Every night, data tables are copied from the server to back-up tables on a separate computer, and these back-up tables are copied to a third back-up file just before being updated. This way, RIMS has two days to discover and correct any data corruption and/or restore the data files.
Every night, new employees and new (postgrad) students are filtered from HRMIS and Student One, screened for doubles and appended to the RIMS Master lists. At the same time, employees’ surname, title, work unit and employment duration are updated.
To enable evaluation of institutional collaborative links, non-JCU co-authors and associate researchers on grant and ethics applications are identified in a separate RIMS table and these individual collaborators are linked to one or more institutions.
Like Research Codes, keywords are collected with each publication and grant application/contract. Weekly all keywords entered are cleansed, standardised and frequently used keywords hierarchically coded.
As database keys are not unique in core central systems, RIMS is required to store and maintain local tables with ‘clean’ filtered data and to create and maintain cross-over tables to track changes to foreign keys over time.
Examples: Some current staff members and even more students have more than one ID. Not all associated staff are added to HR; foreign guest students are not added to Student One.
A lot of time has been spent ‘cleaning-up’ historical data to be able to append these data to the RIMS structure in order to provide historical overviews. RIMS stores unique keys (or ID’s) for employees, students, projects, institutions, journals, etc., uses hierarchically coded lists of funding bodies and their schemes, countries, animals, keywords, research codes, and maintains lists of synonyms for each master list for faster searches and data entry.