minutes of f2f meeting - session 3

Here comes my minutes from session 3 of face 2 face meeting in Cannes

Present: Marc Schröder, Björn Schuller, Alan Giboin (observer), Felix
Burkhardt, Enrico Zovato, Myriam Lamolle

/***** Session 3: plugin of custom vocabulary *****/

Topic 1: QNAMEs
Enrico presents an example to specify a set of values of attribute using
QNAMEs
(voir PLS 1.0 for role attribute)
example :
    <xsd:...
        xmlns:classws="htttp://www.example3.org/classws" />
    <...     role="classws:VVI classws:VVO ... />
where role is a list of attribute values prefixed by a namespace.

Pros : QNAME is define in an unique namespace => no clashes!
    (Felix says the prefix colons are present every time with qualified
name.)

Cons : Myriam says it's very difficult to define a dynamic set of values for
an attribute directly.
One possibility is to declare fixed attribute with an ID in an XSD file and
the XML file references a set of known Ids.
    ex: <xsd:attribute ID="VV0" value="VV0" type ="..." fixed="true"/>
    <xsd:attribute ID="VV1" value="VV1" type ="..." fixed="true"/>
    ...
    <xsd:complexType name="...">
       ...
       <xsd:attribute name="role" type="IDREFS" />
    </xsd:complexType>
and it's impossible to define a set of tags (without type) for a concept
(<xsd:complexType...) and to process a dynamic binding to type these tags
(like Java).

Marc would know how to be sure that XML Input is validated or not...

Remark out of this scope : Björn suggests different scales (unipolar,
bipolar, discrete or not) and the possiblity to custom values for dimensions
and for action-tendencies by the range or discrete values because now it's
just unipolar... abstract description scale?



Topic 2: plugin througth Xproc language (http://www.w3.org/TR/xproc/)
Marc presents a suggestion from Liam Quin (XML Activity Lead at W3C) the
possibilities to use Xproc language (based on XSLT) to dynamically create a
new XSD file (or another Schema, e.g. RelaxNG or Schematron) to validate
EMOML documents.
Principle:
   1) An EmotionML document points to both a generic Schema and a (or
several) custom vocabulary set(-s);
   2) a generic Schema uses the typical format of XML Schema (or Schematron,
or RelaxNG...);
   3) the custom vocabulary could be defined in an XML format to suit our
needs;
   4) a custom EMOML is created by an Xproc script by following the steps
below:
    a) an XProc script would be executed;
    b) the XProc script would look up in the EmotionML document and retrieve
both the generic Schema and the custom vocabulary;
    c) through a suitable merging mechanism (which could be done, e.g., in
XSLT), the generic Schema and the custom vocabulary would be merged into a
custom Schema;

For the purpose of validating a document, the following step would be
performed:
   5) the custom Schema is used in the standard way to validate the
document.

(N.B.: XProc is an XML Pipeline Language, for describing operations to be
performed on XML documents.
An XML Pipeline specifies a sequence of operations to be performed on zero
or more XML documents. Pipelines generally accept zero or more XML documents
as input and produce zero or more XML documents as output. Pipelines are
made up of simple steps which perform atomic operations on XML documents and
constructs similar to conditionals, iteration, and exception handlers which
control which steps are executed.)

Myriam will read the Xproc specifications because it seems very interesting
to create dynamically a new custom EMOML schema to validate a XML document.
On an other hand, Xproc script will be made; we should write all the mapping
rules between generic Schema and each custom vocabulary.

/************ end of minutes ************/


-- 
IUT de Montreuil - Département Informatique
140, rue de la Nouvelle France
93100 Montreuil

tél: +33 (0)1 48 70 34 61