative and computer-aided approaches to content analysis are discussed. Also discussed are problems of linguistic ambiguity and of the computer automation of the technique. While Director of the Experimental Division for the Study of War-Time Communications, Harold Lasswell explained that in the content analysis

Expression of the plum pox virus coat protein region in Escherichia coli.

A cDNA complementary to the 3' end of plum pox virus (PPV) RNA was sequenced. The sequence was investigated for the presumable coat protein cistron by computer-aided translation. A fragment containing the stop codon of the polyprotein gene and a putative virus-specific protease cleavage site was subcloned into an E. coli expression vector. It is shown by immunological analysis that the coat protein cistron is located within the subcloned region.

An Approach to Computer-Aided Translation

As part of a preliminary investigation of computer-aided translation under time-shared data processing, various techniques of computerized dictionary lookup and text file management have been explored under the time-sharing system TSS/360. Computer-aided translation (CAT) is a storage and retrieval operation carried out on line with a computer during the time in which a translation is produced. A system of dictionary access and updating routines, text-processing facilities, and on-line utilities is designed to telescope the delay between the initiation of a translation and its finished printout. The system does not attempt to simulate the human translator by producing an autonomous translation via programmed algorithms; rather, it serves as an extension of the capabilities of the user, who is able to call on the resources of the computer as needed in the process of translation and get an immediate response. Under the system described, users communicate over ordinary telephone lines with the computer by means of remote terminals. In employing the system, the user can switch back and forth as many times as required among human translation, direct dictionary lookup, editing, printing, and system interrogation, and thereby achieve rapid iteration toward the desired goal, i.e., a finished translation.