The HCSS Datalab

Strategic decision-making is inherently subject to significant risk. Uncertainty makes it difficult to engage in long-term decision making, preventing you from taking bold action and realising your desired impact or potential. Within the HCSS Datalab we model complex phenomena, quantify and compare the impact of decisions, to provide insights into such risk. This allows us to gain a better understanding into the everchanging world of geopolitics, defence, and security. Using data and in-depth subject matter knowledge we can help to enhance decision-making and better communicate stories to in turn reach the desired audiences.

What can we do?

The Datalab support HCSS’ research based on the following six pillars

Policy Intervention – By means of optimization, strategy, and causation we can model optimal policy interventions.
Strategy – By leveraging expert surveys for comparison and analysis, we input data into game-theoretic models to simulate strategic interactions between agents and design intervention effectiveness. Additionally we make use of serious gaming setups (link to the gaming webpage).
Causation – Causality explores ‘why things are the way they are’ so that it becomes possible to select the right option, and you can act on that knowledge. Using causal modelling is the first step towards policy evaluation and intervention. Thus, the explanatory power of causal reasoning is potentially a game-changer for policy.
Association – We employ techniques such as indices, machine learning, and ensemble models to get insights into prediction and forecasting.
Dashboarding and Monitoring – Amongst our data visualisation tools we design and build interactive dashboards and monitors. Such as the Climate Security Risk Monitor, The Dutch Foreign Relations Index, Critical Raw Materials Dashboard, and our dashboards related to Cyber Transparency.
Natural Language Processing – Natural Language Processing is a branch of Artificial Intelligence that uses computational techniques in order to learn, understand and produce content based on human languages.

Datalab | News & events

HCSS Datalab introduces GINA – Geopolitics Powered by Data

HCSS Datalab presents: GINA Military

Internship opportunity at the HCSS Datalab

IPMU2024 | HCSS data scientist Maarten Vonk presents paper on Optimizing Causal Interventions in Hybrid Bayesian Networks

HCSS Datalab presents: GINA Economic

Arne Eichholtz on his time as a Datalab intern and finding the perfect place to combine his technical background in a policy setting

Maarten Vonk at EnPax: 3rd International Conference on Environmental Peacebuilding

Meet GINA! HCSS Datalab presents: the Geopolitical Interactive Network Analysis

ICMCIS presentation by Maarten Vonk: A Bayesian Network Approach to Hybrid Threat Modelling

HCSS is hiring an NLP-Datalab Intern!

HCSS Annual Report | The HCSS Datalab

HCSS Annual Report 2023

Internship opportunity: join the HCSS Datalab

WPS Quarterly Update | Droughts, floods and heat contribute to conflict risk in Africa, the Middle East and Asia

Inês Lains on bridging quantitative and qualitative research and her time as an Data Intern

Internship opportunity at the HCSS Datalab (Sept/Oct 2023 starting date)

The Hague Mayor Mr. Jan van Zanen visits HCSS

College of William and Mary Students visit HCSS

Thesis project | Catching the trigger? Predicting interstate conflict

HCSS Datalab paper| Disentangling causality: assumptions in causal discovery and inference

HCSS Internship Program: Data Intern

Diederik Dekkers joins HCSS as a Data Scientist

Saskia Heyster joins HCSS as a Data Scientist

Maarten Vonk | Introducing the HCSS Datalab

Internship opportunity: HCSS Datalab

HCSS is hiring | Interactive Monitor Developer

HCSS seeks a Web Developer

Strategy

We can use a game theoretic mathematical model to simulate the effectiveness of counter-hybrid measures. This is based on information about the costs of imposing counter-hybrid measures, its ability to dissuade the adversary from conducting hybrid operations, and its potential to mitigate the damage of hybrid conducts. When simulating, our model assumes that there is a hybrid threat from an hostile actor looming large specified in a scenario analysis.

Causation

Causal modelling can help us enhance our understanding of the different factors connecting the onset of a climate-related events to inter-communal violence. This figure shows a particular example of a case study for the Dhi Qar province in South Iraq. Case-based studies of causal pathways have three important benefits for policy (1) they facilitate design of preventive action, (2) they contribute to early warning, and (3) they help understand which factors link climate change to these potential outcomes.

Association

We aim to bring empirical rigor to foresight analysis. For instance with this forecast prediction of intra-state conflict until 2029 done in collaboration with the Pardee Centre for International Futures.

Dashboarding and Monitoring

Our interactive dashboards and monitors such are platforms that enable its users to monitor anything from Dutch foreign relations and the level of insecurity caused by climate change to tracking the geographical spread of resources, reserves, extraction and processing capabilities of critical raw materials and our dashboards in collaboration with the CyberPeace Institute on Cyber Transparancy.

Natural Language Processing

We use Natural Language Processing to extract topics (topic modelling) from in-house curated text corpora. But also to conduct Named Entity Recognition, for instance to see how often a country or person is mentioned in combination with weapons or arms trade in the literature. This says something about the state of organised crime with regard to arms trafficking in this country. One possible output is this network graph related to organised crime.