In conditions of an abundance of information, and scientific information in particular, structured knowledge bases are of particular importance, which nowadays are even more important than scientific journals and collections. For specialists in nuclear physics and energy, such a database has existed for about half a century – it is an international nuclear information System created under the auspices of the IAEA – The International Nuclear Information System (INIS). We are talking with Yuri Volkov, an INIS Liaison Officer from the Russian Federation, Associate Professor of the Department of Theoretical and Experimental Physics of Nuclear Reactors at the National Research Nuclear University MEPhI, about how it works, what importance it has for the development of science, and how Russia participates in its work.
The interview was originally published in the «Kommersant» edition
Yuri Volkov. Photo: A. Khomich
– So, what is INIS?
– This is an interesting international initiative. Actually, the main task of the IAEA is to form a collegial opinion on the peaceful development of atomic energy. And in 1970, an initiative was put forward by the USSR and the USA that those publications related to the peaceful uses of atomic energy should be open and accessible. INIS was established in 1970 in accordance with the IAEA's mandate to "facilitate the exchange of scientific and technical information on the peaceful uses of atomic energy." From that moment on, the history of INIS began, and an international nuclear information system appeared, which collects all kinds of publications – but, of course, only open ones.
– Are we talking about publications in scientific journals?
– First of all, these are articles in peer–reviewed journals, but in addition, in the framework of INIS there is such a thing as "gray literature", which includes various reports, conference materials, up to biographies of scientists, in general, everything related to the topics of nuclear science and technology, and this is a fairly broad concept, starting from nuclear reactor physics, particle physics, and ending with geology, uranium mining, and nuclear waste disposal, this is nuclear science in the broadest sense of the word.
– What is the mechanism for replenishing this database?
– First of all, there is the INIS central secretariat at the IAEA headquarters in Vienna, which provides approximately 70% of the revenue. He takes publications from well-known publishers such as Elsivier, Springer, and so on, and enters them into the nuclear information database. Some of the information about it in INIS is contained in metadata format, meaning you can view the output of this publication, but to read the publication itself, you must go to the publisher's website.
– So only the link remains in the database?
– Not only that. It is important that this publication will be indexed in INIS, and an important advantage of the accepted indexing system is that it uses the so-called nuclear thesaurus. A thesaurus is not just a collection of terms, because there are still connections between the terms. For example, a connection can be built: reactor – nuclear fuel – uranium – uranium-235, and so on. The thesaurus has a total of 32,000 terms, and it has been translated into eight languages. The indexing process consists in assigning keywords to each publication according to this thesaurus. And thanks to this, we know that this publication accurately relates to the topic that corresponds to the term from the thesaurus. And that's where the big value lies. In addition to the central Secretariat, there are national INIS offices in the IAEA member countries, which collect information from countries that are not available to open sources, and above all publications in the language of the country concerned - in our case, Russian. And besides, there are individual organizations that upload their information to the database, for example, in Russia it is the Joint Institute for Nuclear Research in Dubna, which has its own liaison officer with INIS. Well, let's say CERN is one of these organizations.
– In Russia, who is the main data provider?
– Previously, the Central Research Institute of Atomininform was engaged in this, and after its reorganization in 2010, the INIS Center at the National Research Nuclear University MEPhI became involved in this. I work there and have been appointed an INIS liaison officer, colloquially an INIS liaison officer. For example, I participate in meetings, and various recommendations are broadcast through me. But, in fact, our main management is carried out in Russia by the Rosatom State Corporation.
– Well, what if we talk not about who downloads, but about what are the "main points" where the nuclear knowledge itself is born, which is uploaded into the system?
– The INIS secretariat does not set any priorities for us. In other words, countries have obligations to upload a certain minimum number of publications to the database, no matter in which areas. Materials from industry and academic conferences are mainly expected from us. We also upload magazines, for example, all issues of the MEPhI Bulletin. But mostly, of course, these are conference materials, because they are often not even published, but abstracts are collected, and they exist only as PDF files, this is exactly what is called "gray" literature.
– In fact, is your task to upload information that cannot be collected simply from open scientific journals?
– Yes, the secretariat collects journal publications independently. Although, if we talk about Russian magazines, then we still need a separate job so that publishers understand the value of getting into INIS.
– Do Russian magazines generally come to the attention of the secretariat?
– It gets what is translated into English. Let's say there is a magazine called Atomic Energy, which is one of the main magazines in the nuclear industry – it translates its articles in full. Other journals, for example, the Bulletin of Atomic Science and Technology, they transfer some of the articles to the journal Physics of Atomic Nuclei, which collects publications on various Russian publications and publishes them with an English translation. The key point for the secretariat is that it should be in English. What comes out in English gets there right away. Much of what is published in Russian is accompanied by annotations translated into English, in principle, this is also sufficient. But here it is necessary to negotiate with the magazines so that they support this system. Some people don't want, perhaps they want readers to come to their website and download from there.
– Is there a large share of Russian content in the total stream that gets into this database?
– Russia and the USSR, which stood at the very origins of INIS, account for about three hundred thousand records, which is quite a serious figure. Last year, we uploaded about three thousand. Although, there are countries that download more, for example, the USA. In America, everything is published in English, and the secretariat takes it automatically.
– Does the language barrier matter?
– Yes, this is a key moment.
– According to your observations, to what extent is this database in demand from the scientific and engineering community?
– Everything has changed here in recent years. Historically, this database was closed. It wasn't publicly available. Until 2010, database instances were distributed to a limited number of organizations on CD-ROMs. But after the situation changed, Google appeared, and publishers began to post their magazines online, it became clear that we need to be open. It has been an open database since 2010. This has changed especially in recent years, and if I type a query into Google regarding my field of activity, say, reactor physics, then somewhere the second or third link will definitely be to INIS.
– In the literature for a scientific article, can a link to INIS replace a link to a scientific journal?
– As a rule, even if we have looked at an article in this database, we still take its source data in the link. But for "gray" literature, it can be.
– Do you think that for a Russian scientist, the INIS hit is evidence of his scientific status?
– Yes and no. Of course, getting into INIS is evidence that this publication is definitely not garbage, if it has entered the database, then there is no frankly unreliable information in it and there is no information not related to this category – and INIS has twenty-seven categories. That is, it fits exactly into one of the twenty-seven areas of nuclear science. On the other hand, this is an uncensored system. That is, there are publications not only from peer-reviewed journals. It's like a huge library. And when we apply for a grant or participate in some kind of competition, publications in the RSCI or SCOPUS are important. It is important for a scientist to increase the number of publications in peer-reviewed publications, and preferably in the top quartiles. And INIS collects everything. But it seriously increases the visibility of the publication. This is important because INIS allows you to access old publications from, say, the sixties and seventies, when many interesting experiments were conducted. At that time, there was a boom in the development of research nuclear reactors, and many interesting experiments were carried out, for example, in materials science. These publications are not available on the websites of the magazines. And if this publication is important to any researcher, the mechanism is as follows. The scientist addresses the INIS officer in his country, who addresses me directly or through the secretariat. And I'm looking for this publication whenever possible.
– Is INIS working on digitizing old publications?
– Such work is underway. For example, we have a good website on the history of Rosatom, which contains a lot of materials on the history of how the nuclear project was created, the earliest publications, and biographies of prominent figures in the nuclear industry. A very good collection of materials. Well, in general, either publishing houses or public libraries are engaged in digitization. This work was, until the last moment, in general, manual. And if we take into account that all materials need to be indexed, then it was quite a lot of work. And automating this process is asking for it. Today, those Russian-language publications that are accompanied by English-language annotations are, in principle, available for automated input. At our INIS center, we are already creating automatic indexing systems based on semantic AI technologies. However, the need for a human specialist still remains. The criterion of truth is still those people who have been doing this for many years, and they are doing it well. But the trend is for this process to be automated.
