• Corpus ID: 246863814

Introducing the ICBe Dataset: Very High Recall and Precision Event Extraction from Narratives about International Crises

  title={Introducing the ICBe Dataset: Very High Recall and Precision Event Extraction from Narratives about International Crises},
  author={Rex W. Douglass and Thomas Leo Scherer and J. Andr{\'e}s Gannon and Erik Gartzke and Jon Lindsay and Shannon P. Carcelli and Jonathan Wilkenfeld and David M. Quinn and Catherine Aiken and Jose Miguel Cabezas Navarro and Neil Lund and Eglė Murauskaitė and Diana Partridge},
How do international crises unfold? We conceive of international affairs as a strategic chess game between adversaries, necessitating a systematic way to measure pieces, moves, and gambits accurately and consistently over different contexts and periods. We develop such a measurement strategy with an ontology of crisis actions and interactions and apply it to a high-quality corpus of crisis narratives recorded by the International Crisis Behavior (ICB) Project. We demonstrate that the ontology has… 


Extracting Political Events from Text Using Syntax and Semantics
This paper formalizes the process of event extraction and introduces a method for identifying the words in text that report who is doing what to whom, along with where, when, why, and as reported by whom.
Near crises in world politics: A new dataset
Crisis escalation to war is a subject of longstanding interest. Case studies, formal models and statistical analysis offer compelling explanations for why some crises escalate to war while others do
Introducing the UCDP Georeferenced Event Dataset
The rationale for the new UCDP GED is introduced, the basic coding procedures as well as the quality controls are explained, and an example of how the data can be used is provided, by illustrating the association between cities and organized violence, taking population density into account.
Integrating domain knowledge: using hierarchies to improve deep classifiers
This paper proposes to make use of preexisting class hierarchies like WordNet to integrate additional domain knowledge into classification by encoding the properties of such a class hierarchy into a probabilistic model and derive a novel label encoding and a corresponding loss function.
The International Crisis Behavior Project
Over the course of more than four decades the International Crisis Behavior (ICB) Project, a major and ongoing data-gathering enterprise in the social sciences, has compiled data that continues to be
Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
Sentence-BERT (SBERT), a modification of the pretrained BERT network that use siamese and triplet network structures to derive semantically meaningful sentence embeddings that can be compared using cosine-similarity is presented.
Introducing ACLED: An Armed Conflict Location and Event Dataset
This article presents ACLED, an Armed Conflict Location and Event Dataset. ACLED codes the actions of rebels, governments, and militias within unstable states, specifying the exact location and date
Crises in World Politics
In examining patterns in international crises, the authors offer one path to a cocerted attack on a central phenomenon in world politics. After surveying the releva literature, including competing
What do we know about crisis, escalation and war? A visual assessment of the International Crisis Behavior Project
  • P. James
  • Art
    Conflict Management and Peace Science
  • 2018
This study uses a visualization technique, systemism, to integrate ICB Project findings about crisis, escalation and war in particular. The domain of the analysis, 1999–2017, is the period following
A Comprehensive Survey on Schema-based Event Extraction with Deep Learning
This paper summarizes the task definition, paradigm, and models of schema-based event extraction and then discusses each of these in detail, focusing on deep learning-based models.