An Approach to Extracting Topic-guided Views from the Sources of a Data Lake

In the last years, data lakes are emerging as an effective and an efficient support for information and knowledge extraction from a huge amount of highly heterogeneous and quickly changing data sources. Data lake management requires the definition of new techniques, very different from the ones adopted for data warehouses in the past. In this scenario, one of the most challenging issues to address consists in the extraction of topic-guided (i.e., thematic) views from the (very heterogeneous and often unstructured) sources of a data lake. In this paper, we propose a new network-based model to uniformly represent structured, semi-structured and unstructured sources of a data lake. Then, we present a new approach to, at least partially, structuring unstructured data. Finally, we define a technique to extract topic-guided views from the sources of a data lake, based on similarity and other semantic relationships among source metadata.

Automated summary

Data lakes have emerged as an effective support for extracting information and knowledge from highly heterogeneous and rapidly changing data sources. However, managing data lakes requires new techniques. This paper introduces a network-based model to represent structured, semi-structured, and unstructured sources of a data lake, and proposes a new approach for extracting topic-guided views.