登入選單
返回Google圖書搜尋
出版Department of Computer Science, Stanford University, 1981
URLhttp://books.google.com.hk/books?id=khd-SgAACAAJ&hl=&source=gbs_api
註釋All A1 programs are essentially reasoning programs. And, to the extent that they reason well about a problem area, all exhibit some expertise at problem solving. Programs that solve the Tower of Hanoi puzzle, for example, reason about the goal state and the initial state in order to find 'expert-level' solutions. Unlike other programs, however, the claims about expert systems are related to questions of usefulness and understandability as well as performance. We can distinguish expert systems from other A1 programs in the following respects: (1) Utility, (2) Performance, and (3) Transparency. Designers of expert systems are motivated to build useful tools in addition to constructing programs that serve as vehicles for A1 research. This is reflected in the tasks chosen. Solving the Tower of Hanoi puzzle, per se, is not a critical bottleneck in any scientific or engineering enterprise. But integrating mathematical expressions and determining molecular structures are important problems for scientists. Utility is the least important of the three criteria and is perhaps less definitional than a personal bias about whether expertise on trivial matters constitutes expertise at all. In some cases a task is chosen just because of its inherent importance. More often than not, a problem's significance for A1 research is also a factor now because expert systems are still constructed by researchers.