Dialogue-Based Research in Man-Machine Communication.

摘要

This paper first surveys current knowledge of human communication from a point of view which seeks to find or develop knowledge that will be useful to computer system designers. The relevant scientific knowledge is found to be fragmentary and hard for designers to use. Next, the problem of complexity is explored. Building a useful knowledge of human communication is an extremely complex task. Controlling this complexity and its effects, without giving up usefulness, is seen as the central problem in designing a research approach. Finally, a new research methodology is presented. It contains some innovations that help control the complexity of the task, and others that make the results useful to designers. The methodology is unique in that it is based on case analysis rather than functional system design, the results are in the form of individual computer algorithms (much smaller than systems), and the algorithms are transferable into useful (nonresearch) systems.