Half-life of expert systems. One interesting method for measuring the life of knowledge assets is to measure the performance of an expert system that has captured these 'assets' in a knowledge base. The time, without maintenance or updating, for the system to regress from excellent to mediocre/poor performance provides an interesting insight into the economic life of
knowledge. For example, DEC's XCON and XSEL product configuration management system has 50% of
its configuration 'rules' change every year, tied directly to the product life cycle (McDermott, 1994). At the other end of the spectrum, the half-life of the knowledge assets in the PBA expert system developed by FMC to monitor the manufacturing of phosphorous is extremely long, because the manufacturing process is well-understood and already 'optimized' (Feigenbaum et al., 1988).
Half-life of expert systems. One interesting method for measuring the life of knowledge assets is to measure the performance of an expert system that has captured these 'assets' in a knowledge base. The time, without maintenance or updating, for the system to regress from excellent to mediocre/poor performance provides an interesting insight into the economic life of
knowledge. For example, DEC's XCON and XSEL product configuration management system has 50% of
its configuration 'rules' change every year, tied directly to the product life cycle (McDermott, 1994). At the other end of the spectrum, the half-life of the knowledge assets in the PBA expert system developed by FMC to monitor the manufacturing of phosphorous is extremely long, because the manufacturing process is well-understood and already 'optimized' (Feigenbaum et al., 1988).
การแปล กรุณารอสักครู่..