The Open Mind Common Sense project is an attempt to construct a database of commonsense knowledge through the collaboration of a distributed community of thousands of non-expert netizens. We give an overview of the project, describe our knowledge acquisition and representation strategy of using natural language rather than formal logic, and demonstrate this strategy with a search engine application that employs simple commonsense reasoning to reformulate problem queries into more effective solution queries.

Earlier work by Newell, Simon, Shaw, and Gelernter in the middle and late 1950s emphasized the heuristic approach, but the weight soon shifted to various syntactic methods culminating in a large effort on resolution type systems in the last half of the 1960s. It was about 1970 when considerable interest was revived in heuristic methods and the use of human supplied, domain dependent, knowledge. It is not my intention here to slight the great names in automatic theorem proving, and their contributions to all we do, but rather to show another side of it. For recent books on automatic theorem proving see Chang and Lee [19], Loveland [44], and Hayes [31]. Also see Nilsson's recent review article [61]. The word "resolution" has come to be associated with general purpose types of theorem provers which use very little domain dependent information and few if any special heuristics besides those of a syntactic nature. It has also connoted the use of clauses and refutation proofs. There was much hope in the late 60's that such systems, especially with various exciting improvements, such as set of support, model elimination, etc., would be powerful provers. But by the early 70's there was emerging a belief that resolution type systems could never really "hack" it, could not prove really hard mathematical theorems, without some extensive changes in philosophy.

Many of the recent expert rule-based systems [Dendral, Mycin, AM, Pecos] have architectures that differ significantly from the simple domainindependent architectures of "pure" production systems. The purpose of this paper is to explore, somewhat more systematically than has been done before, the various ways in which the simplicity constraints can be relaxed, and the benefits of doing so. The most significant benefits arise from three sources: (i) the grain size of a typical rule can be increased until it captures a unit of advice which is meaningful in that system's task domain, (ii) the interpreter can become accessible to the rules and thus become dynamically modifiable, and (iii) meaningful permanent Knowledge can be stored in data memories, not just within productions. Although there are costs associated with relaxing the simplicity constraints, for many task domains the benefits outweigh the costs.

Further progress in the application of computers to many practical fields seems to depend heavily on the success in implementing learning and inductive processes within machines. For example, to develop a consultation system for medical or plant disease diagnosis, prognosis and decision making in general, it is very desirable, perhaps even necessary, to be able to'teach' the system through examples of correct and/or incorrect decisions, rather than by precisely describing the decision process in its full generality and then transforming this description into a computer program. A similar situation exists in computer chess. The development of computer programs playing at the master level (especially the end games) seems to be a formidable task if the programs are not eventually able to learn and improve on their decision making rules through the specific examples of games, rather than by being explicitly told all the rules. Due to easy access to human knowledge about chess and the relative simplicity of testing the results, chess is one of the most attractive testing domains for inductive inference programs.

To use language one must be able to make inferences about the information which language conveys. This is apparent in many ways. For one thing, many of the processes which we typically consider "linguistic" require inference making. For example, structural disambiguation: (1) Waiter, I would like spaghetti with meat sauce and wine. You would not expect to be served a bowl of spaghetti floating in meat sauce and wine. That is, you would expect the meal represented by structure (2) rather than that represented by (3).

Machine intelligence problems are sometimes defined as those problems which (i) computers can't yet do, and (ii) humans can. We shall further consider how much "knowledge" about a finite mathematical function can, on certain assumptions, be credited to a computer program. Although our approach is quite general, we are really only interested in programs which evaluate "semihard" functions, believing that the evaluation of such functions constitutes the defining aspiration of machine intelligence work. If a function is less hard than "semihard," then we can evaluate it by pure algorithm (trading space for time) or by pure lookup (making the opposite trade), with no need to talk of knowledge, advice, machine intelligence, or any of those things. We call such problems "standard." If however the function is "semihard," then we will be driven to construct some form of artful compromise between the two representations: without such a compromise the function will not be evaluable within practical resource limits. If the function is harder than "semihard," i.e. is actually "hard," then no amount of compromise can ever make feasible its evaluation by any terrestrial device.

How can we get a computer to understand natural language? Our view of the problem has progressed over the years to a point where an answer to that question today would look quite different from one given ten or even five years ago. Originally, researchers felt that the most relevant issue was syntax. Later, most people agreed that semantics was the most relevant field of study (although few would have agreed on what semantics was). Five years ago, or so, our research was concentrated on finding an adequate meaning representation for sentences.

The program analyses carefully the initial situation. It creates some plans and tries to execute them. It analyses the situations deeper in the tree only if the plan fails. In that case it generates new plans correcting what is wrong in the old one. So, the program considers only natural branches of the tree. It can find combinations for which it is necessary to look more than twenty ply ahead. The paper describes the methods used for analyzing a situation and for modifying unsuccessful plans. Then we examine some results found by the program.Artificial Intelligence 8 (1977), 275-321

See also: Stanford Heuristic Programming Project Memo HPP-77-25Proc. IJCAI-77: Fifth International Joint Conference on Artificial Intelligence, pp 1014-1029

Austin Tate Department of Artificial Intelligence University of Edinburgh Edinburgh Scotland Abstract Procedures for optimization and resource allocation in Operations Research first require a project network for the task to be specified. The specification of a project network is at present done in an intuitive way. AI work in plan formation has developed formalisms for specifying primitive activities, and recent work by Sacerdoti (1975a) has developed a planner able to generate a plan as a partially ordered network of actions. The "planning: a joint AI/OR approach" project at Edinburgh has extended such work and provided a hierarchic planner which can aid in the generation of project networks. This paper describes the planner (NONLIN) and the Task Formalism (TF) used to hierarchically specify a domain. Current work in Operations Research (OR) and Artificial Intelligence (AI) has concentrated on different aspects of the problem. We have taken an interdisciplinary approach in the hope that this will lead to a development of both these aspects. In the OR approach, the planning process falls into two stages. The constituent "jobs" of a plan are specified together with their precedence relationships (i.e.