NTCIR-2 SUMM (Text Summarization Test Collection)

registration: 2009/02/16 05:12:19, last modified: 2009/02/17 06:17:15
title
  • NTCIR-2 SUMM (Text Summarization Test Collection)
title.abbreviation
title.alternative
  • NTCIR-2 TSC1 Collection
creator
  • developer NTCIR Project, National Institute of Informatics, Tokyo, Japan (http://research.nii.ac.jp/ntcir/index-en.html)
  • researcher Takahiro FUKUSHIMA (Otemon Gakuin University, Japan)
  • researcher Tsutomu HIRAO (NTT, Japan)
  • researcher Hidetsugu NANBA (Hiroshima City University, Japan)
  • researcher Manabu OKUMURA (Tokyo Institute of Technology, Japan)
subject
  • test collection for text summarization system evaluation
subject.linguisticField
  • computational_linguistics
  • discourse_analysis
  • text_and_corpus_linguistics
subject.monoMultilingual
  • monolingual
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 is for the automatic text summarization task and it includes : (1) Document data (Japanese newspaper articles from Mainichi Newspaper) and (2) Model Summaries (Japanese. For each of 180 documents, 7 types of single document summaries prepared in different length by different strategies were prepared by 3 analysts). Note that the NII provides only model summaries, therefore, each purchaser needs to obtain the document data from Mainichi Newspaper Co.
description.price *This metadata is only as a guide
  • Task Participants : Free
  • Non-participant : Not available
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-)
contributor.attribute.author.motherTongue
  • native Mainichi Newspaper Co., Japan (for document data)
contributor.attribute.author.dialect
  • standard_dialect
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
  • 2001-03-30
date.modified
  • 2001-04-09 Summeries for 120 articles were newly added.
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 and their summaries
type.form
  • unfixed
type.sentence
  • short The maximum number of characters in a summry is set.
  • long The maximum number of characters in a summry is set.
type.annotation
  • annotated
type.annotationSample
type.sample
format.extent
  • Model Summaries : 3780 documents (for each of 180 documents, 7 types of single document summaries prepared by 3 analysts)
format.medium
format.encoding
format.markup
format.functionality
  • mai2sgml.pl : Provided by IREX. For transforming the data to IREX IR task format.
  • tscsgml.pl : For transforming the data to the important sentence extraction task format.
identifier
identifier.doi
identifier.islrn
source
  • Newspaper articles from Mainichi Newspaper published in 1994, 1995, and 1998
  • C-001601,CD-Mainichi Shimbun Data Collection
language
  • Japanese
language.area
  • Japan
relation
relation.utilization *This metadata is automatically extracted
coverage.temporal
  • 1994, 1995, and 1998
coverage.spacial
  • Japan
rights
  • To obtain any of the collection, the user must first sign the user agreement with NII.
  • http://research.nii.ac.jp/ntcir/permission/perm-en.html
  • For the document collection (news articles), those who purchase it from Nichigai Associates can obtain Mainichi Shimbun's permission for research use of the dataset.
  • "Research Activities Report" and "Publication Report" should be submitted by the users of NTCIR Test Collections.