Querying and extracting heterogeneous graphs from entreprise data

Nowadays, a huge volume of data is collected and stored daily in enterprizes. To efficiently extract and manage this valuable knowledge helpful in the decision-making process is a hard task. Firstly, because the data can be stored in different ways: in a very structured way like in databases but also in unstructured repositories. Moreover the description of a sole object, person or process can be disseminated in several sources with several structures. Managing these different data models is difficult and makes the extraction of information process not very efficient. Classical query techniques permit to retrieve the set of data which match with specific criteria under a specific model, but richer results could be provided if a unified representation of the disparate data of the enterprise and of their interactions and links could be defined. Graphs can be used for this unified data model, and facilitate the information search as well as the extraction of objects interaction.