One of the today's hot topics for organizations is Knowledge Management (KM). The field of KM came into being because of the need to manage (archive and retrieve) the vast amounts of information now stored on computers. Computers have changed the way we save and store information, and have pushed major components of the economy from the manufacturing sectors into the service/information sectors. Before computers, we stored information primarily on paper as written data. Access and reproduction of this material was limited, so few people within an organization could use it effectively. Access to a specific document required access to the individual who knew where it was stored. Also, because of its form, automated searches were impossible, so historical information about a specific transaction could only be obtained by contacting the individuals personally involved in the events.
The nature, use, and physical form of information are being transformed by widespread adoption of computers. In addition to written documents, we are also storing information as multimedia (video, audio, graphics, and photographs); databases (groups of related data tables), eLearning (instructional material with multimedia), presentations, spreadsheets, white papers, and project plans. The features of digital storage of all this information include easy high fidelity reproduction and ability to carry out automated searches. These result in benefits such as access to the data by an entire company and the ability to find information on specific subjects quickly.
However, as the we store more data, we find it harder to locate particular instances of information. With computers, the entire company can now access all the information, but without adequate management of the data, it remains useless/inaccessible. Many companies are loading endless numbers of memos and policy statements onto their websites. Unfortunately, when we want to find information on a specific topic, we can't because it is buried with mountains of other memos.
In the past, management of data was provided, in effect, by the people who created or stored the specific documents. We now want computers to manage the information. The first step is to place all the data in a repository (web site) and add a search engine. We can easily carry out automated searches based on keywords. For example we could look for documents produced by a particular individual. Usually, though, it is more interesting to find all the documents on a specific transaction. Most documents, unfortunately, do not self-reference the topic in a consistent and unique way such that we can effectively do the automated search. When we attempt to find the documents of interest, we end up without some documents we were looking for, while we get many unrelated ones. Lacking the human touch, automated search engines can do nothing to validate data or put it in context.
The alternate approach - having a librarian who catalogues all the information - provides an improvement on the mechanical search. However, scalability (ability to implement in an environment with rapid production of data) and affordability become problematic. This is where KM systems are being applied. The concept is that once a company deploys a knowledge management system, its employees will be able to find desired information through a combination of automated searches and drill-down/catalogued directory systems. Easy access to information without human intervention will provide an efficient tool and lower costs.
KM systems basically merge these two technologies to provide "intelligent" retrieval of data. These are the same technologies behind Internet search engines like Alta Vista (indexes) and Yahoo! (directories).
When companies set up a KM system, they need to create topic categories and sub-categories like Yahoo! has done. When employees want to publish information to the KM system, they need to fill out a form where they classify the data and add the "self-referencing" information that provides its context. The employee's function as librarians providing data so documents can be automatically searched and categorized. Note that for effective use, the documents must be in a searchable format such as a plain text or web-page (HTML) format. Binary formats (Word, Excel, Adobe PDF) are not searchable by most engines - they can only be reached from the directory category tree.
XML (eXtensible Markup Language) - another hot topic - is being applied to KM systems. XML can be used as a form of identifying documents. It is a technology that lets a content creator (author) imbed tags (hidden information) into a web (HTML) document. These tags identify characteristics of the document such as categories, subcategories, or a business transaction. Many KM systems use XML to identify documents.
Before your organization decides to implement a Knowledge Management system you will need to figure out its goal. Specifically what experience and utility do you want your knowledge management system to provide:
Will you store just operational procedure data (employee benefits, etc.) or will you create repositories for your corporate history (business transactions).
What kinds of data will you store? (Documents, audio, video, graphics, schematics, etc.)
Will you have a department acting as "librarian" or will each employee submitting data act as the "librarian"? Will you limit who is permitted to submit data (to avoid data overload)?
What categories should be set up to provide employees with the most logical access to information. You may want to visit sites like AltaVista and Yahoo! to get ideas since these are basically consumer versions of KM systems. Figure out what you like about each search and see how you can incorporate the best use practices into your Knowledge Management system.
Will you set up a standard searchable tag for each business transaction? For example, you can ask employees writing documents on a particular project to include the project number on all related data and correspondence. Note that this may require additional employee training.
