Resumé

Rita T. Sousa is currently working as a researcher on the KI-Diabetes Detection Project at the Data and Web Science Group, University of Mannheim, Germany. Her research interests include knowledge graphs, machine learning and biomedical applications.

Additional details are available on the CV.

Education

PhD in Informatics

2019 - 2024

Faculty of Sciences of the University of Lisbon, Portugal

International Semantic Web Research Summer School

July 2023

Bertinoro, Italy

Master's Degree in Bioinformatics and Computational Biology

2017 - 2019

Faculty of Sciences of the University of Lisbon, Portugal

Bachelor's Degree in Health Sciences

2014 - 2017

University of Lisbon, Portugal

Professional Experience

Teaching Assistant of Knowledge Graphs

September 2024 - Present

University of Mannheim, Germany

Researcher

January 2024 - Present

Data and Web Science Group, University of Mannheim, Germany

Visiting researcher

September 2022 - November 2022

Data and Web Science Group, University of Mannheim, Germany

Teaching Assistant of Database Management

February 2022 - July 2023

Católica Lisbon School of Business and Economics, Portugal

Researcher

September 2019 - 2023

LASIGE, Faculty of Sciences, University of Lisbon, Portugal

Publications

  • All
  • 2024
  • 2023
  • 2022
  • 2021
  • 2020
  • 2019

Explaining protein–protein interactions with knowledge graph-based semantic similarity

Rita T. Sousa, Sara Silva, Catia Pesquita
Computers in Biology and Medicine
2024

Multi-domain knowledge graph embeddings for gene-disease association prediction

Susana Nunes, Rita T. Sousa, Catia Pesquita
Journal of Biomedical Semantics
2023

Biomedical knowledge graph embeddings with negative statements

Rita T. Sousa, Sara Silva, Heiko Paulheim, Catia Pesquita
ISWC 2023
2023

Explainable representations for relation prediction in knowledge graphs

Rita T. Sousa, Sara Silva, Catia Pesquita
KR 2023
2023

Supervised biomedical semantic similarity

Rita T. Sousa, Sara Silva, Catia Pesquita
IEEE Access
2023

Benchmark datasets for biomedical knowledge graphs with negative statements

Rita T. Sousa, Sara Silva, Catia Pesquita
SeWeBMeDA 2023
2023

Explaining artificial intelligence predictions of disease progression with semantic similarity

Susana Nunes, Rita T. Sousa, Filipa Serrano, Ruben Branco, Diogo F. Soares, Andreia S. Martins, Eleonora Auletta, Eduardo N. Castanho, Sara C. Madeira, Helena Aidos, Catia Pesquita
CLEF
2022

Towards supervised biomedical semantic similarity

Rita T. Sousa, Sara Silva, Catia Pesquita
SeWeBMeDA 2022
2022

The supervised semantic similarity toolkit

Rita T. Sousa, Sara Silva, Catia Pesquita
ESWC 2022 Poster and Demo Track
2022

Explaining protein-protein interaction predictions with Genetic Programming

Rita T. Sousa, Sara Silva, Catia Pesquita
Evo* 2022 Late-Breaking Abstracts
2022

Is there data leakage in protein-protein interaction prediction using knowledge graphs?

Rita T. Sousa, Sara Silva, Catia Pesquita
ISWC Poster and Demo Track
2021

Predicting gene-disease associations with knowledge graph embeddings over multiple ontologies

Susana Nunes, Rita T. Sousa, Catia Pesquita
Bio-Ontologies ISMB
2021

A collection of benchmark data sets for knowledge graph-based similarity in the biomedical domain

Carlota Cardoso,Rita T. Sousa, Sebastian Köhler, Catia Pesquita
Database
2020

evoKGsim+: a framework for tailoring knowledge graph-based similarity for supervised learning

Rita T. Sousa, Sara Silva, Catia Pesquita
ESWC Poster and Demo Track
2020

A collection of benchmark data sets for knowledge graph-based similarity in the biomedical domain

Carlota Cardoso, Rita T. Sousa, Sebastian Köhler, Catia Pesquita
ESWC Poster and Demo Track
2020

Evolving knowledge graph similarity for supervised learning in complex biomedical domains

Rita T. Sousa, Sara Silva, Catia Pesquita
BMC Bioinformatics
2019

Presentations

Serving up-to-date Dynamic Knowledge Graph Embeddings [Slides]

Presentation @ Open Science Day 2024

Integrating gene expression data through knowledge graphs for diabetes prediction [Slides]

Presentation @ SeWeBMeDA 2024

From biomedical data to knowledge [Slides]

Presentation @ MCDS Academic Speed Dating

Biomedical knowledge graph embeddings with negative statements [Slides] [Video]

Presentation @ ISWC 2023

Explainable representations for relation prediction in knowledge graphs [Slides]

Presentation @ KR 2023

Benchmark datasets for biomedical knowledge graphs with negative statements [Slides] [Video]

Presentation @ SeWeBMeDA 2023

Can ChatGP explain protein interactions better than knowledge graphs? [Poster]

Presentation @ Workshop LASIGE 2023

Towards supervised biomedical semantic similarity [Slides]

Presentation @ SeWeBMeDA 2022

The supervised semantic similarity toolkit [Poster]

Presentation @ ESWC 2022

Explaining protein-protein interaction predictions with genetic programming [Slides] [Poster]

Presentation @ Evo* 2022 Late-Breaking Abstracts

Is there data leakage in protein-protein interaction prediction using knowledge graphs? [Poster] [Video]

Presentation @ ISWC 2021

evoKGsim+: a framework for tailoring knowledge graph-based similarity for supervised learning [Poster] [Video]

Presentation @ ESWC 2020

Tools and Datasets

NEGKNOW

Knowledge graph predictions using negative statements challenge.

Benchmark datasets for KG-based similarity

Collection of 21 benchmark data sets for evaluating semantic similarity measures for large biomedical knowledge graphs and ontologies.

Benchmark datasets with negative statements

Collection of datasets for relation prediction tasks that circumvent the difficulties in building benchmarks for knowledge graphs with negative statements.

SEEK

SEEK generates explainable embedding-based representations to support relation prediction in knowledge graphs.

SEEKer

SEEKer is a web-based application designed to facilitate the analysis of SEEK explanations.

TrueWalks

TrueWalks learn knowledge graph embeddings, considering negative statements and their implications.

KGsim2vec

KGsim2vec uses aspect-oriented semantic similarity features to obtain explainable representations of pairs of entities in a knowledge graph for relation prediction.

The supervised semantic similarity toolkit

The toolkit trains supervised ML algorithms to learn a supervised semantic similarity according to the objective similarity.

evoKGsim+

evoKGsim+ applies genetic programming over a set of semantic similarity features to obtain the best combination for a given supervised learning task.

Contacts and Social Media