INTERVIEW WITH MARCO SCHORLEMMER
Staff Scientist at the Artificial Intelligence Research Institute (IIIA-CSIC). Doctor of Informatics by the Polytechnic University of Catalonia (UPC), he carried out his research in the IIIA-CSIC and the UPC at the national level. At the international level, he had research stays in the U.S.A. and the U.K. His research is focused on computational concept systems, i.e. the mathematical, cognitive and computational bases of the concept systems through which we understand and describe the world in which we live and act. This led him to contribute in different areas of computing sciences, such as automated theorem proving, knowledge representation, multi-agent systems and computational creativity.
What is the reason for your visit? Do you come to Brussels often?
I came for the final revision of the European project COINVENT (www.coinvent-project.eu), of which I was the scientific coordinator. We gathered with the rest of the members of the committee the day before at the CSIC Office to prepare the revision.
In the last few years, I have travelled to Brussels perhaps once or twice per year. I had to come for the negotiation and revision of the COINVENT project, but also for the revision of ESSENCE, the Marie Skłodowska Curie Innovative Training Network in which we also participate. Furthermore, one of the partners of ESSENCE is the Free University of Brussels, which is why we have also held some meetings in their facilities.
For the general public, could you explain to us the goals of the project that you have coordinated: “COINVENT: Concept Invention Theory”, funded by the 7th Framework Programme?
In COINVENT we have studied the mathematical models of those general cognitive principals that operate when we create new concepts, so that in the future we can implement computational systems that help humans in tasks that require creative thinking.
The project is focused on the mathematical formalisation and computational implementation of “conceptual blending”, which is a fundamental and intrinsic cognitive operation of ordinary thinking and language. Through conceptual blending, humans combine certain elements and their relationships of different mental spaces into a single mental space, in which new elements and relationships emerge, where also new conclusions can be drawn.
COINVENT has gathered a team of scientists from the top universities and research centres of Europe in the fields of Formal Systems, Cognitive Science, Artificial Intelligence, Computational Creativity, Mathematical Reasoning and Cognitive Musicology, with the aim of developing a creative computational system that has been validated in two representative areas of creativity: mathematics and music.
In your opinion, what is the greatest challenge for a coordinator of European projects?
This has been the first European project I have coordinated, and, for me, the challenge was to be capable of maintaining a global perspective of all the research lines that were tackled in the framework of the project, without losing attention to detail, in order to foster the integration of the work performed by the different partners and achieve synergies that produced a greater collaborative work.
In this sense, I am very satisfied with the work that has been done in the COINVENT project, and after these three years of project not only we have met our scientific objectives, but we have also delved into the collaborations between the institutions that have participated in the project.
Once the difficulties have been overcome, what advice would you give to someone who is going to coordinate a project like this for the first time?
The success of a project resides more in the human quality of the people that lead the teams of each participating institution than in the excellence that these institutions may seem to have. And rather than the popularity of the research topic, it is very important to have a good understanding between the people that work together in the committee, since they must keep an open mind and good communication skills. Creating a committee of great human quality is what I would advise to every future coordinator.
You are the Principal Investigator of a Marie Curie Initial Training Network: ESSENCE (Evolution of Shared Semantics in Computational Environments), also funded by the 7th Framework Programme. What is your role in this project?
The Artificial Intelligence Research Institute (IIIA-CSIC) is one of the main academic partners of this training network. Within the framework of ESSENCE, we trained two PhD students and a young postdoctoral researcher. We lead the work package related to reasoning and knowledge representation techniques; we also manage the communication within the network, and monitor the tasks for the dissemination of their results.
If the design of the next Framework Programme was in your hands tomorrow, what would you add or remove, at both the administrative and scientific levels?
At the scientific level, I have the feeling that the working programs are increasingly fostering a type of research directed towards results that can be applied immediately, to solve very specific problems, and whose usefulness and impact can be directly appreciated. This is a detriment to fundamental research, which does not pursue any particular applicability, only advance in knowledge. In my research area, projects of fundamental research like COINVENT have more and more difficulties to be funded.
At the administrative level, I would further simplify the elaboration and follow-up of research projects. Nowadays, for instance, researchers are asked to define specific tasks and their results with a deadline, but research is dynamic and open, and you don´t know where it will take you. During the preparation of the documents for the final revision, we had to specify with details how many hours we have worked in each work package. How important is this to guarantee the quality of the research conducted? In my opinion, this administrative requirement, just like many others, is totally unnecessary, and it only consumes time and energy that could be invested in the research itself.
Lastly, what have you learned about creativity in the last few years?
I´m not sure if I´ve learned something about creativity, but I´ve learned this about computational creativity: we must adjust computational systems to the cognitive principles that we, humans, use when we think and act in a creative manner. This is the only way we will be able to develop computational systems that really help creativity. After this project, I am less interested in creative machines and more interested in how to make humans and machines create teams that foster creativity.