In this work, we present our approach to find relationships, both taxonomic and non-taxonomic, among the named entities extracted from texts of different genres. Instead of trying to assess the taxonomic relationships among named entities, we have adopted a context word-matching technique to assign separate scores to a different pair of entities, for different taxonomic relations. For non-taxonomic relations, we have mostly focused on verb-based relations and have also proposed a simple system to assign possible sentiment polarity labels among entity-pairs. This is an attempt to build a single unified system which can scan through texts of any genre and provide a fair idea of how the possible named entities are related among each other, assessed on three distinct space-intrinsic property, action, and sentiment.
CITATION STYLE
Maitra, P., & Das, D. (2020). Relation Extraction from Cross-Genre Unstructured Text. In Advances in Intelligent Systems and Computing (Vol. 937, pp. 433–446). Springer Verlag. https://doi.org/10.1007/978-981-13-7403-6_39
Mendeley helps you to discover research relevant for your work.