Abstract
The biomedical domain uses more than 1000 online databases. Crossing and analyzing their content is hard, and users often report the following kind of comment: “There are frequently about 20 windows on my desktop; so I'm lost”. Data crossing and analysis is also mostly done by manually copying and pasting data into a spreadsheet. Data integration community advances do not concretely address the user's needs for better visualization and integration tools. This thesis proposes a biological knowledge mapping environment that simplifies integration and visualization of biological data. I²DEE (an Integrated and Interactive Data Exploration Environment) is based upon a simple graph metamodel. This metamodel confers I²DEE flexibility and extensibility for interoperating with existing data integration approaches. I²DEE provides a visual and homogeneous biological data access and can adapt to specialized user tasks. To demonstrate I²DEE versatility, two applications have been experimented in the context of microarray gene expression data analysis and knowledge engineering. Keywords: Biological data integration, visualization, knowledge engineering, knowledge mapping.