Knowledge-Based Labeling of Semantic Relationships in English

An increasing number of NLP tasks require semantic labels to be assigned, not only to entities that appear in textual elements, but to the relationships between those entities. Interest is growing in shallow semantic role labeling as well as in deep semantic distance metrics grounded in ontologies, as each of these contributes to better understanding and organization of text. In this work I apply knowledgebased techniques to identify and explore deep semantic relationships in several styles of English text: nominal compounds, full sentences in the domain of knowledge acquisition, and phrase-level labels for images in a collection. I also present work on a graphical tool for exploring the relationship between domain text and deep domain knowledge.

• Type: Conference
• Year: 2007
• Where: NAACL
• Authors: Alicia Tribble
• Keywords: Computational Linguistics, Deep Semantic Distance, Deep Semantic Relationships, NAACL 2007, Semantic