United Nations Global Pulse (UN Global Pulse)


United Nations Global Pulse (UN Global Pulse)

Description of Activities on AI

Project 1: Mapping the landscape of artificial intelligence applications against COVID-19

UN Global Pulse worked with researchers from the World Health Organization (WHO) and the MILA- Quebec AI Institute to map the landscape of AI applications that are being built to tackle the COVID-19 pandemic. The research focused on three specific areas: individual patient diagnosis and treatment, protein and drug discovery related research, and the socio-economic impact of the disease. This work also explains main challenges and opportunities for AI cooperation against COVID-19.

Project 2: On the value of ship traffic data for epidemic modeling of diseases

UN Global Pulse set out to explore whether insights from Automated Identification Systems (AIS) data can be included in epidemic modelling of diseases, including COVID-19, to inform more efficient and timely operational responses.

Project 3: Using speech-to-text tech for epidemic intelligence and COVID19 response

Radio remains the most reliable and affordable medium of accessing and sharing information in most of the developing world. Since 2019, UN Global Pulse worked with the WHO to explore the use of data from radio talk shows to signal early warnings of health risks and health-related matters. More recently, our team used the radio monitoring technology it developed to extract transcripts containing COVID19 keywords and to analyse them. The research showed promising results for monitoring the ongoing health emergency, its social and economic impacts, and the spread of infodemics.

Project 4: Developing an AI-powered platform for social listening

A second version of Qatalog—UN Global Pulse’s text analytics tool that can extract, analyse, and visualise data from social media and radio broadcasts—was developed and released. Qatalog is an AI-powered tool that uses speech recognition technology developed by UN Global Pulse to ‘listen’ to public radio talk shows and automate the detection of words spoken during those shows. It also pulls in public Twitter streams, building on one of several partnerships established by UN Global Pulse with private sector data providers on behalf of the UN System. Lastly, it allows users to upload PDF documents for analysis. At the end of 2019, Qatalog offered users the option to choose from 39 different languages to analyse.

Project 5: PulseSatellite: A collaboration tool using human-AI interaction to analyse satellite imagery

UN Global Pulse has been working with UNOSAT over the last three years to build a software tool that leverages AI to identify and count structures from satellite images. This was then expanded to a web-based toolkit that can be easily adapted to other remote sensing applications and which allows for the incorporation of models created by other users. We already have three models loaded into the system – one that allows users to map structures in refugee settlements, a roof density detection model (e.g. for slum mapping), and a flood mapping application. Down the road, we plan to develop more models together with UN partners.

Project 6: Using artificial intelligence to model displacement in Somalia

Predicting refugee and internally displaced persons (IDP) arrivals is of critical interest in humanitarian emergencies since field operations teams must prepare in advance for these arrivals. UN Global Pulse provided support to Project Jetson, an initiative of the United Nations High Commissioner for Refugees (UNHCR), which developed a dashboard early warning system that displays monthly predicted IDP arrivals for regions in Somalia.

Project 7: Understanding population movement after the 2018 Central Sulawesi natural disasters

Our Pulse Lab Jakarta and the International Organization for Migration (IOM) undertook research using data from an Indonesian telecom provider to gather insights on internal displacement of people affected by a tsunami and subsequent landslides in the Central Sulawesi province of Indonesia. An interactive visualisation was designed to make sense of the results. Among its functions, the dashboard communicates the distribution of subscribers, and highlights the most popular destinations where people travelled to after the disasters.

Project 8: Making sense of diplomatic correspondence using big data analytics and visualization

Collaborating with the Indonesian Ministry of Foreign Affairs (MoFA), Pulse Lab Jakarta explored and produced a machine learning visualisation tool to help staff understand diplomatic correspondence between the Ministry and its diplomatic staff abroad. The tool enables MoFA to analyse, locate and make sense of their large volume of correspondence and improves analysts’ abilities to provide relevant, timely and accurate counsel.

Project 9: Supporting development of ethical AI frameworks in the Global South

UN Global Pulse is working both in AI policy and AI innovation in target African countries in order to accelerate the adoption of AI-based innovation to support the achievement of the SDGs. These activities have informed UN Global Pulse’s development of a blueprint to aid African nations in developing national AI strategies. UN Global Pulse will continue hosting national and regional consultations and dialogues with stakeholders in countries such as Ghana and Uganda to develop national AI strategies.

Project 10: Developing ethical frameworks to support human rights-based approaches to AI

UN Global Pulse has established the Expert Group on Governance of Data and AI, an international group of expert stakeholders whose work emerged from the UN Global Pulse Privacy Advisory Group’s work to develop a first draft of a code of ethics for AI focusing on human rights.

Project 11: Deploying AI social listening tools for real-time information to UN country teams

In line with the Secretary-General’s Strategy on New Technologies, UN Global Pulse has assisted UN System Organizations in brokering partnerships with the private sector in order to gain access to AI-based tools. For instance, UN Global Pulse entered into an agreement with Dataminr – the leading AI platform for first response – to provide UN personnel with access to their real-time event detection product for the public sector called First Alert.

Project 12: Assessing the risks, harms and benefits of AI-intensive projects

UN Global Pulse has developed a two-phase “Risk, Harms and Benefits Assessment Tool,” a data privacy, protection, and ethics compliance mechanism designed to help identify and minimize the risks of harms and maximize the positive impacts of data and AI-fueled projects and tools. The assessment can be used at the onset of a project, or when an existing one needs to undergo changes, taking into account every stage of the data life cycle.

Project 13: Developing a Global Data Access Framework

UN Global Pulse is a co-lead of the Global Data Access Framework (GDAF), which, as stated in the Secretary-General’s Roadmap for Digital Cooperation, is a multistakeholder initiative that aims to create a platform for sharing digital public goods in a manner that respects privacy. Its main objective is to enable data sharing across the public and private sector in a privacy-protective manner by helping to develop and scale AI-driven projects. The GDAF will rely upon a state-of-the-art reference architecture that will be developed through a collaborative multi-stakeholder effort and will enable data to be discovered by AI systems more easily.