SemEnglish Primer
Created 2001/1/26 last revised 2001/3/25
pronounced `sem.Eng.lish
Inspired by TimBL's N3 see http://www.w3.org/2000/10/swap/Primer.html
OBJECTIVE: Semenglish is all about bringing to the foreground
the natural language that we all know, love, and understand; and at the
same time not sacrificing the precision that machine agents require one
bit. It is designed to be not only the input from humans to semantic
memory, but also the output from semantic memory to humans. It
will always be possible to unambiguously translate Semenglish to RDF/XML
and visa versa.
-
Every token is either
-
a literal is enclosed in " ..." or in '....'
-
a number
-
a member of the set of punctuation
-
rule: a token cannot end in a punctuation
-
Otherwise it is: a ProperUri of which there are two types
-
fully qualified <NamespacePrefix>:<RelativeFragment>
-
qualified by the namespace defined in the dictionary {see fromDictionary}
-
Phraseology:
-
subject verb object - triple (at top level only?)
-
[verb object] - anomous triple (unnamed with a uri)
-
[subject verb object] - reified triple (subordinate level only?)
-
[statementIdentifier subject verb object] - explicitidly reified triple:
a specific triple in a specific context of utterance
-
All statements build triples for the labeled directed graphs of the RDF
data model.
-
All tokens are either URI fragment terms, URIs or littorals
-
Littorals are either strings enclosed in "..." or numbers
-
a period terminates sentences
-
semicolons mean that subject is to be repeated from prior clause
-
commas mean that subject and verb are to be repeated from prior clause
-
nesting of any subject, verb, or object can be done with [ ...] and express
either anonymous nodes or reified reified statements
-
anonymous nodes have [verb object]
-
reified statements have [subject verb object]
-
{ .... } are statement containers
-
(natural language string) gives the rdfs:label for the anonymous node of
the subject, verb, or object
-
the context of utterance may inherit attributes from the containing XML
element. For example:
-
swag:fromDictionary - the uri of a coherent dictionary containing no homonyms
- defaults to swag playground
-
swag:language - taken as "SemEnglish"
-
swag:context - the uri of a context containing these statements
-
swag:intent - the intent of the author of the utterance
-
A token is that contains a ':' and no embedded spaces is expanded when
translated to RDF
-
the part to the left of the colon becomes the Prefix
-
the prefix is expanded according to {Prefix nsUri 'http://...whatever....#'}
-
in semenglish a uri scheme is just a namespace prefix
00:21:26 <Seth> the namespace uri that is associated with the uri
schema is just itself
00:22:43 <Seth> rdfs nsPrefix http://www.w3.org/2000/01/rdf-schema#
00:22:55 <Seth> http nsPrefix http
00:25:00 <Seth> urn nsPrefix urn
Known problems
-
matters marked with "?"
-
(How to distinguish between explicit and implicit reification ?)
hasBeenFixedBy (Phraseology section above)
-
How to designate questions and procedures? integrate ! and ? in the
language -
Example:
<signature
swag:language="SemEnglish"
swag:talkingIn="http://webns.net/roughterms/"
swag:intent="inform"
swag:context="ThisEmail">
Seth
properName "Seth Russell";
mbox "mailto:seth@robustai.net";
workingOn http://RobustAI.net/MyNetwork/index.html;
workingOn http://robustAI.net/MyNetwork/StickeyCyberMolecules.html
;
wrote http://robustai.net/ai/symknow.htm;
conjectures http://RobustAi.net/Ai/Conjecture.htm;
affiliatedWith http://purl.org/swag/ ;
affiliatedWith http://speaktomecatalog.com ;
lookingFor [Who:Person#x willProgram Sembrowser].
</signature> |