Document Management: Retrieval Issues

Document Retrieval
Documents are available to be retrieved when needed. If they're never going to be retrieved, it would be a meaningless exercise to keep them, wasting space, man hours, and money.

You retrieve documents for personal, business, or legal purposes:

* You might want to re-read an old story or letter, which you've stored in an organized or random manner along with many others. That would be a personal purpose.
* You might want to study the sales trends for a particular line of products you sell. You would then want to retrieve all the sales invoices for the relevant period and summarize the sales patterns of products in that line. This is a typical example of document retrieval with a business purpose.
* An employee is suing you for wrongful dismissal. You would then want to retrieve all documents relating to that employee's employment contract and evidence of breach of contract. The primary purpose here is legal defense.

Document retrieval becomes an issue when the volume of documents is large. In the case of business documents, huge volumes are typical.

Locating a particular document and retrieving it from among the mass of documents then becomes a document management issue. This article seeks to examine how this issue is handled.

Elements of a Document Retrieval System
The three key elements of a document retrieval system are:

A Document Store:
Documents must be stored safely so that they can be retrieved whenever needed. These days, documents are typically stored in computer databases.

A Classification System:
To facilitate later retrieval, the mass of documents must be organized in a meaningful manner that is easy to work with and understand. There would then be an index or catalog that can be referenced for locating the particular document to be retrieved.

Retrieval Request Communication:
The person who needs a document must specify the request in a manner that facilitates retrieval. In the case of computer databases, you can even specify a few relevant words and then ask the system to locate documents containing those words. An interface must be provided to input this request.

An internet search engine is a good example of a document retrieval system. The documents in this case often involve not only traditional Web pages but also items such as books, maps, and images.

What's Involved in Document Retrieval?
Retrieving a document typically involves:

Making Your Request:
You might ask your secretary to find that letter you sent to Mr. Somebody last month. Or you might enter the unique words that letter might contain and ask your e-mail program (with a search function) to find that letter.

Matching Your Query With Documents:
The secretary might ask you whether it was the letter sent in the first week or third week that you want. The e-mail program might bring up a list of mails containing the words you specified. What's involved here is an attempt to match your query with the document. You might have to refine the query by specifying additional criteria before you get exactly what you want (if it exists).

Ranking the Documents in Order of Relevance:
Unlike your secretary, databases can not ask for further clarifications to identify exactly which document you want. Instead, they list all the documents in their index that match your query in some way. It would then become necessary to rank these documents in an order of relevance. Search engines have algorithms to assess the relevance of documents to particular queries and then display them in that order.

Until sometime ago, search applications were normal only on the Internet. These days, such applications are available even for standalone desktops, not to mention your business intranet.

With the help of these applications, you can find what you want, without bothering your secretary. The speed of getting the document you want has become incomparably faster.

Structured Querying
Structured querying can help the process of document retrieval to be both faster and more meaningful.

An e-mail program with search facilities is a good example of structured querying. You can specify name of sender or recipient and date range, in addition to or in place of specific words in the content of the e-mail. In the case of general searches also, you can narrow down your search by refining your search.

Spending a little time on how to refine your search can result in much higher search productivity.

Your Document Management System
All good document management systems have facilities to enable efficient retrieval of the documents stored. Time spent on learning how to use the retrieval tools is a better use of time.

Ultimately, your document management software system should save you time and energy by giving you quick and easy access to your documents.