Acceptability Translation Result of 'KATAKU Version 1.1' and 'Trans Tool 10 Rar'

摘要

This study aimed to analyze the quality of accuracy, readability and acceptability of translation result of "Kataku 1.1" and 'TransTool 10 rar" software programs. The researcher gave 75 questionnaires distributed and 56 respondents completed the forms and returned them to the researcher. The method used to analyze the data is descriptive qualitative research using speadly's component analysis technique. The problem raised focused on the quality of translation from the perspectives of the accuracy, readability and acceptability using Nababan's criteria-based sampling technique for evaluating three aspects of qualification. The texts to be translated are the introductions of twelve texts. The method in this study started from raw data and analyzed them based on their domain, taxonomies, components analysis which were then verified and concluded. Here, investigation was on the translation accuracy, readability and acceptability of 2 language-pair combinations produced by Kataku version 1.1. At a finel level we compared source text with the target texts in explicit marking of error types of translation of individual machine translation systems. Specifically, what aspects of languages of both Tran stools output of target texts had good translations and poor translations; a situation experienced by users who did not understand the source language. The results of this study seen from accuracy level of Kataku V 1.1 and TransTool 10 rar showed that 45 sentences or 24% were accurate, 59 sentences or 32% were less accurate, and 81 sntences or 45% were inaccurate. At the acceptance level, 34 sentences or 18% were acceptable, 58 sentences or 31% were less acceptable, and 93 sentences or 50% were unacceptable. And at the level of legibility, 45 sentences or 24% has a high level of legibility, 59 sentences or 32% have a mid level of readability, and 81 sentences or 45% have a low level of legibility. While "Transtool 10 Rar" showed that 34 sentences or 18% were accurate, 59 sentences or 32% were less accurate, and 93 sentences or 50% were inaccurate. At the acceptance level, 23 sentences or 12% were acceptable, 57 sentences or 31% were less acceptable, 104 sentences or 56% were unacceptable. While the level of readability, 33 sentences or 18% had a high level of legibility, 49 sentences or 32% had a moderate level of readability, and 93 sentences or 50% had a low level of legibility. Viewed from the error level, most errors were sequently done at the level of semantics, syntax, phrase, word order, lexical, lost in contact, and word content. However, from the error rate made by both transtools, "Kataku V 1.1" had less errors than "Transtool 10 Rar". Whereas in terms of the equivalent types of "Kataku V 1.1 and Transtools V 10 rar" mentioned above showed little difference; both of them used more formal equivalents rather than grammatical equivalent. However, "Kataku V 1.1" used one-digit more grammatical equivalents than " Transtools 10 rar". Thus translation using "Kataku V 1.1" is reviewed a little better than " Transtool10 rar". Of the two Tran stools mentioned above Kataku V 1.1 has a higher level of accuracy, acceptability, and readability than Transtools 10 rar: however, both have inaccurate, unacceptable translation and also have low readability of translation.