"Plenary talk" alla conferenza internazionale IPMU 2018 - Information Processing and Management of Uncertainty ABSTRACT: Geo big data are heterogeneous by nature traditionally comprising both georeferenced images acquired by remote sensing and their derived products, and cartographic maps published as open data by public and private organizations. Furthermore, thanks to the Web 2.0 revolution and wide spread diffusion of IoT and smart devices equipped with GNSS sensors, the availability of new and real-time sources of geo big data is rapidly increasing. Let us think at Volunteered Geographic Information (VGI) created by citizens eager to participate in citizen science initiatives, at crowdsourced geotagged posts, created by users of social networks, and at the great variety of low-cost sensor data. This heterogeneous multisource geo data constitutes a challenge for Earth Observation, i.e., for describing the planet Earth's physical, chemical, biological and anthropic systems to monitor and assess the status of and changes in the natural, built and social environment, although to convert data into value, we need to face some open issues related with the effective management of geo big data. Specifically, we need new methods for the representation and discovery of the relevant geo data among huge repositories, the assessment of the questionable geo data quality, and, finally, the cross-analysis and synthesis of geo data to provide decision makers with consistent and comprehensible information. All such tasks involve the management of the imprecision and the uncertainty of both geo data and user needs. In this context fuzzy approaches can provide opportunities to face actual chllanges such as interoperability, quality assurance and assessment and multisource data synthesis.
Geo Big Data for Earth Observation: Challanges and Opportunities of Fuzzy Approaches
Gloria Bordogna
2018
Abstract
"Plenary talk" alla conferenza internazionale IPMU 2018 - Information Processing and Management of Uncertainty ABSTRACT: Geo big data are heterogeneous by nature traditionally comprising both georeferenced images acquired by remote sensing and their derived products, and cartographic maps published as open data by public and private organizations. Furthermore, thanks to the Web 2.0 revolution and wide spread diffusion of IoT and smart devices equipped with GNSS sensors, the availability of new and real-time sources of geo big data is rapidly increasing. Let us think at Volunteered Geographic Information (VGI) created by citizens eager to participate in citizen science initiatives, at crowdsourced geotagged posts, created by users of social networks, and at the great variety of low-cost sensor data. This heterogeneous multisource geo data constitutes a challenge for Earth Observation, i.e., for describing the planet Earth's physical, chemical, biological and anthropic systems to monitor and assess the status of and changes in the natural, built and social environment, although to convert data into value, we need to face some open issues related with the effective management of geo big data. Specifically, we need new methods for the representation and discovery of the relevant geo data among huge repositories, the assessment of the questionable geo data quality, and, finally, the cross-analysis and synthesis of geo data to provide decision makers with consistent and comprehensible information. All such tasks involve the management of the imprecision and the uncertainty of both geo data and user needs. In this context fuzzy approaches can provide opportunities to face actual chllanges such as interoperability, quality assurance and assessment and multisource data synthesis.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.