15th Annual Conference of the International Speech Communication Association

September 14-18, 2014

Fusion of Knowledge-Based and Data-Driven Approaches to Grammar Induction

Spiros Georgiladakis (1), Christina Unger (2), Elias Iosif (1), Sebastian Walter (2), Philipp Cimiano (2), Euripides Petrakis (1), Alexandros Potamianos (3)

(1) Technical University of Crete, Greece
(2) Universität Bielefeld, Germany
(3) NTUA, Greece

Using different sources of information for grammar induction results in grammars that vary in coverage and precision. Fusing such grammars with a strategy that exploits their strengths while minimizing their weaknesses is expected to produce grammars with superior performance. We focus on the fusion of grammars produced using a knowledge-based approach using lexicalized ontologies and a data-driven approach using semantic similarity clustering. We propose various algorithms for finding the mapping between the (non-terminal) rules generated by each grammar induction algorithm, followed by rule fusion. Three fusion approaches are investigated: early, mid and late fusion. Results show that late fusion provides the best relative F-measure performance improvement by 20%.

Full Paper

Bibliographic reference.  Georgiladakis, Spiros / Unger, Christina / Iosif, Elias / Walter, Sebastian / Cimiano, Philipp / Petrakis, Euripides / Potamianos, Alexandros (2014): "Fusion of knowledge-based and data-driven approaches to grammar induction", In INTERSPEECH-2014, 288-292.