graymanicom/catalogue_of_AI_applications_in_government

We curate a catalogue of use cases of AI in government. The starting point is AI watch's catalogue of AI used in the governments of the EU, and the process of curation is intended to find applications that are relevant to South Africa. However, the code will be useful for anyone aiming to curate such a dataset. # Purpose This is a documentation of the process undertaken to curate a catalogue of machine learning tools applied in European governments. The purpose of this catalogue is for the Policy Innovation Lab and its partners in the South African government to have a clear sense of what is possible. In order to achieve this we attempt, where possible, to focus on machine learning tools that can be applied to existing South African datasets, or datasets that we predict can be readily produced. We also prioritise those tools that can be developed easily, bearing in mind South Africa’s current digital infrastructure and technical expertise. Furthermore, we discuss potential challenges, if they exist, to deploying tools that were successful in Europe to the South African context. In summary, we catalogue the low-hanging-fruit of machine learning tools that can be deployed in the South African government, prioritising those that yield a good societal return on investment. # The dataset The dataset that we shall curate is the most recent version of the _Public Sector Tech Watch latest dataset of selected cases_ by the European Commission, Joint Research Centre (JRC) available at This is a catalogue of 2154 examples of emerging digital technologies, including AI and blockchain, deployed in EU governments which is published under a Creative Commons Attribution 4.0 International License. The dataset was created on 2023-10-11 and at the time of our access was last modified on 2024-05-30. We choose this data source for the following reasons: 1. It is an extensive catalogue with a substantial amount of data describing the context, development and use of the tools. 2. The EU contains a diverse group of countries in various stages of development, so although we may have additional work to do in considering the South African context, many of the tools will be applicable. 3. The EU is far ahead of the rest of the world in terms of regulating the use of AI. We therefore may assume …