Managing data related to natural sciences poses new and challenging problems as it is impossible to represent reality on a one-to-one scale, and imprecision has to be taken into account, both in data memorization and in its processing. Machine learning has been a key enabler in the context of information extraction from natural sciences data. However, data-driven results are strongly affected by the volume, the sparsity and different types of imprecision in the available sources. Therefore, it becomes pivotal to associate both to data and to data-driven services information about their quality, in order to effectively interpret the results. Different levels of granularity and multiple data modalities captured from the same processes could coexist, due to technological constraints or other intrinsic limiting factors. In addition, different levels of granularity might be also the result of application requirements, and outcomes at multiple levels of precision needs to be provided. Affinities of quality issues in domains such as chemistry, biology, and geoinformatics are discussed in the paper.

About the Quality of Data and Services in Natural Sciences / Pernici, B.; Ratti, F.; Scalia, G.. - 12521:(2021), pp. 236-248. [10.1007/978-3-030-73203-5_18]

About the Quality of Data and Services in Natural Sciences

Ratti F.;
2021-01-01

Abstract

Managing data related to natural sciences poses new and challenging problems as it is impossible to represent reality on a one-to-one scale, and imprecision has to be taken into account, both in data memorization and in its processing. Machine learning has been a key enabler in the context of information extraction from natural sciences data. However, data-driven results are strongly affected by the volume, the sparsity and different types of imprecision in the available sources. Therefore, it becomes pivotal to associate both to data and to data-driven services information about their quality, in order to effectively interpret the results. Different levels of granularity and multiple data modalities captured from the same processes could coexist, due to technological constraints or other intrinsic limiting factors. In addition, different levels of granularity might be also the result of application requirements, and outcomes at multiple levels of precision needs to be provided. Affinities of quality issues in domains such as chemistry, biology, and geoinformatics are discussed in the paper.
2021
Inglese
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Springer Science and Business Media Deutschland GmbH
12521
236
248
13
Esperti anonimi
Internazionale
Goal 3: Good health and well-being
Data quality
Machine learning
Quality of service
No
2 Contributo in Volume::2.1 Contributo in volume (Capitolo o Saggio)
3
268
About the Quality of Data and Services in Natural Sciences / Pernici, B.; Ratti, F.; Scalia, G.. - 12521:(2021), pp. 236-248. [10.1007/978-3-030-73203-5_18]
none
Pernici, B.; Ratti, F.; Scalia, G.
info:eu-repo/semantics/bookPart
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11768/199436
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? ND
social impact