[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: RE: [humanmarkup-comment] Base Schema-channel
Len, Rob and Rex, (humanmarkup-comment) In my group's work we are looking at something interesting. It is the invariance (patterns of occurrence and co-occurrence) in log files from Intrusion Detection Systems. The idea is that these log files can be convolved by some frame (with slots and fillers) that picks up a part of a pattern that indicates one of say 301 event types.. (with the possibility that a new event may also occur.) We hope to establish a connection between the algorithmic convolution and the actual fashion that the cortical layers of the human perception system work. A second application of the same methodology is used in text understand and image understanding. This process is COMPLETELY in the data space. The events themselves are in the data space. What is not in the data space is the human interaction for information produced from the "synthetic perception" of these data space events. What I am looking for is two things: 1) a process of encoding the event descriptions and relationships into Topic Maps 2) a process of understanding the human analysis and dissemination - including a process that will help the intelligence vetting avoid natural tendencies that lead to false senseMaking. the strategy is that by adopting a tri-level architecture (based on relative stratified complexity and the biological model of intelligence) we develop an agile perceptional system, and a cognitive repository. What I hope to see from the human markup committee is a notational language or process methodology that allows me to not only explain human information interaction, but to simulate human information interaction as part of the NIMA work. http://www.ontologystream.com/cA/papers/cA-SPS.htm -----Original Message----- From: Bullard, Claude L (Len) [mailto:clbullar@ingr.com] Sent: Tuesday, June 04, 2002 5:14 PM To: 'Rob Nixon' Cc: paul; Rex Brooks; 'Norm Badler'; allbeck@graphics.cis.upenn.edu; humanmarkup-comment@lists.oasis-open.org; cognite@zianet.com Subject: RE: [humanmarkup-comment] Base Schema-channel As soon as we have completed the schema draft review and have something we are roughly consenting to, or even before, someone should be looking at topic maps built over this. Relationships are indeed key for simulation modeling as well as other applications. We have the same problems in public safety and data mining. Identifying the data is hard, but building a set of data dictionary topics backed by precise queries is a heckuva lot harder and very much determined by the point of view of the analyst and the question itself. len "... neck deep in the Big Muddy and the ol' fool says to push on" Pete Segar From: Rob Nixon [mailto:rnixon@qdyn.com] Ditto Len, Your box scenario is fine by me, it's the overall collection of boxes and how they "overlap" with each other in a somewhat fuzzy way that I'm talking about. Trying to find the appropriate box "interfaces" or "relationship mappings" that will make HumanMarkup useful. But this will be an evolutionary process. Thanks for the comments.
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC