Knowledge extraction in the Environment and Health domains is certainly an important asset for both scientific research and decision support. However, these strategic domains are characterized by a significant heterogeneity of structured and unstructured documents that do not allow a complete transfer of knowledge. Not infrequently, this critical point emerges during the implementation of research projects as an obstacle to the capitalization of information useful for the management of territories or in more recent times to the fight against the pandemic. This paper aims to achieve an analysis of the different forms of knowledge that characterize the scientific production in these specific fields trying to take a holistic approach to text management, tables and graphs through a multidisciplinary logic with the aim of making the knowledge accessible through a geolocalized representation of sites at risk. A case study on a specific corpus of documents is provided.
Analysis, evaluation and comparison of knowledge extraction tools in the Environmental and Health domain. A holistic approach.
Anna Rovella;Eugenio Cesario;Martin Critelli;Armando Bartucci;Francesca M. C. Messiniti
2022-01-01
Abstract
Knowledge extraction in the Environment and Health domains is certainly an important asset for both scientific research and decision support. However, these strategic domains are characterized by a significant heterogeneity of structured and unstructured documents that do not allow a complete transfer of knowledge. Not infrequently, this critical point emerges during the implementation of research projects as an obstacle to the capitalization of information useful for the management of territories or in more recent times to the fight against the pandemic. This paper aims to achieve an analysis of the different forms of knowledge that characterize the scientific production in these specific fields trying to take a holistic approach to text management, tables and graphs through a multidisciplinary logic with the aim of making the knowledge accessible through a geolocalized representation of sites at risk. A case study on a specific corpus of documents is provided.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.