Talking to GDELT Through Knowledge Graphs
Paper • 2503.07584 • Published
page_content stringlengths 1.48k 5.24k | metadata dict |
|---|---|
arXiv:2503.07584v3 [cs.IR] 24 Jun 2025
Talking to GDELT Through Knowledge Graphs
Audun Myers, Max Vargas, Sinan G. Aksoy, Cliff Joslyn, Benjamin Wilson,
Lee Burke, Tom Grimes
Abstract
In this work we study various Retrieval Augmented Regeneration (RAG) approaches to gain an
understanding of the strengths and weakness... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
[23, 28].
Throughout this work we use the GDELT dataset as a case study.
GDELT is a massive collection of
news reports that provide a real-time computational record of global events that is published every 15
minutes. It aggregates information from various news sources, blogs, and social media platforms to construct
a ... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
2
Constructing a Knowledge Graph for GDELT
As previously mentioned, while the GDELT-GKG2 dataset is not actually natively in the form of a knowledge
graph, it is advertised and frequently cited as being one. We believe that we are making a distinct contribution
to the research community by converting the very popular G... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
Figure 3: GDELT GKG 2.0 ontology relating articles and events..
Methods to automatically determine the graphical form of a relational database are widely known [21]. Most
naturally, consider a table T with m rows T[j], 1 ≤j ≤m and n columns T.i, 1 ≤i ≤n. Then each of the
m rows T[j] is represented as a node in one meta... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
• Solid edges indicate a field in a relational table and are labeled with the type of semantic relation.
• Dashed and bold edges indicate the structural, one-to-many relations in the relational schema.
The naming convention also captures the unique identifier for these csv files, so that ⟨GLOBALEVENTID⟩
identifies uniq... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
(a) Example DKG constructed from
ontology with no labels, but color
coding set to match ontology.
(b) Example LKG constructed cor-
pus of text using LlamaIndex.
(c) Example GRKG constructed cor-
pus of text using GraphRAG, re-
moving all isolated nodes.
Large
nodes have degree ≥25.
Figure 4: KG formations from GDELT Da... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
we have five pipelines on how we implement an LLM for talking to GDELT. These pipelines are shown
in Fig. 1, where in each case, we use an LLM to produce a final answer from the information obtained
from each retrieval method. We note that this final processing is done with models with 7-8B parameters.
With the GraphRA... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
Graph Query
on DKG
G-Retriever
on DKG
RAG using
Vector Store
G-Retriever
on LKG
GraphRAG Q&A
on GRKG
What is the name of
the Bridge that
collapsed and what
river was it on?
The Francis Scott Key
Bridge and it was on
the Patapsco River
in Maryland.
The bridge is located
in Sri Lanka.
However, there is no
explicit mentio... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
stores reveals that the LLM independently recognizes Brandon Scott as the mayor of Baltimore. For all
other questions, the language model cannot answer the posed questions by itself due to the recency of the
bridge collapse.
To quantitatively evaluate the quality of the answers generated by our different question-answe... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
Figure 5: Box plots comparing the cosine similarity scores of different question-answering methods applied
to the GDELT data concerning the Baltimore bridge collapse. Higher cosine similarity indicates a greater
semantic similarity between the predicted and actual answers.
raw documents or a curated KG should not be of... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
References
[1] Knowledge
graph
index.
https://docs.llamaindex.ai/en/stable/examples/index_structs/
knowledge_graph/KnowledgeGraphDemo/. Accessed: 2024-07-22.
[2] Rawan Alamro, Andrew McCarren, and Amal Al-Rasheed. Predicting saudi stock market index by in-
corporating gdelt using multivariate time series modelling. In ... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
[15] Costas Mavromatis and George Karypis. Gnn-rag: Graph neural retrieval for large language model
reasoning, 2024.
[16] Innocensia Owuor and Hartwig H Hochmair. Temporal relationship between daily reports of covid-19
infections and related gdelt and tweet mentions. Geographies, 3(3):584–609, 2023.
[17] Innocensia Owu... | {
"author": "Audun Myers; Max Vargas; Sinan G. Aksoy; Cliff Joslyn; Benjamin Wilson; Lee Burke; Tom Grimes",
"creationDate": "",
"creationdate": "",
"creator": "arXiv GenPDF (tex2pdf:)",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/2503.07584v3.pdf",
"format": "PDF 1.5",
"keywor... |
THE GDELT GLOBAL KNOWLEDGE GRAPH (GKG)
DATA FORMAT CODEBOOK V2.1
2/19/2015
http://gdeltproject.org/
INTRODUCTION
This codebook introduces the GDELT Global Knowledge Graph (GKG) Version 2.1, which expands
GDELT’s ability to quantify global human society beyond cataloging physical occurrences towards
actually... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
adds a series of new capabilities that greatly enhance what can be done with the GKG data, opening
entirely new analytic opportunities. Some of the most significant changes:
Realtime Measurement of 2,300 Emotions and Themes. The GDELT Global Content Analysis
Measures (GCAM) module represents what we believe i... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
of contextual details from the news, encoding not only the people, organizations, locations and
events driving the news, but also functional roles and underlying thematic context. However,
with the previous GKG system it was difficult to associate those various data points together.
For example, an article might ... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
Single Data File. Previously there were two separate GKG data files, one containing Counts only
and one containing the full GKG file. The original rationale for having two separate files was that
users interested only in counts could download a much smaller daily file, but in practice nearly
all applications us... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
or mentions an argument over constitutionalism or a forthcoming policy announcement, will now be
included in the GKG stream. Similarly, an article that has no recognizable metadata, but does yield
GCAM emotional/thematic scores will also be included. When processing GKG 2.1 files, users should
therefore be careful... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
V2SOURCECOLLECTIONIDENTIFIER. (integer) This is a numeric identifier that refers to the
source collection the document came from and is used to interpret the DocumentIdentifier in
the next column. In essence, it specifies how to interpret the DocumentIdentifier to locate the
actual document. At present, it ca... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
the fields within a Count are separated by the pound symbol (“#”). Unlike the primary GDELT
event stream, these records are not issued unique identifier numbers, nor are they dated. As an
example of how to interpret this file, an entry with CountType=KILL, Number=47,
ObjectType=”jihadists” indicates that the articl... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
proximity to it. If a theme is mentioned multiple times in a document, each mention will appear
separately in this field.
V1LOCATIONS. (semicolon-delimited blocks, with pound symbol (“#”) delimited fields) This is a
list of all locations found in the text, extracted through the Leetaru (2012) algorithm. 2 The... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
shortform of the state’s name (such as “TX” for Texas). Note: see the notice above for
CountryCode regarding the FIPS10-4 / GENC transition. Note: to obtain ADM2 (district-
level) assignments for locations, you can either perform a spatial join against a ShapeFile
template in any GIS software, or cross-walk the Fea... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
or council. This engine is highly adaptive and is currently tuned to err on the side of inclusion
when it is less confident about a match to ensure maximal recall of smaller organizations around
the world that are of especial interest to many users of the GKG. Conversely, certain smaller
companies with names and c... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
o Word Count. (integer) This is the total number of words in the document. This field
was added in version 1.5 of the format.
V2.1ENHANCEDDATES. (semicolon-delimited blocks, with comma-delimited fields) This
contains a list of all date references in the document, along with the character offsets of
approxim... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
the score of any word-count field to convert to a percentage density score. As an example,
assume a document with 125 words. The General Inquirer dictionary has been assigned the
DictionaryID of 2 and its “Bodypt” dimension has a DimensionID of 21. SentiWordNet has a
DictionaryID of 10 and its “Positive” dimensio... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
significance, credibly, and/or interest to their audiences. Only image-based embedded posts are
included in this field – videos are identified in the following field.
V2.1SOCIALVIDEOEMBEDS. (semicolon-delimited list of URLs). News websites are increasingly
embedding videos inline in their articles to illustr... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
offsets of approximately where in the document they were found. Its primary role is to allow
for rapid numeric assessment of evolving situations (such as mentions of everything from the
number of affected households to the estimated dollar amount of damage to the number of
relief trucks and troops being sent into t... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
indicates that the Stanford Chinese Word Segmenter 10 was used to segment the text
into individual words and sentences, which were then translated by GDELT
Translingual’s own version 1.0 Chinese (Traditional or Simplified) translation and
language models.
V2EXTRASXML. (special XML formatted) This field is rese... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150301133731-05'00'",
"creationdate": "2015-03-01T13:37:31-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/gdelt-gkg-2-1.pdf",
"format": "PDF 1.6",
"keywords": "",
"modDate": "D:202502010... |
THE GDELT EVENT DATABASE
DATA FORMAT CODEBOOK V2.0
2/19/2015
http://gdeltproject.org/
INTRODUCTION
This codebook provides a quick overview of the fields in the GDELT Event file format and their
descriptions. GDELT Event records are stored in an expanded version of the dyadic CAMEO format,
capturing two act... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
EVENT TABLE
EVENTID AND DATE ATTRIBUTES
The first few fields of an event record capture its globally unique identifier number, the date the event
took place on, and several alternatively formatted versions of the date designed to make it easier to
work with the event records in different analytical software pro... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
terrorists,” but the TABARI ACTORS dictionary labels that group as “Insurgents,” the latter label will be
used. Use the GDELT Global Knowledge Graph to enrich actors with additional information from the rest
of the article. NOTE: the CountryCode field reflects a combination of information from the TABARI
ACTORS dic... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
organizational classes like Non-Governmental Movement. Special codes such as Moderate and
Radical may refer to the operational strategy of a group.
Actor1Type2Code. (string) If multiple type/role codes are specified for Actor1, this returns the
second code.
Actor1Type3Code. (string) If multiple type/role c... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
Conflict. This field specifies this primary classification for the event type, allowing analysis at the
highest level of aggregation. The numeric codes in this field map to the Quad Classes as follows:
1=Verbal Cooperation, 2=Material Cooperation, 3=Verbal Conflict, 4=Material Conflict.
GoldsteinScale. (floati... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
occurrence described in the context of a more positive narrative (such as a report of an attack
occurring in a discussion of improving conditions on the ground in a country and how the
number of attacks per day has been greatly reduced). NOTE: this field refers only to the first
news report to mention an event and ... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
Actor1Geo_Type. (integer) This field specifies the geographic resolution of the match type and
holds one of the following values: 1=COUNTRY (match was at the country level), 2=USSTATE
(match was to a US state), 3=USCITY (match was to a US city or landmark), 4=WORLDCITY
(match was to a city or landmark outside t... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
Finally, a set of fields at the end of the record provide additional data management information for the
event record.
DATEADDED. (integer) This field stores the date the event was added to the master database
in YYYYMMDDHHMMSS format in the UTC timezone. For those needing to access events at 15
minute reso... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
records GDELT’s confidence in its extraction of that event from that particular article. This field is a
percent, ranging from 10 to 100% and indicates how aggressively GDELT had to perform tasks like
coreference or grammatical restructuring to extract the event from that article. Sorting all mentions of
an event ... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
MentionIdentifier. (integer) This is the unique external identifier for the source document. It
can be used to uniquely identify the document and access it if you have the necessary
subscriptions or authorizations and/or the document is public access. This field can contain a
range of values, from URLs of ope... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
the MentionIdentifier field above to merge the Mentions table with the GKG table to access the
complete set of 2,300 emotions and themes from the GCAM system.
MentionDocTranslationInfo. (string) This field is internally delimited by semicolons and is used
to record provenance information for machine translat... | {
"author": "Kalev Leetaru",
"creationDate": "D:20150219133950-05'00'",
"creationdate": "2015-02-19T13:39:50-05:00",
"creator": "Microsoft® Word 2010",
"file_path": "/home/donbr/don-aie-cohort8/gdelt-knowledge-base/data/raw/GDELT-Event_Codebook-V2.0.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": ... |
GDELT RAG Source Documents
Dataset Description
This dataset contains source documents extracted from the research paper "Talking to GDELT Through Knowledge Graphs" (arXiv:2503.07584v3). The documents are used as the knowledge base for a Retrieval-Augmented Generation (RAG) system focused on GDELT (Global Database of Events, Language, and Tone) analysis.
Dataset Summary
- Total Documents: 38 pages
- Source: Research paper on GDELT Knowledge Graphs
- Format: PDF pages with extracted text and metadata
- Use Case: RAG knowledge base for GDELT-related queries
Data Fields
page_content(string): Extracted text content from the PDF pagemetadata(dict): Document metadata including:title: Paper titleauthor: Paper authorspage: Page numbertotal_pages: Total pages in source documentfile_path: Original file pathformat: Document format (PDF)producer,creator: PDF metadata- Other PDF metadata fields
Data Splits
This dataset contains a single split with all 38 documents.
Source Data
The source material is the research paper:
- Title: "Talking to GDELT Through Knowledge Graphs"
- Authors: Audun Myers, Max Vargas, Sinan G. Aksoy, Cliff Joslyn, Benjamin Wilson, Lee Burke, Tom Grimes
- arXiv ID: 2503.07584v3
Licensing
This dataset is released under the Apache 2.0 license.
Citation
If you use this dataset, please cite the original paper:
@article{myers2025talking,
title={Talking to GDELT Through Knowledge Graphs},
author={Myers, Audun and Vargas, Max and Aksoy, Sinan G and Joslyn, Cliff and Wilson, Benjamin and Burke, Lee and Grimes, Tom},
journal={arXiv preprint arXiv:2503.07584},
year={2025}
}
Dataset Creation
This dataset was created as part of the AI Engineering Bootcamp Cohort 8 certification challenge project.
- Downloads last month
- 7