Cicero Nogueira dos Santos is a Research Staff Member in the Natural Resources Analytics group at IBM Research – Brazil since March 2012. His current research focus is on the development and improvement of machine learning and data mining approaches to solve problems in the areas of natural language processing and natural resources. In 2012 he was part of the team who developed the best performing system in the CoNLL 2012 shared task: Modeling Multilingual Unrestricted Coreference in OntoNotes.
Prior to joining IBM, he worked as an Assistant Professor for two years and a half at Fortaleza University (UNIFOR), Brazil. There he developed research and teaching activities in the computer science graduate and undergraduate programs.
During his PhD, he worked as a Graduate Research Assistant in the Laboratory of Algorithms Engineering and Neural Networks (LEARN) at PUC-Rio. In his PhD thesis, Cicero proposed the Entropy Guided Transformation Learning (ETL) algorithm, which is a learning strategy suitable for sequential classification. He also worked with collaborative filtering algorithms for recommendation systems; as well as machine learning algorithms for stock market forecasting.