[ChimeText] 24 Jul (next Thursday): Ming Zhou and Chin Yew-Lin / Generating Chinese Couplets using a Statistical MT Approach AND Web ScaleQuestion Answering -- SQuAD

[Min-Yen Kan]

Hi CHIMETEXT members:

We are fortunate to have Ming Zhou also be a part of the MSRA NLP
visit on the 24th (SIGIR workshop day).  Please attend the talks if
you are free.
Together they will present a holistic view of the NLP research being
done in one of Asia's finest research labs.

Hope to see you there!  -Min

----

Microsoft Research Asia Lab talks / Recent Research in NLP at MSRA

Venue: SR7 (COM1 02-07)

Talk Overviews:
3:00-4:00 - Ming Zhou / Generating Chinese Couplets using a
Statistical MT Approach
4:00-5:00 - Chin-Yew Lin / Web Scale Question Answering -- SQuAD

ABSTRACTS:

   1. Ming Zhou

      Title: Generating Chinese Couplets using a Statistical MT Approach

      Part of the unique cultural heritage of China is the game of
Chinese couplets (duìlián) One person challenges the other person with
a sentence (first sentence). The other person then replies with a
sentence (second sentence), in a way that corresponding words in the
two sentences match each other by obeying certain constraints on
semantic, syntactic, and lexical relatedness. This task is viewed as a
difficult problem in AI and has not been explored in the research
community.

      In this paper, we regard this task as a kind of machine
translation process. We present a phrase-based SMT approach to
generate the second sentence. First, the system takes as input the
first sentence and generates as output an N-best list of proposed
second sentences using a phrase-based SMT decoder. Then, a set of
filters is used to remove candidates violating linguistic constraints.
Finally, a Ranking SVM is applied to rerank the candidates. A
comprehensive evaluation, using both human judgments and BLEU scores,
has been conducted, and the results demonstrate that this approach is
very successful.

      You can view this interesting AI gaming at
http://duilian.msra.cn/ which has become very popular in China.

      Bio: Ming Zhou, research manager of Natutal Language Computing
Group at Microsoft Research Asia (MSRA). As one of the first group in
MSRA, this group has been working on machine translation, information
retrieval, question answering and language gaming and has contributed
many technologies to MS products such as Chinese/Japanese IME, Chinese
word breaker, English writing assistant, search engine speller,
multi-language search and keyword bidding, text mining, etc.

      Ming developed the China's first Chinese-English machine system
CEMT-I in 1988 which set up the foundation of machine translation
research of Harbin Institute of Technology. He is the inventor of
J-Beijing Chinese-Japanese machine translation system, a famous MT
product in Japan which has taken the 62% market share for 10 years
since it was launched in 1998. Ming Zhou got his PhD degree at Harbin
Institute of Technology in 1991. Then he had his post-doc in Tsinghua
University in 1991-1993. He then became an associate professort at the
same university untill 1999 when he joined MSRA.

   2. Chin-Yew Lin

      Title: Web Scale Question Answering -- SQuAD

      Abstract: Question answering has been a very active research
field in information retrieval and natural language processing.
Despite the success of TREC QA track, large scale robust QA systems
are still yet to be found in the real world. In this talk, I will
briefly introduce recent progress on SQuAD --a question and answering
project aiming to crawl, index, and serve all question and answer
pairs existing on the web. I will address six main challenges of the
project and then focus on the topic of question search and
recommendation. Three demos will be shown to highlight how SQuAD
technologies can be used in different scenarios.

      Bio: Dr. Chin-Yew LIN is a lead researcher and research manager
at Microsoft Research Asia. Before joining Microsoft in 2006, he was a
senior research scientist at the Information Sciences Institute at
University of Southern California (USC/ISI) where he worked in the
Natural Language Processing and Machine Translation group since 1997.
His research interests are automated summarization, opinion analysis,
question answering, computational advertising, community intelligence,
machine translation, and machine learning.

      Recently, his main focus is developing scalable automatic
question answering and distillation system -- SQuAD. He also developed
automatic evaluation technologies for summarization, QA, and MT. In
particular, he created the ROUGE automatic summarization evaluation
package. It has become the de facto standard in summarization
evaluations. More than 200 research sites worldwide have downloaded
this package.

Upcoming Talks:
24 Jul:  MSRA NLP Research Labs talks: 2 talks on
1) Ming Zhou / Generating Chinese Couplets using a Statistical MT Approach
2) Chin Yew-Lin / Web Scale Question Answering  -- SQuAD
25 Jul: Yahoo! Research Labs talks: 5 talks on
1) Ricardo Baeza-Yates / Distributed Information Retrieval
2) Evgeniy Gabrilovich / Overview of Computationa lAdvertising
3) Rosie Jones / Geography in Web Search
4) Donald Metzler / Predicting when (not) to Advertise
5) Vanessa Murdock / Diversifying Image Search with User Generated Content
28 Jul: Qiu Long / Context for Semantic Similarity