Use Cases
How can the VIEWS forecasts be leveraged to support peacebuilding and atrocity prevention across the triple nexus of peace, development, and humanitarian assistance?
Browse the use case gallery below to learn more about some of the current applications of our data.
UN Kenya
Peace and Development Unit (PDU) at the UN Kenya Country Office
A 2022-2023 collaboration to explore the potential and feasibility of using big data and advanced machine learning techniques to forecast political violence in the Kenyan context. Browse project highlights in the slide carousel.
Leveraging VIEWS data for anticipatory risk analysis
Motivated by the UN Secretary General and UNDP data strategies’ call for increased use of data analytics in UN activities, the UN Kenya PDU team partnered with VIEWS to explore the feasibility of using big data and advanced machine learning techniques to forecast political violence in the Kenyan context.
Part of an internal feasibility study, VIEWS forecasts were fed into the national UNDP Crisis Risk Dashboard, where they were post-processed by the PDU team into a format suitable for the specific needs of the Kenyan field office and made readily available to the end users.
The product served as a tool for anticipatory risk analysis in the ongoing project on prevention and inclusive development in North-Eastern Kenya, supporting analytical efforts and providing insights for decision-making and strategic planning at the field level.
Sneak peak of the UNDP Crisis Risk Dashboard, displaying the custom integration of VIEWS data on the right.
Post-processing VIEWS data to tackle challenges and meet the needs of the end users
To meet the needs of the UN Kenya Country Office, the VIEWS predictions were adapted to a new level of analysis (Admin 1), aggregated to a quarterly temporal resolution, and post-processed into custom conflict risk indices. The customized solution also addressed data challenges specific to the Kenyan context, such as the uncertainty that stems from the low incidence of conflict events at the sub-national level in the country.
VIEWS-adaptation of UCDP-recorded state-based political violence in Kenya over Oct 2017-Oct 2022. Post-processed and plotted by UN Kenya.
Customization of VIEWS forecasts for state-based violence in Kenya in Q1-2023. Post-processed and plotted by UN Kenya.
How well does the model predict state-based conflict in Kenya?
Evaluation of how well the VIEWS model predicted state-based conflict in Kenya for each month in 2022, when looking at 1-month-ahead and 3-months-ahead forecasts, respectively.
For the 1-month-ahead plot, the prediction for Jan 2022 is drawn from the Dec 2021 dataset, the prediction for Feb 2022 from the Jan 2022 dataset, and so forth. For the 3-month-ahead plot, the prediction
for Jan 2022 is drawn from the Oct 2021 dataset, the prediction for Feb 2022 from the Nov 2021 dataset, etc.
Evaluation of predictive performance in Kenya for Jan 2022 – Dec 2022, when forecasting one month into the future. Analyzed and plotted by UN Kenya.
Evaluation of predictive performance in Kenya for Jan 2022 – Dec 2022, when forecasting three months into the future. Analyzed and plotted by UN Kenya.
UNESCWA
United Nations Economic and Social Commission for Western Asia
A 2020-2022 collaboration to develop a bespoke forecasting model specialised in predicting armed conflict in the Arab region.
Leveraging a bespoke prediction model to anticipate hot-spots of emergency in the Arab region
Grounded in commissioned research to study the drivers of armed conflict in the Arab region with a particular focus on the linkages between climate and conflict, the project resulted in the launch of the bespoke ViEWS-ESCWA conflict
prediction model.
The model’s predictions are now fed into the internal ESCWA conflict risk dashboard, where they provide a visual quantification of the risk of state-based deadly violence arising from multiple drivers to help identify the hot-spots of emergency that may lay ahead in the Arab region.
Beyond the model itself, this collaboration laid the groundwork for the later expansion also of the main VIEWS model from Africa to
the Middle East, and funded the development of the first iteration of the VIEWS API that facilitates live integration of VIEWS data into in-house data dashboards.
Sneak peak of the ESCWA risk dashboard showcasing results from the bespoke ViEWS-ESCWA model.
Snapshot of the custom adaptation and visualization of conflict predictions from the ViEWS-ESCWA model in the ESCWA risk dashboard. The predictions have been normalized to a risk scale from low to high to better suit the needs of the end users.
UNHCR
United Nations High Commissioner for Refugees
UNHCR uses the conflict predictions from the VIEWS project for our own predictive work regarding the predictions of refugee flows between country pairs 36 months ahead.
As most forced displacement is conflict related, the VIEWS predictions help us to generate better predictions for our planning cycle.
