Archival Filing, Retrieval, and Management

The basics of intellectual control of records are similar in concept to the levels of archival description with record groups, subgroups, series, and items.

Archivists and records managers need to understand the following to manage and provide access to the records:

• The functional activities of the organization and the business processes that create records
• The records created by those processes, who creates them, who uses them, and how the records relate to one another
• The characteristics of those records, such as retention requirements, vital records status, and restrictions on access
• The custodians and locations of the records 

Improving an organization’s ability to access its information is one of the key contributions information professionals can make to the engineering or re-engineering of business processes. For example, retrieval can be improved by drafting a taxonomy, a classification according to a predetermined system. However, enterprise-wide taxonomies can be challenging and expensive to implement in any case.

Taxonomy Development

Developing a taxonomy must be done in cooperation with the people who use it. Given the amount of work and difficulty in getting user acceptance, the approach must be well thought through before beginning and then reexamined periodically as the work progresses. The work can take years in a medium to large organization, and even in a small organization, a year would not be unusual.

Information professionals should be clear about what problems they are trying to solve with a taxonomy. To what extent does the organization need a common taxonomy across the enterprise? At what level is a common taxonomy appropriate? Here is where the typical archivist’s and records manager’s desires for uniformity and common access points meet each organization’s desire to do what they do and as easily as possible.

Taxonomies are seen as an antidote to the Google approach to information retrieval: index everything and retrieve with some learned capability to rank responses. Google provides information but not context, which is essential for records. Taxonomies provide context by grouping “like” things together. The question is, how does one determine like? For example, in an organization, the news releases may be filed together, the project documents for environmental remediation may be filed together, and remediation production projects may be filed together because each contains records about the same business function. If a user wants to review all news releases (as evidence of business transactions), they will be able to find them. But if they want everything about a remediation site in New Jersey, the taxonomy does not help as much. Most people are searching for information, not transaction records.

Taxonomies work well when looking for evidence or transactions or predicting the requests for information received. If neither of those is true, putting things into categories slows down the search. Information professionals may provide more help with better metadata and better training for users to search effectively.

Metadata Creation

Metadata is crucial to managing the lifecycle of a record. Often metadata schema looks at one aspect of a record or information asset. Dublin Core, for example, is focused on making information assets widely available over the web. Its metadata helps identify and describe digital objects but is less valuable when describing records and their relationships to one another.

The problem with metadata is that it is not automatic; it must be collected and entered. People use Dublin Core to describe things that they want to share. They want the document or resource to be found. Hence they may do the work to enter the metadata. Most records creators can find their documents and are not interested in putting in more work to help someone else find them on their own. Archivists must find a way to harvest the metadata for all documents that need to be managed. Records creators will not do it and believe that there needs to be an automated way to do it that requires little human intervention.

The Pay Off

There is a “pay me now or pay me later” aspect to metadata. Everyone understands that adding metadata makes the long-term management of records easier. It is upfront work for long-term gain. Most users opt for paying later–when they are searching–because they do not want to apply metadata now. Partly this is a question of time, but it is more a result of the inability to make adding metadata simple. Records managers, especially, have held on to a model of taxonomy that was implemented in paper by trained staff who did nothing but file. Today information professionals are asking everyone to know enough about records management filing procedures to file correctly. That will not happen. Even automated systems do much better with simple filing schemes.

Good Enough Metadata

Archivists and records managers need to determine good enough metadata for management and the easiest way to collect it. Of course, much of the ongoing management metadata can be collected automatically from the systems themselves. Much of it could be system generated. But that still leaves a lot missing, especially for unstructured data. In time, one hopes that collecting and preserving metadata will become easier so that the filing, retrieval, and management of archival materials will become more effortless.