|
Back to Event List
The Future of Machine Translation
This is a PAST event. See "Meeting Notes" section for audio, video, documents and other information.
Original event date/time: Friday February 24th, 2006, 7:00 pm to 9:00 pm Amanda Hargis will be leading an informal talk on "The Future of Machine Translation" -- recent technological breakthroughs in the ability of machines to translate from one language to another.
Abstract:
Amanda Hargis will be leading an informal talk on "The Future of Machine Translation" -- recent technological breakthroughs in the ability of machines to translate from one language to another.
Here are some articles and web pages on the subject. Reading them all before the meeting is mandatory. No, just kidding. But you might
enjoy reading up before coming to the Future Salon.
Repetez, en anglais, s'il vous plait
While the quality of computer-rendered translations has improved
greatly over the past 20 years, some results are still just as
grammatically goofy as the instructions on a chopstick wrapper. With
Language Weaver researchers approach the problem differently. Instead
of following rigid grammatical rules, Language Weaver matches correct
words and phrases across languages based on the probability that such
words and phrases are correct in a given context.
Google dominates in machine translation tests
Search giant Google's ambitions to make the Web more
international has gotten a slight boost from a U.S.
government-run test in which its translation software beat out
technology from IBM and academia. Google scored the highest in
Arabic-to-English and Chinese-to-English translation tests
conducted by the National Institute of Science and Technology.
Software learns to translate by reading up
Translation software that develops an understanding of
languages by scanning through thousands of previously
translated documents has been released by US researchers. They
offer technology that can translate to or from English with
four languages - Arabic, Chinese, French and Spanish.
New algorithm for learning languages
Cornell University and Tel Aviv University researchers have
developed a method for enabling a computer program to scan text
in any of a number of languages, including English and Chinese,
and autonomously and without previous information infer the
underlying rules of grammar. The rules can then be used to
generate new and meaningful sentences. The method also works
for such data as sheet music or protein sequences.
And in case you really want to know the guts of how these systems
work, here's the research paper describing the Cornell system. There
will be a pop quiz on the implementation details at the rmfuture
meeting. No, just kidding. Actually I couldn't quiz you on it because
I haven't read it myself. That's why you should read it, so you can
explain it all to me (and everyone else who comes to Friday's
meeting).
Unsupervised learning of natural languages
Given a corpus of text, our unsupervised algorithm recursively
distills from it hierarchically structured patterns. The
algorithm relies on a statistical method for pattern extraction
and on structured generalization, two processes that have been
implicated in language acquisition. It has been evaluated on
languages as diverse as English and Chinese.
This is a past event.
Meeting Notes:
Attendance was light, so we had a fascinating
two-person discussion. If you'd like to follow the
discussion, download the handouts
and audio.
NIST 2005 Machine Translation Evaluation Official Results
LinguisticsTest.doc
machinetrans.txt
translation_presentation.doc
rmfuture_2006_02_24_02.mp3
|