SANTO: A Web-based Annotation Tool for Ontology-driven Slot Filling

摘要: Supervised machine learning algorithms require training data whose generation for complex relation extraction tasks tends to be difficult. Being optimized at sentence level, many annotation tools lack in facilitating the of relational structures that are widely spread across text. This leads non-intuitive and cumbersome visualizations, making process unnecessarily time-consuming. We propose SANTO, an easy-to-use, domain-adaptive tool specialized slot filling which may involve problems cardinality referential grounding. The web-based architecture enables fast clearly structured multiple users parallel. Relational formulated as templates following conceptualization underlying ontology. Further, import export procedures standard formats enable interoperability with external sources tools.