Communications of the ACM
How to Support Research in Computer Science via Russian Scientific Foundations
It is well known that the disciplinary structure of Russian science is highly heterogeneous. Along with the research areas that are traditionally developed in Russia, there are also directions that are frankly failing. The fields of knowledge, the development of which traditionally exceeds the average world level, include physics and astronomy, earth and the planetary sciences, as well as chemistry and mathematics. Successes in these fields of knowledge are due to the inheritance from the Soviet Union. These areas of research were on a special account and given priority attention and funding. Over the decades, scientific schools have been formed in these areas whose activities still bring strong results. In addition, support for these fields is provided by the structure of secondary and higher education. These areas are traditionally distinguished by strong mass training due to the excellent methodological support.
At the same time, a number of new fields of science, which define the advanced edge of research, in Russia are lagging far behind the world average. Such areas of knowledge include medicine, agricultural sciences, biotechnology, pharmacology and social sciences. Computer science can also be included in this list.
However, compared with medicine and pharmacology, the computer industry has significant differences. Russia has its own powerful computer industry, connected both with software development and the provision of real services to the state, businesses, and the public. Companies such as Yandex, Kaspersky Lab, Security Code, Playrix, and many others have multi-million-dollar turnovers, and these companies cost billions of dollars. The situation has improved even more with the provision of modern high-speed telecommunications services to the public. Russia is among the leaders in terms of the share of the population whose households are connected to the Internet via fiber optic technologies, as well as having the ability to use 4G LTE. In addition, the cost of such services is one of the lowest in the world, which indicates the highest competition. However, in the field of medical services, things are far from being rosy, as evidenced by the nation's low average life expectancy.
All the factors for the successful development of computer science are obvious. However, the dynamics of publication activity, citation, number of patents, and sales of intellectual property do not at all indicate any significant development in research activity. In this research, an attempt will be made to understand this state of affairs and to propose a number of measures to overcome the situation.
In Russia, almost all research funding is provided by the state. Financial support is carried out both directly, in the form of state assignments to universities and institutes of the Russian Academy of Sciences (RAS), and through research funding. The state assignment differs for universities and institutes of RAS, for which the government assignment is their main budget item, covering all expenditures, as well as maintenance of the premises and utilities, payments for administration and support staff, equipment purchasing, etc. For universities, it is completely competitive and involves carrying out initiative projects. For institutes of RAS, it requires the fulfillment of some indicators of the effectiveness of research. Scientific funding is almost entirely provided by the state, and there is almost no private funding. Two large groups can be distinguished among the scientific foundations: the first aims to finance basic and exploratory research, while the second supports applied development.
The first group should include the two largest funds, the Russian Foundation for Basic Research (RFBR) and the Russian Science Foundation (RSF). The RFBR is older, having been established in 1992. Until recently, it mainly provided support to initiative projects of individual researchers or small groups; up to 60% of its funding was devoted to such projects. Research topics were proposed by applicants, and the small size of its grants made it possible to finance many projects and make its research support quite massive, as every second or third application was supported. At present, the support of initiative projects has been greatly reduced, and if part of the funding was transferred to support youth initiative projects, most of the funds went to thematic projects. Moreover, the themes of such projects are determined by the fund's management behind the scenes, without wide discussion.
When founding the RSF in 2013, constituents were said to support large, ambitious projects. Since then, however, the dollar has changed in value, while the amount of research support provided in rubles has remained the same, about $100,000 a year for a group of 10 people. This is the largest grant of about 600 dispersed each year, to all branches of science. There is also support for laboratories with up to 30 people and a total of $500,000 a year. Grants typically are allocated for three years in the RFBR, although there are also two-year grants.
The second type of funding support for applied research and technology deployment is based on previously obtained results. There are several support channels, but they all involve co-financing from independent companies. At least 20% of co-financing is required by the federal targeted programs, which involve carrying out applied research and copyright registration of their results. Moreover, a private company involved in co-financing will be in charge of the resulting intellectual property. Support for applied projects is also provided by the Advanced Research Foundation (Russian Foundation for Advanced Research Projects), which is the Russian equivalent of DARPA (the U.S. Defense Advanced Research Projects Agency). The Advanced Research Foundation exclusively supports technology to enhance defenses.
The Innovation Promotion Foundation (the Foundation for Assistance to Small Innovative Enterprises in the Scientific and Technical Sphere) aims to create small enterprises that manufacture products based on the prototypes created. This foundation often collaborates with the Skolkovo Innovation Center, where new enterprises can be registered and receive substantial tax breaks. These funds also help to find consumers for products and develop sales technologies.
It would seem that there are enough support channels; nevertheless, Russia's success in the development of high technologies leaves much to be desired. If there are individual successes, then they are in no way connected with the activity of state funds. The reason for the low efficiency of the above funds is quite banal and associated with low-quality expertise.
Moreover, the situation of each of the funds is individual, but generally not favourable for computer science. The difficult situation in the RFBR is explained by the fact that the composition of the expert councils has ceased to be updated, and a stable group has been formed that makes decisions. At the same time, ordinary experts are loaded unevenly. Most experts do not have access to publications in the first quartile of journals and presentations at leading IEEE and ACM conferences. The citation of expert papers is often limited in Russia, as they have no international recognition. Therefore, experts do not understand what an international level is and how difficult it is to pass an examination for publications in journals with a high impact factor.
Note that in Russia, there is no computer science journal that would be included not only in the first, but also in the second quartile of Web of Science. Only two journals are included in the second quartile. That is why reviewers begin their reviews with the words, "Expected results may be of some value and be published in the proceedings of international conferences and even in prestigious international journals. However, I do not think that they will correspond to the level of world achievements in this field."
The scientific foundation RSF is much stricter regarding reviewers. It has approved an ethics clause for experts, and the principle of rotation applies. Nevertheless, the support of computer science from this fund is minimal. The reason for this is that computer science is in a single block with mathematics, mechanics, and control theory. Funding is allocated to the blocks in proportion to the number of applications submitted, but the proportions within the block are not respected. Traditionally, representatives of computer science submit a significant number of applications, but all of them are redistributed by the joint expert council in favor of mathematics. Experts without a solid track record in the field of computer science are involved in the review.
To support projects that can serve as the basis for new technologies, a new type of research has been specifically introduced in Russia. This is so-called exploratory research, the task of which is to discover new principles for creating ideas and technologies. It is fundamentally impossible to formulate precisely the scientific novelty of such studies; it is possible to designate only the direction of the search and describe the patterns that should be sought. It is impossible to approach search research with fundamental measures, as this immediately dooms the research to failure. Moreover, to formulate the scientific novelty before obtaining research results fundamentally cannot be done. Unfortunately, current RSF experts simply do not understand the specifics of exploratory research, which is why I personally do not know of any solid, practical application of results obtained with RSF funding in the field of computer science.
Similarly, it is difficult to find experts to evaluate applied and prototype projects. It should be noted that, according to Russian legislation, the RAS carries out expert functions in relation to all projects financed from the budget. At the same time, computer science is one of the few industries where universities have better results than the RAS.
Yet objectively, neither the academy nor the universities will be able to carry out a large project, such as creating their own operating system or developing a new generation of telecommunications equipment. The Russian Federation is currently focusing on the implementation of big scientific and technical projects, which is why the development of computer science requires a separate development program.
In my opinion, the basis of such a program could be the transfer of expert functions to specialists from successful Russian IT companies. All projects supporting computer science should be consistent with these companies. For the initial stage at which the development of principles of interaction between companies, research groups, and government organizations could take place, it is necessary to choose a small pilot project.
Such a pilot project could involve creating R&D departments of large IT companies based on leading Russian universities in the field of computer science. The leading universities should include those that are included in the rankings of the leading agencies, Times Higher Education (THE) and Quacquarelli Symonds (QS). This amounts to fewer than 20 universities.
The criterion for inclusion in these rankings of a large company is its registration in Russia and the number of employees working it, as well as the company's turnover. For example, a company should employ more than 300 employees, and its annual turnover should be more than $10 million. As is known, many Russian companies cut costs on R&D departments; therefore, the financing of such departments from scientific funds could serve as an excellent experience for IT companies.
The optimal organizational form could be a state assignment for Russian universities, or a special competition within the framework of a federal targeted program. Such a project could include the creation of 20 R&D units in IT companies based on leading universities. The research program of these units should be formed by IT companies, and they should receive all the intellectual property created in the course of such a project. The number of employees should not exceed 10 people, and at least half of them should be masters and postgraduate students. In this case, funding should not exceed $120,000 per year for one project. The project duration is three years. Thus, the total value of all projects will not exceed $7 million, but the results obtained may change the structure of the IT industry.
Guest blogger Andrei Sukhov is a professor in the Department of Supercomputers and General Informatics, and in the Department of Information Systems and Technologies, of Samara University, Russia.
No entries found