On February 22-23-24th, twelve MOOD experts came together at the headquarter offices in Lisbon of the Instituto de Engenharia de Sistemas e Computadores – Investigação e Desenvolviment (INESC-ID), MOOD Portugues partner, for a 3-day, face to face workshop dedicated to text mining. Withing MOOD, text mining activities are mainly covered by the TETIS group, part of CIRAD, INRAE, AgroParisTech (France) and INESC-ID (Portugal).
At this meeting, partners conducted in-depth scientific and technical discussions regarding text-mining activities within MOOD. During the first day, the group outlined an overview of MOOD text-mining activities, summarizing the work done and that currently in progress through a series of seven online presentations by MOOD postdoctoral, PhD and MSc researchers.
In the following day, the workshop focussed on “brainstorming dealing with different research topics as Quality (sources, content, etc.), thematic entity recognition, spatial entity recognition, linking on heterogeneous sources, linking on covariates and disease data.” explained Mathieu Roche, co-organizer of the workshop. The discussion touched upon the MOOD hackathon dedicated to Antimicrobial resistance that will be conducted over the next years of the project in order to answer the most intrigued issues regarding AMR surveillance.
“Of course, it is important to organize concrete discussions [i.e. face to face meetings] to focus on specific issues and computer science challenges for WP2 and WP3. Different action points have been identified, e.g. write a document on quality features and how to implement them, dataset preparation with geocoding (how to associate polygons to spatial entities), improving keyphrase extraction with discriminative measures, etc.” reflected Mathieu.
The group is planning to organise specific meetings related to action points (remotely) and a general meeting in person to present the progress of text-mining tasks within MOOD.