We know that Automated Decision-Making (ADM) is currently changing industry, thus people and countries started to be concerned about the impact that may have on everyone lives. The GDPR stresses the importance of a Right to Explanation (e.g., art. 22, artt. 13-14-15, recital 71), requiring the AI industry to adapt consequently, thus giving rise to eXplainable AI (XAI). Modern XAI proposes some solutions to make ADM more transparent following the principle included in the GDPR (art. 5), but many researchers criticize XAI to provide little justification for choosing different explanation types or representations. In this paper we propose a new model of an explanatory process based on the idea of explanatory narratives, claiming that it is powerful enough to allow many possible types of explanations including causal, contrastive, justificatory and other types of non-causal explanations.
Francesco Sovrano, F.V. (2019). The difference between Explainable and Explaining: requirements and challenges under the GDPR. Aachen : CEUR-WS.org.
The difference between Explainable and Explaining: requirements and challenges under the GDPR
Francesco Sovrano
;Fabio Vitali;Monica Palmirani
2019
Abstract
We know that Automated Decision-Making (ADM) is currently changing industry, thus people and countries started to be concerned about the impact that may have on everyone lives. The GDPR stresses the importance of a Right to Explanation (e.g., art. 22, artt. 13-14-15, recital 71), requiring the AI industry to adapt consequently, thus giving rise to eXplainable AI (XAI). Modern XAI proposes some solutions to make ADM more transparent following the principle included in the GDPR (art. 5), but many researchers criticize XAI to provide little justification for choosing different explanation types or representations. In this paper we propose a new model of an explanatory process based on the idea of explanatory narratives, claiming that it is powerful enough to allow many possible types of explanations including causal, contrastive, justificatory and other types of non-causal explanations.File | Dimensione | Formato | |
---|---|---|---|
xaila2019-paper1.pdf
accesso aperto
Tipo:
Versione (PDF) editoriale
Licenza:
Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY)
Dimensione
1.48 MB
Formato
Adobe PDF
|
1.48 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.