From unstructured data included in real-estate listings to information systems over ontological graphs
Krzysztof Pancerz, Piotr Grochowalski · 2017 International Conference on Information and Digital Technologies (IDT) · 2017
In the paper, we consider the problem of automatic transformation of unstructured data included in real-estate listings into data arranged in a tabular form, as the so-called information tables. Transformation is an important preprocessing stage enabling us to obtain data in a form accepted by many data mining and machine learning tools. In the presented approach, information tables represent information systems over ontological graphs. In such systems, the domain knowledge covering some data semantics is directly incorporated. In the paper, we describe important aspects of a specialized computer tool created by us to automatize the transformation process.