Development of semantic feature engineering for statistical analysis on parse ranking

• Main Author: Yin, Xiaocheng.
• Other Authors: Koe Choon Chiaw, Lawrence
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Computer science and engineering::Information systems::Information storage and retrieval; DRNTU::Humanities::Linguistics::Semantics
• Online Access: http://hdl.handle.net/10356/52309
• Institution: Nanyang Technological University

In this report, the use of semantic information in parse selection is investigated. It is shown that increasing sense-based semantic features based on deep linguistic processing directly helps improving the effectiveness of parse selection. A Python software model was implemented to carry out feature engineering on semantic parsing results by parsers and the data was from SemCor corpus. Different types of semantic features are generated using the model and training and testing was conducted using a maximum entropy model TADM. Also, baseline features are generalized upwards in the WordNet hierarchy to help investigate the effectiveness of disambiguation in parse selection. Generalized features provide better parse selection accuracy than more specific features.