Channels: General Business · Business Technology · E-Business · Knowledge Management Resources: News · World · Information! · Conferences · Careers · e-Zine · Analyses · Books · Community: Join the Network! · Online Community · Sponsored Events · Executive Jobs Congratulations!! Winners of $390,000 in 'Giveaway for Success' Click to Enter!!	SEARCH [HELP]

PROCESS TO PRODUCT:
CREATING TOOLS FOR KNOWLEDGE MANAGEMENT

5. ARCHITECTURAL APPROACHES

A variety of architectural approaches are evident when studying todayís software technologies that support knowledge management. The following section describes and diagrams some of those approaches. It not an exhaustive depiction but representative of some of the general approaches that are most prevalent today.

5.1 MCC InfoSleuth

The first example represents an architecture formulated from an ontological approach to organizing knowledge. This type of approach is generally more prevalent in academic research environments and has spawned international standards such as the Knowledge Information Format (KIF), Knowledge Query and Manipulation Language (KQML), and the Interlingua Working Group (IWG). With a basis in artificial intelligence, this method of approach is centered upon understanding and duplicating some elements of human cognition.

The example shown in below (Figure 2) is from the Microelectronic Computer Corporation (MCC), a national research center created in the late 1980ís. The architecture includes modern technology, in the form of JAVA applets, integrated with the older programmatic interface of KQML. It also includes multiple, rather than a single type, of intelligent agent and, like most products, provides linkages to databases.

 

Figure 2

5.2 IBM Agent Builder

Displaying an approach similar to the one of MCC, IBM has designed a broad architecture (Figure 3) for developing agents but is focused primarily on those connections needed to integrate legacy systems to new architectures. They have therefore spent considerable effort on developing libraries for format translation as well as adapter technology to provide connections between those libraries and various object technologies.

Figure 3

5.3 Autonomy AgentWare

A more recent approach, one following by several companies, is the development of architectures similar to that shown by Autonomyís AgentWare shown in Figure 4. These are generally more focused on capturing information from the Internet based on user preference but lack may be less devoted to libraries or ontologies. In the case of Autonomy, there is much greater focus on attempting to capture tacit knowledge and convert it to explicit form. This is one area that many newer knowledge management architectures are trying to include.

Figure 4

 

There are a variety of architectural perspectives that have not been explored in this paper. The interconnection of additional devices such as hand held computers, packet networks, televisions, and telephones is not included in the architectures shown previously. There does appear, however, to be a general attempt to converge distributed local and wide area computer networks and the Internet into seamless information systems. At the same time, there are many differing approaches to the gathering, storing, and displaying of information.