A NLG framework for user tailoring and profiling in healthcare

Autor: Simone Balloccu, Steffen Pauws, Ehud Reiter
Rok vydání: 2020
DOI: 10.5281/zenodo.4777689
Popis: Communication in healthcare can improve therapy adherence and patient engagement. Research into healthcare-oriented Natural Language Generation (NLG) systems suggests that tailoring to user profile can improve overall effectiveness. However lots of systems adopts a single or small group of user profiles, thus overlooking that user subsets may have different needs and that these could evolve over time. In this paper we conceptualize a framework that can produce Customised healthcare reports by extracting meaningful data insights and producing a final text which varies in content and terminology according to the user profile and traits. The dietary domain will be used to show a working example.
Databáze: OpenAIRE