Showing 1–2 of 2 results for author: Gusmão, A C

Interpreting Embedding Models of Knowledge Bases: A Pedagogical Approach

Authors: Arthur Colombini Gusmão, Alvaro Henrique Chaim Correia, Glauber De Bona, Fabio Gagliardi Cozman

Abstract: Knowledge bases are employed in a variety of applications from natural language processing to semantic web search; alas, in practice their usefulness is hurt by their incompleteness. Embedding models attain state-of-the-art accuracy in knowledge base completion, but their predictions are notoriously hard to interpret. In this paper, we adapt "pedagogical approaches" (from the literature on neural… ▽ More Knowledge bases are employed in a variety of applications from natural language processing to semantic web search; alas, in practice their usefulness is hurt by their incompleteness. Embedding models attain state-of-the-art accuracy in knowledge base completion, but their predictions are notoriously hard to interpret. In this paper, we adapt "pedagogical approaches" (from the literature on neural networks) so as to interpret embedding models by extracting weighted Horn rules from them. We show how pedagogical approaches have to be adapted to take upon the large-scale relational aspects of knowledge bases and show experimentally their strengths and weaknesses. △ Less

Submitted 20 June, 2018; originally announced June 2018.

Comments: presented at 2018 ICML Workshop on Human Interpretability in Machine Learning (WHI 2018), Stockholm, Sweden

Speeding-up ProbLog's Parameter Learning

Authors: Francisco H. O. V. de Faria, Arthur C. Gusmão, Fabio G. Cozman, Denis D. Mauá

Abstract: ProbLog is a state-of-art combination of logic programming and probabilities; in particular ProbLog offers parameter learning through a variant of the EM algorithm. However, the resulting learning algorithm is rather slow, even when the data are complete. In this short paper we offer some insights that lead to orders of magnitude improvements in ProbLog's parameter learning speed with complete dat… ▽ More ProbLog is a state-of-art combination of logic programming and probabilities; in particular ProbLog offers parameter learning through a variant of the EM algorithm. However, the resulting learning algorithm is rather slow, even when the data are complete. In this short paper we offer some insights that lead to orders of magnitude improvements in ProbLog's parameter learning speed with complete data. △ Less

Submitted 1 August, 2017; v1 submitted 25 July, 2017; originally announced July 2017.

Comments: StarAI - International Workshop on Statistical Relational AI

MSC Class: 97R40