[tangent4ronpaul;1886963]
OCH = Operation Cat Herder
http://operationcatherder.com/index.php/Main_Page
This is a new tool to me. Does not seem to fit these needs.
It sounds like you want to re-create the semantic web (still very much in it's infancy) from scratch:
http://en.wikipedia.org/wiki/Semantic_Web
Definitely not trying to use or re-create semantic web, nor to re-invent.
It kinda sounds like you are describing what we have in this forum software but with umpteen zillion sub forums, of sub forums of sub forums and voting on individual posts. Or more like a Wiki... with a message forum and voting system attached to each section of each topic.
This has been a first cut. I'm not sufficiently familiar with forums and Wiki to relate well to their usage. Actually, I'm making this up as we go!
From the Wiki standpoint - you have a couple of problems: 1) your users need to learn or know the syntax and 2) they have to agree to abide by formatting, classification and reference rules. (They also need to be able to spell and use the correct keywords - GOOD LUCK!). Wiki sorta works because other people can come along and edit your posts and there is a small army of editors lurking in the background that step in and really know that they are doing.
For a "Constitution" forum the breakdown is either Article/Section/Clause or is Amendment/Clause (or something similar). Each clause has fixed and variable topics of interest to be discussed. Fixed data, including an abstract, poll Q & A's, and links to reference material would be preserved in text format. Variable topics are a grey area needing to be explored further, i.e. what is the purpose of the clause (say, eminent domain), range of interpretations, practical limits of application, punishments and/or penalties, etc. This requires considerable discussion whose purpose should be to translate these parameter variables (previously called facets) into cryptic tokens (voting ballots) that each carry an array of fixed data (i.e. voter ID and parameer ID), and an array of variable data from voter input regarding the parameter (such as a poll result, a scaled value of 1 thru 9, or whatever) formatted in such a way as to be weighed or tallied with other same parameter token variables (perhaps in linked lists containing up to thousands of tokens representing one for each voter?).
As you are coming from a one post per topic, per person standpoint - what makes Wiki work breaks down. Also, you really haven't reduced the amount of information posted on a topic - just forced each contributor to post in one place. You may still have 500 contributors now making really loooong posts on a specific topic. And like Wiki, there is still a need for discussion about what is posted. At least with a Wiki- you get ONE article about a specific topic, not one for each poster. Or am I not understanding what you are trying to do. Your description was a bit confusing.
Unfortunately I do not know what pigeon holes exist, therefore cannot at this time describe my pigeon in terms acceptable to existing holes. I can only describe the desired result (a culmination of voter preferences on partiular legal issues) and a possible method of arriving there. I have some familiarity with linked lists of tokens, which can be very versitile in dedicated computer applications. Here we are trying to use existing web tools. I cannot even speak HTML, so I know not what can readily be done on existing website systems.
Are you thinking of a Daily Kaos(sp?) or Digg approach where articles get voted up or down? (but with the overhead of all these levels?) - That's still lossy, as those with an agenda can bury good info and inflate bad info - also, an article can be garbage but have one or two great links and get buried.
I dunno.
Also, this should be / shouldn't be law bit and the really fine grain level voting system... - How does that relate to what you were describing before? Who is the intended audience? and have you got any idea how many variables you are talking about and what kind of taxonomy this is going to require? - it sounds like a DB nightmare!
Intended audience is the large community of voters, of course, their legislatures, and possibly judges and scholars.
It is probable that the variable portion of each legal issue token would be unique in length and content, in which case there would need to have a corresponding description for that token's layout resident in the database.
Before you were talking about documenting our representatives voting records and summarizing / providing links to specific issues.
I do not recall doing that. However, I did suggest a comparable forum for candidates, but comparable perhaps only from the standpoint of the user.
Now your talking about voting individuals posts up and down to determine if the topic should become law? What are you trying to describe here?
First (pilot) application would be the U.S. Constitution, for its universal appeal especially to to Ron Paul types, but also to all stripes of voters. Legislatures, etc., would need to be able to discern from the resulting data the public's consent to the purpose of each clause (say, eminent domain), range of interpretations, practical limits of application, punishments and/or penalties, etc.
And what happens to the votes if someone changes their post?
Sorry for the confusion. Hopefully a voter could retrieve his token(s) and change them to suit changing events and times; but that such changes would not affect the apparent number of voters, or whatever.
For all you are describing - why not just crawl Thomas and organize the lower level, only based on their tags as an extension - linking to them for content, beyond a basic bill title, number and status? Become an extension of their system, without recreating and re-organizing everything.
Not sure I follow you. Who is Thomas. Would he get ticked off if I crawl him? Seriously, if this can be done at all, I'm open to whatever tools will do the job. I can only hope to steer towards obtaining useful results.
As to the data/text mining I was referring to the point is to aggregate the data and create a single search point in order to make the process less labor intensive.
Thought you meant manual data/text mining. Again, whatever works.
