📋 Tender Overview
The Patterns Of Conflict Emergence: Developing An Automated Pattern Recognition System For Conflict
Do Past Events Return As Events Of The Present?History, It Has Been Widely Said, Has A Tendency To Repeat Itself. What About In The Case Of War? The Eu-Funded Pace Project Will Study Recurring Patterns In The Escalation And Emergence Of Wars. It Will Combine New Methods (Using The Shape Of The Sequence Of Events Rather Than Its Raw Values) And Data On Conflict To Extract Typical Pre-War Motifs. The Project Will Identify Patterns From The Onset Of World War I To Hamas’ Rocket Launches. It Will Study The Perceptions Using Data From Financial Markets, News Articles And Diplomatic Documents. Pace Will Also Assess Possible Fundamental Limits To The Predictability Of Conflict Events.
Funded Under
Excellent Science - European Research Council (Erc)
Total Cost
€ 1 952 848,00
Eu Contribution
€ 1 952 848,00
Coordinated By
The Provost, Fellows, Foundation Scholars & The Other Members Of Board, Of The College Of The Holy & Undivided Trinity Of Queen Elizabeth Near Dublin
Ireland
Start Date
1 January 2022
End Date
31 December 2026
Objective
Are there recurring patterns in the escalation and emergence of wars? The idea that history may repeat itself is old. But
recent advances overcoming methodological and data barriers present an opportunity to identify these recurrences
empirically and to examine whether these patterns can be classified to improve forecasts and inform theories of conflict. I
propose to combine new methods—using the shape of the sequence of events rather than its raw values—and novel data
on conflict from finance, diplomatic cables, and newspapers, to extract typical pre-war motifs. Just as DNA sequencing has
been critical to medical diagnoses, PaCE aims to diagnose international politics by uncovering the relevant patterns in the
area of conflict. Our goals are to:
(i) Identify patterns in the pre-conflict actions using data on conflict events—from the onset of WWI to Hamas’s rocket
launches—and in their perceptions using data from financial markets (the crowd’s perception), news articles (the experts ),
and diplomatic documents (the policy-makers). This will allow us to evaluate the patterns of escalation over different
timescales—from the decade to the minute. The similarity between temporal sequences will be measured using algorithms
which allow for flexible matching, such as Dynamic Time Warping.
(ii) Evaluate the utility of these patterns to improve forecasts of conflict with both historical and live out-of-sample
predictions. Our results, using shape-based classification methods, will be made public and evaluated in real time. Moreover,
using new measures of complexity to distinguish regular, chaotic, and random behavior, I will measure possible fundamental
limits to the predictability of conflict events.
(iii) Summarize the core features of dangerous patterns into motifs—recurring patterns—that can
