[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: RE: HM.applications-Profiling-Level of Details/Abstraction
Matching glue in many systems like this is done by matching. For example, in the SOAR System, something like this (abstracted from "Why Is a Raven Like a Writing Desk?" Lessons in Interface Consistency and Analogical Reasoning from Two Cognitive Architectures John Rieman*, Clayton Lewis*, Richard M. Young , and Peter G. Polson* *Institute of Cognitive Science University of Colorado, Boulder, CO 80309-0344 MRC Applied Psychology Unit 15 Chaucer Road, Cambridge CB2 2EF, UK rieman@cs.colorado.edu **Knowledge is represented as propositions about objects, where each proposition has a subject, a relation, and an object Represents certain basic procedural knowledge as productions. Productions fire when their condition side matches the current goal. For most productions, the result of firing is to add new information to Working Memory. Many productions also alter the goal or the stack of goals, and a few productions have the special function of causing physical movements that interact with the outside world. Look for matches between objects and elements of instructions. Instructions are represented in subject-relation-object form, Identify an overlap, Propose identified object as a candidate for action.** You'll find most chaining systems work something like that. Len http://www.mp3.com/LenBullard Ekam sat.h, Vipraah bahudhaa vadanti. Daamyata. Datta. Dayadhvam.h -----Original Message----- From: Mark Brownell [mailto:gizmotron@earthlink.net] Sent: Friday, September 07, 2001 11:20 AM To: humanmarkup-comment@lists.oasis-open.org Subject: Re: HM.applications-Profiling-Level of Details/Abstraction ----- Original Message ----- From: Bullard, Claude L (Len) > > - If object is this.object and goal is this.goal; use this.action. > > This requires the system (eg, the Human object) to identify > the meaningful attributes. > > SELECT objects WHERE object is-a objectType AND has-name (property = > somevalue) > > The facts may be episodic (was started) instead of categorical > (is-a object). > > It may need chains of facts between the example task and the example action, > > to make features of the action meaningful. What identifies the correct > action/control in a set (the item selection problem)? > > Scenario: a human unfamiliar with a car's controls wants to > cool the car. > > Goal is "cold" > > Given > > Episodic fact: (this.action(push red button) -> this.state(heater.On) -> > Result = "hot" ) > Context fact: (blue = "cold") > > Learns: > > - rules for learning > (this.action (this.object(controls) -> this.state(system.on)) > - rules for specific results > (this.action (push blue button) -> this.state (airConditioner.on) -> > Result="cold") > > and these become part of the knowledge base. > > Sean, how would RDF represent this? Then we should inquire as to how a > human object uses the knowledge base. > > Len Hi All, To me, this form of explaining situations that then can be written into a RDF are very helpful in understanding HumanML discussion. Something that jumps out at me is the association of ideas with respect to other forms of data that can also be associated with (I guess here) the first idea, or topic. So if there is going to be a RDF that attaches associations of a primary object to specified sub-objects, (secondary objects), then how will (a smart glue) associate beyond a RDF description of an associative primary object. Especially when secondary objects have associative properties that relate to secondary objects in other primary objects. In my own opinion it is the presence of context that some how establishes the characteristics of a magic glue for this kind of purpose. So far I have learned that an application using the RDF will establish its own rule base given that the RDF is loosely floating somehow in a gravity free environment, where objects obtain connective properties by the application's task oriented interface to either machine or human need. Please be patient with me here. I just woke from one dream where I saw a connectivity of information as a linking mechanism. I'm now trying to contemplate the existence of a magic glue. If you get what I'm saying here, can you explain magic glue to me. For instance from the above example ["red" - turns on the heater] secondarily relates to ["blue" - turns on the car's cooler] with both being part of the car's dashboard apparatus. Now all are part of the properties of a car, where "red" is also a indicator light that both verifies that the heater is on and an indicator light, from the idiot light group, of the instrument panel, of the dashboard, that reports that the engine is seizing up as a result of the oil falling out of the car's engine. It seams that meaningfulness expressed in RDF form will need some form of AI to determine its associative connection to the request being made to it. In other words, a magic glue. Best regards, Mark Brownell ---------------------------------------------------------------- To subscribe or unsubscribe from this elist use the subscription manager: <http://lists.oasis-open.org/ob/adm.pl>
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC