Transcript Research Information System for Materials

Research Information System for
Materials - Database, Simulation
and Knowledge
Toshihiro Ashino
Toyo University
([email protected])
Outline of Presentation
Role of Material Information System
Material Data Integration
Computer Simulation
Material knowledge on Semantic Web
Conclusion
Role of Material Research
Information System
Archive, Retrieve
Experimental Data
Fact DB,
Bibliography, Patent, etc.
STM, AFM, etc.
Information System
Understanding, Modeling
Data Visualization,
Data Mining, etc.
Computer Simulation
Band Calculation,
Molecular Dynamics, etc.
Material Data Integration
Standardized data representation is required for
Automated data acquisition with
computerized measurement equipments.
Development of Analysis, Visualization
tools.
Interoperability of material databases,
hyperlinks, data retrieval.
MatML (Materials Propertiy Data Markup Language)
MatML – Materials Propertiy Data
Markup Language
Material data exchange format based
on XML
http://www.matml.org/
By NIST (US), NIMS (Japan), etc.
Defined with DTD and XML Schema
<MatML_Doc>
<Material>
<BulkDetails>
<Name>1350</Name>
<Class>metal</Class>
<Subclass>
alluminum alloy
</Subclass>
<Specification>
ASTM B230
</Specification>
<Form>
Rolled rod and shapes
</Form>
....
More primitive than metadata..
Standard to display scientific data.
Standard to display and exchange formulae, equation.
- MathML can be solution?
Computer Simulation
Importance of Computer Simulation is increasing
Simulation of multi-scale (space and time)
and complexed phoenomena (multiple
models) is required to material design
Modularization of Simulation Codes
Integration with Databases, Visualization
or Modeling tools
Dependencies of Material Data
and Simulation Models
Virtual Laboratory for material
design
Modular Simuation Framework
Modularized Elemental Models (FEM, Dislocation
Dynamics, etc).
Exchange Data in XML format.
Module Description with RDF.
Module Integration with Scripting Language
Integrating Simulation and
Database
Material Simulation uses so many common basic data,
- e.g. Crystal Structure, Atomic Number etc.
Automated input data creation from databases.
Store simulated results in reusable and retrievable format.
Computational Combinatrial
Chemistory
There are over 5,000 binary
system entries in crystallographic
database.
Computational Combinatrial
Chemistory
Execute computer simulation for
all binary combination of elements
and make derived database from
results.
Automatic generation of input data
from database.
Computational Combinatrial
Chemistory
Electron band status suggests
some property of materials.
One of band calculation
program LMTO is not accurate
but light-weight. It will take few
months with 100PC's.
Continuous improving
process/mechanism for models,
data, programs - is required.
Material Knowledge handling on
Semantic Web
Semantic Web, the “Next generation Web”, to caputure
semantics.
Re-implementation of knowledge technology on the
Internet. - “Smart” search engine, “Semantic” link, etc.
Use XML Schema for data exchange.
Current focus is OWL (Web Ontology Language).
How to take advantage of Semantic Web technology to
manage material knowledge?
Material Thesaurus
ASM material thesaurus
includes over 6000 words
(concepts) and defines
upper-lower, relate-to relation
between them.
Thesaurus is a kind of
ontology which has much
restricted descriptive power.
Material Thesaurus into Ontology
Is there a way to create
OWL scelton from
thesaurus?
Multiple inheritance.
Destriction condition
description with relation.
Clear definition of Subclass
and Instantiation.
Concluding Remarks
There are so many models and data sources and they
are isolated.
Information system is expected to integrate, but it seems
not to be successful for material research unlike in
bioinformatics.
There are so many old data, knowledge and legacy
programs written in Fotran 77, etc..
Information technology like semantic web or
collaboration tools can help re-organize such resources?