Details

Autor(en) / Beteiligte

• Bednár, Peter
• Paralič, Ján
• Babič, František
• Sarnovský, Martin
• Hartono, Pitoyo
• Mařík, Vladimír
• Sinčák, Peter

Titel

Knowledge-Based Approaches to Intelligent Data Analysis

Ist Teil von

• Towards Digital Intelligence Society, p.75-97

Ort / Verlag

Cham: Springer International Publishing

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• One of the most fundamental phenomena heavily influencing the digital society is Big Data. It is crucial not only to collect and analyze vast amounts of data but do it in an intelligent way. We believe that in order to do so, there needs to be a suitable interplay between the knowledge already known in the given application domain (background knowledge) and the knowledge inductively gained from data utilizing various data analysis techniques. We call it a knowledge-based approach to data analysis or intelligent data analysis. In this chapter, we will focus on two main types of the knowledge-based approach to data analysis. We start with the introduction of the semantic modelling of data analytics processes, which can efficiently cover an explicit form of background knowledge. The main focus here will be on the conceptualization of domain knowledge shared between the domain expert and data scientist and modelling of data mining workflows in order to achieve reproducibility and reusability. The second situation is typical for medical application, where the prevalent amount of background knowledge tends to stay tacit. In such a situation, the human-in-the-loop approach is a way how to perform data analysis intelligently. For both of these types of knowledge-based data analysis, specific case studies are presented to show how intelligent data analysis works in practice.