D-LeafIT Corpus is made up of 1819 Italian drug package leaflets crawled from the AIFA Database, among which 1439 refer to generic drugs and 380 to class A drugs. The generic drug leaflets amount to 6,154,007 tokens while the class A to 1,650,879 tokens, for a total amount of 7,804,886 tokens. Each entry in the corpus contains a unique numerical identifier, the drug name, the MA code, the drug class (generic or class A), the text of the whole leaflet, four paragraphs extracted (i. drug definition & intended use, ii. contraindications, iii. dosage & administration and iv. adverse drug reactions) the active ingredient and specifications concerning the packaging and format (e.g. number of tablets and unit of measurement). Furthermore, we also provide PoS tagging and relevant NER tagging annotations.
For further information and to cite this work refer to:
Giordano, L., & di Buono, M. P. (2024). Large Language Models as Drug Information Providers for Patients. In Proceedings of CL4HEALTH (PATIENT-ORIENTED LANGUAGE PROCESSING) Workshop @ LREC-COLING 2024.