Automatic Meta Data Generation for High Performance Database System

Maintaining the consistency of meta data in large, high performance database systems is a challenging task, often requiring manual correction and insertion of meta data. Identifying changes in the meta data manually is however a costly and error-prone process.

In order to make this process more efficient, regular meta data records retrieved from international organizations need to mined for changes, comparing them against the records in the database. Such changes must then be propagated to the database in a traceable and efficient manner. Issues include the problem of minimizing the load of routine maintenance on a high performance system, as well as how to deal with differences in the meta data.

The task of this exjobb is to develop an automatic meta data update system, that takes as input data files from external systems, extracts the relevant meta data, compares it with the data in the operational database, and - if there are changes - automatically produces updates for the meta data in the system.