From archive to live

Important results from the project

** Denna text är maskinöversatt ** The project "From archive to live" aimed to use an AI tool to automatically read and categorize large amounts of complex text. The AI tool was thus expected to improve the quality of RFSU´s ability to work with UN documents (texts). By using data from the UN Human Rights Council´s archive, this AI tool was expected to improve RFSU´s work with analysis of international regulations, which was also expected to strengthen RFSU´s collaboration with relevant actors, including SDG 3 and 5 and more. The project goal was achieved.

Expected long term effects

** Denna text är maskinöversatt ** The project led to expected results. The project developed an IT tool that managed to collect large amounts of online texts and ask existing AI models to read and categorize them according to given theoretical frameworks. AI-processed responses were saved in a database for end-user use. Obtained AI responses contributed to streamlining the process and improving quality in the analysis of complex political-legal documents relevant to influencing and following up political objective in, among others, the SDGs 3 and 5.

Approach and implementation

** Denna text är maskinöversatt ** The project had 4 parts: method for collection (1), development of thematic reference for AI (2), information management (3) and visualization of AI results, storage and "interface" for end users (4) The project was carried out largely according to plan. Unexpected challenges, e.g. inconsistency in the document source as well as irrelevant answers if too many criteria were included in the question. Time saved from phase-1 enabled more time laid on phase-2 and 3. Clear, more defined questions are lessons for future AI-related work at RFSU