[Date Prev][Date Next][Date Index]

[ntcir:354] Seminar on "User behavior analysis for IR" Dec.17, 13:30-15:00


Dear all,

We will have a seminar talk on "User behavior analysis for
information retrieval" by Dr. Min Zhang, Tsinghua University,
on December 17th, tomorrow.

You are most welcome to join us with the seminar and discussion!

If you will attend the seminar, please fill in the form at the
end of the English message and email us.

(Japanese message will be find after the English one)


**********************************************************
Seminar on "User behavior analysis for information retrieval"
                                    by Dr. Min Zhang
**********************************************************

Date/Time:
  Thursday, December 17, 2009.
  13:30 - 15:00

Venue:
  NII, Lecture Room 2008 (20th floor)

Title:
  User behavior analysis for information retrieval

Abstract:
  In this talk I'll introduce our research progress on the user
behavior analysis for information retrieval. First, I'll talk
on the search engine re-ranking model based on user behavior
information. According to the wisdom of the crowd, we propose
a user click information fusion algorithm with the multiply
search engine user logs. Evaluation on the real world search
engine results shows that consistent and significant improvement
have been achieved. Second, we also do not give up the study
on long-tail and rare queries. I'll introduce our research on
exploring reliable and relevant user clicks (RCs) in
click-through data. We proposed several features to separate
RCs from click noises. A learning algorithm is presented to
estimate the quality of user clicks. Experimental results on
large scale dataset shows the significant effectiveness of
the proposed model. This work is the first, to the best of our
knowledge, to interpret user click based on individual user
behavior. And it gives an initial solution to click quality
estimation for both hot and long-tail queries. Finally,
a large scale Web information retrieval evaluation dataset
will be introduced, which contains more than 10 billion web
pages with 5 Terabytes, 10,000 user queries and the related
answers.


Bio:
  Dr. Min Zhang is an associate professor in the Department of
Computer Science & Technology (DCST), Tsinghua University.
She received her Bachelor and PhD degrees from DCST at
Tsinghua University in 1999 and 2003, respectively. Dr. Zhang
specializes in information retrieval, Web user behavior analysis
and machine learning. She has published more than 50 papers on
journals and conferences, such as JASIST, IR, SIGIR, WWW, CIKM,
etc. She has participated in TREC (Text REtrieval Conference)
as the team leader since 2002. Her team has continuously achieved
multiple top performances during the past 8 years. Dr. Zhang is
also the vice director of Tsinghua-Sohu Joint research Laboratory
of Search Technology, and has served as the PC member in WWW,
KDD, WSDM, EMNLP, IJCNLP, AIRS, etc.

--
Academic Host: Noriko Kando, Professor, NII (ex. 2529)
***********************************************************

********************************
To: kando-secr@xxxxxxxxx

Application form to attend the seminar talk on Dec. 17, 2009
TITLE: "User behavior analysis for information retrieval"

I will attend the seminar talk at 13:30-15:00, Dec. 17, 2009
・Your name:　　　
・Affliation:
・E-mail address:
********************************


*** Japanese message ***

中国の清華大学から、注目の女性若手研究者 Min Zhang博士を
お招きして講演会を開催します。中国の人気ポータルサイト捜狗(Sofu)
との共同研究による膨大なクリックログを用いた検索ランキングの
改善と利用者行動の分析に関する非常に興味深い研究成果について
ご講演くださる予定です。

直前のお知らせとなってしまい、はなはだ恐縮ですが、ぜひ、
奮ってご参加ください。

みなさまのご参加をお待ちしております。

周囲のご関心の方々にも、ご紹介くださると幸いです。

ご参加くださる方は、お手数ですが、資料の準備の都合上、
可能ならば末尾の書式でお申込みください。
お申込みなしの当日参加も歓迎です。

**********************************************************
Seminar on "User behavior analysis for information retrieval"
                                    by Dr. Min Zhang
**********************************************************

日時:　2009年12月17日（木）13:30 - 15:00
場所：国立情報学研究所 20階　講義室2008

講演タイトル:
  User behavior analysis for information retrieval

Abstract:
  In this talk I'll introduce our research progress on the user
behavior analysis for information retrieval. First, I'll talk
on the search engine re-ranking model based on user behavior
information. According to the wisdom of the crowd, we propose
a user click information fusion algorithm with the multiply
search engine user logs. Evaluation on the real world search
engine results shows that consistent and significant improvement
have been achieved. Second, we also do not give up the study
on long-tail and rare queries. I'll introduce our research on
exploring reliable and relevant user clicks (RCs) in
click-through data. We proposed several features to separate
RCs from click noises. A learning algorithm is presented to
estimate the quality of user clicks. Experimental results on
large scale dataset shows the significant effectiveness of
the proposed model. This work is the first, to the best of our
knowledge, to interpret user click based on individual user
behavior. And it gives an initial solution to click quality
estimation for both hot and long-tail queries. Finally,
a large scale Web information retrieval evaluation dataset
will be introduced, which contains more than 10 billion web
pages with 5 Terabytes, 10,000 user queries and the related
answers.


Bio:
  Dr. Min Zhang is an associate professor in the Department of
Computer Science & Technology (DCST), Tsinghua University.
She received her Bachelor and PhD degrees from DCST at
Tsinghua University in 1999 and 2003, respectively. Dr. Zhang
specializes in information retrieval, Web user behavior analysis
and machine learning. She has published more than 50 papers on
journals and conferences, such as JASIST, IR, SIGIR, WWW, CIKM,
etc. She has participated in TREC (Text REtrieval Conference)
as the team leader since 2002. Her team has continuously achieved
multiple top performances during the past 8 years. Dr. Zhang is
also the vice director of Tsinghua-Sohu Joint research Laboratory
of Search Technology, and has served as the PC member in WWW,
KDD, WSDM, EMNLP, IJCNLP, AIRS, etc.

--
Academic Host: Noriko Kando, Professor, NII (ex. 2529)
***********************************************************


********************************
To: kando-secr@xxxxxxxxx

12月17日講演会　参加申し込み
TITLE: "User behavior analysis for information retrieval"

12月17日（木）13:30-15:00　の講演会に参加します。
・お名前　　　
・ご所属
・Email
********************************



------------------------------------------
This message is brought by NTCIR and distributed through the mailing list:
ntcir [at] nii.ac.jp for announcements and updates about NTCIR.
* To unsubscribe, send mail to ntc-secretariat [at] nii.ac.jp,
with the subject:ntcirML UNSUBSCRIBE [your e-mail address]
* For more info, visit:
http://research.nii.ac.jp/ntcir/
------------------------------------------
Noriko Kando, Leader, NTCIR Project
ntc-admin [at] nii.ac.jp

NTCIR Project Office
ntc-secretariat [at] nii.ac.jp

National Institute of Informatics
2-1-2 Hitotsubashi,
Chiyoda-ku, Tokyo 101-8430, Japan
---------------------------------------------