Document Retrieval

DOE-STD-3024-98

DOCUMENT RETRIEVAL

Efforts to retrieve design information have sometimes been limited to search only for specific types of

information under the perception that such a limitation will lead to cost-effective results. Although general

decisions regarding how far to look for information can be made and are valid, experience has shown that a

priori limiting the search to certain types of information may not be effective nor efficient.

Reasonable general limitations can be placed on the search for design information. A recommended approach

is to try to identify the most promising locations and search only those places. This is the concept of a "smart

search." With a very small effort, it is usually possible to identify those locations that are most likely to have

the desired information. A few telephone calls could be made to long-term employees to find out where the

most fruitful places to look are. For example, someone may remember a packrat engineer who never throws

away anything and his file cabinets are rich; someone in program management who has the files on old

projects; or someone in document control center. When these few most productive locations have been

searched, the "smart search" is complete.

An important management consideration is to know when to stop. How will we know when we have searched

long enough? To facilitate answering this question, it is necessary to understand the information that is

needed and the types of documents that are likely to contain the information. The information required is

indicated by "shall" statements in the body of this standard. Engineers who have had design experience with

large engineering firms can usually tell us what types of documents are typically produced for different types

of systems and components. The search can be stopped upon completion of reasonable efforts to retrieve the

types of documents that would be expected to contain the information.

REVIEWING RETRIEVED DOCUMENTS

The system engineer (cognizant facility engineer) can easily and quickly ascertain which of the retrieved

documents contain information that is applicable to the current configuration of the facility. This can

sometimes be accomplished on the basis of the date of the document and a knowledge of previous system

modifications.

A question arises about what would we do with the documents that are not related to the current SDD task.

Some suggest merely casting those documents aside or sending them to Records Management for storage. A

better approach might be to make a list of the documents found, to avoid having to re-scan this location

during future tasks. For example, when the next SDD is being developed, we might need some more

information from that location. An even better approach is to conduct the search/retrieval process for all the

SDD systems as one consolidated task and to keep an inventory of other documents found for broader future

uses.

Having retrieved some documents related to the system at hand, how might we sort that information? One

approach is to sort the documents by system, then by type of document (reports and studies, analysis,

calculations, drawings, specifications, procurement documents), and finally by document date (with the most