About UNDP
As the United Nations lead agency on international development, UNDP works in 170 countries and territories to eradicate poverty and reduce inequality. We help countries to develop policies, leadership skills, partnering abilities, institutional capabilities, and to build resilience to achieve the Sustainable Development Goals. Our work is concentrated in three focus areas; sustainable development, democratic governance and peace building, and climate and disaster resilience.Activities by UNDP
Crisis Risk Dashboard: Predictive Analytics and Multi-dimensional Analysis
Africa DjiboutiEritreaEthiopiaSomalia
The Risk Anticipation Hub is developing predictive analytics components to the Crisis Risk Dashboards and related products. This includes:
1. Better anticipation of conflicts & impacts of prevention
We leverage the most recent advances in machine learning to obtain conflict forecasts and conflict prevention impact estimates to improve risk anticipation. For this we combine the world-leading conflict forecasting systems in a new data product that will enable forward-looking risk analysis and decision-support, including a prioritisation of countries for prevention activities and sub-national forecasts of conflict risks.
2. Conflict forecasting using news data
We utilise GDELT’s high frequency, text-based data to capture the “chatter” of what is happening in local contexts to see if it is predictive of conflict. We limited the geographic scope to the Horn of Africa and aggregated to each country’s admin level 1 and to a 1-2 week time interval. With an escalation of ACLED fatalities as the outcome to predict, we utilise article document embeddings as the main predictors, and conflict history and some static features (population density, etc.) as additional features.
3. Multi-dimensional analysis
We utilise AI and machine learning-based technologies for in-depth multi-dimensional analyses. The types of analyses include:
Natural language processing and text analysis - Given a body of texts, we can find the topics and representative keywords, commonly used terms and phrases, and sentiments and emotions reflected in the content. We can also categorise texts into predefined groups to find trends and summarise and distil text into its most essential content. This includes analysis of social media data.
Hotspot and spatial analysis - Given data on events or phenomena, hotspot analysis can find patterns and uncover relationships between spatial information. We can identify geographic areas with significantly high (or low) levels of incidents for further investigation
Outlier and anomaly detection - Given data on events or measurements, various algorithms and methods can be used to find unexpected values that differ significantly from the norm or usual trend. This can alert us to further monitoring or investigation.
Time series analysis - Given data recorded over time at regular intervals, we can identify underlying trends and detect seasonality and repeating patterns to better understand a context.
Digital Innovation: Women and AI (DIWA)
Asia and the Pacific Not ApplicableThis program aims to empower women and girls in Asia and the Pacific to access, shape, and lead in the digital economy. DIWA provides training in AI, data science, and digital branding, and supports women in sectors like AgriTech and energy. It also works to deploy AI-powered tools to detect online harassment and support legal reforms.
Activity Type Policy/Regulatory GuidanceAwareness/AdvocacyFuture Signals Platform AI Extension
Global Not ApplicableThis AI extension implements a feature that ingests articles and processes them into trends across a variety of natural-language topics. It outputs recommendations in the Signal Scanning form to be filled out by human users, supporting the Signals & Trends System for foresight analysis.
Activity Type AI Tools/SolutionsAI for people-centred smart cities workstream in the UN inter-agency working group on AI
GlobalA new workstream was established to understand the needs of cities and the capabilities of UN System organizations in relation to AI. The objectives are to leverage AI for people-centered smart cities, promote ethical guidelines, and strengthen inter-agency cooperation. The project is kicking off a global assessment of responsible AI use in cities.
Activity Type Research/Reports/AssessmentsNetworks/Mentorship/ExchangePolicy/Regulatory Guidance