Safe and preserve personalization in module-based data management

In many application domains (e.g., medicine or biology), comprehensive schemas resulting from collaborative initiatives are made available. For instance, SNOMED is an ontological schema containing more than 400.000 concept names covering various areas such as anatomy, diseases, medication, and even geographic locations. Such well-established schemas are often associated with reliable data that have been carefully collected, cleansed, and verified, thus providing reference ontology-based data management systems (DMSs) in different application domains. The efforts made to design reference DMSs whenever we have to develop our own DMS with specific needs. A way to do this is to extract from the reference DMS the piece of schema relevant to our application needs, possibly to personalize it with extra-constraints w.r.t. our application under construction, and then to manage our own dataset using the resulting schema. Recent work in description logics (DLs) provides different solutions to achieve such a reuse of a reference ontology-based DMS. Indeed, modern ontological languages like the W3C recommendations RDFS, OWL, and OWL2 – are actually XML-based syntactic variants of well-known DLs. Existing definitions of modules in the literature basically resort to the notion of (deductive) conservative extension of a schema or of uniform interpolant of a schema, a.k.a. forgetting about non-interesting relations of a schema. A reference ontology-based DMS in order to build a new DMS with specific needs. One step further by not only considering the design of a module-based DMS. A module-based DMS can benefit from the reference DMS to enhance its own data management skills.