Comparative Study of Pre-Editing and Post-Editing

Name

Institution

Comparative Study代写 In this respect, different people from different parts of the world speak different language thus translating communications fundamental

 

Abstract

In the modern globalization, sharing of information is crucial. In this respect, different people from different parts of the world speak different language thus translating communications fundamental. This led to the emergence of translation tool when human being became expensive and taking much time to convert the text from one language to another. Although the machine translations have not entirely replaced humans, they make translation work easier and fast. Therefore, human use machine translation output as the basis on which to start their translation.

Pre-editing and post-editing have become critical points of reference. In this paper, the research proposal, the researcher seek to compare the impact of pre-editing and post-editing machine translations with and without a guideline. At the end of the research, it will seek to answer the question on the significance of the pre-editing and post-editing in translation works. Comparative Study代写**范文

In the achievement of this objective, the paper is divided into sections. The first section explores the general introduction to the research question. The second section provides the theoretical background of the research followed by the third section on the rationale for the study. The paper continues with the fourth section, fifth, and sixth outlining literature reviews, methodologies and anticipated outcomes.


A Comparative Study by Performing Pre-editing and Post-editing to the Inputs of Machine Translation and Translation Memory with and without Guidelines  Comparative Study代写

Introduction

The emergence of the internet and globalization resulted in the high demand for translation services.

The evolution resurgence machine translation (MT) systems were highly dependent on the need for fast translation tools. As a result, machine translation post-editing, which is used to correct the MT error and make the text to the standard, has become one of the ways in facing the demand for the multilingual translation and communications.

As we know Machine Translation (MT) by itself cannot produce good quality and readable output. Even the Translation Memory (TM) sometimes need human intervention. Therefore, Post Editing is a crucial method to be used to remedy the MT and TM outputs. The Post Editing process consumes a lot of time and effort, and sometimes it is better to translate from scratch than applying Post Editing. Fortunately, the Pre-Editing method has come to use where it takes place on the source text before being processed in the MT. Comparative Study代写**范文

It is believed that Pre-Editing can reduce time and effort that is usually spent in the Post Editing phase. This research aims to investigate whether or not Pre-Editing has a positive impact on MT and TM Arabic-English outputs. Also, the research aims to classify and measure the impact of Pre-Editing over the Post Editing process and the satisfaction of Post editors.

The research will be conducted by analyzing three texts through three different stages.  Comparative Study代写

The researcher will produce the first text. In the text, the researcher will try to challenge the machine translator by creating ambiguity in the text through sentence structure and word choices. The second text will be political in which the language and word choice will be specifically those used in the political field. Moreover, the third will be a medical text, again like with the political text; word choice will be specifically those used in the medical field. The aim will be to identify how machine translation can translate different genres ranging from science to humanities. It will also evaluate how it can be effectively compared to human translators. Comparative Study代写**范文

Stage 1 will contain testing all the texts above by applying Pre-Editing to them. In this stage, the Pre-Editing will be conducted through following some guidelines. That will be set by the researcher or use existing guidelines such those presented by Hidalgo Ríos (2015). The Pre-Edited versions of the texts will then undergo MT and TM. The participant will the student taking English linguistic studies at Michigan University and use Arabic as the first language. In this case, they will be able to adjust the text semantically and linguistically to that the machine has been trained to translate.

At stage 2 the same texts examined above will be tested again in this stage but without guidelines. That means all three types of texts will be pre-edited by the same participants but without guidelines. Then, those texts will be undergone MT and TM.

Finally, at stage 3, the same texts will be put into the MT and the CAT tool software (TM) without any pre-editing. In this stage, there will be no participants needed because the text will be input unchanged into the MT CAT tool.

The text which has been processed by the MT and CAT tool will then be taken to the post-editing teams.

At this stage, the team will be required to edit while identifying the discrepancies between the initial text and the post-edited text. The post-editing will need to be detailed and specific as per the topic and context of the text.

Theoretical Background  Comparative Study代写

The use of computers for translation purposed was envisioned since the early 1940s. The work of translation using machines has proven to be challenging and sophisticated practice. Although at some points isolated cases of sentence structures in some texts may appear to be of high quality, no machine has ever closed the full-text translation of quality like that of human translators. Although the use of full machine translation may be unattainable, there approaches used to make the method effective and efficient (Modh, & Saini, 2018).

One of the most important approaches is pre-editing to remove the text ambiguity and sentence structures which may not match the translation tool instructions. Another one is to use the translation tool interactively and solve the issues as they arise in the translation process. The next one is for the MT to generate the general idea of the text, which is the most preferred method where the high-quality text is not required. The last approach uses the MT output as the raw text for editing. Comparative Study代写**范文

The adoption of machine translation varies from one country to the other and from one language to the other depending on the language pairing. The translation from one language to the other is mostly dependent on the usability of the two-language involved interchangeably. For instance, the translation from French to Spanish and French to Italian has been found to have a high rate of reusable sentences which makes the translation process easy to adopt (Leal Fontes, 2013).

Comparative Study代写
Comparative Study代写

Post-editing  Comparative Study代写

In understanding the post-editing of machine translation, it is essential to understand both direct human translation and machine translation as well as their differences and similarities. Post-editing is “… scenarios, humans continuously edit machine translation output into production quality translations, providing an additional, constant stream of data absent in batch translation” (Denkowski, Dyer, & Lavie, 2014). Comparative Study代写**范文

In other words, post-editing involves a rework on the machine translation output to make it semantically and linguistically sensible and error free. The primary reason for post-editing is to ensure quality as the machine translation is not entirely efficient in producing an error-free full translation. In the context of this research, the three texts will undergo post-editing with the linguistic professionals in order to measure the time taken in making error-free texts as well as the discrepancies in the text depending on the MT input text.

Pre-editing  Comparative Study代写

The approach to machine translation requires pre-editing which is mainly aimed at reducing the structural and word ambiguities as well as ensure the source text matches the translation tool guideline (Štefčik, 2015). Different tools have different translation guideline in the bid to ensuring the quality output and also the reduce the amount of time taken to in post-editing. Pre-editing is relying on the set rules of editing the source text to fit the requirement of the system. Comparative Study代写**范文

In the context of this study, both pre-edited and none edited source texts will be fed into the translation tool and compare the output during the post-editing phase. It is expected that pre-editing reduces on time needed to post-edit because the input was adjusted to remove the language ambiguities and structural complexities to suit the translation module used.

The rationale of the Study  Comparative Study代写

Quality and time-saving in translation process notwithstanding, the emergence of the machine translation has elicited divergence concerns. Each machine translation tool has the guideline which the pre-editors use to prepare the text for the translation process. (Štefčik, 2015) It has been argued that pre-editing and post-editing consume much time instead direct human editing can be preferred. A capability and efficiency test of the translation tools is vital to measure the dependency of the machine translation output of the translation tool. That is the time taken to post-edit the machine output with and without the guideline. Comparative Study代写**范文

The findings will be significant in developing free translation tools which will not need pre-editing or post-editing post content usability, semantic meaning, and correctness. The research is essential in identifying the efficiency and effectiveness of the pre-editing according to the guidelines as per the machine language translation instructions. As such, it will identify the time taken to post-edit the texts which have undergone through the machine translation tool with and without a guideline.

Research Question  Comparative Study代写

Question: What are the differences in pre-editing and post-editing of the Arabic-English machine translation and translation memory with and without guidelines.

Definitions:

  • Pre-editing:Involve processing the text before it is input for machine translation. The step corrects errors in the source text while trying to remove ambiguities and structural complexities in the text.
  • Post-editing:The step involves professional translators or linguists who work on the output of the machine translation in order to remove semantic and linguistic errors.
  • Machine translation:Is the process of using machine tools and software to translate a text from one language to another. For instance, from Arabic to English.
  • Translation memory:Is a database that store the source text and the translated text which the professional translators use for post-editing.

The Objective of The Research  Comparative Study代写

The research will seek to compare the time taken and impact of performing pre-editing and post-editing to the inputs of machine translation. And the translation memory with and without the guidelines. The research will also identify the discrepancies between the two in the two processes.

Hypothesis  Comparative Study代写**范文

The texts in the post-editing will be less in the first stage and the second stage because the texts in the first two stages had undergone pre-editing stage before the machine translation processed them.

Research Questions  Comparative Study代写

The research will have several objectives which it will strive to achieve:

  • What is the impact of applying Pre-Editing (with and without guidelines) over Arabic-English machine translation output and post-edited output, and to what extent?
  • Does the Pre-Editing function correctly with MT, TM or both?
  • Are Post-editors of MT and TM pre-edited outputs satisfied through reducing their effort and time?

Conceptual Framework  Comparative Study代写

The research will use both quantitative and qualitative approach to data collection and analysis. In this case, the dependent variables will be the input text, pre-editing, and post-editing of the source text. The independent variable will be the quality of the resulting machine translation. In gathering the research data, a group of translators will work on pre-editing and post-editing the MT text. Each of the participants will comment on the comparisons with the source text and the time taken and complexity of post-editing. The result will be recorded then qualitatively and quantitatively analyzed to conclude. The results will be evaluated by the impact of pre-editing made on the Arabic text which will be translated into English.

Literature Review  Comparative Study代写

Machine translation has been in evolution since the early 1940s (Stein, 2018).

The necessity for having machine translation is the globalization and integration of various languages in transactions. Although the technology has not been fully effective and efficient, combined with various approaches, it becomes usable and efficient. Moreover, the growth in the machine translation is dependent on the development of technologies which are emerging like Artificial Intelligence. However, machine translation is unlikely to attain the quality level of human translation.

The work of translation is the interaction between humans and the machines.  Comparative Study代写

In this case, various modules of the translation tools are integrated to form a full process that includes the human pre-editors, post-editors, and evaluators (Seretan, Roturier, Silva, & Bouillon, 2014). The three groups of humans play a crucial role in ensuring the success of the translation work. Therefore, it can be concluded that translation is not an isolated machine affair but a human-machine interaction which ultimately produce quality results fast and effectively.

Some recent studies have discussed the concept of Pre-Editing of MT inputs.  Comparative Study代写**范文

Štefčik (2015) investigate the impact of Pre-Editing on statistical MT and Post Editing. The content analyzed belong to user-generated content that is found in forums. Also, the language pair used in this study is English-French. It was found that pre-editing reduces not only the post-editing time but also the quality of the machine translation output (Siegel, 2018). The time taken to post-edit the MT output is less compared to the text which was not pre-edited before the translation process.

On the other hand, post-editing issue has gain prominence on the quality and the impact of processing pre-edited texts.  Comparative Study代写

There are questions raised as for the need of the post-editing and whether it is worth (Koponen, 2016). It was found that post-editing took less time when there was pre-editing before machine translation. This study validated the importance of post-editing in reducing both semantic and linguistic errors. The machine translation is prone to various language errors depending on the genre and complexity of the paper.

Methodology  Comparative Study代写

The research will be a process-oriented. It will focus on data elicitation, keystroke logging, screen recording. The participants will take part in the pre-editing and post-editing then record their comments on the output texts. The study will use various translation tools such as BBFlash Back which will be used to record the keystroke logging and time, LISA QA. Which is the external professional translator and reviewer, and Acrolinx IQ Engine. The participant’s opinions will then be evaluated quantitatively and qualitatively to arrive at a conclusive finding of the research. The data will be evaluated comparatively on the impact of text pre-editing and post-editing with and without the guideline.

Expected Outcomes  Comparative Study代写

Stage 1: The text will be translated from Arabic to English. Compared to stage 3, the post-editing took less time.

Stage 2: The text examined in stage 1 is re-examined. In this case, the text will be translated from English to Arabic. It was found to much time to correct the errors which were higher compared to stage 1.

Stage 3: The text is passed without following the guideline. Much time was taken to do post-editing.

Comparative Study代写
Comparative Study代写

References  Comparative Study代写

Denkowski, M., Dyer, C., & Lavie, A. (2014). Learning from post-editing: Online model adaptation for statistical machine translation. In Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics (pp. 395-404).

Hidalgo Rios, R. (2015). Analysis of the Impact of the Use of Post-editing Guidelines in Raw Machine Translation Output (Doctoral dissertation, University of Geneva).

Koponen, M. (2016). Is machine translation post-editing worth the effort? A survey of research into post-editing and effort. JoSTrans, 25, 131-148.

Leal Fontes, H. (2013). Evaluating Machine Translation: preliminary findings from the first DGT-wide translators’ survey. Languages and Translation, 6:10– 11.

Modh, J. C., & Saini, J. R. (2018). A study of machine translation approaches for the Gujarati language. International Journal of Advanced Research in Computer Science, 9(1).

Seretan, V., Roturier, J., Silva, D., & Bouillon, P. (2014). The ACCEPT Portal: An online framework for the pre-editing and post-editing of user-generated content. In Proceedings of the EACL 2014 Workshop on Humans and Computer-assisted Translation (pp. 66-71).

Stein, D. (2018). Machine translation: Past, present and future. Language technologies for a multilingual Europe, 4, 5.

Siegel, M. (2018). Authoring support for controlled language and machine translation: A report from practice. Language technologies for a multilingual Europe, 4, 85.

Štefčik, J. (2015). Evaluating machine translation quality: A case study of a translation of a verbatim transcription from slovak into german. Vertimo studijos, 8(8), 139-153.

 

更多其他:代写案例 Case study代写 网课代修 数据分析代写 Review代写 文学论文代写 Report代写 Proposal代写 Essay代写 数据分析代写 Academic代写 代写CS 润色修改

合作平台:天才代写 幽灵代写  写手招聘

此站点使用Akismet来减少垃圾评论。了解我们如何处理您的评论数据

 
QQ在线咨询
售前咨询热线
QQ:3554475127
售后咨询热线
QQ:3042439236