OntoNotes Release 2.0

registration: 2008/03/10 08:36:35, last modified: 2016/01/17 10:21:15
title
  • OntoNotes Release 2.0
title.abbreviation
title.alternative
creator
  • developer BBN Technologies (coreference annotation)
  • developer University of Colorado (Chinese Treebank, Chinese/English PropBank, word sense for verb)
  • developer University of Pennsylvania (English Treebank and English PropBank)
  • developer University of Southern California's Information Sciences Institute (word sense for nouns and ontology)
  • researcher Ralph Weischedel, BBN Technologies (coreference annotation)
  • researcher Lance Ramshaw, BBN Technologies (coreference annotation)
  • researcher Sameer Pradhan, BBN Technologies
  • researcher Martha Palmer, University of Colorado (Chinese Treebank, Chinese/English PropBank, word sense for verb)
  • researcher Nianwen Xue, University of Colorado (Chinese Treebank, Chinese PropBank)
  • researcher Mitchell Marcus, University of Pennsylvania (English Treebank)
  • researcher Ann Taylor, University of Pennsylvania
  • researcher Craig Greenberg, University of Pennsylvania
  • researcher Eduard Hovy, University of Southern California's Information Sciences Institute (word sense for nouns and ontology)
  • researcher Robert Belvin, University of Southern California's Information Sciences Institute
  • researcher Ann Houston, Grammarsmith
subject
  • syntactically annotated corpus
  • semantically annotated corpus
  • named entity tagged corpus
  • word sense annotated corpus
subject.linguisticField
  • computational_linguistics
  • semantics
  • syntax
  • text_and_corpus_linguistics
subject.monoMultilingual
  • multilingual
subject.resourceSubject
  • corpus
description
  • *Introduction*

    The OntoNotes project is a collaborative effort between BBN Technologies, the University of Colorado, the University of Pennsylvania, and the University of Southern California's Information Sciences Institute. The goal of the project is to annotate a large corpus comprising various genres of text (news, conversational telephone speech, weblogs, use net, broadcast, talk shows) in three languages (English, Chinese, and Arabic) with structural information (syntax and predicate argument structure) and shallow semantics (word sense linked to an ontology and coreference). OntoNotes Release 2.0 is a continuation of the OntoNotes project and is supported by the Defense Advanced Research Projects Agency, GALE Program Contract No. HR0011-06-C-0022.

    OntoNotes Release 1.0 (LDC2007T21) contains 400k words of Chinese newswire data (from Xinhua News Agency and Sinorama Magazine) and 300k words of English newswire data (from the Wall Street Journal). OntoNotes Release 2.0 adds the following to the corpus: 274k words of Chinese broadcast news data (from China Broadcating System, China Central TV, China National Radio, China Television System and Voice of America); and 200k words of English broadcast news data (from ABC, CNN, NBC, Public Radio International and Voice of America).

    Natural language applications like machine translation, question answering, and summarization currently are forced to depend on impoverished text models like bags of words or n-grams, while the decisions that they are making ought to be based on the meanings of those words in context. That lack of semantics causes problems throughout the applications. Misinterpreting the meaning of an ambiguous word results in failing to extract data, incorrect alignments for translation, and ambiguous language models. Incorrect coreference resolution results in missed information (because a connection is not made) or incorrectly conflated information (due to false connections). OntoNotes builds on two time-tested resources, following the Penn Treebank for syntax and the Penn PropBank for predicate-argument structure. Its semantic representation will include word sense disambiguation for nouns and verbs, with each word sense connected to an ontology, and coreference. The current goals call for annotation of over a million words each of English and Chinese, and half a million words of Arabic over five years.

    The authors wish to make this resource available to the natural language research community so that decoders for these phenomena can be trained to generate the same structure in new documents. Lessons learned over the years have shown that the quality of annotation is crucial if it is going to be used for training machine learning algorithms. Taking this cue, each layer of annotation in OntoNotes will have at least 90% inter-annotator agreement. Pilot studies have shown that predicate structure, word sense, ontology linking, and coreference can all be annotated rapidly and with better than 90% consistency.

    *Samples*

    For an example of the data in this corpus, please examine the following samples

    * Chinese
    * English

    *Sponsorship*

    This work is supported in part by the Defense Advanced Research Projects Agency, GALE Program Grant No. HR0011-06-1-0003. The content of this publication does not necessarily reflect the position or policy of the Government, and no official endorsement should be inferred.

    The World is a co-production of Public Radio International and the British Broadcasting Corporation and is produced at WGBH Boston.
description.price *This metadata is only as a guide
description.language
  • English
description.inputDevice
description.inputEnvironment
description.speakingStyle
description.speechMode
description.samplingRate
description.additionalData
publisher
contributor
  • sponsor Defense Research Advanced Projects Agency, GALE Program Grant No. HR0011-06-C-0022
contributor.attribute.author.motherTongue
  • native Journalists and writers at Wall Street Journal, Sinorama Magazine and Xinhua News Agency for newswire and script writers for broadcast news
contributor.attribute.author.dialect
  • standard_dialect
contributor.attribute.author.level
  • professional Journalists and writers at Wall Street Journal, Sinorama Magazine and Xinhua News Agency for newswire and script writers for broadcast news
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
  • 2007-00-00 0000-00-00
date.issued
  • 2008-02-18
date.modified
type
  • Text
type.discourseType
  • report
type.linguisticType
  • primary_text
type.purpose
  • analysis machine translation, natural language processing
  • developing_technologies machine translation, natural language processing
type.style
  • written newswires
  • written newspaper
  • speech broadcastnews
type.form
  • unfixed
type.sentence
  • short
  • long
type.annotation
  • annotated coreference based on the BBN Technologies' co-reference guidelines : http://www.ldc.upenn.edu/Catalog/docs/LDC2008T04/coreference/
  • annotated syntactic structure and part-of-speech based on the Penn Treebank II guidelines (http://www.cis.upenn.edu/~treebank/) with addendum (http://www.ldc.upenn.edu/Catalog/docs/LDC2008T04/treebank/)
  • annotated PropBank annotation (predicate-argument structure) : http://www.ldc.upenn.edu/Catalog/docs/LDC2008T04/propbank/
  • annotated word sense annotation carried out at the University of Colorado and Information Sciences Institute
  • annotated ontology based on Omega ontology : http://omega.isi.edu/
  • annotated entity names (11 entity names and 7 value types)
type.annotationSample
type.sample
  • [English Tree and Sense Predicate Structure] :

    http://www.ldc.upenn.edu/Catalog/desc/addenda/LDC2008T04.jpg
  • [Chinese Tree and Sense Predicate Structure] :

    http://www.ldc.upenn.edu/Catalog/desc/addenda/LDC2008T04_ch.jpg
  • [Word Sense]:

    wsj/02/wsj_0242@wsj@en@on 0 4 complain-v ?,? 1
    wsj/02/wsj_0242@wsj@en@on 0 9 push-v ?,? 1
    wsj/02/wsj_0242@wsj@en@on 0 15 create-v ?,? 1
    wsj/02/wsj_0242@wsj@en@on 0 25 affect-v ?,? 1
    wsj/02/wsj_0242@wsj@en@on 1 3 pace-v ?,? 1
    wsj/02/wsj_0242@wsj@en@on 1 11 aim-v ?,? 2
    wsj/02/wsj_0242@wsj@en@on 1 24 reduction-n ?,? 1
    wsj/02/wsj_0242@wsj@en@on 1 36 register-v ?,? 1
    wsj/02/wsj_0242@wsj@en@on 2 3 call-v ?,? 3
    wsj/02/wsj_0242@wsj@en@on 2 6 agreement-v ?,? 1

    “OntoNotes Release 2.0 with OntoNotes DB Tool v. 0.92 beta and OntoViewer v. 0.9 beta” (2007), pp. 39.
    http://www.ldc.upenn.edu/Catalog/docs/LDC2008T04/OntoNotes-Release-2.0.pdf
  • [Coreference]:

    wsj_0037.mrg
    * Judging from the American in Haruki Murakami 's `` A Wild Sheep Chase '' -LRB- Kodansha , 320 pages , $18.95 *U* -RRB- , baby boomers on both sides of the Pacific have a lot in common .
    Although *-2 set *-1 in Japan , the novel 's texture is almost entirely Western , especially American .
    Characters drink Salty Dogs , whistle `` Johnny B. Goode '' and watch Bugs Bunny reruns .
    They read Mickey Spillane and talk about Groucho and Harpo .
    They worry about their careers , drink too much and suffer through broken marriage and desultory affairs .

    The rest is omitted.

    “OntoNotes Release 2.0 with OntoNotes DB Tool v. 0.92 beta and OntoViewer v. 0.9 beta” (2007), pp. 39.
    http://www.ldc.upenn.edu/Catalog/docs/LDC2008T04/OntoNotes-Release-2.0.pdf
  • [Entity Names]:


    wsj/02/wsj_0242@wsj@en@on
    Some U.S. allies are complaining that President Bush is pushing conventional-arms talks too quickly , creating a risk that negotiators will make errors that could affect the security of Western Europe for years .
    Concerns about the pace of the Vienna talks -- which are aimed at the destruction of some 100,000 weapons , as well as major reductions and realignments of troops in central Europe -- also are being registered at the Pentagon .
    Mr. Bush has called for an agreement by next September at the latest .
    But some American defense officials believe the North Atlantic Treaty Organization should take more time to examine the long-term implications of the options being considered .

    “OntoNotes Release 2.0 with OntoNotes DB Tool v. 0.92 beta and OntoViewer v. 0.9 beta” (2007), pp. 40.
    http://www.ldc.upenn.edu/Catalog/docs/LDC2008T04/OntoNotes-Release-2.0.pdf
format.extent
  • Medium: 1 DVD
  • Words (English broadcast) : 200K words
  • Words (Chinese broadcast) : 274K words
  • Words (English newswire) : 300K words
  • Words (Chinese newswire) : 400K words
  • Corpus size: 1258291 KB
format.medium
  • DVD
format.encoding
format.markup
  • XML
  • ENAMEX (for entity names annotation)
  • Treebank bracketing
format.functionality
  • "OntoNotes DB Tool v0.92 beta":
    For assembling the database from the original annotation files and for deriving various views of the data from the database. Can provide an API that can implement new queries or views. Included in the package.
  • "Ontoviewer v0.9 beta":
    For viewing the various levels of OntoNotes annotation. Included in the package.
  • "Graphviz (version 2.14)":
    An open source graphics package for generating and displaying the ontology as an actual tree structure from the database. Available at: www.graphviz.org.
identifier
identifier.doi
identifier.islrn
  • 563-087-655-210-0
source
  • English newswire : C-001546,Treebank-2 (newspaper articles of late 1980's from Wall Street Journal)
  • Chinese newswire : C-000695,Chinese Treebank 5.0 (newspaper articles and newswires from Xinhua, Ingormation Service Department of HKSAR and Sinorama magazine)
  • English / Chinese broadcast news transcripts : C-001297,TDT4 Multilingual Text and Annotations (news programs by China Broadcating System, China Central TV, China National Radio, China Television System and Voice of America for Chinese data, and for English data, ABC's "World News Tonight," CNN's "Headline News," NBC's "NBC Nightly News," Public Radio International's "The World" and news programs by Voice of America)
language
  • English American English
  • Chinese
  • Mandarin Chinese
language.area
  • America United States, North America
  • Asia China
relation
relation.utilization *This metadata is automatically extracted
coverage.temporal
  • 1989 (English newswire materials)
  • 1994 - 2001 (Chinese newswire materials)
  • 2000-10 - 2001-01 (Broadcast materials)
coverage.spacial
  • United States
  • China
rights
  • User Agreement for Non-Members: http://www.ldc.upenn.edu/Catalog/nonmem_agree/genericlicense.html
  • Licensing information for OntoNotes DB Tool package and Ontoviewer (included in the package) : Included in the corpus package
  • Portions Copyright : 2000-2001 American Broadcasting Company
  • Portions Copyright : 2000-2001 Cable News Network, LP, LLLP
  • Portions Copyright : 2000-2001 China Broadcasting System
  • Portions Copyright : 2000-2001 China Central TV
  • Portions Copyright : 2000-2001 China National Radio
  • Portions Copyright : 2000-2001 China Television System
  • Portions Copyright : 1989 Dow Jones & Company, Inc.
  • Portions Copyright : 2000-2001 National Broadcasting, Company, Inc.
  • Portions Copyright : 2000-2001 Public Radio International
  • Portions Copyright : 1996-2001 Sinorama Magazine
  • Portions Copyright : 1994-1998 Xinhua News Agency
  • Portions Copyright : 1995, 2005, 2006, 2007, 2008 Trustees of the University of Pennsylvania
  • Licensing Instructions for Subscription & Standard Members, and Non-Members: http://www.ldc.upenn.edu/language-resources/data/obtaining