I am a lecturer in the School of Computer Science and Informatics, which I joined in 2018 as a Research Associate, to work on the ERC-funded Starting Grant project FLEXILOG, alongside prof. Steven Schockaert. Previously, I was a Natural Language Processing (NLP) scientist at Savana (a Madrid-based company focused on delivering data-driven healthcare solutions), and before that I was a PhD student at Pompeu Fabra University (Barcelona). I have a background in English Linguistics and I have worked several years in second language teaching (English and Spanish). My research, in the area of Artificial Intelligence and NLP, is centered on meaning representation, computational semantics, multilingual NLP and modeling social media interactions. I am a Fulbright and laCaixa fellow and Erasmus Mundus alumni.
Currently, my teaching focuses on data manipulation and exploitation, databases (relational and not relational) and NLP. I currently teach at Cardiff University's School of Computer Science and at the National Software Academy.
I was a postdoc working in the ERC-funded FLEXILOG project, with prof. Steven Schockaert.
I was an NLP scientist, helped developing Savana's EH Read technology, an NLP system designed for extracting meaningful information from Electronic Health Records.
I did my PhD under the supervision of Horacio Saggion. I did a research stay at the Linguistic Computing Laboratory, led by Roberto Navigli. My PhD dissertation can be found here. I collaborated with the FP-7 Dr. Inventor project, as well as María de Maeztu's funded Music Meets NLP project.
I did my PhD on the interplay between computational lexicography and Natural Language Processing. During these years, I also taught academic communication for PhD and undergraduate students.
Funded by the laCaixa foundation, I studied a 2-year program in NLP at UAB and University of Wolverhampton. Some of the modules were Programming for Corpus Linguistics, Computational Linguistics, Speech Technologies and MT and Other NLP applications.
Semantic Vector Networks, a Python framework for learning unsupervised relation vectors. Published at Coling 2018.View Project
A neural network architecture implemented in SpaCy (for using syntactic features) and Keras with state-of-the-art results in the task of Definition Extraction (identifying definitional text fragments in corpora). Published at NAACL 2018.View Project
A knowledge-based Python system for learning hypernymic relations (e.g., Crane -> writer, author, human, animal, lifting device, equipment, etc.). Published at EMNLP 2016.View Project