Publication
Deal with it! Towards Self-Organizing Data Schemas from Semi-Structured Inserts
Benjamin Hättasch; Leon Krüger; Carsten Binnig
In: Proceedings of the Ninth International Workshop on Exploiting Artificial Intelligence Techniques for Data Management, aiDM 2026, Bengaluru, India, 31 May 2026 - 5 June 2026. International Workshop on Exploiting Artificial Intelligence Techniques for Data Management (aiDM), Pages 31-40, ACM, 2026.
Abstract
In scenarios where the structure of information to store is not
known upfront or changes over time, neither traditional relational
databases nor schemaless approaches are well-suited. Relational
databases are great for data analysis and exploration, but require
a carefully crafted schema, which causes high manual overhead.
Entities not considered during schema design cannot be stored. In
contrast, schemaless approaches allow users to store all kinds of
data without the need for a schema, but require schema-checking
on read to ensure that queries can read certain attributes. Therefore,
we present an approach that works on top of existing relational
databases and autonomously organizes the schema based on incoming
data. We call our system JUSTINE (JUST INsert Engine).
Users can store data while omitting table or column names or using
the wording of their mental schema, which might deviate from the
real one in the database. Our system maps the incoming data to
the existing schema and adjusts the schema where needed (i.e., by
adding columns or new tables) to handle the insertion of attributes
or entities not represented in the schema yet. To ensure a highquality
schema, our system provides operations to periodically fix
mistakes and optimize the schema (e.g., by combining tables). We
propose a way to evaluate such systems, even though there is no
single gold standard for a suitable schema, and use this for an initial
evaluation of our approach. Our approach should be seen as a first
step towards self-organizing data schemas, therefore, we provide
an extensive roadmap for further research.