CHILDHUB DEEP LEARNING AND VISUALISATION
The proof-of-concept project “ChildHub Deep Learning and Visualisation” has been developed in collaboration with Terre des hommes, the leading Swiss organisation for children’s aid. Terre des hommes launched an Innovation Grant to elicit intelligence out of various data in the ChildHub platform. In response to users’ needs, the Innovation Grant focused on the design and development of digitally savvy tools to foster more efficient knowledge management and maximise opportunities through networking strategies. GluoNNet was awarded the Innovation Grant thanks to its expertise in state-of-the-art data science and technology combined with the ability to provide strategic foresight.
We developed so-called ‘document galaxies’ which allow the user making sense of the large online library of documents in ChildHub which encompasses different types of publications with an eventual country or regional perspective.
The traditional search engine of the ChildHub online library allows users to find documents by keywords, which limits its functionality. This shortcoming undermines the full exploitation of knowledge in this vast library, making it complicated to discover interesting documents beyond typical search terms.
We created a tool, dubbed Document Galaxy, which makes the library more immediate and intuitive. The tool is based on advanced Machine Learning and Artificial Intelligence techniques to create an abstract map where documents are placed closer if they contain more keywords in common.
THE ADDED BENEFIT
The Document Galaxy method is a powerful tool that provides a structured visualisation of vast amounts of items in a document library. It allows its user to have a better overview, finding crucial information that otherwise would have been concealed.