Title: Complex Data Mining in a GeoSpatial Context

Organizers:

• Cyril de Runz: CReSTIC, Univ. Reims, France, Cet adresse mail est protégé contre les spambots. Vous avez d'activer le javascript pour la visualiser.
  

• Thomas Devogele: LI, Univ. Tours, France, Cet adresse mail est protégé contre les spambots. Vous avez d'activer le javascript pour la visualiser.

• Julien Perret: COGIT, IGN, France, Cet adresse mail est protégé contre les spambots. Vous avez d'activer le javascript pour la visualiser.

Url: http://sites.google.com/site/cdmgsc2012/

Description:

This workshop would be a framework for collaboration between researchers/ industrialists in order to share their practices in the field of complex data mining for geoapplications. For instance, complex data in a geospatial context may be non-structured data such as web data about location. In many fields, such as geodatabases, geographical web semantic, and of course in GIS, the amount of available data grows larger every day. Therefore, this often leads us to handle data (i) from multiple sources such as sensors, or web research and survey; (ii) representing the same information (location, territory, land) at different dates (periods of time); (iii) grouping information from different types (images, texts, coordinates, shapes,...) or of different natures (web, ontology, text mining,...), etc.
In the geospatial context, the complexity may concern the processes (data acquisition, data structuring, knowledge discovery) or the data theirselves (lack of information, uncertain data, serial data, etc.). Therefore, complex data mining in a geospatial context should structure complex data before extracting geopatterns. The consideration of the data mining step from the preprocessing to the utility of the extracted geopatterns is one of the workshop goals.

The main topics will include:
●Complex geodata preprocessing and structuring,
●Spatiotemporal data mining methods and processes,
●Clustering, learning and merging of multi-sources and distributed data in geoapplications,
●Experiment feedback from knowledge data discovery from complex data in a geospatial context,
●Uncertain spatial data mining ,
●etc.

By organizing this workshop, we wish to stimulate discussions/debates from theoretical works to more experimental ones.
Relevance of the workshop to AGILE: This workshop is in the continuity of the AGILE topics (spatio(temporal) data reasoning, analysis, mining, etc.) and possibly makes a bridge between them with a focus on spatio(temporal) data complexity (multi-sources, uncertainty, etc.). Therefore, it could be a good introduction to the AGILE main conference and may interest researchers from many communities (GIS, Computer science, Geography, etc.).

Length: Full-day

Planned outcome: Potential participants are invited to submit extended abstracts (around 1000 words) written in English and addressing some of the above mentioned topics. Abstracts will be reviewed and some will be selected for a paper presentation during the workshop.
Abstracts may present either a finished work, a work in progress, or a reasoning about complex data mining in a geospatial context. Discussions about work made across disciplines are also welcome. Contributions are welcomed both from senior researchers and PhD students.
The abstracts and the presentations will be published online.

Programme: (here)

Dates:

● Deadline extended abstracts (max. 1500 words): march, 2^nd, 2012
● Notification of acceptance: march, 12^th, 2012
● Workshop date: april, 24^th,. 2012

Submission:

● Article submissions : http://www.easychair.org/conferences/?conf=cdmgscagile2012
● Questions are sent to: Cet adresse mail est protégé contre les spambots. Vous avez d'activer le javascript pour la visualiser.
● Registration for the workshop through the AGILE conference registration system is
required.