Australian Researchers’ Mobility Portal

Wednesday 12th September 2007

“The adventure of working on an EU-funded project completely changed my professional life,” says Professor Jean-François Boulicaut from the National Institute for Applied Sciences (INSA) in Lyon, France.

“If I hadn’t had the idea to set up the project, with very good partners, then I could not have developed the theme and I would not be where I am today, doing what I am doing,” he said.

The theme in question is data mining. Where he is today is a well-respected member of the international data mining community.

It was on his return from a sabbatical year, working with one of the best teams in the world at the University of Helsinki in Finland, that Professor Boulicaut just knew he wanted to pursue his new found interest in the subject.

In 1998, the Professor decided to submit a project proposal under the Future and Emerging Technologies (FET) section of the Information Society Technologies (IST) strand of the EU’s Fifth Framework Programme (FP5). The “Consortium on discovering knowledge with inductive queries” (cInQ) project gave him the opportunity to rub shoulders with Professor Mannila, one of the pioneers in the field of data mining: 'the science of extracting useful information from large data sets or databases'.

Between 2001 and 2004, Professor Boulicaut led the project, working on the concept of inductive queries, which use data mining as a sophisticated tool for database queries to identify meaningful data or patterns. Although the science is often used by business intelligence organisations and financial analysts, the Professor decided early on that he wanted to apply data mining algorithms to the plethora of biomedical data, and especially gene expression data.

As the cInQ project progressed, the researchers became pioneers in developing algorithms that can quickly calculate actionable patterns from huge databases.

His second and current project, IQ, funded under the IST/FET part of the Sixth Framework Programme (FP6), began in 2005 with the goal of developing further the framework of inductive querying that will enable the development of effective inductive database systems. Targeted applications in bioinformatics have been selected to demonstrate the applicability of the algorithmic results.

Working on two EU-funded projects has proven to be a career enhancing, if not always an easy experience for the researcher. “On the one hand, I have had the great opportunity of cooperating with the best teams in the field of data mining in Europe, which has tremendous value in my eyes. On the other hand, I would say that the administrative obligations of setting up an EU-funded project are much too heavy and time consuming,” Professor Boulicaut says, adding that the situation is however improving.

Professor Boulicaut compares the administrative burden of an EU-funded project with a project funded by the French National Agency for Research (ANR), and claims that there is no “common measure” in terms of submitting a project, the obligatory follow-up and the sums of money involved. “This puts a serious brake on researchers who want to set up an EU-funded project,” he believes.

However, according to the Professor, processes are improving as more and more support and help are being offered to researchers to guide them as they submit a project proposal. He also points to the added value of working in a network of great researchers and the international visibility of EU-funded research projects as reasons for getting involved in a second European project.

“It was my first European project that allowed me to undertake quality research, helped by my partners in the project, who were already well known and established in the data mining world,” says Professor Boulicaut. “Thanks to this first experience, I immediately became quite well known in the field of data mining at the European level and the project emerged very quickly onto the European scene.”

Towards future generations of researchers interested in working on EU-funded projects, Professor Boulicaut is nothing but encouraging. If it changed his professional life for the better, then it is safe to say it could change any young, up and coming researcher’s life, he says. If the proof of the pudding is in the eating, the fact that Professor Boulicaut is about to move on to his third helping is telling. He is currently working on a proposal related to data mining for systems biology.

The researcher would like to see the EU doing more to support research into data mining, which he believes to be a hugely important field, now beyond being defined as a “future emerging technology”. “There is a colossal effort at the European level to produce data, but that effort is not proportionate to the exploitation of that data,” Professor Boulicaut says.

And yet this is precisely where data mining is of most relevance, according to the computer scientist, “for behind data mining lies the idea of bringing out potentially interesting hypotheses when we are not quite sure of what exactly we want to find from our mass of data”. The challenge is indeed to discover knowledge from data.

Professor Boulicaut also links his perceived lack of support for data mining research with a general malaise over the direction of research, and basic research especially, primarily in France, but also in Europe at large.

“The problem as I see it is that public-private partnerships in research are supplanting academic research, which has an unfortunate knock-on effect on both basic research and the economy of the continent in the face of globalisation,” the Professor says. “I really think we should be supporting quality research at the academic level and also defending the idea of basic research where scientists are not asked to provide an application every three months, but where we are given the licence to work on something which could be transferred towards a start-up or a company later, without any strong hypothesis about how long it should take.”

Some instruments exist for supporting basic research, including the EU’s European Research Council, established in 2007. But more and more funding schemes are targeted towards public-private research, believes Professor Boulicaut. This situation could lead to a brain drain that could irrevocably affect Europe’s economy in the space of five to 10 years. Instead, the researcher would like to see support for quality basic research at the academic level, and suggests that those receiving research funding should not have to fit into one of the following two categories: public basic research carried out by potential Nobel Prize winning scientists, or applied research carried out by private sector companies.

“I would like to see a kind of intermediary vision where not only the handful of potential Nobel Prize winners receive funding under the terms of excellent research,” says Professor Boulicaut, “but also the whole layer of researchers who do long-term quality basic research without private partners on important topics for their country, the European continent and industry.”

However, Professor Boulicaut, who himself is active in long-term research, has never failed to win EU funding for his projects. In fact, his experiences have turned him into a pro-European with a firm belief that his fellow European researchers can achieve great results, and not just in the field of data mining.