The tasks of entity recognition through ontological commitment, fact extraction and organization with respect to a target schema, and entity deduplication have all been examined in recent years, and systems exist that can perform each individual task. A framework combining all these tasks, however, is still needed to accomplish the goal of automatically extracting and organizing biographical facts about persons found in historical documents into disambiguated entity records. We introduce FROntIER (Fact Recognizer for Ontologies with Inference and Entity Resolution) as the framework to recognize and extract facts using an ontology and organize facts of interest through inferring implicit facts using inference rules, a target ontology, and entity resolution. We give two case studies of FROntIER's performance over a few select pages from The Ely Ancestry [BEV02] and Index to The Register of Marriages and Baptisms in the Parish of Kilbarchan, 1649-1772 [Gra12].
College and Department
Physical and Mathematical Sciences; Computer Science
BYU ScholarsArchive Citation
Park, Joseph, "FROntIER: A Framework for Extracting and Organizing Biographical Facts in Historical Documents" (2015). Theses and Dissertations. 4368.
information extraction, inference, entity resolution