Personally, I don't see the data quality or formatting issues that Lianne does and have been transferring data from other systems into RE since 1994. The bigger problem for me used to be that, because our organisation creates RE records as students (and academic staff) arrive, we have different departments changing the student record and RE record independently. For the past 3 years, however, I have frequent snapshots of both the student and RE records in a Data Warehouse in which I can run difference checks. Those checks not only report the differences for manual correction, but also export RE import files for batch-wise correction. I believe that some organisations have had RE plug-ins written to allow direct updating of the RE data, but I haven't found that necessary.
We don't use SITS, and I have no direct contact with anyone who does, although I am aware of its market position. My approach isn't vendor-specific, although one has to understand both the SR and RE data structures, but one has to do that to utilise the data in a Warehouse anyway.
I've just started in my post at the University of Leicester, where we use SITS. The problem I can see, at the moment, is that SITS doesn't appear to have the ability to record City, Postcode etc - it appears to record the addresses as Address line 1, Address line 2, Address line3 etc, with City or Postcode in any one of those fiels (well, apart from Address line1 hopefully).
So it isn't easy to automatically extract the City/postcode/country etc without manual intervention (plus the address line 1, 2 etc need to be concatenated with "\n" to allow import into RE). So the problem (for us at least) doesn't seem to be data entry quality, rather the ability to distinguish between fields adequetely...
I'm thinking of producing an Excel spreadsheet to try to (semi) automatically rearrange the fields, but I think this will involve a "learning" process for the spreadsheet to learn to distinguish the Cities, Postcodes etc.
Any other good solutions gladly accepted!
Thanks
Jon
Data Management Officer, Development and Alumni Relations Office
University of Leicester
Hi Jon,
at Loughborough I have a similair issue to use - we have a bespoke SRS and whilst most of the transfer is straightforward I do have issues in the structure with addresses... in the SRS they have line1, line2, line 3 etc with no separate field for country or city. At present I cleanse the data using Experian to get the UK addresses into a standard format, and then feed the data into an access database where I've got a series of pre-wrtten queries to play "hunt the country" by going backwards through the data to identify the last line of the address and formatting it on the basis that's the country. It still needs a bit of manual manipulation at the end but it does resolve most of the addresses.
Richard