您在這裡

Case Study

8 九月, 2015 - 10:56

The first step in the process of knowledge harvesting, focus, involves determining what knowledge to harvest. Many firms determine what knowledge to harvest by having the harvesters meet with management. Note that the harvesters can either be employees from inside the company or external consultants. Through the meeting, the harvesters can determine the target audience and choose to harvest knowledge based on the target audience’s needs (Eisenhart, 50). CALL takes this process a step further, by not only deciding upon the content to harvest based on the target audience, but also upon the significance of the data to be harvested--“potential for generating…future strategic value” (Chua, 254). This means that valuations of knowledge will have to be determined; CALL works with senior Army officers to identify specific learning objectives that the harvested knowledge needs to fulfill. This latter method resolves, from the very start, the potential for information overload due to harvesting too much knowledge and overwhelming target audiences with superfluous data (Chua, 257). However, in addition to avoiding information overload, excessive focus can also cause CALL to ‘miss’ information; high-potential information is lost due to its lack of fitting the established criteria (fitting a team’s needs and fulfilling a learning objective).

Depending on the chosen method, once the knowledge to be harvested has been determined, the next step is to ‘find and elicit’ the knowledge. Find and elicit refers to discovering the experts in a knowledge area and then interviewing or extracting the desired knowledge from them. Common strategies for the find and elicit step employ a first-person or second-person approach. Cerabyte Inc. produces software, called Infinos System, which supports a first-person style. The system asks experts a series of leading questions while he or she works through a process in order to obtain best practices and procedures knowledge. “The questions are designed to help employees clearly articulate best practices by encouraging them to think, clarify, and record their actions” (Duffy, 60). Georgia-Pacific, a forest products company, uses a second- person format of extracting knowledge. The company first directly interviews both identified experts and the target audience to determine gaps in knowledge. The harvester then interviews the experts again to fill in these gaps (Eisenhart, 50). CALL’s method to find and elicit knowledge utilizes a unique third-person approach of gathering knowledge, which places less emphasis on the ‘expert’ and more emphasis on outside observers. To gather information from identified experts – soldiers in the field – CALL first assembles a data collection team of about eight to fifty “subject matter experts…from various units across the Army” (Chua, 254). The team members are selected specifically for their removal from the event being studied, in order to promote objectivity. The team members are also selected such that many different fields of study are represented, which “enables deep knowledge to be collected for each event” and enhances “the reliability and validity” of gathered knowledge (Chua, 255). Additionally, the use of outsiders from different specialties allows for “fresh ideas” to be added to analysis (Chua, 258). The CALL data collectors are then dispatched alongside troops to observe, capture, and document in the midst of the event (Chua, 257). The collectors document with a variety of media in order to capture the reality of the situation as fully as possible, usually employing video and photographs, as well as diagrams and written descriptions. This detail not only allows end-users to gain from the knowledge by ‘reliving’ the experience (Chua, 255), but also to retain the context of the event (Chua, 257). Occasionally, the collectors will also interview a wide range of personnel to gather feedback and interpretations in order to get more perspectives on the event. Note that a peripheral benefit of this method is that it is not affected by an organization’s culture of sharing. CALL’s method is not centered on the individual expert and his or her desire to share knowledge; rather, CALL focuses on third-party observers to make unbiased observations from multiple viewpoints. Despite its benefits, CALL’s avoidance of the expert is also the third-person approach’s biggest shortcoming. Many companies have found that their knowledge harvesting projects are reliant upon the fact that their harvested knowledge has come directly from experts. Employees are more receptive of harvested knowledge that come first- or second-hand from experts. They claim that information is more “memorable” and “educational” if the sources are fellow employee-experts (Keelan).

Organize, the third step in the Knowledge Harvesting process, refers to categorizing, fixing, adding, or pruning the gathered information. In this stage, the harvester asks, “If I was in this situation, would I get to the same place [the experts] did [using this gathered knowledge]?” If the answer is no, then the harvester needs to better organize or edit the gathered knowledge (Eisenhart, 52). CALL utilizes a very extensive human-centered approach to the organize step. First, the data collection team communicates within itself to organize individual insights. These insights are then sent to CALL headquarters to be analyzed by another group, the analysts, who act a conduit, “seek[ing] input and insights from other Army experts.” The analysts are ultimately responsible for further organizing and editing in an attempt to construct new knowledge from the disparate pieces of knowledge sent in from the data collectors. After indexing this new knowledge, it is transmitted electronically for review by other professionals in the Army. Finally, after review by Army professionals – upon which the knowledge is codified and divided into ‘lessons’ – the data collectors return to CALL headquarters and review all of the compiled lessons. Through a process called the ‘murder board,’ the data collectors decide which lessons are thrown out and which are important enough to be packaged for distribution (Chua, 255-256). Despite this exhaustive and complete approach to organizing, the amount of work that must be done – in addition to the lack of use of technology – to reach this point of completeness dramatically increases the delay in getting harvested information to the target audience. This could pose a major problem if the harvested knowledge is time- sensitive in the environment (who wants obsolete information?) or within the target audience (what team wants knowledge that it needed last week?).