We present a toolkit and dataset for entity-aspect linking. The tool takes as input a sentence and provides the most relevant aspect for each mentioned entity; it is implemented in Python and available as a script and via an online demo. It is accompanied by the first large dataset of entity-aspects, comprising more than 20,000 entities manually linked to the most relevant aspect, given a sentence as context. Each is expressed in structured manner as Open Information Extraction (OIE) triples (Subject, Relation, Object), having semantic information for polarity, modality, quantity and attributions.
Dieser Eintrag ist Teil der Universitätsbibliographie.
Das Dokument wird vom Publikationsserver der Universitätsbibliothek Mannheim bereitgestellt.
Creative Commons Attribution 4.0 International (CC BY 4.0)
Search Authors in
