< Back to previous page

Project

Assistance to the Assessment and Methodological support Unit for the provision of services to EFSA on the use of machine learning techniques for literature reviews and systematic reviews in EFSA. (R-7363)

The awarded project focuses on machine learning techniques that could be used by the European Food Safety Authority (EFSA) during the production of literature reviews and/or systematic reviews in order to harmonise and streamline the process, saving time and resources. Therefore the scientific team (Prof. Marc Aerts, Dr. Stijn Jaspers), the advisory team (Prof. Tomasz Burzykowski and Prof. Mieke Aerts) and supporting team (Dr. Sarah Vercruysse and Ms Chantal Brody) will work towards the following general and specific objectives (SO), each corresponding to specific work packages to be completed in a total project time of 18 months: General objective is to obtain selected Machine Learning Techniques (MLTs) for the routine generation of literature reviews and systematic reviews in EFSA. SO1 Identify and describe the steps of literature searches and/or systematic reviews, based on the paper by Tsafnat et al. (2014). For each of these steps, one should assess if the task can be conducted using MLTs. SO2 Evaluate potential benefits and challenges associated with the MLTs introduced in the first specific objective. SO3 Identify specific links, areas for integration of resources with related EFSA on going activities and investigate the integration of already existing tools (e.g. DistillER) with new identified promising MLTs in diverse EFSA scientific areas SO4 Test/pilot the application of the most reliable MLTs on already produced systematic reviews, with a minimum of 2 case-studies and a minimum of 2 MLTs to be tested SO5 Produce a final report including SWOT analyses, sensitivity and specificity levels and recommendations regarding the use of tested methodologies/technologies and provide a fully documented list of MLTs, R codes and algorithms that can be readily used in EFSA routinely.
Date:14 Sep 2016 →  13 Mar 2018
Keywords:MISSING VALUES, MULTIVARIATE CATEGORICAL DATA
Disciplines:Public health care, Public health sciences, Public health services