Knowledge driven decision support system for provenance models in relational database

Data provenance is one of the most discussed subjects in the era of database analytics. Most of the database applications store a huge volume of useful data on several distributed servers, which can be acquired or reiterate for some decision making aspects such as what is the source of resultant data, why this data is generated, how this data is derived, who has created this data etc. This need leads to capture provenance information and querying. Thus an efficient provenance model has become the necessity for database analytics that can provide a number of characteristics such as debugging, backtracking, trust management, update management etc. Various provenance models for relational database analytics such as DB-Notes, Mondrian, Perm, Trio, Tramp, and Orchestra are existed in literature for capturing and querying the provenance information. Although each model has its own benefits and limitations, yet no one is efficient in all respect. Therefore, in this paper we have suggested a knowledge driven decision support system based on Analytic Hierarchy Process (AHP) to evaluate the performance of existing provenance models in relational database. For this purpose diverse decision parameters such as Audit Trail, Annotation, inversion, Schema Mapping, Overheads etc. are examined appropriately.