Explainable Artificial Intelligence in Data Science

Abstract A widespread need to explain the behavior and outcomes of AI-based systems has emerged, due their ubiquitous presence. Thus, providing renewed momentum relatively new research area eXplainable AI (XAI). Nowadays, importance XAI lies in fact that increasing control transference this kind system for decision making -or, at least, its use assisting executive stakeholders- already affects many sensitive realms (as Politics, Social Sciences, or Law). The decision-making power handover opaque makes mandatory explaining those, primarily application scenarios where stakeholders are unaware both high technology applied basic principles governing technological solutions. issue should not be reduced a merely technical problem; explainer would compelled transmit richer knowledge about (including role within informational ecosystem he/she works). To achieve such an aim, could exploit, if necessary, practices from other scientific humanistic areas. first aim paper is emphasize justify multidisciplinary approach beneficiated part philosophical corpus on Explaining, underscoring particular nuances field Data Science. second objective develop some arguments justifying authors’ bet by more relevant ideas inspired by, one hand, formal techniques Knowledge Representation Reasoning, modeling human reasoning when facing explanation. This way, seek sound balance between pure justification explainer-explainee agreement.