Transcript Data management, cross analysis and database (WP5, 5.1, 5.2)

WP 5 Data management & analysis
• Michel Bohms and Philomena M. Bluyssen – TNO
• Isabella Annesi-Maesano - UMPC Paris 06
• Aileen Yang and Alena Bartonova – NILU
Objectives
• 1. To build a common database on IAQ and other
environmental parameters and related health effects
• 2. To facilitate the coordination of the different groups
that will use the experimental data obtained on both
school environmental characteristics and health
WP 5 Data management & analysis
• Task 1 Specification Database structures and
functionalities (months: 3-6)
• Task 2 Implementation Database structures,
content creation and functionalities (months: 7-14)
(old task 2 and 3)
• Task 3 Analysis (old task 4 Source analysis and 5
Epidemiological Analysis?) (months: 14-20)
• Task 6. Recommendation-design tool (with WP7)
(months: 18-22) transfer to WP7???
Type of data in database
-building and HVAC characteristics of the
investigated schools (including inventory of
possible sources, design criteria, maintenance and
occupancy);
- outdoor situation and outdoor air conditions;
- indoor air conditions (including content) and
ventilation situations;
- health and comfort of pupils and teachers;
- potential confounders of the relationships
between school environment and health.
Type of analysis
-Analysis of the data resulting in possible links
between exposure concentrations of air pollutants
and sources and activities (old task 4)
-Analysis of the data resulting in possible links
between health in children and teachers and
sources and activities (old task 5)
Task 1 Specification Database structures
and functionalities (months: 3-6)
• Data/parameters for structure (input): WP2, 3, 4
• Data/parameters for analysis (output): WP5 and WP6
• Functionalities for analysis: WP2, WP5, WP6
• Functionalities for other purposes: WP2, 8?
• What do we put in and what do we want to get out?
• How do we prefer to put it in and get it out?
WP 5 Who will do what? Person-months?
Co-leaders: TNO, UMPC Paris 06 and NILU?
Task 1: to provide data and functionalities for
structure from WP leaders or directly from
partners?
Task 2 Implementation database structures and
functionalities – TNO?
Task 3 Analysis – TNO, NILU, and ????
Relation with other WP’s? Will there be
different databases?
WP2 will be enhanced by an information web-based system
providing Internet access as well as a search function to:
-Data as available both from the reviewed studies, publications &
reports as well as current study later on as they become available.
WP 3.2 Task 5: Database on IAQ in European schools
The collected data in the main and the case studies will be reported
in a database, which will be easily accessible for the project
partners, in order to facilitate the application of the data in the next
WPs.
WP 3.3 European wide dataset for input in WP5 (cross-analyses)
WP8 web-site on IAQ
Perhaps two databases?
A database for the public
and
a database for SINPHONIE to make analysis.
Possible approaches data management
1. Semantic web approach and open source
implementation
• Standard way of saving and retrieving data on
Internet via a Web server, open to every one.
• User control is possible.
2. Closed database: assessed via login etc. (i.e.
oracle, SQL server)
3. Or something even more innovative: Google..
Next steps
• Partners: Who is involved?
• Task distribution:
• Who will supply which data?
• Who will do which analysis?
• Database:
• Which type of data and functionalities?
• What type of database(s)?
WP5 Meeting – Medio January – in Delft?
Semantic web approach and open source
implementation
• W3C Semantic Web technologies
• Web Ontology Language (OWL)
• On top of RDF/RDFS
• SPARQL Quering
• OSS Joseki Semantic Web (RDF) Server
• With Triple DataBase (TDB) back-end
Benefits of proposed approach
• 100% ‘Open Standards’-based
• For Data Storage & Access
• Independent of any particular platform, software vendor or
user group
• 100% Open Source Software (OSS) implementation (of those
open standards)
• 100% free of charge
•
•
•
•
•
•
Semantic (model-driven)
Flexible
Generic
International
Web-based (distributed, universal access, …)
Already successfully applied in other EU initiatives like IntUBE
An Example
Informal
• Sinphonie Ontology, small part as example:
• Sites (150x, 5 per country)
• Environment
• Outdoor Air Quality Measurements (18x)
• CO2 in ppm
• RelativeHumidity in %
• School
• ClassRooms (3x)
• RoomType
• Indoor Air Quality Measurments (18x)
• CO2 in ppm
• RelativeHumidity in %
• (Health/Well-being measurements)
An Example
Formalisation
XLS
Sinphonie
BIM/SIM/PIM
Workbook
OWL
Sinphonie
BIM/SIM/PIM
Ontology
(Excel-2-OWL Converter)
Semantic
Content *
OWL Configurator
* Joseki RDF Server/TDB
Example:
Excel workbook /1
Example:
Excel workbook /2
Example:
Excel workbook /3
Example:
Excel workbook /4
Example:
Excel workbook /5
Example:
Excel-2-owl mapping
Example:
Manual
Instantiation
with
PMO
Configurator
saved as:
Instance_School_1.owl
at:
www.bimtoolset.org/ontologies
Example:
Put in on the Joseki RDF Server (in TDB data store)
• SPARQL Update query:
• LOAD <http://www.bimtoolset.org/ontologies/Instance_School_1.owl>
Example:
Query the data /1
####################################################
# Get all triples
####################################################
SELECT *
{
GRAPH <http://www.bimtoolset.org/ontologies/IntUBE-EnergyBIM-Data>
{ ?x ?y ?z }
}
Example:
Query the data /2
Example:
Query the data /3
• ##########################################################
• ## Get all triples of the bim_data
• ##########################################################
• PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
• PREFIX sin: <http://www.bimtoolset.org/ontologies/Sinphonie.owl#>
•
•
•
•
SELECT *
{ GRAPH <http://www.bimtoolset.org/ontologies/IntUBE-EnergyBIM-Data>
{?x a sin:School}
}
Example: Sensor data
Query the data /4
SPARQLer Query Results
x
<http://www.bimtoolset.org/ontologies/Instance_School_1.owl#Inst
ance_School_1>
Example:
Query the data /5
• ################################################
• # Get the amount of rooms
• ################################################
• PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
• PREFIX sin: <http://www.bimtoolset.org/ontologies/Sinphonie.owl#>
•
•
•
•
SELECT count (?y)
WHERE
{
GRAPH <http://www.bimtoolset.org/ontologies/IntUBE-EnergyBIMData>
• {?y a sin:Room}
•}
Example:
Query the data /6
SPARQLer Query Results
.1
"4" ^^<http://www.w3.org/2001/XMLSchema#integer>
The four non-optional rooms for the one school instantiated
(without site context).
Final example
################################################
# Get all relative humidities per room, per school, per site
################################################
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX sin: <http://www.bimtoolset.org/ontologies/Sinphonie.owl#>
SELECT ?x ?y ?z ?z2
WHERE
{
GRAPH <http://www.bimtoolset.org/ontologies/IntUBE-EnergyBIM-Data>
{
OPTIONAL {?x a sin:Site}.
OPTIONAL {?y a sin:School}.
OPTIONAL {?z a sin:Room}.
OPTIONAL {?x sin:School ?y}.
OPTIONAL {?y sin:Room ?z}.
{?z sin:RelativeHumidity-Measurement ?z1}.
{?z1 sin:RelativeHumidity-value ?z2}.
}}
Final example
Query Result, now in JSON syntax output
{
"head": {
"vars": [ "x" , "y" , "z" , "z2" ]
},
"results": {
"bindings": [
{
"y": { "type": "uri" , "value":
"http://www.bimtoolset.org/ontologies/Instance_School_1.owl#Instance_School_1" } ,
"z": { "type": "uri" , "value":
"http://www.bimtoolset.org/ontologies/Instance_School_1.owl#Instance_Room_4" } ,
"z2": { "datatype": "xsd:float" , "type": "typed-literal" , "value": "90." }
} ] }}
Example:
More
• Much more beyond Sparql Select:
• Construct/Describe
• No “XML Result” results but RDF/XML code
• Sparql Update (beyond “load”)
• Insert new triples
• Jena ARQ beyond standard Sparql
• LET ()
• Various functionalities / “magic properties”
• Like Membership of SET
• Very powerfull ! Especially in combination with SPIN
• > OO Ontologies