799

In the lead article, Paul Cohen analyzes over 1.50 papers that were presented at the national conference last summer. Based on this analysis, he makes some interesting observations on the types of research in which we currently engage. Most research (or at least most research considered worthy of presentation by the AAAI-90 Program Committee) follows one of two strategies, according to Cohen's statistical analysis. One strategy is model oriented; that is, formal models of symbolic problem solving are hypothesized to be applicable to particular situations and then often tested on toy problems. The second strategy is system oriented; that is, it emphasizes the building of systems to solve difficult real-world problems.