Saturday, February 19, 2005
On the Nature of science and
cultural Polemics à
On the discussion of data
regularity à
On the Stephenson Cyber Attack
Taxonomy à
On the evolution of the
digital Glass Bead Game à
Communication from InOrb Technologies about d-GBG
Paul;
I’ve been analyzing the glass bead game formats etc. to see if there was any way I could at least semi-automate some metadata extraction based on some common formatting among the different threads and beads within a thread. It is becoming apparent that it is not feasible at this time to do so. There is just very little commonality among the beads.
So, the only way to really fill out things such as titles, author, etc. is to go it manually. As of current, there are 524+ html pages alone that I would have to manually add this information to, which doesn’t even include the word files etc.
What I DO feel we can do relatively quickly is utilize output from VisualText to assign nouns and concepts to each bead. I have been looking through the VisualText passes used on the beads, and am working on understanding them better to enable me to modify them to use on the beadgames.
Ultimately, the rest of the metadata would take a number of months of work to assign.
In the meantime, a new online form is going to be made available which allows the user to enter their own metadata as required (there’s a subject field, author field, etc..). Initially this will just be your standard web form (using php), but eventually I will begin to add authentication and moderation features.
<Future link to
the d-GBG>
As for now, my priority is on getting the attack data and investigation/interview note repository stood up.
I have encountered a number of problems with Greenstone open source document repository system, and have gotten in touch with the Greenstone developers. Accordingly, a new version is being released next week to address some of these issues, as well as a number of others. Currently, I am testing out a pre-release version for any further bugs and working on trying to get it into a position that suits us best.
Once the new version comes out, and I get some example data from Peter, than I can begin working on standing up a demo.
The basic principle for the first portion of the repository is to create an interface for the navigation and analysis of risk domains as described within the process, and which Peter and I have discussed briefly. Basically, we use XML tagging, at this point, to create linkage between the attack data and the data entered by the investigator into the system.
The data allows us to see what type of information exchange is occurring on the network. This is where the event compression we have discussed will be used to create an output of only distinct events so as to be better viewed and analyzed by the investigator to complete a portion of the baseline risk assessment.
The Orbs will also be factored in on a secondary level. This will allow us to move away from Greenstone down the road if we so wish, while allowing us to test out Greenstone’s capabilities with regards to what we’re trying to do.
From what I’ve gathered from the Greenstone community, what we’re trying to do hasn’t really be done in any way (in other domains/fields or whatever) by anyone as yet. So, it should be interesting to see what we can and cannot do with the system.
Yours truly,
Nathan Einwechter