The Digital Workplace: Organisational Metadata and Recordkeeping 6 June 2007 Noni Oldfield.

Post on 17-Dec-2015

213 views 0 download

Tags:

transcript

The Digital Workplace:Organisational Metadata and Recordkeeping

6 June 2007

Noni Oldfield

Disclaimer

This is my Personal opinion only

• It is not the official view of Inland Revenue or other members of the Records and Document Management Team at Inland Revenue

• I am not an academic – I don’t always keep track of where the information that informs my knowledge comes from!

So you are newly arrived at …

You know what you want

Lots of information neatly organised, easily accessible, and busily being used

But a quick look around reveals …

The “records” are just piled up and most aren’t being used because they can’t be found.

You are sure that there is a lot of duplication too.

You dig deeper and unwittingly reveal …

This!

Metadata ALREADY exists in the organisation … lots of it!

• Organisational metadata is kept in silos – each different from the rest

• Most information is not considered to be “a record”.

• It is not standardised across the organisation, but some parts use external standards, such as NZGLS, MARC, and XBRL.

Standardising organisational metadata

The need to standardise metadata is known• Has been known by the Library and Records

professions for more than 50 years

• Has been known by IT departments since the 1970’s

• Has been known by Senior Management only recently and reluctantly

What have organisations done about standardising metadata?

• Data warehouses – 1970s• Metadata repositories and standards – 1980s-1990s• Metadata models and architectures – 1990s-2000s• XML languages built-in to databases and the web

– 2000s

Is it fixed? No its not.

What are our problems?

1. Organisations still can’t manage all their metadata

How do Records fit in to this?

Most organisations think of “records” as “unstructured information”– Shared drives– Personal drives– Email system– File cabinets– CD-ROMs– Plans– Etc

To the organisation

Recordkeeping metadata is just another silo

• It is applied to unstructured information only

• It is generic and only has a loose connection with the business of the organisation

• That business connection is via “functions” and “activities” (i.e. tied in with business processes – which are usually not very well documented either).

So what about unstructured information?

This is the biggest chunk of information without good metadata in most cases.

If you are lucky …

… this kind of information is already being captured into an electronic document and records management system (eDRMS).

– This almost certainly won’t use the same metadata definitions as the transactional information

– It may (or may not) conform to existing recordkeeping metadata standards

But probably …

… it is not. You are limited to the metadata that your software can add.– The operating software and the office software can both

add metadata automatically.– If it is not configured properly, then much of this

automatic metadata is meaningless.– Recordkeeping metadata standards assume far more

metadata can be easily added than is actually the case.

What about recordkeepingmetadata standards?

These are most useful for specifying the functionality for new eDRM systems

Existing systems are better evaluated against ISO 15489 to see if they meet the requirements

What are our problems?

1. Organisations still can’t manage all their metadata

2. Records are perceived to be just another silo

So what is recordkeeping?

• PRA 2005 definition covers too much.

• Recordkeeping should be about keeping records, not about keeping everything.- There is too much information now to keep everything

(just in case)- The recordkeeping profession is getting swamped by

waves of new kinds of information- Not all information should be treated in the same way

Recordkeeping is facing waves of information

• Email – the first wave• World wide web – the second wave• Dynamic, intelligent, data (data wrapped in its own

definitions – data from applications and data wrapped in XML language tags) – the current wave

• Non-written records (phone calls, video surveillance, etc) – the approaching wave

• Ubiquitous computing – the distant wave

Tsunami

Recordkeeping needs to go back to its’ roots

• Define classes of information that ARE records

• Use appraisal to find the information we need to keep and make it very easy to discard the rest

• Train our end-users to create summaries of their conversations (file notes, minutes, etc) and keep those instead of keeping the “raw data” – i.e. everything.

What are our problems?

• Organisations still can’t manage all their metadata

2. Records are perceived to be just another silo

3. Recordkeeping is being drowned in too much information

Of course, for every problem there is an opportunity …

Opportunity 1 – The metadata solution

Let Records take over all the metadata management for the organisation – business metadata included.

• A business owner of the big picture is badly needed (and no-one else really wants it).

• Centralisation/standardisation of metadata is useful for the whole organisation and enables better records to be kept

• Needs to be done in conjunction with other metadata custodians in the organisation.

Opportunity 2 –Metadata standards

Let us create recordkeeping metadata standards at different levels

1. A simple standard, like Dublin Core, for people who don’t have an eDRMS

2. An XML-based language that allows interchange of standard recordkeeping metadata anywhere in the world

3. Get everyone using the same standard for eDRMS – possibly based on ISO 23081.

Opportunity 3 – Back to the future

• Let us define the “records” we want to keep, instead of defining those that we can discard.

• Let us really push to create summaries (records) of information instead of keeping all the data.

If we can’t do that, then we must define an automatic (or at least really easy) process to allow people to discard/delete information such as emails.

Opportunity 4 – the ECM solution

• If we are going to treat all information as records then we must manage all information.

• This is too important to be left to anyone else.– This is a trend in the USA, driven by e-Discovery– It could happen here too– Forget “records”, treat everything as content and

manage the whole lifecycle

• We will still need an easy way to delete unwanted information!

The Good News …

The only way to go is up!

And metadata may be the way to go