Engineers and linguists once again entered into dialogue

At the beginning of June, the international conference on computational linguistics Dialogue , which our company has been doing for many years, was held in the Moscow-based “Bekasovo”. We wrote about what constitutes a “Dialogue” here , so we will not repeat, but tell you what was new.

Perhaps one of the most important events for us - the collection of scientific papers of the Dialogue will finally be indexed by SCOPUS , the largest international citation system. This system is not just an index of scientific articles, it is a kind of authoritative confirmation of the scientific importance of the publication included in SCOPUS. Why is it important? To confirm their status, all scientists should have a set of publications in recognized authoritative publications. Moreover, to confirm their own scientific authority abroad and promote their own ideas and developments, it is important to have publications in sources that are recognized by international indices, such as SCOPUS. In addition, having entered into SCOPUS publications, they automatically fall into the list of the Higher Attestation Commission (Higher Attestation Commission). This enables Russian scientists to have publications that are recognized as authoritative both in the world and in Russia. The Scopus’s status of the Dialogue compilation gives the conference a higher status and we are pleased that now participation in the Dialogue will give more “weight” to the scientific works of our speakers.

This year one of the dominants of Dialogue was computational semantics, the first day of the conference was devoted to it. This field of computational linguistics studies various ways of computer modeling of the meanings of words, phrases, sentences, whole texts. The problem is important to solve and very difficult, there are different approaches. For example, the newest one is the distribution semantics, in which the meaning is described on the basis of the word compatibility statistics calculated by large text boxes. In ontological semantics, emphasis is placed on taking into account knowledge of the world, formal semantics widely uses mathematical logic to describe the meanings of sentences, operational semantics emphasizes the instructive aspects of the language: interprets phrases as commands that control some actuator, for example, a robot. Finally, the traditional “explanatory” semantics tries to fully describe the meaning with the help of special interpretation languages.
')
The conference was attended by famous scientists, leaders of these areas of computer semantics: Marco Baroni, Barbara Party, Sergey Nirenburg, Konstantin Vorontsov, Elena Paducheva. In the afternoon there were interesting reports, and in the evening a special round table was held, at which it was possible to compare different models, to see not only the differences, but also the possibilities of interaction between these approaches. It seems that the idea of ​​a “review of semantic approaches” has succeeded, and we can expect the emergence of new research and researchers in this area.

We see that year by year computational linguistics is beginning to influence the industry more and more. The use of advanced linguistic technologies becomes the basis for the competitive advantage of solutions in the field of electronic document circulation, processing, search and storage of information. Therefore, we decided to hold on the second day of the conference “Technology Day” - in fact, this is an attempt to visually show potential customers the best practical solutions in the field of computational linguistics and get their objective comparative assessment. Invited presentations were made by representatives of companies and organizations leading in the development of linguistic technologies: RCO, ABBYY, Yandex, IBM, scientific groups of the Moscow State University, the Research Institute for System Programming, the Research Institute for System Analysis. Representatives of all organizations and departments that are most interested in using linguistic technologies to solve their problems came to listen to them. In the evening, a round table was held at which the problems of interaction between science, engineering and business in our country were discussed. According to the testimony of the representatives of the departments, "it was VERY interesting to them, they learned a lot of new things." This “Technology Day” was the first, and the experiment can be considered successful: in the next “Dialogue” it will be possible to repeat this event.

And at the conference were summed up the competition for the automatic resolution of anaphora and coreference. Such competitions for the Russian language were held for the first time. Seven developers of automatic text analysis systems took part in the contest: ABBYY, RCO, SemSyn, Open Corpora (St.Petersburg), Mail.ru, Institute for Systems Analysis of the Russian Academy of Sciences, Sergey Ponomarev. More on this in the second part of the article.