Applied AI, Data Science, Deep Learning, Machine Learning, Natural Language Processing, Information Retrieval, Health Informatics
PhD in Computer Science (2011 – 2019)
Adviser: Allan Hanbury
Co-Adviser: Guido Zuccon
Title: Understandability and Expertise in Consumer Health Search – Retrieving topically relevant and understandable health information on the Web.
MSc in Computer Science (2010 – 2011)
Adviser: Gisele Pappa
Developed a method to assign weights to features for text classification based on Genetic Programming.
- TrecTools is an open-source Python library for assisting Information Retrieval (IR) practitioners with TREC-like campaigns. It provides an interface to repetitive and common tasks such as analyzing and evaluating runs, running traditional IR frameworks like Indri and Terrier with different baselines, analyzing results, or even fusing ranking lists to create a more robust run.
- Sleep-Wake Benchmark is a toolkit to process actigraphy data for sleep analysis. It provides the required procedures for data manipulation and a set of traditional and machine learning algorithms for sleep-wake sleep stage classification.
- ReadabilityCalculator is an open-source library that implements the most traditional readability formulas and procedures widely used to estimate the readability of a text.
- Machine learning and data analysis: Python, Scikit-learn, Pytorch, TensorFlow, NumPy, SciPy, Pandas, matplotlib, NLTK.
- Information retrieval frameworks: ElasticSearch/Lucene/SOLR, Indri/Lemur, Terrier.
- Favorite text editors: Vim, Jupyter.
- System evaluation: A/B testing, interleaving, IR evaluation in TREC/CLEF.