Friday, January 27, 2006
[145] ß [parallel discussion on generative
methodology
Experience in Reasoning in OWL and the BioPAX activity
Communication from Andrea Splendiani (Interspersed with text from Prueitt) (footnotes by Prueitt)
Allow me to comment some point here... (Prueitt: I'll need a
little of interspersed text for this. Sorry, but you'll see this may be the
case for an exception). [1]
Paul (Prueitt) said (regarding the purpose of BioPAX
and others):
1) integrating massive relational data bases (SQL)
into a single sharable
information system based on the use of OWL Full.
To be precise, the goal is the development of an exchange format, that entails the ability to unify data (if enough semantics is added). [2]
However, it is a stated subgoal to provide an exchange format only.
Paul continued:
2) assisting natural scientists to conduct
investigations about the nature
and function of cell and gene expression.
This may be one of the usage of BioPAX. BioPAX is thought to
define pathways. It's up to the user whether
these are to be related with gene expression, other data, or not data at all. [3]
Paul continued:
In so many cases, what the biologist thinks she/he
is getting is not at all
what the computer will actually retrieve. BioPAX has gone a long ways in
determining what it is that the biologist wants. By using OWL Full in a
sometimes non-standard way, they are getting some of this.
*) We are addressing OWL-DL, not OWL-Full (unless I've lost something here...). [4]
*) I would not say that BioPAX has gone a long way in determining what biologist want. [5] It has gone a long way in defining a common conceptualization (allow me the term...) on pathway among the community of pathway information producers (largely) and users (less).
*) I'm assuming that in "what the biology thinks she/he
is getting" you don't reference to OWA/CWA issues. [6]
Paul continued:
But in each case, one gets retrieval of information
due to some specific
programmatic functions, not inference; where
inference sameAs human inference
This is for sure. [7]
Paul continued:
The point is that the type of deductive mechanism
developed in OWL Full (or
other "expressive" logic) is merely a complicated retrieval mechanism.
It is a representation system. A way to formalize a
conceptualization (allow me some term also here...).
Why do you say "retrieval system" ? [8]
Paul continued:
As Alan remarked, correctly, not long ago; the
purpose of BioPAX is to move
bioinformatics information from relational
databases, where the schema are
very brittle and inflexible, to XML information
bases with organizing schema
based on the RDF model. RDF is used to express class
- sub-class
hierarchies,
I'm not the expert here. But I would say class/subClass
concepts are in RDFS and OWL, non RDF. [9]
Paul continued:
and these serve in a classical "taxonomy"
fashion to organize
how data is stored and retrieved.
I would say this is a little bit reductive. Relations among
classes are inferred through their definitions. [10]
Paul continued:
RDF used in this fashion blocks off the
development of n-ary representations, such as what
Drs Krieg and Ballard
(and others) have developed:
This may be. But RDF is kept simple. Its first goal is to
annotate resources on the web. Not to build a "centralized" knowledge
base...[11]
Paul continued:
The RDF/OWL evolution is thus seen as a evolutionary
dead end, but full of
utility as a step beyond the relational (SQL) database. The more that is
invested in this RDF/OWL evolutionary path, the
harder it will become to get
back onto a pathway towards true human computer
interactions based on an
understanding of respective natures.
I don't really agree on this.
First, what you suggest seems to me something that could be developed "on-top" of an OWL/RDF like level.
Second, RDF/OWL are not only meant for a "database
perspective", but for a "web perspective". It may be not the
best for all representation styles... but there are tradeoffs. [12]
Andrea
[1] Prueitt: Of course, there
are many points to discussion.
[2] Prueitt: Why do we add the term “semantics”
here. What does this mean to you,
Andrea, and what does it mean to others?
Is there a communication of anything here, other than a communication of
a deep question as to what the meaning of the term might mean to you and to
others? IN most cases, the work becomes
a buzz word that does not have content.
In other cases, the word’s meaning varies from one leading expert to
another (John Sowa to Jim Hendler, for example)
[3] Prueitt: YES, and this
pathway ontology is generalizable to “social expression” and to “business
expression”. Right? (smiles)
[4] This would be my mistake, I
thought BioPAX was OWL Full… the differences are often defined in ways that I
do not follow, so I really do not know what the difference is. This does not mean that I could not
immediately understand if it were intelligently explained without using
misleading terms.
[5] Prueitt: Of course, but
there is an important leadership process.
[6] Prueitt: I do not know what
these acronyms stand for.
[7] Prueitt: So why use the
term “inference” when “retrieval” is more correct? (Rhetorically asked, of course.
I think that Andrea agrees with the point that mislabeling key concepts
in the Semantic Web literature is very damaging. )
[8] Prueitt: I say retrieval
because even though the conceptual schema is based on a conceptual
representation of an object of investigation, there is not the complete sense
of “human thinking” that the term “inference” brings to mind.
[9] Prueitt: Yes, absolutely
and this is a major point to think about.
[10] Prueitt: Ahhh, finally. This is where the language has lead us to make mistakes. The fact is that you do not agree with your
statement, (I conjecture). What you
meant was, “Relations among classes are traversed through their definitions”. The inference was prior to the knowledge
engineer defining the classes, properties, and attributes and is made in the
mind of the human. The OWL language
then allows the Knowledge engineer to construct a retrieval mechanism.
[11] Prueitt: There is nothing simple about what is
layered on top of RDF.
[12] Prueitt: Both of the last
two statements are open questions. I
will not disagree. The demonstration of
what can be built on top of OWL/RDF has not been outlined. I agree that in the same way that one can
transform SQL relational data schema into OWL…. That one can transform OWL into other things….
I support completely the working concepts of BioPAX, based on my understanding of the biological issues as well as my understanding of the individual motivation.