Marcos Fernández Pichel
Marcos holds a degree in Computer Engineering (2019) and he was awarded with the Extraordinary Award for his Diploma Studies. The Galician Regional Government also awarded him with a BsC Studies Award. His predoctoral research in Information Retrieval (IR) had a high impact at the international level, especially in the misinformation detection community. In fact, his first contribution was published in the prestigious international conference ECIR 2021 (conference A, CORE). This conference is the usual vehicle of communication of the advances in the field of IR in Europe. Marcos also participated in the TREC Health Misinformation detection shared-task campaigns for three years in a row: 2020, 2021 and 2022. Marcos’ team obtained a meritorious third position in the 2021 campaign, while competing against some of the top universities in the IR field, such as the University of Waterloo. Derived from these successful participations, Marcos developed a health misinformation detection system whose design, architecture and evaluation were published as a journal publication in Engineering Applications in Artificial Intelligence (EAAI), Q1 in Computer Science and Engineering. The Galician Royal Academy of Sciences (RAGC) awarded Marcos for this publication with the Ernesto Vieitez Cortizo 2023 Award, which recognises the best science research project conducted by a young researcher in Galicia.
As part of his predoctoral studies, he also led several open source projects, yielding libraries and datasets that are widely used in both the Natural Language Processing (NLP) and IR communities. For instance, Marcos and his peers developed Pyplexity, a multi-purpose text pre-processing Python library that currently has more than 20,000 downloads. This pre-processing method was published in the journal Natural Language Engineering (NLE), Q1 in Linguistics JCR. Overall, Marcos contributed with 15 publications during his predoctoral stage, presented in both international conferences and journals. These articles already have a high number of external citations. Marcos also participated in the programme committee of several international conferences: ECIR (2023 and 2024 editions), LREC 2024, and the Conference on Information Knowledge and Management (CiKM) 2023, among others.
During his predoctoral studies he made a research stay at the prestigious Chair of Information Science at the Universität Regensburg in Germany. This research stay strengthened the collaboration between both research groups, also deriving in international publications like one in the ACM Conference on Conversational User Interfaces (CUI). Marcos also actively collaborated with other research groups, like the IRLab in the University of Coruña (UDC) or the Grupo Novos Medios from the Faculty of Communication at the University of Santiago de Compostela. The first collaboration is framed within the Early Risk Detection Initiative (eRisk), which is a shared-task that both universities organise every year. The collaboration with Grupo Novos Medios is multidisciplinary and focuses on online misinformation detection related to radon gas (where Marcos is responsible for the core technical part).
During his doctoral studies, Marcos also acted as teaching assistant in the BsC in Computer Science of the USC and the MsC in Big Data Analytics at the same university. He also successfully supervised 3 BsC Degree Final Projects and 2 MsC Degree Final Projects.
Currently, Marcos is an Assistant Professor at the Escola Politécnica Superior de Enxeñaría (EPS) of the University of Santiago de Compostela in Lugo’s campus. He has teaching duties at the BsC in Management and Technology (“Empresa e Tecnoloxía”) and the BsC in Robotics (“Robótica”). Among their most recent publications is one that evaluates how reliable both search engines and LLMs are in answering medical questions, which was published in NPJ Digital Medicine (a Q1 journal in Medical Informatics). Additionally, they have just published two papers at the prestigious SIGIR conference (CORE A*): one evaluating how users can better search for web information to combat misinformation, and another proposing an automatic query reformulation strategy to improve search result quality. Among his most recent publications is one that evaluates how reliable both search engines and LLMs are in answering medical questions, which was published in NPJ Digital Medicine (a D1 journal in Medical Informatics). Additionally, he has just published two papers at the prestigious SIGIR conference (CORE A*): one evaluating how users can better search for web information to combat misinformation, and another proposing an automatic query reformulation strategy to improve search result quality. He also regularly participates in dissemination activities oriented to the society, organised both by public and private institutions.
Collaborator Researchers
/team/marcos-fernandez-pichel
team_en