By Eurecat – Technology Centre of Catalonia
MediaFutures has started its 2nd support programme of projects focused on shedding light on the phenomena of misinformation and disinformation. Start-ups and artists will use data-driven innovative approaches to handle controversial information. MediaFutures makes open datasets and open-source tools available to participants and the community to handle data based on the participants’ expertise levels.
In the repository, participants can find relevant datasets to study information dynamics. The repository will be updated with other original and unique data created by the participants of the programme. Furthermore, we provide a selection of free open source tools for data collection, analysis and visualization; these include digital methods tools that help to manipulate and visualize data through a visual interface without coding, and data science tools, primarily Python packages that can help to process data and extract information, covering the most common libraries currently used in innovative data-driven projects.
Algorithmic Fairness
MediaFutures also recommends the use of tools to check bias and discrimination in machine learning models, which become essential to create trustworthy AI systems. In this case, we will propose tools such as Aequitas, a bias and fairness audit toolkit for machine learning; AI Fairness 360, a toolkit to help detect and remove bias in machine learning models, or What-if tool, a visual tool to test machine learning models, among others.
1st cohort tools
Finally, the MediaFutures data repository contains a summary of data tools used during the execution of the 1st cohort projects. We also propose some open-source tools with similar functionalities as an alternative to the proprietary tools used by participants, with the objective to advocate for open-source software.
During this 2nd programme, Eurecat – Technology Centre of Catalonia, will offer support to the selected participants in data science-related aspects and design experiments with a short selection of projects in order to perform research on topics such as citizen engagement, misinformation, controversy analysis, etc., thus adding value to the projects and generating knowledge about these issues that are becoming so relevant in our society.