Statistical Machine Translation lecture at Kasetsart University

บรรยาย: การแปลภาษาด้วยเครื่องด้วยวิธีทางสถิติ: อะไรที่เป็นไปได้ในวันนี้?

โดย ฟิลิปป์ เคิห์น มหาวิทยาลัยเอดินบะระ สก็อตแลนด์

วันจันทร์ที่ 18 ธันวาคม 2549 – 9:30-11:30 น.

ห้อง 204 ตึกวิศวกรรมคอมพิวเตอร์ (ตึก 15) มหาวิทยาลัยเกษตรศาสตร์ บางเขน

ลงทะเบียน


Lecture: Statistical Machine Translation: What is possible today?

by Philipp Koehn, University of Edinburgh, Scotland

Monday, December 18, 2006 – 9:30-11:30 am

Room 204, Computer Engineering Building (Building 15), Kasetsart University

Register


Synopsis:

Philipp will give an overview of the current state of the art and the research challenges in machine translation. Translating text from one language to another by computer is one of the oldest challenges in artificial intelligence research, but for many years machine translation (MT) has been considered a dirty word by many for failing to deliver. Today, with the use of novel methods, fast computers and access to large amounts of translated material, we are coming closer to a solution to this problem and it is now possible to build machine translation system within a few hours.

The statistical approach to machine translation provides a set of techniques for automatically learning translation knowledge from existing human translations (bilingual data), and applying that knowledge to translate previously-unseen sentences. When it was first introduced, statistical MT was far too slow and inaccurate to be useful — it was an interesting lab experiment. Now, statistical MT significantly outperforms other methods in many language pairs and domains, at speeds permitting commercial applications like foreign news broadcast translation.

Key issues addressed include:

  • What changes have occurred that now make this possible?
  • How good is the quality and what remains to be done?
  • How has use of phrasal and syntactic knowledge helped?
  • What are the major technical advances of the past few years, and known limitations?
  • What new techniques are now being applied to machine translation to improve quality even further?

Presenter Bio:

Philipp Koehn is a lecturer (Assistant Professor) at the University of Edinburgh. He received his PhD from the University of Southern California, where he was a research assistant at the Information Sciences Institute (ISI) from 1997 to 2003. He was a postdoctoral research associate at the Massachusetts Institute of Technology (MIT) in 2004, and joined the University of Edinburgh as a lecturer in 2005. His research centers on statistical machine translation, but he has also worked on speech in 1999 at AT&T Research Labs and text classification in 2000 at Whizbang Labs. Philipp is a co-founder of Getprice, a German price comparison Internet company, where he acted as CTO from 2000-2005.

technorati tags: 

Published by

bact

bact' is a name

Leave a Reply