My name is Eliezer Silva, I am a PhD Research Fellow in the Data and Artificial Intelligence Group, at the Department of Computer Science, The Norwegian University of Science and Technology. In my doctoral studies I am researching probabilistic / Bayesian / statistical modelling approaches and scalable (approximate) inference algorithms for personalization problems (Collaborative Filtering, Recommender Systems, Link Prediction, etc), under the supervision of Prof. Helge Langseth and Prof. Heri Ramampiaro.
I obtained my M.Sc. degree in Computer Engineering from the School of Electrical and Computer Engineering (Unicamp – State University of Campinas) under the supervision of Prof. Eduardo Valle, with a dissertation with the title “Metric space indexing for nearest neighbour search in multimedia context” (slides, fulltext). This research project focused on trade-offs (quality vs. complexity, for example) in the design of data structures and algorithms for similarity search (k-Nearest Neighbor), with a practical interest in multimedia applications (image retrieval, deduplication, video indexing). Our main approach was to design data-structures and algorithms that generalized Locality-Sensitive Hashing for any metric space using partitioning methods.
I’ve got a B.Sc. in Computer Engineering at the Federal University of Espirito Santo, in 2011. My undergraduate dissertation was on content-based text retrieval using Latent Semantic Indexing (LSI) and Vector Space Model. During my bachelor degree, I’ve worked also with Classical Mechanics Simulation using Functional Programming Language (Scheme).
My general research interest is focused on designing efficient algorithmic methods for Machine Learning and Information Retrieval problems, including search, inference and modelling, with a special interest in probabilistic data structures, probabilistic algorithms and probabilistic models.
Feel free to browse through this website and contact me if you have interest in anything I’ve been working. I try (and mostly fail) to update my research blog, discussing new ideas, methods and stuff.