NTCIR-4 CLIR (IR Test Collection)

registration: 2009/02/16 05:26:34, last modified: 2009/02/20 02:20:42
title
  • NTCIR-4 CLIR (IR Test Collection)
title.abbreviation
title.alternative
  • NTCIR-4 CLIR言語横断検索テストコレクション
creator
  • developer NTCIR Project, National Institute of Informatics, Tokyo, Japan (http://research.nii.ac.jp/ntcir/index-en.html)
subject
  • test collection for information retrieval system testing and evaluation
subject.linguisticField
  • computational_linguistics
  • discourse_analysis
  • text_and_corpus_linguistics
  • translating_and_interpreting
subject.monoMultilingual
  • multilingual
subject.resourceSubject
  • corpus
description
  • The NTCIR project is a series of on-going evaluation workshops designed to enhance research in Information Access (IA) technologies including information retrieval, question answering, and text summarization/extraction. The Collection can be used for experiments of cross-lingual information retrieval between Chinese, Japanese, Korean and English. It includes (1) Document data (news articles, Chinese/Korean/Japanese/English), (2) Search topics (Chinese/Korean/Japanese/English), and (3) Relevance Judgements. Note that the NII does not provide the entire collection. Some of document data will be provided only to the workshop participants and some will have to be obtained from the original publishers.
description.price *This metadata is only as a guide
  • Topic and Relevance Judgment Data : Free for research use
  • English Document Data (Mainichi Daily News, Korea Times, and Hon Kong Standard) : Free for research use
  • English Document Data (Xinhua portion) : Available from Linguistic Data Consortium (LDC) in the US. (http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2002T31)
  • Chinese Document Data ("CIRB020" containing United Daily and other papers) : Free for research use
  • Japanese Document Data (Mainichi portion) : Available from Nichigai Associates, Inc. in Japan.
  • Japanese Document Data (Yomiuri portion) : Available from Nihon Database Kaihatsu Co., Ltd (NDK) in Japan
  • Other Document Data : Free for NTCIR Workshop participants (N/A for non-participants)
description.language
  • Japanese
  • English
description.inputDevice
description.inputEnvironment
description.speakingStyle
description.speechMode
description.samplingRate
description.additionalData
publisher
contributor
  • sponsor Research for Future Program, Japan Society for Promotion of Science (JSPS), Japan (1997-)
  • sponsor National Center for Science Information Systems (NACSIS), Japan (1997-)
  • sponsor Research Center for Information Resources at National Institute of Informatics (RCIR/NII), Japan (2000-)
  • sponsor MEXT Grant-in-Aid for Scientific Research on Priority Areas of "Informatics" (#13224087), Japan (2001-)
  • editor Nichigai Associates, Inc., Japan (http://www.nichigai.co.jp/)
  • editor Nihon Database Kaihatsu Co., Ltd (NDK), Japan (http://www.ndk.co.jp/index.html)
  • editor Linguistic Data Consortium (LDC), US (http://www.ldc.upenn.edu/)
contributor.attribute.author.motherTongue
  • native Mainichi Newspaper Co., Japan (for Japanese document data)
  • native The Yomiuri Shimbun, Japan (for Japanese document data)
  • native United Daily News, Taiwan (for Chinese document data)
  • native Economic Daily News, Taiwan (for Chinese document data)
  • native Min Sheng Daily, Taiwan (for Chinese document data)
  • native United Evening News, Taiwan (for Chinese document data)
  • native Star News, Taiwan (for Chinese document data)
  • native China Times, Taiwan (for Chinese document data)
  • native China Times Express, Taiwan (for Chinese document data)
  • native Commercial Times, Taiwan (for Chinese document data)
  • native China Daily News, Taiwan (for Chinese document data)
  • native Central Daily News, Taiwan (for Chinese document data)
  • native Hankookilbo, South Korea (for Korean document data)
  • native Chosunilbo, South Korea (for Korean document data)
  • non_native Taiwan News, Taiwan (for English document data)
  • non_native China Times English News, Taiwan (for English document data)
  • non_native Mainichi Daily News, Japan (for English document data)
  • non_native Korea Times, South Korea (for English document data)
  • non_native Xinhua News Agency, China (for English document data)
  • non_native Hong Kong Standard (The Standard), Hong Kong, China (for English document data)
contributor.attribute.author.dialect
  • standard_dialect Japanese, Korean, and English
  • dialect Chinese (Taiwanese Mandarin)
contributor.attribute.author.level
  • professional
contributor.attribute.author.age
  • adult
contributor.attribute.author.gender
  • male
  • female
contributor.attribute.speaker.motherTongue
contributor.attribute.speaker.dialect
contributor.attribute.speaker.level
contributor.attribute.speaker.age
contributor.attribute.speaker.gender
contributor.attribute.speaker.number
date.created
date.issued
  • 2003-00-00
date.modified
type
  • Text
type.discourseType
  • report
type.linguisticType
  • primary_text
type.purpose
  • developing_technologies To construct a large-scale Japanese-language test collections for providing sound infrastructure to evaluate the search effectiveness of information retrieval systems and for facilitating IR researches.
type.style
  • written newspaper articles
type.form
  • unfixed
type.sentence
  • short
  • long
type.annotation
  • annotated
type.annotationSample
  • Relevance Judgment File Format:

    Each of the following fields is separated by a TAB:
    (topic-ID) (dummy) (document-ID) (relevance judgment "0" or "1") (comment)

    http://research.nii.ac.jp/ntcir/permission/ntcir-4/perm-en-CLIR.html
  • Tags Used in Topics:

    : The tag for each topic
    : Topic identifier
    : Source language code: CH, EN, JA, KR
    : Target language code: CH, EN, JA, KR
    : The concise representation of information request, which is composed of noun or noun phrase.
    : A short description of the topic.The brief description of information need, which is composed of one or two sentences.
    : A much longer description of topic.The has to be detailed, like the further interpretation to the request and proper nouns, the list of relevant or irrelevant items, the specific requirements or limitations of relevant documents, and so on.
    : The keywords relevant.

    http://research.nii.ac.jp/ntcir/permission/ntcir-4/perm-ja-CLIR.html
  • Tags Used in Document Data:

    Mandatory tags:
    : The tag for each document
    : Document identifier
    : Language code: CH, EN, JA, KR
    : Title of this news article
    : Issue date
    : Text of news article
    Optional tags：

    : Paragraph marker
    : Section identifier in original newspapers
    : Contain figures or not
    : Number of words in 2 bytes (for Mainichi Newspaper)

    http://research.nii.ac.jp/ntcir/permission/ntcir-4/perm-ja-CLIR.html
type.sample
  • Topic Sample:


    013
    CH
    EN
    NBA labor dispute

    To retrieve the labor dispute between the two parties of the US National Basketball Association at the end of 1998 and the agreement that they reached.


    The content of the related documents should include the causes of NBA labor dispute, the relations between the players and the management, main controversial issues of both sides, compromises after negotiation and content of the new agreement, etc. The document will be regarded as irrelevant if it only touched upon the influences of closing the court on each game of the season.


    The rest is omitted.

    http://research.nii.ac.jp/ntcir/permission/ntcir-4/perm-ja-CLIR.html
format.extent
  • Number of Search Topics (Japanese/Korean/Chinese/English) : 60
  • Total Number and Size of the Documents ：1,579,727 (about 3GB)
  • Japanese Documents [1] (Mainichi) : 220,078 docs
  • Japanese Documents [2] (Yomiuri) : 375,980 docs
  • Chinese Documents [1] (United Daily News and others) : 249,508 docs (CIRB020)
  • Chinese Documents [2] (China Times and others) : 132,173 docs
  • Korean Documents [1] (Hankookilbo) : 149,921 docs
  • Korean Documents [2] (Chosunilbo) : 104,517 docs
  • English Documents [1] (Taiwan News) : 7,489 docs
  • English Documents [2] (China Times English News) : 2,715 docs
  • English Documents [3] (Mainichi Daily News) : 12,723 docs
  • English Documents [4] (Korea Times) : 19,599 docs
  • English Documents [5] (Xinhua) : 208,168 docs
  • English Documents [6] (Hong Kong Standard) : 96,856 docs
format.medium
format.encoding
format.markup
format.functionality
identifier
identifier.doi
identifier.islrn
source
  • News articles from Mainichi Newspaper, Japan (1998-1999) (C-001601,CD-Mainichi Shimbun Data Collection)
  • News articles from Yomiuri Newspaper, Japan (1998-1999) (C-001632,Yomiuri Shimbun Articles Data(Japanese))
  • News articles from United Daily News, Taiwan (1998-1999)
  • News articles from Economic Daily News, Taiwan (1998-1999)
  • News articles from Min Sheng Daily, Taiwan (1998-1999)
  • News articles from United Evening News, Taiwan (1998-1999)
  • News articles from Star News, Taiwan (1998-1999)
  • News articles from China Times, Taiwan (1998-1999)
  • News articles from China Times Express, Taiwan (1998-1999)
  • News articles from Commercial Times, Taiwan (1998-1999)
  • News articles from China Daily News, Taiwan (1998-1999)
  • News articles from Central Daily News, Taiwan (1998-1999)
  • News articles from Taiwan News, Taiwan (1998-1999)
  • News articles from China Times English News, Taiwan (1998-1999)
  • News articles from Mainichi Daily News, Japan (1998-1999)
  • News articles from Korea Times, South Korea (1998-1999)
  • Xinhua News texts, China (1998-1999) (C-001315,The AQUAINT Corpus of English News Text)
  • News articles from Hong Kong Standard (The Standard), Hong Kong, China (1998-1999)
  • News articles from Hankookilbo, South Korea (1998-1999)
  • News articles from Chosunilbo, South Korea (1998-1999)
  • C-003744,NTCIR-3 CLIR: IR/CLIR Test Collection
language
  • Japanese
  • English
  • Korean
  • Mandarin Chinese Taiwanese Mandarin written in traditional Chinese characters (繁体字)
language.area
  • Japan
  • Asia Taiwan, Hong Kong, China, North and South Koreas
  • Europe
  • America
  • Oceania
relation
relation.utilization *This metadata is automatically extracted
coverage.temporal
  • 1998 - 1999
coverage.spacial
  • Japan
  • South Korea
  • Taiwan
  • Hong Kong, China
  • Beijing, China
rights