Schema discovery in Document-Oriented databases

In the software engineering community, NoSQL databases are gaining more and more popularity each year. One of their most appealing features is flexibility, which is characterised by the absence of an explicitly declared data schema. On one hand, this provides easier storage for high volumes of heterogeneous data. On the other, the lack of any schema definition brings additional difficulties to several software development and maintenance processes. Thus, in past years, there have been introduced various approaches for discovering and defining the implicit schema behind the data stored in NoSQL databases. However, none of them has become a recognized standard and many are still lacking in various ways. In the thesis, we present a systematic mapping study of currently available approaches for schema discovery in JSON-based document-oriented NoSQL databases. We map the current state-of-the-art in the field, cluster the separate features and capabilities of the existing approaches and finally explain the leading motives and potential gaps in current research. The work aims to provide a secondary study, which can be used as a comprehensive knowledge base for further research or just interest in the field.