NTCIR, Noriko Kando, NII, Japan

advertisement
What is happening at NTCIR
Thanks for Teruko Mitamura, Tetsuya Sakai, Fred Gey,
Yohei Seki, Daisuke Ishikawa, Atsushi Fujii, Hidetsugu
Nanba, Terumasa Ehara for preparing slides
Noriko Kando
National Institute of Informatics, Japan
http://research.nii.ac.jp/ntcir/
kando@nii. ac. jp
ntcir at clef 2010-09-21
Noriko Kando
1
NTCIR:
NII Test Collection for Information Retrieval
Research Infrastructure for Evaluating IA
A series of evaluation workshops designed to
enhance research in information-access technologies by
providing an infrastructure for large-scale evaluations.
■Data sets, evaluation methodologies, and forum
Project started in late 1997
Once every 18 months
Data sets (Test collections or TCs)
Scientific, news, patents, and web
Chinese, Korean, Japanese, and English
Tasks (Research Areas)
IR: Cross-lingual tasks, patents, web, Geo
QA:Monolingual tasks, cross-lingual tasks
Summarization, trend info., patent maps
Opinion analysis, text mining
Community-based Research Activities
ntcir at clef 2010-09-21
Noriko Kando
2
Tasks at Past NTCIRs
NTCIR
1
2
3
4
5
6
7
8
'99 '01 '02 '04 '05 '07 '08 '09User Generated
■ Community QA
Contents
■ ■ ■ Opinion Analysis
■ ■ Cross-Lingual QA + IR
Module-Based
IR for Focused
Domain
■ Geo Temporal
■ ■ ■ ■ □
Patent
■ ■ ■ Complex/ Any Types
Question
Answering
Summarization /
Consolidation
Web
Crosslingual
Retrieval
Text Retrieval
■ ■
Dialog
■ ■ ■ ■ Cross-Lingual
Factoid, List
■ ■ ■ ■
■ ■ ■ ■ ■ Text Mining / Classification
Trend Info Visualization
■ ■ ■
■ ■ ■
■ ■ ■
Text Summarization
Web
■ ■ Statistical MT
■ ■ ■ ■ ■ ■ ■ ■ Cross-Lingual IR
■ ■ ■ ■ ■ ■ ■ ■ Non-English Search
■ ■ ■ ■ ■ ■ ■ ■ Ad Hoc IR, IR for QA
The Years the meetings were held. The tasks started 18 months before
ntcir at clef 2010-09-21
Noriko Kando
3
NTCIR-8 Tasks (2008.07—2009.06)
rd Int’l W on Evaluating
The
Information Access (EVIA ) refereed
3
1. Advanced CL Info Access
- QA(CsCtJE->CsCtJ) ※Any Types of
Questions
GeoTime
(E, (CsCtJE->CsCtJ
J) Geo Temporal
Information
- IR for QA
)
2. User Generated Contents (CGM)
- Opinion Analysis (Multilingual News) (Blog)
- [Pilot] Community QA (Yahoo! Chiebukuro)
3. Focused Domain:Patent
-Patent Translation; English -> Japanese
New
New
※Statistical MT, The World-Largest training data (J-E sentence
alignment), Summer School, Extrinsic eval by CLIR
-Patent Mining papers -> IPC
-Evaluation of SMT
ntcir at clef 2010-09-21
New
Noriko Kando
4
NTCIR-7 & -8 Program Committee
Mark Sanderson, Doug Oard, Atsushi Fujii, Tatsunori Mori,
Fred Gey, Noriko Kando
(and Ellen Voorhees, Sung Hyun Myaeng, Hsin-Hsi Chen,
Tetsuya Sakai)
ntcir at clef 2010-09-21
Noriko Kando
5
NTCIR-8 Coordination
• NTCIR-8 is coordinated by NTCIR Project at NII,
Japan. The following organizations contribute to the
organization of NTCIR-8 as Task Organizers
- Academia Sinica
- Carnegie Mellon Univ
- Chinese Academy of Science
- Hiroshima City University
- Hitachi, Co Ltd.
- Hokkai Gakuen University
- IBM
- Microsoft Research Asia
- National Institute of
Information and Communication
Technology
- National Institute of
Informatics
- National Taiwan Univ
ntcir at clef 2010-09-21
-National Taiwan Ocean Univ
- Oki Electonic Co.
- Tokyo Institute of
Technology
- Tokyo Univ
- Toyohashi Univ of
Technology and Science
- Univ of California Barkeley
-Univ of Tsukuba
- Yamanashi Eiwa College
- Yokohama National
University
Noriko Kando
6
[GeoTime]
Dublin City Univ
Carnegie Mellon Univ
Hokkaido Univ
Dalian Univ of Technology
INESC-ID, Porugal
National Taiwan Ocean Univ
International Inst of Technology,
Shenyan Institute of
Hyderbad
Aeronautical Engineering Kieo Univ
Univ of Tokushima
Nataional Inst of Materials Science
Wuhan Univ
Osaka Kyoiku Univ
Univ California, Berkeley
[IR4QA]
Univ of Iowa
Carnegie Mellon Univ
Univ of Lisbon
Chaoyang Univ of Technology Yokohama City Univ
Dalian Univ of Technology
Dublin City Univ
[MOAT]
Inner Mongolia Univ
Beijing Uni of Posts and
Queensland Univ of
Telecommunications
Technology
Chaoyang Univ of Technology
Shenyan Inst of Aeronautical Chinese Univ of HK+ Tsinghua Univ
Engineering
City Univ of Hong Kong (2 groups)
Trinity College Dublin
Hong Kong Polytechnic Univ
Univ California, Berkeley
KAIST
Wuhan Univ
National Taiwan Univ
Wuhan Univ (Computer
NEC Laboratories China
School)
Peking Univ
Wuhan Univ of Science and
Pohang Univ of Sci and Tech
Technology
SICS
Toyohashi Univ of Technology
Univ of Alicante
Univ of Neuchatel
NTCIR-8 Active
Yuan Ze Univ
Participants
ntcir at clef 2010-09-21
Noriko Kando
[CCLQA]
[Patent Mining]
Hiroshima City Univ
Hitachi, Ltd.
IBM Japan, Ltd.
Institute of Scientific and
Technical Information of China
KAIST
National Univ of Singapore
NEC
Shanghai Jiao Tong Univ
Shenyang Institute of Aeronautical
Engineering
Toyohashi Univ of Technology
Univ of Applied Sciences - UNIGE
[Patent Translation]
Dublin City University, CNGL
Hiroshima City University
Kyoto University
NiCT
Pohang Univ of Sci and Tech
tottori university
Toyohashi University of Technology
Yamanashi Eiwa College
[Community QA]
Microsoft Research Asia
National Institute of Informatics
Shirayuri College
7
Complex Cross-lingual Question Answering
(CCLQA) Task
Different teams
can exchange
and create a
“dream-team”
QA system
Small teams that
do not possess an
entire QA system
can contribute
IR and QA communities can collaborate
http://aclia.lti.cs.cmu.edu/ntcir8/
ntcir at clef 2010-09-21
Noriko Kando
8
Evaluation Topics
– any types of questions Type
#
Example Question
Related Past
NTCIR Task
DEFINITION
10
What is the Human Genome Project?
ACLIA
BIOGRAPHY
10
Who is Howard Dean?
ACLIA
RELATIONSHIP
20
What is the relationship between
Saddam Hussein and Jacques Chirac?
ACLIA
EVENT
20
What are the major conflicts
between India and China on border
issues?
ACLIA
WHY
20
Why doesn't U.S. ratify the Kyoto
Protocol?
QAC-4
PERSON
5
Who is the Finland's first woman
president?
QAC 1-3, CLQA
1,2
ORGANIZATION
5
What is the name of the company
that produced the first Fairtrade
coffee?
QAC 1-3, CLQA
1,2
LOCATION
5
What is the name of the river that
separates North Korea from China?
QAC 1-3, CLQA
1,2
DATE
5
When did Queen Victoria die?
QAC 1-3, CLQA
1,2
9
ntcir
at clef 2010-09-21
Noriko Kando
CT/JA-T IR4QA run rankings
ntcir at clef 2010-09-21
Mean Q
Mean nDCG
Mean Q
Mean nDCG
Noriko Kando
10
CCLQA Human Evaluation Preliminary Results: JA-JA
JA-JA Runs
ALL
LTI-JA-JA-01-T
0.1069
LTI-JA-JA-02-T
0.1443
LTI-JA-JA-03-T
0.1438
JA-JA automatic evaluation
JA-JA Runs
ALL
LTI-JA-JA-01-T
0.2024
LTI-JA-JA-02-T
0.2259
LTI-JA-JA-03-T
0.2252
11 at clef 2010-09-21
ntcir
Noriko Kando
Effect of Combination;
IR4QA+CCLQA JA-JA Collaboration Track: F3 score
based on automatic evaluation
CCLQA
LTI
IR BRKLY-JA-JA-01-DN
4 BRKLY-JA-JA-02-T
Q
BRKLY-JA-JA-03-DN
A
BRKLY-JA-JA-04-DN
BRKLY-JA-JA-05-T
12 at clef 2010-09-21
ntcir
0.2934
0.2686
0.2074
0.3000
0.2746
Noriko Kando
NTCIR-GEOTIME
GEOTEMPORAL INFORMATION RETRIEVAL
(New Track in NTCIR Workshop 8)
Fredric Gey and Ray Larson and Noriko Kando
Relevance judgments system by Jorge Machado and Hideki Shima
Evaluation : Tetsuya Sakai
Judgments: U Iowa, U Lisbon, U California Barkelay, NII
Search with a specific focus on Geography + To distinguish
from past GIR evaluations, we introduced a temporal
component
Asian language geographic search has not previously been
evaluated, even though about 50 percent of the NTCIR-6
Cross-Language topics had a geographic component (usually a
restriction to a particular country).
ntcir at clef 2010-09-21
Noriko Kando
13
NTCIR-GeoTime
PARTICIPANTS
• Japanese Runs submitted by eight groups (two anonymous
pooling supporters)
Team Name
Organization
Anonymous
Anonymous
BRKLY
University of California, Berkeley,
USA
FORST
Yokohama National University, JAPAN
HU-KB
Hokkaido University, JAPAN
KOLIS
Keio University, JAPAN
ANON2
Anonymous submission, group 2
M
National Institute of Materials
Science, JAPAN
OKSAT
Osaka Kyoiku University, JAPAN
ntcir at clef 2010-09-21
Noriko Kando
14
NTCIR-GeoTime
PARTICIPANTS
• English Runs submitted by six groups
Team Name
Organization
BRKLY
University of California, Berkeley,
USA
DCU
Dublin City University, IRELAND
IITH
International Institute of Technology,
Hyderbad, INDIA
INESC
National Institute of Electroniques
and Computer Systems, Lisbon,
PORTUGAL
UIOWA
University of Iowa, USA
XLDB
University of Lisbon, PORTUGAL
ntcir at clef 2010-09-21
Noriko Kando
15
NTCIR-GeoTime Approached
• BRKLY: baseline approach, probablistic + psued
relevance feedback
•DCU, IITH, XLDB (U Lisbon) : geographic
enhancements
•KOLIS (Keio U) : counting the number of
geographic and temporal expression in top-ranked
docs in initial search, then re-rank
•FORST (Yokohama Nat U): utilize factoid QA
technique to question decomposition
•HU-KB (Hokkaido U), U Iowa: Hybrid approach
combinging probablistic model and weighted
boolean
query formulation
ntcir
at clef 2010-09-21
Noriko Kando
16
NTCIR-GeoTime: ENGLISH TOPIC DIFFICULTY by Average
Precision
Per-topic AP, Q and nDCG averaged over 25 English runs
for 25 topics sorted by topic difficulty (AP ascending)
1
0.9
0.8
0.7
0.6
AP
Q
0.5
nDCG
0.4
0.3
0.2
0.1
0
21 18 22 15 10 12 20 5 11 25 17 4 24 2 14 9 16 23 13 6
8 19 3
7
1
Most Difficult English topic
(21): When and where were 17
Noriko Kando
the 2010 Winter Olympics host city location announced?
ntcir at clef 2010-09-21
NTCIR-GeoTime: JAPANESE TOPIC DIFFICULTY by Average
Precision
Per-topic AP, Q and nDCG averaged over 34 Japanes
runs for 24 topics sorted by topic difficulty (AP ascending)
0.9
0.8
0.7
0.6
0.5
AP
0.4
Q
0.3
nDCG
0.2
0.1
0
1814 25222012 15231311 24 4 2 8 191016 5 21 6 1 9 3 7
Most Difficult Japanese topic 18: What date was a
ntcir at clef 2010-09-21
country was invaded by Noriko
the Kando
United States in 2002?
18
NTCIR-GeoTime: TOPIC per TOPIC Analysis on Japanese
1.00
HU-KB-JA-JA03-D
0.80
KOLIS-JA-JA04-D
FORST-JA-JA04-D
nDCG
0.60
BRKLY-JA-JA02-D
0.40
0.20
0.00
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 18 19 20 21 22 23 24 25
Topic ID
ntcir at clef 2010-09-21
(#17 is missing)
Noriko Kando
19
NTCIR-GeoTime CHALLENGES
• Geographic reference resolution is difficult
enough, but
• More difficult to process temporal expression
(“last Wednesday”) references
•Can indefinite answers be accepted (“a few hours”)?
•Need Japanese Gazetteers
•Need NE annotated corpus for further refinement
ntcir at clef 2010-09-21
Noriko Kando
20
Multilingual Opinion Analysis
(MAOT)
CO2 Reduction?
Lehman
shock?
ntcir at clef 2010-09-21
Noriko Kando
21
Opinion Question List
ID
N01
(N03)
N04
N05
N06
N07
N08
N11
N13
N14
N16
N17
Opinion Question
What negative prospects were discussed about the Euro when it was introduced in January of 2002?
(What reasons were discussed about Bomb Terror in Bali Island in October, 2002?)
What reasons have been given for the Space Shuttle Columbia accident in February, 2002?
What negative comments were discussed about Bush's decision to start Iraq war in March, 2003?
What negative prospects and opinions were discussed about SARS which started spreading in March, 2003?
What reasons are given for the blackout around North America in August, 2003?
What reasons and background information was discussed about the terrorist train bombing that happened in Madrid in March, 2004?
Why did supporters want to elect George W. Bush in the November 2004 American Presidential Election?
What positive comments were discussed to help the victims from earthquake and tsunami in Sumatera, Indonesia in December,
2004?
What objections are given for the US opposition to the Kyoto Protocol that was enacted in February 2005?
What reasons have been given for the anti-Japanese demonstrations that took place in April, 2005 in Peking and Shanghai in
China?
In July 2005 there were terrorist bombings in London. What reasons and background were given, and what controversies were
discussed?
N18
What actions by President George Bush were criticized in response to Hurricane Katrina's August 2005 landing?
N20
What negative opinions and discussion happened about the Bird Flu that started spreading in October, 2005?
Identify opinions that indicate that Arnold Schwarzenegger is a bad choice to be elected the new governor of California in the
October 2003 election.
Find positive opinions about the reaction of Nuclear and Industrial Safety Agency officials to the Mihama nuclear powerplant
accident in August 2004.
What were the advantages and disadvantages of the direct flight between Taiwan and Mainland China commercially?
What are good and bad approaches to losing weight?
What are complaints about XIX Olympic Winter Games that were held in and around Salt Lake City, Utah, United States in 2002?
What are the comments about China's first manned space flight which happened successfully in October 2003?
What negative comments were discussed when in April 2004 CBS made public pictures showing cruel U.S. military abuse of Iraqi
prisoners of war?
N24
N26
N27
N32
N36
N39
N41
ntcir at clef 2010-09-21
Noriko Kando
22
Effective approaches
• The following teams attained good results
with accomplished feature filtering, hot
machine learning, and rich lexicon resources.
TeamID
UNINE
PKUTM
CityUHK
Lang
EN
SC
TC
Feature Filtering
Macihne Learing
Z score
logistic regression
Iterative classifier
SVM (better than NB, ME, DT)
Supervised Lexicon Ensemble
ntcir at clef 2010-09-21
Noriko Kando
Lexicon Resource
SentiWordNet
In House, NTU, and Jun LI's lexicon
NTUSD, LCPW, LCNW, CPWP, SKPI
23
Community QA Pilot Task
Rank all posted answers by answer quality
(as estimated by system) for every question.
Training
Best Answer:
Questioner
selected
Test
Test
collection
for CQA:
1500
questions
Yahoo Chiebukuro
data version 1.0:
3 million questions
1500
questions
selected at
random
http://research.nii.ac.jp/ntcir/ntcirws8/yahoo/index-en.html
ntcir at clef 2010-09-21
24
Good
Answers:
4 assessors
individually
assessed
4 university students
Noriko Kando
GA-Hit@1
not useful
1
0.9
GA-nDCG
and Q
similar
5th run from MSRA
0.8 used the BA info for
test
0.7 Qs directly so does
not represent
0.6Practical performance
In general,
BA-Hit@1 and
GA metrics agree
with each other.
0.5
0.4
0.3
0.2
0.1
BA-Hit@1
GA-Hit@1
GA-nG@1
GA-nDCG
GA-Q
Runs
sorted by
GA-nG@1
ntcir at clef 2010-09-21
Noriko Kando
25
120
#questions
100
LOVE is HARD
according to BA!
Systems can’t find
the asker’s BAs!
BA-
80
60
easy
medium
40
hard
20
0
ntcir at clef 2010-09-21
Noriko Kando
26
120
#questions
100
80
LOVE is EASY according to
GA!
GA-nG@1
Systems can find many good
answers that are not BA!
(BA-evaluation not good enough
for social questions)
60
easy
medium
40
hard
GA-based evaluation alleviates
the bias and nonexhaustiveness
problems of BA.
GA-based evaluation is more
discriminative - detects system
differences that BA overlooks.
20
0
ntcir at clef 2010-09-21
Noriko Kando
27
Goal: Automatic Creation of technical
trend maps from a set of research
papers and patents.
Effect 1
Technology [AAA 1993]
1
[US Pat. XXXXX]
Technology
2
Technology
[CCC 2000]
3
Effect 3
Effect 2
[BBB 2002]
[DDD 2001]
[US Pat. YY-YYY]
[US Pat. ZZZZZ]
[US Pat. WWW]
Research papers and patents are classified in
terms of elemental technologies and their effects.
ntcir at clef 2010-09-21
Noriko Kando
28
Evaluation
Subtask 1 (Research Paper Classification)
Metrics: Mean Average Precision (MAP)
• k-NN based approach is superior to machine learning
approach.
• Re-ranking of IPC codes is effective.
Subtask 2: Technical Trend Map Creation
Metrics: Recall, Precision, and F-measure
• Top systems employed CRF, and the following
features are effective.
– Dependency structure
– Document structure
– Domain adaptation
ntcir at clef 2010-09-21
Noriko Kando
29
History of Patent IR at NTCIR
• NTCIR-3 (2001-2002)
– Technology survey
• Applied conventional IR
problems to patent data
2 years of JPO
patent applications
* JPO = Japan Patent Offic
• NTCIR-4 (2003-2004)
5 years of JPO
– Invalidity search
patent
applications
Both
document
sets were
• Addressed patent-specific published in 1993-2002
IR problems
• NTCIR-5 (2004-2005)
10 years of JPO
– Enlarged invalidity search patent applications
• NTCIR-6 (2006-2007)
– Added English patents
ntcir at clef 2010-09-21
10 years of USPTO
patents granted
* USPTO = US Patent & Trademark Offi
Noriko Kando
30
Extrinsic E-J: BLEU & IR measures
0.9
MAP & Recall@N
0.8
MAP,Recall@
0.7
Recall@1000 (R = 0.77)
Recall@500 (R = 0.84)
Recall@200 (R = 0.86)
Recall@100 (R = 0.86)
MAP (R = 0.77)
0.6
0.5
Recall@100,200 are highly
correlated with BLEU (R = 0.86)
0.4
0.3
0.2
0.1
12.00
14.00
16.00
18.00
20.00
22.00
24.00
26.00
BLEU
ntcir at clef 2010-09-21
Noriko Kando
31
Overall structure
General
Chairs
Evaluation
Task Chairs
+ PCs
Core
Challenge
Task
Grand
Challenge
Task
EVIA Chairs
+ PCs
???
New NTCIR Structure
NTCIR general chairs:
Noriko Kando (NII), Eiichiro Sumita (NICT), Tsuneaki Ka
NTCIR evaluation chairs:
Hideo Joho (U of Tsukuba)
Tetsuya Sakai (MSRA)
Task Selection
Committee
Grand challenges
Pilot task
Pilot task
Core challenge ta
Core challenge ta
EVIA chairs:
Mark Sanderson
(RMIT)
William Webber (U
EVIA
Program + 1
of Melbourne)
Committee
Refereed papers on
Evaluating
Information Access
Tasks accepted for NTCIR-9
CORE TASKS
• Intent (with One-Click Access subtask)
• Recognizing Natural Language Inference
• Geotemporal information retrieval (GeoTime2)
• IR for Spoken Documents
PILOT TASKS
• Cross-lingual Link Discovery
• Evaluation Framework of Interactive Information
Access
• Patent Machine Translation (PATMT)
INTENT (w 1CLICK) task (CS,
J)
Organisers: Min Zhang, Yiqun Liu (Tsinghua U), Ruihua
Song, Tetsuya Sakai, Youngin Song (MSRA), Makoto
Kato (Kyoto U)
Harry
Potter
• Subtopic mining
book
Find different intents given an
book
book
ambiguous/underspecified query
movie
• Ranking
Selectively diversity Web search results
• One click access (1CLICK)
Satisfy information need with the system’s first
result page (X-byte text, not document list)
No need to click after
clicking on SEARCH
Recognizing Natural Language
Inference
(CS, CT, J)
Organisers: Hideki Shima,
Teruko Mitamura (CMU),
Chuan-Jie Lin (NTOU),
Cheng-Wei Lee (Academia Si
• YES/NO task
Does text 1 entail text 2?
text 1 ⇒ text 2
• Main task
Given text 1 and text 2, choose from
(1) Forward entailment (2) backward entailment
(3) equivalent (4) contradict (5) independent
• Extrinsic evaluation via CLIA will be
conducted
GeoTime2 (E, J, K?)
Organisers: Fred Gey, Ray Larson (UCB), Noriko Kando
(NII)
• Second round of ad hoc IR for WHEN and
WHERE
• GeoTime1 topic: How old was Max Schmeling
when he died, and where did he die?
At GeoTime1, docs that contain both WHEN and
WHERE were treated as relevant;
Those that contain only WHEN or only WHERE
info were treated as partially relevant.
Can we do better?
Can we create more realistic
IR for Spoken Documents (J)
Organisers: Tomoyosi Akiba, Seiichi Nakagawa, Kiyoaki Aikawa (Toyohashi U of
Technology), Yoshiaki Itoh (Iwate Prefectural U), Tatsuya Kawahara (Kyoto U),
Xinhui Hu (NICT), Hiroaki Nanjo (Ryukoku University), Hiromitsu Nishizaki (U of
Yamanashi), Tomoko Matsui (Institute of Statistical Mathematics), Yoichi Yamashita
(Ritsumeikan U)
• Handling spontaneous speech data (spoken lectures)
• Spoken term detection
Find the position of a given query term within SD
• Spoken document retrieval
Find passages in SD data that match a given query
Reference speech recognition results
provided (non-speech people can
easily participate)
Cross-lingual Link Discovery (E>C,K,J)
Organisers: Ling-Xiang Tang, Shlomo Geva, Andrew Trotman, Yue Xu,
(Queesland U of Technology), Andrew Trotman (U of
Otago) INEX
Linking
and NTCIR!
• Given an English Wikipedia page,
(1)Identify anchors; and
(2)For each anchor, provide a list of C/K/J
documents to be linked
太陽
Solar eclipse
(1)
sun
(2)
日食
Corresponding
Lunar eclipse W entry
太陽
月食
月食
Evaluation Framework of Interactive
Information Access (J)
Organisers: Tsuneaki Kato (U Tokyo) and Mitsunori
Matsushita (Kansai U)
• Explore evaluation of interactive and
exploratory information access
• Shared modules, shared interfaces between
modules
• Interactive complex question answering
subtask (visual/textual output)
• Trend information summarization subtask
(e.g. cabinet support rate change)
PATMT (C-E, J-E, E-J)
Organisers: Benjamin Tsou, Bin Lu (CUHK), Isao Goto
(NICT)
• C-E MT (new at NTCIR-9), J-E MT, E-J
MT
Manual and automatic evaluation (BLEU,
NIST)
• PATMT@NTCIR-10
will include extrinsic
High
Low
Human
MAP
BLEU
evaluation
by CLIR
rating
High
• Lessons from PATMT@NTCIR-7,8:
with multiple reference translations (RTs)
TALK OUTLINE
•
•
•
•
About myself
Looking Back
NTCIR
NTCIR-5,6 CLIR (as a participant)
NTCIR-7,8 ACLIA
NTCIR-8 GeoTime
Looking Forward
NTCIR Grand Challenges
NTCIR-9 Tasks
Summary
TALK SUMMARY
• Looking back (1999-2010)
- More teams are now using straight MT for CLIR.
Often over 90% of monolingual performance.
What are CLIR researchers doing now? CLIA?
• Looking forward (2010-)
- Grand challenges [tentative] = NN2S + E2E =
No Need To Search + Easy To Explore.
- Multilinguality will still be important especially in
the system2user direction.
Please participate in NTCIR-9 tasks!
NTCIR-9 Important Dates
(tentative)
1st Oct 2010 first call for participation
Winter-Autumn 2011
tasks run
November 2011
camera-ready papers due
December 2011
NTCIR-9@NII, Tokyo
Watch http://research.nii.ac.jp/ntcir
Thanks
Merci
Danke schön
Gracie
Gracias Ta! Tack
Köszönöm
Kiitos
Terima Kasih
Khap Khun
Ahsante
Tak
謝謝
ありがとう
http://research.nii.ac.jp/ntcir/
Will be moved to: http://ntcir.nii.ac.jp/
ntcir at clef 2010-09-21
Noriko Kando
45
Download