What? And, for what?

The sheer amount of numerical data, textual documents, images, video, Web sites, available today overwhelming, and cannot satisfy, per se, the emerging knowledge society, which we are entering in. It is indeed necessary to extract, from this wealth of information, the knowledge hidden inside. Only this ability could guarantee a better future to the individuals and the society, as well as a sustainable economical development and competitiveness.

To locate useful information, to transform it into actionable knowledge, and to manage its use for decision-making can be accomplished through the exploitation of methodologies and tools of Data Mining and Knowledge Management (DMKM). Notwithstanding the availability on the market and in academic environments of advanced solutions and systems, DMKM still calls for further research and developments to face new important challenges. In particular, some hot issues are still to be tackled, such as the following ones:

• To face the exponential increase of data it is not sufficient to rely on larger storing devices and/or faster computers. New intelligent approaches are to be designed to tame the very size of data.
• Data assume different modalities, such as numbers, texts, audio-video, sensor signals, and so on. Integrating into a unique system such complex data is still a challenge. Also, spatial distribution of data (for instance, on several Web sites or different data bases) is a source of difficulty for integration.

More importantly, there is still a "semantic gap" between the form in which the data are represented and used by a computer and their "meaning" for a human user. The new emerging techniques for the Semantic Web are trying to close this gap.

From the socio-economic point of view, DMKM is the future of the domain of the Business Intelligence (BI). The specific domain of BI (data warehouse, OLAP cube, data mining, knowledge management...) is still growing very fast in the economy. In economic terms, we can highlight that during the period 2006-2016, in the USA, the number of persons working in the domain of computing will go up from 542,000 to 742,000 corresponding to a growth rate of 37%.

In EU, the first EU forecast of skill needs is published and shows the same trends. In France, the number of persons working in computing will grow from 200,000 (1985) to 600,0000 (2015).

In the specific domain of BI:

Gartner estimates that by the end of 2009, 90% of Global 2000 companies will have mission-critical BI and data warehouse systems – up from less than 25 percent in 2007. Current trends reveal emerging shift to information management.

The Cambridge, Mass.-based research firm expects the BI software market to reach nearly $13 billion in revenue by 2014, up from $8.5 billion last year (2008).

Beyond these economic, scientific and technological challenges, the reasons that have motivated the members of the consortium to propose this Erasmus Mundus Master's degree are based on multiple observations:

• There is a need for new skills and knowledge: From an academic point of view, Data Mining and Knowledge Management is part of the field of "Discovery Informatics". This is considered by one of the two main international computer science associations (the Association for Computer Machinery, ACM) as the major new field in computer science. It aims to develop knowledge and skills in the field of "Business Intelligence", which is currently in a boom period in companies, as seen in the wide range of publications by the main analysis groups for economic conjuncture. Specialists in Data Mining and Knowledge Management are being recruited in all economic sectors: banking, insurance, health, industry and education…
• The absence of specialized and integrated training: Most European higher education computer science and applied mathematics programmes include certain teaching modules connected with Data Mining and/or Knowledge Management. Reflections and studies regarding teaching and practicing Data Mining and Knowledge Management have been carried out. It has thus been shown that at present there is no common effort coordinated at the European levels for developing this technology in education, research and industry. The challenges and stakes associated with this field, however, require means and skills that have to be federated at the European level. For the purpose of this project, we have reviewed almost all the programmes offered throughout the universities worldwide.
• Pedagogical know-how of the partners: Three members of the consortium have been co-hosting a Master's level course for the last 9 years (http://dea-ecd.univ-lyon2.fr) in the field of Knowledge Extraction from Data (Extraction des Connaissances à partir des Données ECD). This Master has experience of organizing courses in multiple locations as it operates in three countries: France (Lyon, 1999 – Nantes, 1999 – Paris, 1999), Romania (Bucharest, 2002) and Italy (Vercelli, 2006). This Master program is integrated in a graduate program of the University of Canto (Vietnam) since 2007. Moreover, professors from the other partners of the Consortium (UPC, UPMO) had participated as professors in it. From 25 to 35 international students are regularly enrolled each year in the master of ECD.
• Scientific expertise: The six research teams serving as the basis for the consortium have expertise that is recognized all over the world in the fields of Data Mining and Knowledge Management. They participate in international networks of excellence, and collaborate in research via common projects and/or co-direction of doctoral theses.

These four observations have resulted in the project for the consortium's Erasmus Mundus Master taking shape in order to construct a specialized offer in Data Mining and Knowledge Management.