=Paper=
{{Paper
|id=Vol-1169/CLEF2003wn-other-Kando2003
|storemode=property
|title=Evaluation of Information Access Technologies at NTCIR Workshop
|pdfUrl=https://ceur-ws.org/Vol-1169/CLEF2003wn-other-Kando2003.pdf
|volume=Vol-1169
|dblpUrl=https://dblp.org/rec/conf/clef/Kando03a
}}
==Evaluation of Information Access Technologies at NTCIR Workshop==
https://ceur-ws.org/Vol-1169/CLEF2003wn-other-Kando2003.pdf
Evaluation of Information Access Technologies at NTCIR Workshop
Noriko Kando
National Institute of Informatics (NII), Tokyo
kando@nii.ac.jp
Abstract: This paper introduces the NTCIR Workshops, a series of evaluation workshops that are designed to
enhance research in information access technologies, such as information retrieval, text summarization, question
answering, text mining, etc., by providing infrastructure of large-scale evaluation. A brief history, test collections,
and recent progress after the previous CLEF Workshop are described with highlighting the difference from
CLEF in this paper. To conclude, some thoughts on future directions are suggested.
1 Introduction
The NTCIR Workshops [1]1 are a series of evaluation workshops designed to enhance research in information access (IA)
technologies including information retrieval (IR), cross-lingual information retrieval (CLIR), automatic text summarization,
question answering, text mining, etc.
The aims of the NTCIR project are:
1. to encourage research in information access technologies by providing large-scale test collections reusable for
experiments,
2. to provide a forum for research groups interested in cross-system comparisons and exchanging research ideas in an
informal atmosphere, and
3. to investigate methodologies and metrics for evaluation of information access technologies and methods for
constructing large-scale reusable test collections.
That is to say, the main goal of the NTCIR project is to provide infrastructure of large-scale evaluation. The importance of
large-scale evaluation infrastructure in IA research has been widely recognized. Fundamental text processing procedures for IA
such as stemming and indexing include language-dependent procedures. In particular, processing texts written in Japanese or
other East Asian languages such as Chinese is quite different from processing English, French or other European languages,
because there are no explicit boundaries (i.e., no spaces) between words in a sentence. The NTCIR project therefore started in
late 1997 with emphasis on, but not limited to, Japanese or other East Asian languages, and its series of workshops has attracted
international participation.
1.1 Information Access
The term “information access” (IA) includes a whole process to make information in the documents usable for the user who has
problems or information needs. A traditional IR system returns a ranked list of retrieved documents that are likely to contain
information relevant to the user’s needs. This is one of the most fundamental and core processes of IA. It is however not the end
of the story for the users. After obtaining a ranked list of retrieved documents, the user skims the documents, performs relevance
judgments, locates the relevant information, reads, analyses, compares the contents with other documents, integrates, summarizes
1 NTCIR-3 and 4 are sponsored by the National Institute of Informatics (NII) and Japanese MEXT Grant-in-Aid for Scientific Research on
Informatics (#13224087) in and after FY2001. Patent task is organized by collaboration with Japan Intellectual Property Right Association and
NII, and CLIR Task is organized by collaboration with National Taiwan University, Korean Institute for Scientific and Technological
Information (KISTI).
and performs information-based work such as decision making, problem solving, writing, etc., based on the information obtained
from the retrieved documents. We have looked at IA technologies to help users utilize the information in large-scale document
collections. IR, summarization, question answering, etc are a “family”, in which the same target is aimed while each of the
technologies has been investigated by different communities with least interaction2.
1.2 Focus of the NTCIR
As shown in Figure 1, we have looked at both traditional laboratory-type IR system testing and the evaluation of challenging
technologies. For the laboratory-type testing, we placed emphasis on IR and CLIR with Japanese or other Asian languages and
testing on various document genres. For the challenging issues, the targets are the shift from document retrieval to technologies
that utilize “information” in documents, and investigation of methodologies and metrics for more realistic and reliable evaluation.
For the latter, we have paid attention to users’ information seeking task in the experiment design. These two directions have been
supported by a forum of researchers and discussion among them.
From the beginning, CLIR has been one of the central interests of the NTCIR, because CLIR between English and
own-languages is critical for international information transfer in Asian countries, and it was challenging to perform CLIR
between languages with completely different structures and origins such as English and Chinese or English and Japanese.
Focus of NTCIR
Lab-typed IR Test New Challenges
Asian Languages/cross-language Intersection of IR + NLP
Variety of Genre To make information in the
documents more usable for
Parallel/comparable Corpus users!
Realistic eval/user task
Forum of Researchers
Idea Exchange
Discussion/Investigation on
Evaluation methods/metrics
Figure 1. Focus of NTCIR
In the following, the next section provides a brief history of NTCIR. Section 3 describes NTCIR Test Collections, Section 4
reports recent progress after our reports at the previous CLFEs [2-4], and Section 5 outlines the features of the coming NTCIR
Workshop, NTCIR-4. Section 6 is summary.
2 NTCIR
2.1 History of NTCIR
In the NTCIR, a workshop is held once per about one and half years. Since we respect the interaction between participants, we
call a whole the process from document release to the final meeting as “workshop”. Each workshop selects several research areas
called “Task”, or "Challenge" for more challenging task. Each task has been organized by the researchers of the domain and a
task may consist of more than one subtasks. Figure 2 shows the evolution of the tasks in the NTCIR Workshops and Table 1 is a
list of subtasks and test collections used in the tasks [5-7].
As shown in Table 1, the 4th NTCIR Workshop hosts 5 tasks, CLIR, Patent Retrieval Task (PATENT), Question Answering
Challenge (QAC), Text Summarization Challenge (TSC), and WEB Task (WEB) and their sub-tasks.
2 In addition to the above, how to define the question of the user before the retrieval is also included in the scope of the IA although it has not
been explicitly investigated in NTCIR.
Tasks (Research Areas) of NTCIR Workshops
1st 2nd 3rd 4th
Japanese IR Start！
t Cross-lingual IR
a
s Patent Retrieval
k
s Web Retrieval
Term Extraction/
Role Analysis
Question
Answering
Text
Summarization
Nov 98 Apr 2003
About once per 1 ½ years
Figure 2. Tasks of NTCIR Workshops
Table 1. History of NTCIR Workshops
Period Tasks Subtasks Test collections
Ad Hoc IR J-JE
Nov.1998-
1 CLIR J-E NTCIR-1
Sept.1999
Term Extraction Term Extraction/ Role Analysis
Chiniese IR: C-C
Chinese Text Retrieval CIRB010
CLIR: E-C
June 2000-
Monolingual IR: J-J, E-E
2 March Japanese&English IR NTCIR-1, -2
CLIR: J-E, E-J, J-JE, E-JE
2001
Intrinsic - Extraction/Free generated
Text Summarization NTCIR-2Summ
Extrinsic - IR task-based
Single Language IR:C-C,K-K,J-J
CLIR Bilingual CLIR:x-J,x-C, x-K NTCIR-3CLIR
Mulilingual CLIR:x-CJE
Cross Genre w/ or wo CLIR CCKE-J
Patent NTCIR-3 PATENT
[Optional] Alianment, RST Analysis of Claims
Subtask-1: Five Possible Answers
Oct. 2001-
3 Question Answering Subtask-2: One Set of All the Answers NTCIR-3QA
Oct. 2002
Subtask-3: Series of Questions
Single Document Summarization
Text Summarization NTCIR-3 SUMM
Multi-document Summarization
Survey Retrieval
Web Retrieval Target Retrieval NTCIR-3 WEB
[Optional] Speech-Driven
Single Language IR:C-C,K-K,J-J
Bilingual CLIR:x-J,x-C, x-K
CLIR NTCIR-4CLIR
Pivoted Bilingual CLIR
Mulilingual CLIR:x-CKJE
"Invalidity Search"= Search Patents by a Patent
Patent NTCIR-4 PATENT
[Feasibility] Automatic Patent Map Creation
Apr. 2003 - Subtask-1: Five Possible Answers
4
June 2004 Question Answering Subtask-2: One Set of All the Answers NTCIR-4 QA
Subtask-3: Series of Questions
Text Summarization Multi-document Summarization NTCIR-4 SUMM
Informational Retrieval
Navigational Retrieval
Web Retrieval NTCIR-4 WEB
[Pilot] Geographical Information
[Pilot] (Search Results) Topical Classification
n-m: n=query language, m=document language(s), J:Japanese, E:English, C:Chinese, K:Korean, x
*: number of active participating groups that submitted task results
2.2 Participants
As shown in Figures 3 and 4, the number of participants has been gradually increasing. Different tasks attracted different research
groups although many are overlapped, or changed the participating tasks over workshops. Many international participants were
enrolled to CLIR. Patent Retrieval task attracted many participants from company research laboratories and “veteran” NTCIR
participants. WEB task has participants from various research communities like machine learning, DBMS, and so on. The
number of collaborating teams across different organizations is increasing in recent NTCIRs.
Participants/Tasks
NTCIR workshop:
120
Number of Participating Groups
100 QA
# of ParticipatingGroups
4th workshop Summarization
*The number of the
Registration
80
Term Extraction
3rd workshop Web Retrieval
60
Patent Retrieval
2st workshop 40 chinese NonJapanese IR
Chinese .Korean
CLIR
# of group J!E,E!J、
20 J!E Japanese IR
IR
1st workshop # of country E!C x!CJEK
0
0 20 40 60 80 100 120 1st(1998-9)
1st(1998-9) 2nd(2000-1)
2nd(2000-1) 3rd
3rd (2001-2)
(2001-2) 4th(2003-4)
4th(2003-4)
Fig 3 Number of Participating Groups Fig 4 Participating Groups per Task
3 Test Collections
The test collections constructed for the NTCIR Workshops are listed in Table 2. In the NTCIR project the term “test collection” is
used for any kind of data set usable for system testing and experiments although it often means IR test collections used in search
experiments. One of our interests is to prepare realistic evaluation infrastructure, and those efforts include scaling up the
document collection, document genres, languages, topic structure and relevance judgments.
3.1 Documents
Documents were collected from various domains or genres. Format of the documents are basically the same as TREC or CLEF
and are plain text with SGML-like tags. Each of the specialized document genre collections contained characteristic fields for the
genre – Web collection contains html tags, hyperlinks, URL of the document, etc., and patent collection has tags indicating
document structure of patent, and both patent and scientific document collections have parallel corpora of English and Japanese
abstracts. The task (experiment) design and relevance judgment criteria were set according to the nature of the document
collection and user community who use the type of documents in their everyday tasks.
3.2 Topics.
A sample topic record is shown in Figure. 5. Topics are defined as statements of “user’s requests” rather than “queries”, which
are the strings actually submitted to the system, because we wish to allow both manual and automatic query construction from the
topics. Emphasis has been shifted towards the topic structure capable more realistic experiments as well as to see the effect of
background information of the topic. The characteristics are summarized as followings;
Topic Structure: Topic Structure has slightly changed in each NTCIR. A topic basically consists of a , a
description <DESC>, and a detailed narrative <NARR> of the search request as similar to those used in CLEF and TREC. It
may contain additional fields as shown in Table 3. Most of NTCIR collections contain a list of concepts <CONC>, but they are
not heavily used by participants.
Table 2. Test collections constructed through NTCIR
NTCIR Test Collections; IR and QA
Task data
documents topic/ relevnce
collection task question judge
genre filename lang year # of doc size lang #
ntc1-je JE 339,483 577MB
83
IR ntc1-j J 332,918 312MB 3 grades
sci. ntc1-e E 187,080 218MB 60
NTCIR-1 Term 1988-1997 J
abstract
extraction/
ntc1-tmrc J 2,000 - - -
role
analysis
CIRB010 IR news CIRB010 Ct 1998-1999 132220 132MB Ct E 50 4 grades
sci. ntc2-j J 400,248 600MB
NTCIR-2 IR 1986-1999** JE 49 4 grades
abstract ntc2-e E 134,978 200MB
news KEIB010 K 1994 66,146 74MB CtKJE 30 4 grades
CIRB011 Ct 132,173
NTCIR-3 CIRB020 249,508
IR
CLIR news Mainichi J 1998-1999 220,078 870MB CtKJE 50 4 grades
EIRB010 10,204
E
Mainichi Daily 12,723
patent full kkh *3 J 1998-1999 697,262 18GB C C KJ
NTCIR-3 t s
IR abstract jsh *3 J 1995-1999 1,706,154 1,883MB 31 3 grades
PATENT E
abstract paj *3 E 1995-1999 1,701,339 2,711MB
NTCIR-3 exact
QA news Mainichi J 1998-1999 220,078 282MB J* 1200
QA answer
NTCIR-3 Web NW100G-01 multiple crawled in 11,038,720 100GB 4grades
IR J* 47
WEB (html/tex NW10G-01 *4 2001 - 10GB + relative
CIRB011 132,173
Ct
CIRB020 249,508
Hankookilbo + 220,078
K
Chosenilbo + 149,498
NTCIR-4 Mainichi J 105,517
CLIR
IR news
EIRB010
1998-1999
10,204
ca.2.7GB CtKJE 60 4 grades
Mainichi Daily 12,723
Korea Times + E 21,377
Hong Kong Standard + 96,856
Xinhua (AQUAINT) + 208,168
Publication of ca.
patent full J 1993-2002 ca.45GB
NTCIR-4 unexamined patent 3,500,000 CtCsKJ
IR
PATENT Patnet Abstracts of ca. E
abstract E 1993-2002 ca.10GB
Japan (PAJ) + 3,500,000
NTCIR-4 Mainichi 220,078 ca.776M
QA news J 1998-1999 J*
QA Yomiuri + ca. 340,000 B
NTCIR-4 Web multiple crawled in
IR NW100G-01 11,038,720 100GB J*
WEB (html/text) *4 2001
J:Japanese, E:English, C:Chinese (Ct:Traditional Chinese, Cs: Simplified Chinese), K:Korean;
"+" indicates the document collection newly added for NTCIR-4
* English translation is available ** gakkai subfiles: 1997-1999, kaken subfiles: 1986-1997
*3: kkh : Publication of unexamined patent application, jsh: Japanese abstract, paj: English translation of jsh
*4: almost Japanese or English (some in other languages)
NTCIR Text Summarization
documents summaries
collection task
genre filename lang year # of doc types analysts total#
NTCIR-2 1994.1995
single doc news Mainichi J 180 doc 7 3 3780
SUMM .1998
NTCIR-2 1000
single doc news Mainichi J 1998 2 1 2000
TAO doc
NTCIR-3 single doc Mainichi J 1998- 60 docs 7 3 1260
news
SUMM multi doc Mainichi J 1999 50 sets 2 3 300
NTCIR-4 Mainichi 1998-
multi doc news J
SUMM Yomiuri 1999
Sample Topic
written statement of user’s needs
<TOPIC>
<NUM>0010</NUM> purpose/background
<TITLE CASE="b">Aurora, conditions, observation
I want to know the conditions that give rise to an aurora for
observation purposes
I want to observe an aurora so I want to know the
conditions necessary for its occurrence and the mechanism
Relevance
judgment behind it.Aurora observation records, etc. list
criteria the place and time so only documents that provide additional
information such as the weather and temperature at the time of
occurrence are relevant.
Aurora, occurrence, conditions, observation, given rel docs
mechanism
NW003201843, NW001129327, NW002699585
1st year Master’s student, female, 2.5 years search
experience
user attribute
Figure 5. Sample topic (NTCIR-3 WEB)
Table 3 Topic fields
Topic Structure of NTCIR IR Test Collections
NTCIR-3 NTCIR-3 NTCIR-3 NTCIR-4 NTCIR-4 NTCIR-4
NTCIR-1 NTCIR-2 CIRB010
CLIR PATENT WEB CLIR PATENT WEB
ad hoc, ad hoc, ad hoc, Cross-genre, ad hoc,
Task CLIR ad hoc CLIR invalidity
CLIR CLIR CLIR CLIR other
T-only, T-only, CLAIM- T-only,
Mandatory Run * D-only D-only N/A D-only S+A
D-only D-only only D-only
Topic Field
TITLE ** very short very short very short very short very short query query very short query
DESC yes yes yes yes yes yes yes yes yes
NARR (unstructured yes yes yes yes yes yes
NARR (structured) yes yes yes
NARR. BACK *10 yes yes yes
NARR. RELE *10 yes yes yes
NARR. TERM *10 yes yes yes
PURPOSE *7 yes
CONC yes yes yes yes yes yes yes
FIELDS yes yes
TLANG / LANG *3 yes yes
SLANG *3 yes yes
RDOC *4 yes yes
PI *4
USER *5 yes yes
ARTICLE *6 yes
DOC *9 yes
SUPPLEMENT *6 yes
CLAIM *8 yes
COMP *8 yes
COMP. CNUM *8 yes
*: D-only=DESC only, T-only=TITLE only, A+S= run using ARTICLE and SUPPLEMENT only
**: "very short"=very short description of search request; "qeury"=comma separated term list
*3: TLANG/LANG=target language, the language of the topic; SLANG=source language, the language the topic
originally constructed.
*4: RDOC=known relevant documents; PI=the patent for the invention mentioned in the news articles.
*5: USER=users' attribute
*6: ARTICLE=a news article reporting an invention; SUPPLEMENT=memorandam to focus the issues in the article
relevant to the user's needs; if a human knowledgeable searcher reads ARTICLE and SUPPLEMENT, he/she
understand the user's search request as specif
*7: Purpose of search (only "invalidity search" for NTCIR-4 PATENT)
*8: CLAIM=Target claim in the query patent. It was used as qeury of the search and may consists of multiple
components; COMP=Component of a claim; CNUM=Claim component ID
*9: Query patent fulltext (fulltext of a patent that is used as a query of the search)
*10: BACK=Background knowledge/purpose of search; RELE=relevance judgment criteria; TERM=term definitions
as Query: A title is originally defined as a very short description, or “nickname” of the topic, and, since NTCIR-3
3
WEB , changed to be a "query", a string put into a search engine by users and defined as a comma-separated term lists up to three terms.
Structured <NARR>: Originally a narrative <NARR> was defined and instructed to the topic authors that it may contain
background knowledge, purpose of the search, detailed explanation of the topic, criteria for relevance judgment, term definitions,
etc. Since NTCIR-3 WEB, such information categories in <NARR> explicitly marked by tags like <BACK>, <RELE>, etc. as
Figure 5. The purpose of this change is to examine the effect of additional information on the search effectiveness explicitly.
Mandatory runs: Any combination of topic fields is allowed to use in experiments for research purpose. In the Workshop,
the Mandatory Runs are defined in each task, and every participant must submit at least one mandatory run using the specified
topic field only. The purpose of this is to enhance the cross-system comparison based on the common condition and see the
effectiveness of the additional information over it. Mandatory runs are originally “<DESC> only”, then gradually shift to
“<TITLE> only as well as <DESC> only”.
3.3 Relevance Judgments
Relevance judgments are done by pooling, and the format and methods are basically the same as other evaluation projects
including CLEF and TREC. The differences shall be summarized as follows;
1. Pooling strategies are slightly different according to each of the task
・ Additional interactive recall-oriented searches are done to improve the exhaustivity (NTCIR-1,-2) [8]
・ Additional interactive recall-oriented search are done by professional patent intermediaries (PATENT) [9]
・ “One-click distance model”, in which hyperlinked documents are allowed to see in WEB [10]
・ Cross-lingual pooling for parallel or quasi-parallel documents (NTCIR-1,-2)[8]
・ Graded-depth pooling: pool creating top10, 11-20, 21-30, 31-41, (PATENT) [9]
2. Multi-grade and relative relevance judgments
・ Highly Relevant, Relevant, Partially Relevant [5-7], Irrelevant; Best Relevant, 2nd Best, 3rd Best, etc. [10]
3. Judgments includes the extracted passages to show the reason why the assessors assessed the documents as
“relevant”
4. Pooled document lists to be judged are sorted in descending order of likelihood to be relevant (not the order of the
document IDs)
5. Relevance judgment files may be prepared to each of the target language document sub-collections in CLIR
For 4, it helps assessors to judge consistently over a long list of pooled documents to be judged (typically 2000 - 3000
documents). Relevance judgments may change over assessors and over time. If relevant documents are appeared intensively in
the first part of the list, it is easier for the non-professional assessors to set and confirm their criteria for relevance judgments, and
then they can always refer those documents to re-confirm their own relevance judgment criteria when they go down to the lower
ranked document. We understand they may be suffered by “order effect” of the ranked list of pooled documents in judgments,
but we intentionally have used this strategy as practical and most effective one in our environment based on the comparative tests
and interviews with assessors.
For 5, in multilingual CLIR, a topic can not always obtain sufficient number of relevant documents on every language
document sub-collection, and this is the natural situation in multi-lingual CLIR. As a result, some topics can not be usable
experiments on specific language documents. We can not find the way to manage this issue and only strategy we could take in
NTCIR-4 CLIR is to increase the number of topics, so that larger number of topics can be used common across the document
sub-collections and then improve the stableness of the evaluation.
Assessors are users of the document genre, judgments are done by the topic author except CLIR in NTCIR-3 and -4 since
topics are created in cooperation of multiple countries, and then translated into each language and tested usability on each
3 Topic authors are instructed to sort the terms in <TITLE> in descending order of importance to express the search request resembling the way
language document sub-collection. Judging other users’ topics is sometimes hard for users and take longer time.
First two NTCIRs used two assessors per topic then tested inter-assessors consistency and found that the inconsistency
among multiple assessors on a topic does not affect the stableness of the evaluation when tested on sufficient number of topics.
Based on this, single assessor per topic is used in and after NTCIR-3.
3.4 Evaluation
For the evaluation, trec-eval program [11] is used by setting two threshold of the levels of relevance judgments, i.e. “Rigid
Relevance” for “Relevant” or higher, “Relaxed Relevant” for “Partial Relevant” or higher ranked relevance for IR experiments.
As additional metrics, several metrics for multi-grade relevance judgments are proposed including weighted mean average
precision (wMAP), weighted mean reciprocal rank (wMRR, for WEB task), and used decline cumulated gain (DCG) [12-13].
For Question Answering, MRR is used for subtask-1, return 5 possible answers and no penalty for wrong answers, and
F-measure for subtask-2, return one set of all the answers and penalty will be given for wrong answers, and subtask-3, series of
question. For Text Summarization, content based and readability based intrinsic evaluation was done in NTCIR-3 for both single
document and multi-document summarization, and proposed new evaluation methodology based on revision (edit distance) on
system summaries by professional analysts who created the model summaries.
4. Further Analysis of NTCIR-3
After our previous reports at CLEFs [2-4] and the overview papers in the Proceedings of the NTCIR-3 [7], several additional
analyses were done on the NTCIR-3 results and collection.
For PATENT retrieval task, though a new strategy for cross-genre retrieval called “term distillation” was proposed by Ricoh
group and worked well on the collection, many research questions regarding patent retrieval were remained unsolved in
NTCIR-3. The questions are, for example;
1. Is there any particular IR model (or weighting scheme) specifically effective on Patent?
2. Influence of the wide variation of document length (from 100 words to 30,000 word tokens in a document!)
3. Indexing (Character bi-gram vs. Word-based)
4. Target document collections: Fulltext vs. abstract (many commercialized systems used abstracts only)
For 1., it has been reported that tf is not effective on Patent at the SIGIR 2000Workshop on Patent Retrieval, but we could not find
the concrete answers to the question through the NTCIR-3.
To answer these question, the NTCIR-3 Patent Task organizers conducted additional experiments on the patent collection
and newspaper collection, and tested 8 different weighting schemes including both vector space as well as probabilistic models,
on 6 different document collections, using 4 different indexing strategies, character bi-gram, word, compound terms, hybrid of
character bi-gram and word; and 3 different topic length on a system. The results will be reported in [14].
For WEB, one participating group was consisted as a collaboration of research groups with strong background of
content-based text retrieval and of web-link analysis, worked well at NTCIR-3 WEB. Further analysis on the effect of link on
WEB collection, link-based approaches are generally worked well especially on the short queries like using TITLE only, or
more specifically the first term of the TITLE, i.e. the most important terms for the users (topic authors) [15].
in which the users input the terms as queries. The relation between the terms is specified as an attribute of the <TITLE> in WEB Task.
5. Challenges at NTCIR-4
As shown in Table 1, the 4th NTCIR Workshop hosts 5 tasks, CLIR, PATENT, QAC, TSC, and WEB and their sub-tasks.
Evaluation schedule varies according to each task.
April 2003: Document Release
June – September 2003: Dry Run
October – December 2003: Formal Run
20 February 2004: Evaluation Results Release
2-5 June 2004: Workshop Meeting at NII, Tokyo Japan
For the further information including late registration of the task participation, please consult NTCIR web sites at;
http://research.nii.ac.jp/ntcir and http://research.nii.ac.jp/ntcir/ntc-ws4, or contact the author.
5.1 NTCIR-4CLIR
Since this is the second multilingual CLIR at NTCIR, the same task design will be continued from the previous one. Minor
revision was made only to solve the major problems raised in the assessment on the NTCIR-3 as follows;
・ Enlarge the English and Korea document collections comparable to Chinese and Japanese. 2.7GB in total.
・ New sub-task of Pivot Language Bilingual CLIR
・ Restrict the pair of topic and document languages, so that comparison will be done in fruitfully
・ Set T-only run as mandatory as well as D-only run
・ Question type – topics were categorized according to the nature and types of the answers in order to take a
good balance of the topic set.
The new sub-task, pivot CLIR uses English as a Pivot language, then test the effectiveness of the transitive CLIR. It is one of the
practical approaches of Multilingual CLIR in the environment with less availability of the direct translation resources but rich in
those between each of the languages and English.
5.2 NTCIR-4 Question Answering (QAC) and Text Summarization (TSC)
QAC plans three subtasks as previous one at NTCIR-3. Among the three, subtask-1 and -2 will be done without major change.
Only exceptions are; use different question sets for each of subtask-1 and -2, and increase the number of topics containing
multiple answers. It was decided to avoid overestimate of the groups ignoring the possibility of multiple answers and returning
the first priority answer only to the every question in subtask 2.
QAC subtask-3, answering to the series of question, is one of the major focus of the NTCIR-4 QAC. We plan to increase
the number of sequence as well as task design aiming to tackle the problems resembling the real-world “Report Writing” task
based on a set of relevant documents. The task design also related to the TSC, content-based evaluation of multi-document
summarization will be done by set of questions. This is, more fundamentally, what kind of aspects of an event or topic that users
want to know. Some of the questions may be more appropriate for the current factoid –oriented QA and others may covered by
summarization. IR covered both and those focus of QAC and TSC has many intersection of the focus of the CLIR to see the
categorization of question types.
5.3 Specialized Genre Related Tasks at NTCIR-4: Patent and WEB
Both PATENT and WEB plan (1) Main task(s) and (2) Feasibility or Pilot studies for more challenging tasks as follows;
PATENT- Main: Invalidity Task:
To search patents to invalidate the query patents. Claims of the query patents are used as query and they are
segmented into components of the invention or technologies consisting of the investigation, then search related
patents. A patent may be invalidated by one patent or by combination of multiple patents. Return document IDs as
well as relevant passages.
PATENT - Feasibility: Long term research plan over NTCIR4-5. Automatic Patent Map Creation
A kind of Text mining -- Detect sets of technologies used in a set of patents, extract them, and make a table to show
the relationship between technologies and patents, and evolution or trends among them.
WEB - Main: Informational Search and Navigation Oriented Search, in which find most informative and reliable page
WEB – Pilot: Geographical oriented and Topical Classification of the Search results
For the details, please visit the website of each task, which are linked from the NTCIR’s main web site.
6. Summary
Brief history of NTCIR and recent progress after NTCIR-3 are reported in this paper. One of the characteristic features of the
NTCIR is targeting “Information Access” technologies, in which a whole process for users to obtain and utilize the information
in the documents are interested in and see the intersection between all the related technologies including IR, Summarization, QA,
Text mining, etc., and treat them as like a “family”. Other aspects are, for see the users’ information task behind the
laboratory-typed testing. We are in the process of the fourth-iteration in a series. Evaluation must be changed according to the
technologies evolution and change of the social needs. We have been and are struggling for this. Collaboration and any leads and
advices are always more than welcome.
References
1. NTCIR Project: http://research.nii.ac.jp/ntcir/.
2. Kando, N. "NTCIR Workshop: Japanese- and Chinese-English cross-lingual information retrieval and multi-grade relevance judgments
".In Proceedings of the first Cross-Language Evaluation Forum (CLEF2000), Lisbon, Portugal, Sept.17-22, 2000 Springer, 2001,
pp.24-33 (Lecture Notes in Computer Science; 2069)
3. Kando, N. " CLIR system evaluation at the second NTCIR workshop", In Proceedings of the second Cross-Language Evaluation Forum
(CLEF2001) , Darmstadt, German, Sept 3-4, 2001, Springer, 2002, pp.371-388 (Lecture Notes in Computer Science; 2406)
4. Kando, N. "CLIR at NTCIR Workshop 3; Cross-Language and Cross-Genre Retrieval” In Proceedings of 3rd Cross-Language
Evaluation Forum (CLEF2002), Rome, Italy, Sept. 19-20, 2002 , Springer (Lecture Note in Computer Science) (to appear)
5. NTCIR Workshop 1: Proceedings of the First NTCIR Workshop on Research in Japanese Text Retrieval and Term Recognition. Tokyo
Japan, 30 Aug.–1 Sept., 1999. ISBN 4-924600-77-6. (http://research.nii.ac.jp/ntcir/workshop/OnlineProceedings/).
6. NTCIR Workshop 2: Proceedings of the Second NTCIR Workshop on Research in Chinese and Japanese Text Retrieval and Text
Summarization. Tokyo Japan, June 2000 – March 2001. ISBN 4-924600-96-2.
7. NTCIR Workshop 3: Proceedings of the Third NTCIR Workshop on Research in Information Retrieval, Question Answering and
Summarization, Tokyo Japan, Oct. 2001 – Oct. 2002 . ISBN-4-86049-016-9 .
8. Kuriyama, K., Yoshioka, M., Kando, N. "Construction of a Large Scale Test Collection NTCIR-2: The Effect of Additional Interactive
Search and Cross-Lingual Pooling", IPSJ Transactions on Databases, Vol.43, No. SIG2 (TOD13), pp.48-59, March 2002.(in Japanese)
9. Iwayama, M., Fujii, A., Kando, N., Takano, A. "Overview of Patent Retrieval Task at NTCIR-3". In NTCIR Workshop 3: Proceedings of
the Third NTCIR Workshop on Research in Information Retrieval, Question Answering and Summarization, Tokyo Japan, Oct. 2001 –
Oct. 2002 . (to appear).
10. Eguchi, K., Oyama, K., Ishida, E., Kando, N., Kuriyama, K. "Overview of Web Retrieval Task at the Third NTCIR Workshop", In
NTCIR Workshop 3: Proceedings of the Third NTCIR Workshop on Research in Information Retrieval, Question Answering and
Summarization, Tokyo Japan, Oct. 2001 – Oct. 2002 . (to appear).
11. Buckley, C. trec-eval IR evaluation package. Available from ftp://ftp.cs.cornell.edu/pub/smart.
12. Jarvelin, K., Kekalainen, J. “IR evaluation methods for retrieving highly relevant documents”. In: Proceedings of the 23rd Annual
International ACM-SIGIR Conference on Research and Development in Information Retrieval. Athens Greece, July 2000, 41–48.
13. Voorhees，E.M. “Evaluation by Highly Relevant Documents”，in Proceedings of the 24th Annual International ACM SIGIR Conference
on Research and Development in Information Retrieval (SIGIR 2001)，New Orleans, Sept. 2001, pp. 74-82.
14. Iwayama, M., Fujii, A., Kando, N., Marukawa, K. “Empirical study on retrieval models for different document genres: Patents and
newspaper articles. In Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in
Information Retrieval (SIGIR 2003), Toronto, Canada, July 2003, (to appear)
15. Eguchi, K., Oyama, K., Ishida, E., Kando, N., and Kuriyama K., “Evaluation Methods for Web Retrieval Tasks
Considering Hyperlink Structure”, IEICE Transaction on Information and Systems. (Sep. 2003, in Japanese, to appear).
</pre>