EU Project PARMENIDES
Ontology-driven temporal text mining
The EU-funded project PARMENIDES (IST-2001-39023), Ontology Driven Temporal Text Mining on Organisational Data for Extracting Temporal Valid Knowledge, developed text mining methods that extract temporal ontologies semi-automatically from very large text collections and keep them up to date.
As a research associate, I developed text mining methods for ontology enrichment, among them the topic-discovery approach RELFIN. Project partners included Unilever, Biovista, Wordmap, Neurosoft, Otto-von-Guericke University Magdeburg, the University of Zurich and the University of Manchester.
Twenty years later, ontology learning from text is again at the centre of my work, now with large language models: see ontology-based knowledge synthesis.