A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....which can be used to specify a range of cooperation strategies. Although ACLs can be seen as analogous to coordination languages in distributed computing, they should be situated at a higher level of abstraction, as their aim is to support coordination not at the symbol but at the knowledge level [31]. That is, they should provide communication primitives which support the use, request, and supply of knowledge independently from implementation related aspects. Our aim in this paper is to analyse the communication primitives used in ACLs and the implications on the underlying levels ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, 1993.

....6 those that occur in other software engineering paradigms. Firstly, agent oriented interactions generally occur through a high level (declarative) agent communication language (typically based on speech act theory [1] Consequently, interactions are usually conducted at the knowledge level [25]: in terms of which goals should be followed, at what time, and by whom (cf. method invocation or function calls that operate at a purely syntactic level) Secondly, as agents are flexible problem solvers, operating in an environment over which they have only partial control and observability, ....

A. Newell, (1993) "Reflections on the Knowledge Level" Artificial Intelligence 59 31-38.

....tools listed in x2.4.2) Meta knowledge will be discussed below under the headings: problems solving methods (PSMs) x2.4.1) quality knowledge (x2.4.2) fix knowledge (x2.4.3) and social knowledge (x2.4.4) 2.4. 1 PSMs: Problem Solving Methods 275 In Newell s knowledge level modeling KL approach [81, 82], intelligence is modeled as a search for appropriate operators that convert some current state to a goal state. Domain specific knowledge in used to select the operators according to the principle of rationality; i.e. an intelligent agent will select an operator which its knowledge tells it will ....

....PSMs may be expressed graphically such as in Figure 14 (ovals are functions, rectangles are data structures) 2. In the minority KL A view, KBs contain wordK, sentenceK, and a single PSM. In Newell s operationalisation of KL, this single PSM is the 295 problem space computational model (PSCM) [82,124,125]. Programming the PSCM involves the consideration of multiple, nested problem spaces. Whenever a don t know what to do state is reached, a new problem space is forked to solve that problem. The observation that a PSCM system is performing (e.g. classification is a user interpretation of a ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993.

.... 10 quisition, Modeling, and Management Sant Feliu de Guixols, Catalonia, Spain October 15 18, 1997 1 Introduction Conventional software descriptions can be divided into several perspectives that include the events and activities in which data is processed (x2.2) Knowledge 15 level modeling [20, 21] adds an extra perspective; i.e. the goal of the program and the options which must be reflected on while exploring that goal. This article argues that these perspectives can be connected as follows. Roughly speaking, conventional software notations will be used to define some road map and ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993. 475

....SOAR using the PSCM involves the consideration of multiple, nested problem spaces. Whenever a don t know what to do state is reached, a new problem space is forked to solve that problem. Newell concluded that the PSCM was the bridge between SOAR and true KL modeling [Newell et al. 1991,Newell, 1993] There is a difference between PSCM (hereafter, KLA ) and KLB , a KL modeling variant which groups together a set of authors who argue for basically the same technique; i.e. Clancey s model construction operators [Clancey, 1992] Steels components of expertise [Steels, 1990] Chandrasekaran s ....

Newell, A. (1993). Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38.

....University, Caulfield East, Melbourne, Melbourne, Australia, 3185. email: timm insect.sd.monash.edu.au. Past papers available from http: www.sd.monash.edu.au timm pub docs papersonly.html. This paper is Dept. of Software Engineering technical report TR95 23. WpRef: 95 abkl ing approach [38, 43, 45, 44]. The fundamental premise of KLB is that a knowledge base should be divided into domain specific facts and domain independent abstract problem solving inference procedures (e.g. Clancey s model construction operators [8] Steels components of expertise [62] Chandrasekaran s task analysis, SPARK ....

.... facts and domain independent abstract problem solving inference procedures (e.g. Clancey s model construction operators [8] Steels components of expertise [62] Chandrasekaran s task analysis, SPARK BURN FIREFIGHTER [33] and KADS [68] KLA refers to Newell s research on the knowledge level [43, 45, 44] and the SOAR project [58, 69] KLA does not explicitly model problem solving procedures. The observation that a KLA system such as SOAR is performing (e.g. classification is a user interpretation of 1. The application of domain specific knowledge controlling. 2. a single inference ....

[Article contains additional citation context not shown here]

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993.

....listed in x3.4.2) Meta knowledge will be discussed below under the headings: problems solving methods (PSMs) x3.4.1) quality knowledge (x3.4.2) fix knowledge (x3.4.3) and social knowledge (x3.4.4) 420 3.4. 1 PSMs: Problem Solving Methods In Newell s knowledge level modeling KL approach [88,89], intelligence is modeled as a search for appropriate operators that convert some current state to a goal state. Domain specific knowledge in used to select the operators according to the principle of rationality; i.e. an intelligent agent will select an operator which 425 its knowledge tells it ....

....PSMs may be expressed graphically such as in Figure 15 (ovals are functions, rectangles are data structures) 2. In the minority KL A view, KBs contain wordK, sentenceK, and a single 440 PSM. In Newell s operationalisation of KL, this single PSM is the problemspace computational model (PSCM) [89,131,132]. Programming the PSCM involves the consideration of multiple, nested problem spaces. Whenever a don t know what to do state is reached, a new problem space is forked to solve that problem. The observation that a PSCM system is performing 445 (e.g. classification is a user interpretation of a ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993. 1765

....(PSCM) 61] Programming SOAR using the PSCM involves the consideration of multiple, nested problem spaces. Whenever a don t know what to do state is reached, a new problem space is forked to solve that problem. Newell concluded that the PSCM was the bridge between SOAR and true KL modeling [38, 39]. We distinguish between PSCM (which we term KL A) and KL B, a KL modeling variant which groups together a set of authors who argue for basically the same technique; e.g. Clancey s model construction operators [6] Steels components of expertise [58] Chandrasekaran s task analysis, SPARK BURN ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993.

....can be used to specify a range of cooperation strategies. Knowledge level coordination languages are situated at a higher level of abstraction with respect normal coordination languages of distributed computing, as they support coordination not at the symbol level but at the knowledge level [16]. A previous paper by one of the authors [8] represents a preliminary attempt towards a formal analysis of the interaction problems in this class of languages. In particular, a formal operational semantics for a subset of KQML is provided, which helped in showing relations between knowledge level ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, 1993.

....In the following we will apply the KL hypothesis to the phenomenon of co ordination. 3.1 Co ordination and the Knowledge level Newell s KL hypothesis has been shown to be useful in many areas of Artificial Intelligence, such as knowledge acquisition, knowledge based systems construction etc. [17]. Dietterich first applied it to machine learning. He distinguishes Knowledge level learning, where a system acquires a new KL description, from Symbol level learning, where the system learns only to evoke the knowledge it already has faster and more reliable [8] Schreiber and colleagues transfer ....

A. Newell. "Reflections on the Knowledge Level". Artificial Intelligence 59, 1993, p. 31-38

....Melbourne, Australia, 3185. email: timm insect.sd.monash.edu.au. Past papers available from http: www.sd.monash.edu.au timm pub docs papersonly.html. This paper is Dept. of Software Engineering technical report TR95 23. WpRef: 95 abkl Newell s knowledge level (KL) modeling approach [45, 52, 54, 53]. The fundamental premise of KLB is that a knowledge base should be divided into domain specific facts and domainindependent abstract problem solving inference procedures (e.g. Clancey s model construction operators [11] Steels components of expertise [76] Chandrasekaran s task analysis, ....

.... facts and domainindependent abstract problem solving inference procedures (e.g. Clancey s model construction operators [11] Steels components of expertise [76] Chandrasekaran s task analysis, SPARK BURN FIREFIGHTER [41] and KADS [83] KLA refers to Newell s research on the knowledge level [52, 54, 53] and the SOAR project [72, 84] KLA does not explicitly model problem solving procedures. The observation that a KLA system such as SOAR is performing (e.g. classification is a userinterpretation of 1. The application of domain specific knowledge controlling. 2. a single inference procedure ....

[Article contains additional citation context not shown here]

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993.

....[7] A new expert system design approach (which has come to dominant the knowledge acquisition (KA) field) is the the search for reusable abstract domain independent problem solving strategies. We call this approach KLB since it is a variant of Newell s knowledge level (KL) modeling approach [35, 37, 36]. KADS is a KLB variant used by many commercial expert systems practioners 1 . Wielinga et al. note that, as of 1992, KADS has been used in some 40 to 50 KBS projects, 17 of which are described in published papers [59] Amongst AI 94 participants, interest in KADS was high. Despite the current ....

....(PSCM) 61] Programming SOAR using the PSCM involves the consideration of multiple, nested problem spaces. Whenever a don t know what to do state is reached, a new problem space is forked to solve that problem. Newell concluded that the PSCM was the bridge between SOAR and true KL modeling [37, 36]. We distinguish between PSCM (which we term KLA ) and KLB , a KL modeling variant which groups together a set of authors who argue for basically the same technique; i.e. Clancey s model construction operators [11] Steels components of expertise [51] Chandrasekaran s task analysis, SPARK BURN ....

A. Newell. Reflections on the knowledge level. Artificial Intelligence, 59:31--38, Feburary 1993.

....extensively in AI. The scarcity of citations on this issue in a recent survey of work on mental states within AI by Shoham and Cousins [58] attests to this fact. 1 Similarly, although Newell s paper on the Knowledge Level [45] is among the most referenced AI papers [5] in his perspective paper [46], Newell laments the lack of work following up on these ideas by the logicist community. He mentions Levesque s [40] as the only exception. Given this situation, it is worth clarifying what we view as the four central questions in mentallevel modeling. They are: 1) Structure what class of ....

....a number of ideas for formalizing it. Newell also advocated this idea, stressing the need for an abstract level of representation of programs and machines, which he called the knowledge level [45] However, the tone of both papers is, in general, intuitive, informal, and motivational. In fact, in [46], Newell laments the lack of attempts to pursue this approach within the logicist community . We believe this to be the first work within AI to provide formal treatment of these ideas. Our approach makes a number of semantic contributions: We proposed a structure for a mental level description of ....

A. Newell. Reflections on the knowledge level. Artificial Intelligence, pages 31--38, 1993.

....that literature. 5.1 KL= KLA and KLB We divide KL into KLB and KLA [17] The fundamental premise of KLB is that a knowledge base should be divided into domain specific facts and domain independent abstract problem solving patterns. KLA refers to Newell s research on the knowledge level [23, 25, 24] and the SOAR project [28, 35] SOAR models intelligence as the search for appropriate operators that convert some current state to a goal state. Domain specific knowledge as used to select the operators according to the principle of rationality; i.e. an intelligent agent will select an operator ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993.

....can be used to specify a range of cooperation strategies. Knowledge level coordination languages are situated at a higher level of abstraction with respect normal coordination languages of distributed computing, as they support coordination not at the symbol level but at the knowledge level [16]. A previous paper by one of the authors [8] represents a preliminary attempt towards a formal analysis of the interaction problems in this class of languages. In particular, a formal operational semantics for a subset of KQML is provided, which helped in showing relations between knowledge level ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, 1993.

....of mental components such as beliefs, capabilities, choices, commitments etc. We take this description to be a helpful abstract model for viewing software agents, even if their actual implementation does not make claims to such ambitious concepts. Agents then reside at the knowledge level [24, 25] and are not well served by general languages and protocols developed for distributed computing. Such languages and protocols focus on processes rather than on the programs or collection of programs that constitute the agents. As a result, a communication language should be powerful enough to ....

Allen Newell. Reflections on the knowledge level. Artificial Intelligence, 59:31--38, 1993.

....campos labein.es Abstract This paper presents a high level framework for analysing and designing intelligent agents. The framework s key abstraction mechanism is a new computer level called the Social Level. The Social Level sits immediately above the Knowledge Level, as defined by Allen Newell, and is concerned with the inherently social aspects of multiple agent systems. To illustrate the working of this framework, an important new class of agent is identified and then specified. Socially responsible agents retain their local autonomy but still draw from, and provide resources to, the ....

....solving paradigms, etc. Agent problem solving behaviour can then be characterised through the Principle of Rationality: if an agent has knowledge that one of its actions will lead to one of its goals, then the agent will select that action (Newell, 1982) Given the success of the KL approach (Newell, 1993), it seemed natural to see whether it could be used directly to describe problem solving in multi agent contexts. Aitken et al. 1995) attempt to do precisely this and present a KL analysis of a number of multi agent systems (focusing in particular on the problem of global coherence) However ....

Newell, A., (1993) "Reflections on the Knowledge Level" Artificial Intelligence 59 31-38.

....of mental components such as beliefs, capabilities, choices, commitments etc. We take this description to be a helpful abstract model for viewing software agents, even if their actual implementation does not make claims to such ambitious concepts. Agents then reside at the knowledge level [19, 20] and cannot therefore be accommodated by languages or protocols that appear in Distributed Computing and focus on processes rather than on programs or collection of programs that constitute the agents. As a result, a communication language should be powerful enough to support communication between ....

Allen Newell. Reflections on the knowledge level. Artificial Intelligence, 59:31--38, 1993.

....[5] A new expert system design approach (which has come to dominant the knowledge acquisition (KA) field) is the the search for reusable abstract domain independent problem solving strategies. We call this approach KLB since it is a variant of Newell s knowledge level (KL) modeling approach [28, 30, 29]. KADS is a KLB variant used by many commercial expert systems practioners 1 . Wielinga et al. note that, as of 1992, KADS has been used in some 40 to 50 KBS projects, 17 of which are described in published papers [46] Amongst AI 94 participants, interest in KADS was high. Despite the current ....

....(PSCM) 48] Programming SOAR using the PSCM involves the consideration of multiple, nested problem spaces. Whenever a don t know what to do state is reached, a new problem space is forked to solve that problem. Newell concluded that the PSCM was the bridge between SOAR and true KL modeling [30, 29]. We distinguish between PSCM (which we term KLA ) and KLB , a KL modeling variant which groups together a set of authors who argue for basically the same technique; i.e. Clancey s model construction operators [9] Steels components of expertise [40] Chandrasekaran s task analysis, SPARK BURN ....

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, Feburary 1993.

No context found.

A. Newell. Reflections on the Knowledge Level. Artificial Intelligence, 59:31--38, 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC