Using Markov logic for relation learning in the never-ending learning

• Grant number: 12/21792-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: January 01, 2013
• End date: December 31, 2013
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Estevam Rafael Hruschka Júnior
• Grantee: Raphael Gianotti Serrano dos Santos
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

A recent and relevant research topic in Machine Learning is the Never-Ending Learning approach that focus on proposing algorithms and models to build learning systems that learn cumulatively forever, using what they have learned yesterday to improve their ability to learn better today, and keep learning indefinitely. The first Never-Ending Learning system described in the literature is called NELL (Never-Ending Language Learner). Markov Logics is an approach that combines first order logics and the ability to cope with uncertainty. In this project, the main goals are: I) investigate the use of Markov Logics to learn rules for the NELL system; II) investigate, propose and implement new approaches that allow the use of Markov Logic Networks algorithms to be applied to NELL's knowledge base; and III) study ways of using NELL's knowledge base to supervise the learning of new rules based on markov Logics. (AU)