The setting in network of information management devices, which can be either of small size (sensor - device) or of large size (cluster) is going along and accelerate the emergence of data processing in a ubiquity way. This essential trend leads to an explosion, of the quantity as well as the diversity of available information. The meaning of information is often ignored by processing operated in lower layers of information managing systems. In a context, where information sources become superabundant, the execution of the most elementary operation, bearing on most elementary informations, needs to be adapted to the meaning associated with the data In this document, developed topics, are the seek for meaning in texts and data management in large scale ubiquitous systems. This research has been done at the LIG laboratory in the SIGMA team, Joseph Fourier University Grenoble.
