Engineers and computer linguists - to the "Dialogue"

From May 30 to June 2, the 24th International Conference on Computational Linguistics “ Dialogue ” will be held at the Russian State University for the Humanities (RSUH). We have already talked about what “Dialogue” is and why ABBYY is its main organizer. In this post, we announce the main topics of the conference, key speakers and their presentations, publish information about competitions on resolving lexical ambiguity in Russian, Dialogue Evaluation, and other details about the conference.

This year the main topics of the “Dialogue” will be:

• Linguistic interpretation of the results of deep machine learning ("black boxes"). The principal feature of today's incredibly popular multilayer neural networks is that few people understand what factors actually influenced the outcome of their work. They are a kind of "black boxes". On the one hand, it gives rise to some pessimism among some linguists about the prospects of their science, and on the other, makes us think about the possibility of using deep learning for linguistic research proper and about a meaningful interpretation of its results. Perhaps this understanding will help improve the work of the neural networks themselves.
• Methods of applying big data analysis technologies to tasks for which such data is not enough. It is no secret that for many text analytics tasks it is very difficult to find adequate training datasets. And I want to solve them. There are different approaches to this problem, from automatic generation of training data to technologies for transferring deep learning results from tasks where there is a lot of data, to tasks where there is not enough (so-called TransferLearning).
• Multimodal communication. This is the study of all spheres of speech act, language, intonation, facial expressions and gestures, emotional and communicative behavior. The conference will discuss new results and their practical application, for example, for dialogue agents or robots.

In the "Dialogue" will be attended by well-known experts in computational linguistics :
')
Alexander Panchenko from the University of Hamburg (Germany). This young Russian scientist successfully deals with problems in the field of computer lexical semantics. His invited report is devoted to linguistic interpretation of the results of deep learning, May 31, Thursday, 15: 30-16: 40.

Andrei Kibrik , Director of the Institute of Linguistics of the Russian Academy of Sciences, Professor of Moscow State University. He will make a report (together with a senior researcher at the Laboratory of Cognitive Studies of the Faculty of Psychology, ION RANEPA, Olga Fedorova) on the interaction of the processes of generation and understanding in face-to-face multimodal communication, May 30, Wednesday, 11: 30-13: 00 .

Sergey Sharov , Professor, Department of Translation Studies, University of Leeds (UK). One of the most authoritative experts in the world in the field of online text analysis. At the conference, he will talk about experiments using machine translation for the tasks of multilingual genre classification of texts, June 1, Friday, 10: 30-13: 30.

Igor Boguslavsky , a professor at the Madrid University of Technology and head of the Laboratory of Computational Linguistics at the IITP RAS, one of the largest specialists in ontological semantics. His report will focus on how to teach a computer to understand football reports, May 30, Wednesday, 11: 30-13: 00.

Natalia Lukashevich from the Moscow State University Computer Center, a leading researcher in the field of creating computer thesauri and their application to various tasks of NLP. Will tell on the "Dialogue" about the new direction in the so-called. Sentiment analysis: revealing a complex picture of opinions expressed in the texts, May 31, Thursday, 12: 30-14: 00.

• On May 31, at 19:15, a round table on “black boxes” will take place. If you are interested in what opportunities do deep learning methods open for linguistics, we recommend you to attend the conference on this day.
• June 1, at 17:00, we invite you to the BigData vs. Round Table SmallData. Engineers and linguists will discuss how to solve the problem of lack of training data.

The conference materials will be published in the yearbook “ Computer linguistics and intellectual technologies ”, which is included in the international citation system Scopus . On June 2, from 17:00 to 19:00, a student session will also be held on the main areas of the Dialogue. The best works of the participants will receive prizes from the Association of Teachers of Fundamental and Applied Linguistics ( APIFIPL ).

One of the key events of the Dialogue is the holding of technological competitions for developers of systems for linguistic analysis of texts, Dialogue Evaluation . This year, for the first time, they are devoted to the automatic determination of the meanings of multi-valued words in Russian texts. This is the first such competition for the Russian language. In more detail about this competition we already told on Habré. The results of the Dialogue Evaluation will be announced at the “Dialogue” on May 31, on Thursday, at 17:00. At the end of the conference we will publish them in this blog, stay tuned.

Conference working languages ​​are Russian and English. The detailed program of the conference is posted here .

You can register here , registration takes place until May 29. Terms of participation .

Elizaveta Titarenko, editor of the corporate blog ABBYY