Artificial Intelligence applications to data-driven groundwater resources management: the KINDRA example

Marco Petitta¹, Leila Jane Di Giacomo² & Sathish Sadhasivam^1
1sup ²Brooklyn Technical High School, 29 Fort Greene Pl, Brooklyn, NY 11217, United States.
Corresponding author e-mail: marco.petitta@uniroma1.it

The European Union Project Knowledge Inventory for Hydrogeology Research (KINDRA https://kindraproject.eu/) built a metadata database very useful to be tested by using a tool of Artificial Intelligence across the knowledge inventory for hydrogeology research, with the ultimate goal of helping the implementation of policies and optimization of water management in Europe. As a common repository of research across all European countries, KINDRA has faced three challenges that we tested the application of AI to help resolve. First and foremost, standardization of the translation into a common language (English) of contributions in multiple languages. An additional challenge was the generation and continuous optimization of keywords as the database grew. The use of AI has proven effective at identifying optimal keywords both in an unconstrained environment and in a constrained one, among a predetermined list. As all research data repositories, the third challenge faced by KINDRA was its ease of use by the research community at large, testing translation in other languages than english.

groundwater, metadata, artificial intelligence, European waters, FAIR data.