This is a copy of a query put on the relevant esd tollkit forum (in case 
anyone recognises it). I am just opening the question out to a wider 
pool of expertise.

We are in the process of implementing and EDRMS [Wisdom by 
Diagonal (now Morse)] in a local government borough. 
At the moment we have a file plan based on the LGCS but I am a little 
concerned about how the file structure should be set up below this 
level. At present each individual service area which works on address-
based cases seems to be creating a separate folder for each address in 
the borough which seems a little ungainly to me. I have heard mention 
of possibly putting the address level of the structure above the service 
level, but I think this would interfere with the LGCS and would still 
represent a huge file structure. All the reading I have done on EDRMS 
etc indicates that the metadata is critical and that without it an EDRMS 
will become unusable.
I already notice a significant amount of irrelevant returns on searches 
from a system which contains relatively little data as yet, but this could 
be due to not having narrowed down the search string properly.
 
I have noted comments which concur with my own understanding that if 
you have good metadata then the file structure becomes largely 
irrelevant, but the problem with the creation of all the required 
metadata is that it probably even more onerous than creating large 
numbers of folders. 
I wonder if someone could answer a few questions for me: 

1. Is it necessary to have a standardised metadata scheme across all 
departments contributing to the EDRMS data and are there automated 
or semi automated ways of applying it to the documents? 
2. Would it be possible to extract the metadata automatically from a 
well structured file name protocol? 
3. What is the basic minimum metadata required for functions such as 
Housing, Environmental Health and Planning/ Building Control? 
4. Is the extent of such metadata laid down in standards? (ie: do eGov 
metadata standards extend to this level of detail?) 
5. If not has anyone developed a standardised set and what are they? 
6. Below LGCS, is a simple folder structure based on "Year" a 
reasonable solution or should it be more complex and if so how and 
why?
7. Does anyone please have an example of their own or of best practice 
relating to these questions?

Many thanks in advance, 
Martin

For any technical queries re JISC please email [log in to unmask]
For any content based queries, please email [log in to unmask]