CoolFace
Datasetpublic

dwb2023/gdelt-gkg-march2020

Dataset Card for dwb2023/gdelt-gkg-march2020 Dataset Details Dataset Description This dataset contains GDELT Global Knowledge Graph (GKG) data for March 16-22, 2020, capturing the complex network of events, actors, and relationships during a critical week of the COVID-19 pandemic. The data structure supports temporal, spatial, and contextual queries across multiple dimensions of global crisis response. Curated by: dwb2023 Language(s): Multilingual… See the full description on the dataset page: https://huggingface.co/datasets/dwb2023/gdelt-gkg-march2020.

sourceHugging Facecc-by-4.0updated 2y agoView on Hugging Face
0likes553downloads
Dataset Card

Dataset Card for dwb2023/gdelt-gkg-march2020

Dataset Details

Dataset Description

This dataset contains GDELT Global Knowledge Graph (GKG) data for March 16-22, 2020, capturing the complex network of events, actors, and relationships during a critical week of the COVID-19 pandemic. The data structure supports temporal, spatial, and contextual queries across multiple dimensions of global crisis response.

  • —Curated by: dwb2023
  • —Language(s): Multilingual (primary English with machine translations)
  • —License: Creative Commons Attribution 4.0 International License (CC BY 4.0)

Competency Questions Support

The dataset structure supports answering complex queries such as:

  1. 1.Temporal Evolution:
  2. 2."How did air cargo disruptions evolve over the week of March 16-22?"
  3. 3."What were the cascading effects of specific airport closures?"
  1. 1.Actor Relationships:
  2. 2."Which organizations were involved in coordinating PPE shipments?"
  3. 3."How did different stakeholders respond to supply chain bottlenecks?"
  1. 1.Geographical Patterns:
  2. 2."Where did major transportation bottlenecks emerge?"
  3. 3."How did regional responses differ across continents?"
  1. 1.Event Causality:
  2. 2."What were the immediate effects of border closure announcements?"
  3. 3."How did policy changes impact cargo movement?"

Dataset Structure

Field Categories and Relationships

  • —this is a learning exercise to confirm the ability of the GKG data to answer the competency questions in the last column
FieldVersion NotesCategoryIntentSource TypeDescriptionExample Competency Questions
Identity & Metadata
GKGRECORDIDCore FieldIdentityRecord TrackingSystem GeneratedUnique identifier (YYYYMMDDHHMMSS-X)"Find all events from a specific 15-minute window"
DATELatest VersionTemporalChronological OrderingPublicationPublication timestamp"What was the sequence of announcements about border closures?"
SourceCollectionIdentifierCore FieldProvenanceSource ValidationSystemType of source (1=web, 2=citation)"Which sources first reported specific disruptions?"
SourceCommonNameCore Field
DocumentIdentifierCore Field
Event Description
V1CountsOriginal GKG 1.0QuantitativeEvent ScaleContent AnalysisNumeric mentions (protesters, casualties)"How many flights were cancelled at specific hubs?"
V2.1CountsLatest Version (Feb 2015)QuantitativeEvent ContextEnhanced AnalysisCounts with positional context and location"What was the scale of disruption at different locations?"
V2.1AmountsLatest VersionQuantitativeResource TrackingContent AnalysisNumeric amounts with context"What were the reported PPE shipment volumes?"
Geographic Context
V1LocationsOriginal GKG 1.0SpatialGeographic ReferenceLocation AnalysisBasic location extraction"Where were the major supply chain disruptions?"
V2EnhancedLocationsVersion 2.0SpatialGeographic ContextEnhanced AnalysisLocations with proximity context"How did disruptions spread geographically?"
Actor Network
V1PersonsOriginal GKG 1.0EntityActor IdentificationName RecognitionPerson name extraction"Who were the key decision makers?"
V1OrganizationsOriginal GKG 1.0EntityOrganization TrackingName RecognitionOrganization extraction"Which agencies coordinated responses?"
V2EnhancedPersonsVersion 2.0EntityActor ContextEnhanced AnalysisPersons with role context"What roles did specific actors play?"
V2EnhancedOrganizationsVersion 2.0EntityOrganization ContextEnhanced AnalysisOrganizations with relationship context"How did organizations collaborate?"
Thematic Analysis
V1ThemesOriginal GKG 1.0SemanticTopic ClassificationTheme DetectionBasic theme categorization"What were the main types of disruptions?"
V2EnhancedThemesVersion 2.0SemanticTheme ContextEnhanced AnalysisThemes with relationships"How did different issues interconnect?"
V1.5ToneIncremental UpdateSentimentEmotional ContextTone AnalysisSix emotional dimensions plus word count"How did response sentiment change over time?"
Enhanced Context
V2GCAMVersion 2.0AnalyticalDeep Content AnalysisML Analysis2,300+ content dimensions"What were the underlying patterns in coverage?"
V2.1EnhancedDatesLatest VersionTemporalTime ReferenceDate AnalysisReferenced dates and context"What future events were being planned?"
V2.1QuotationsLatest VersionAttributionDirect EvidenceQuote ExtractionAttributed statements"How did stakeholders explain their decisions?"

Ontological Relationships

The dataset supports three primary types of relationships:

  1. 1.Temporal Relationships:
  2. 2.Event sequences
  3. 3.Cause-effect chains
  4. 4.Response timelines
  1. 1.Spatial Relationships:
  2. 2.Geographic clustering
  3. 3.Spread patterns
  4. 4.Regional interconnections
  1. 1.Actor Relationships:
  2. 2.Organizational networks
  3. 3.Response coordination
  4. 4.Stakeholder interactions

Attribution

This dataset contains data from the GDELT Project (https://www.gdeltproject.org).