發表文章

Advanced Translation Technology

  Welcome!   Please visit the following articles for the projects published for this course: 1) Bubble Milk Tea MT Project  MT Project 2) CCC Turkey Evaluation TMS Project TMS Project The first project includes training Microsoft Custom Translator for Chinese recipes to English and the second project focused on an evaluation for a gaming company to select a suitable TMS.  Please feel free to visit the articles and contact me for anything that might interest you.

Bubble Milk Tea MT Project

圖片
Overview This pilot project aims to estimate the work involved in training a neural machine translation (NMT)  engine for Lee Kum Kee (LKK) recipes from Chinese into English. Post-editing machine translations  (PEMT) should meet the criteria as follows 1) Efficiency 2) Reduce Cost 3) High Quality Team Members Our team members are Chih-Yuan Chen, Ruby Lee,  Hannah Liu, and myself. For this project, our team formed an imaginary company called Bubble Milk Tea.  Project Specifications For this project, our team used Microsoft Custom Translator. We were  trying to train this neural machine translation (NMT)  engine for Lee Kum Kee (LKK) recipes from Chinese into English . We did a total of 10 rounds of training.  Our goal is that p ost-editing machine translations (PEMT) should be  30% faster than human translation, 30% cheaper than HT and make the  recipes  s imple, clear, and user-friendly.  Proposal Feel Free to Download the Proposal...

CCC Turkey Evaluation TMS Project

Overview For this project, our team focused on a gaming company (ABC Games) and evaluate the best suitable TMS for this company to obtain.  T hree   TMS that our team evaluated for this project 1) Crowdin 2) Transifex 3) XTM Team Members Our team members are Chih-Yuan Chen, Ruby Lee, and myself. For this project, our team formed an imaginary company called CCC Turkey Evaluation .  Project Specifications The imaginary company- ABC Games was based on an actual gaming company. This gaming company focused on game developing and publishing as well as IP licensing. This company has currently released 36 games in the market (computer + mobile).  Based on our knowledge and experience, we decided on seven essential criteria for a TMS to be counted as suitable for a gaming company.  These criteria include: 1) Voting for the best translation 2) QA check for consistency 3) Access to terminology 4) Security 5) Price 6) Visual Feedback 7) Live Participation Furthermore, based...

CAT Introduction (Lessons Learned)

The purpose of this portfolio is to introduce the basics of CAT tools.  While there are multiple CAT tools online, the CAT tool used in this portfolio is SDL TRADOS. The ultimate goal of translation technology is to reach a FA, HQ and UT translation. However, we are only able to achieve any of these two combinations.   There are two parts to this portfolio. The first is "Useful Tips for Regex with SDL Trados (QA Checker)" in which Language-Specific Regex for SDL Trados QA Checker/Find&Replace were written and verified to increase the quality and speed of the translation.  The second part is the "Team CAT Translation Project " in which we simulated a real-life scenario between the language service provider (LSP) and the client. We are required to provide a translation quote  (proposal) and attend the start-up meeting with the client. ♔Key Takeaways after Reviewing this Portfolio 1) Anticipate questions from the client and interact  2) Actively engag...

Useful Tips for Regex with SDL Trados (QA Checker)

圖片
  Language-Specific Regex for SDL Trados QA Checker/Find&Replace 1.   Replace the date format in Chinese content. In Chinese, we tend to prefer  年( year )月( month )日( date ) to show date. So for certain dates, we would like to write 2020/10/09 as 2020 年 10 月 9 日 . The first one is to identify these unchanged dates in target content and replace them. Regular Expression : (\d\d\d\d|\d\d)\/(\d\d|\d)\/(\d\d|\d) Replace:$1 年 $2 月 $3 日 2.  From English to Chinese, sometimes, numbers are expressed as k/m/b. However, that’s not the case in Chinese. We tend to use 万( 10k )百万( m )千万( 10m )亿( 100m ) The following expressions will send a warning while detected m/k/b numbers in Chinese for a doublecheck. Regular Expression: (\d{1,}[k|K])|(\d{1,}[m|M]|\d{1,}[b|B]) 3.   This Regex is to identify the English punctuation marks in Chinese content, and sends a warning when verifying translation. Regular Expression: [.,?!:“”?] ...

Team CAT Translation Project

圖片
Overview The purpose of this project is to simulate a real-life scenario between  the language service provider (LSP) and the client. We are required to provide a translation quote  (proposal) and attend the start-up meeting with the client (Prof. Wooten). This stimulation project was conducted under SDL Trados for translation.  Team Members Our team members are  Chih-Yuan Chen, Ruby Lee and myself. For this project, our team formed an imaginary translation company called CCC Turkey Translate.  Project Specifications For this project, we imagined our client to be the American Institute in Taiwan (AIT). Our goal is to translate the report on importing U.S. pork and beef that contains ractopamine. The translation for this project is from  Traditional Chinese into English and because the purpose of this translation is to be published on the official  website for the English-speaking public to read everything must be at top standard.  Proposal ✅ Downl...