From: cvs@openprivacy.orgCVS update: openprivacy/htdocs/notes/reputations
Date: Monday January 29, 19101 @ 12:37
Author: fen
CVSWEB Options: -------------------
Main CVSWeb: http://openprivacy.org/cgi-bin/cvsweb/cvsweb.cgi
View this module: http://openprivacy.org/cgi-bin/cvsweb/cvsweb.cgi/openprivacy/htdocs/notes/reputations
-----------------------------------
Update of /usr/local/cvs/public/openprivacy/htdocs/notes/reputations
In directory giga:/home/fen/projects/openprivacy/htdocs/notes/reputations
Modified Files:
protocol.txt
Log Message:
some thoughts added
*****************************************************************
File: openprivacy/htdocs/notes/reputations/protocol.txt
CVSWEB Options: -------------------
CVSWeb: Annotate this file: http://openprivacy.org/cgi-bin/cvsweb/cvsweb.cgi/openprivacy/htdocs/notes/reputations/protocol.txt?annotate=1.18
CVSWeb: View this file: http://openprivacy.org/cgi-bin/cvsweb/cvsweb.cgi/openprivacy/htdocs/notes/reputations/protocol.txt?rev=1.18&content-type=text/x-cvsweb-markup
CVSWeb: Diff to previous version: http://openprivacy.org/cgi-bin/cvsweb/cvsweb.cgi/openprivacy/htdocs/notes/reputations/protocol.txt.diff?r1=1.18&r2=1.17
-----------------------------------
Index: openprivacy/htdocs/notes/reputations/protocol.txt
diff -u openprivacy/htdocs/notes/reputations/protocol.txt:1.17 openprivacy/htdocs/notes/reputations/protocol.txt:1.18
--- openprivacy/htdocs/notes/reputations/protocol.txt:1.17 Tue Jan 23 14:20:55 2001
+++ openprivacy/htdocs/notes/reputations/protocol.txt Mon Jan 29 12:37:42 2001
@@ -1,4 +1,4 @@
-$Id: protocol.txt,v 1.17 2001/01/23 22:20:55 burton Exp $
+$Id: protocol.txt,v 1.18 2001/01/29 20:37:42 fen Exp $
protocol requirements:
@@ -52,11 +52,13 @@
- In order to build a tree structure we need a way to 'audit' the Reputation
that was returned. This would be the second node level.. This could be a
lot of data so we need a fast way to return it. (this could be XPATH, we
- could return a count() first
-- We need to have an optional <description> field. This is meta-info for
+ could return a count() first
+
+- We need to have an optional <description> field. This is meta-info for
human agents. This might just be accomplished via additional RDF. A good
example of why this is necessary is that a human agent might add reasoning
to an opinion they create:
+
Northwest Airlines -> Reliability -> BAD
<description>
@@ -65,41 +67,48 @@
Of course this data is repudiatable but is still valid if you really trusts
its creator.
-
- another way to do this could be to have all the information exposed and
- XQL/XPATH queries ran... of course this is a data problem... we wouldn't want
- to just spit this data back because it could be a lot of data...
+ XQL/XPATH queries ran... of course this is a data problem... we wouldn't
+ want to just spit this data back because it could be a lot of data...
-- I think that we need might need to define some basic API calls getReputation()
- etc but also have a base method like query() and pass it an XPath query and
- this returns an XML Document. Reputations should be forced into an XMLSchema
- so that the query maps correctly. We might need to create our own XPath
- impl... (or reuse one).
+- I think that we need might need to define some basic API calls
+ getReputation() etc but also have a base method like query() and pass it
+ an XPath query and this returns an XML Document. Reputations should be
+ forced into an XMLSchema so that the query maps correctly. We might need
+ to create our own XPath impl... (or reuse one).
- seems like a cool way to use HTTP and XPointer...
-
- TIME??? How do we handle this??? This currently isn't in HTTP so maybe we
- shouldn't worry..
-
-
-
+ shouldn't worry..
-
-
+- should the "popularity" of a URL also be published? This could be the
+ total number of opinions cast... this might need to have a start and end
+ date.
+
+- ability to calculate reputation sets (arrays) based on threshold... "show
+ me all Reputation objects for URI1 with a opinion > X..
+
+ - Given an ontology... determine the X best (or worst) things... sorted.
+ A good example. Show me the top 10 Open Source projects, Show me the
+ best RSS channels. It would be nice if http://apps.kde.com had this.
+=== problems to be addressed === 2001.01.29 fen
+Ontology
-
-
-- should the "popularity" of a URL also be published? This could be the total
- number of opinions cast... this might need to have a start and end date.
-
-- ability to calculate reputation sets (arrays) based on threshold... "show me
- all Reputation objects for URI1 with a opinion > X..
-
- - Given an ontology... determine the X best (or worst) things... sorted. A
- good example. Show me the top 10 Open Source projects, Show me the best
- RSS channels. It would be nice if http://apps.kde.com had this.
-
+- object may have multiple ontologies
+- an RCE might be needed to help determine ontologies
+- object ontologies might have fuzziness
+
+OpenPrivacy thrives in a multiplicity of opinion
+
+prob[
+ opinion[
+ uri: U
+ val: V (wrt ontology O)
+ ]
+ pub: J1
+ sig: J1(hash(opinion))
+]
This archive was generated by hypermail 2b30 : Mon Jan 29 2001 - 12:37:42 PST