It is our pleasure to present you the proceedings of the Workshop Conference of COLINS 2021, the fourth edition of the International Conference on Computational Linguistics and Intelligent Systems, held in Lviv (Ukraine) on April 22-23, 2021. The main purpose of the CoLInS conference is a discussion of the recent research results in all areas of Natural Language Processing and Intelligent Systems Development.
Computational linguistics and intelligent systems. – Lviv, 2021. – Volume 2 : Proceedings of the 5nd International conference, COLINS 2020. Workshop, Lviv, Ukraine April 22–23, 2021. – 131 p. – URL: http://ceur-ws.org/Vol-2870/
(2021-05-04) Vozniak, Ihor; Lviv Polytechnic National University
Development of professional and scholar terminology as well as its lexicographical
inventorization remain one of the biggest challenges for modern Ukrainian linguistics. This
study aims therefore to investigate theoretical basics of creating specialized terminological
electronic dictionaries since electronic dictionaries are the most convenient way for both
professionals and linguists to learn, study, and use the language. The notion and main
characteristics of electronic dictionaries were analyzed. Specifics of terminological
dictionaries and their differences from learner’s dictionaries were described. Corpus-based
approach was investigated as basis for the highest level of dictionary’s objectivity and
representativeness. The results of this research attest a huge potential for practical application
of the described methodology.
(2021-05-04) Tkachenko, Olha; Tkachenko, Kostiantyn; Tkachenko, Oleksandr; State University of Infrastructure and Technology; National Aviation University
The article discusses the problems of designing linguistic ontologies for educational
information systems. An approach to the formalized description of linguistic ontologies is
considered, taking into account the concepts of subject areas of training information systems
and the relationship between these concepts. The thesaurus of the training information
system, built on the basis of linguistic ontologies, is considered.
(2021-05-04) Kulyna, Olha; Lviv Polytechnic National University
A Last Will and Testament as a legal document of Inheritance Law is of particular importance
for the life of modern societies of all developed and underdeveloped countries. The research
focuses on the complex analysis of the study of English Last Will and Testament as a social
and communicative phenomenon which is a repetitive speech act that generates a typical
linguistic layout of the content to meet the communicative needs of a testator/testatrix on the
issue of the inheritance of property and money after their death in the situation of bequest. The
corpus of the research contains 400 wills written in England between 1837 and 2015 (525 023
characters). Attention is paid to discourse markers which provide structural integrity of the
text in wills. The main aim of this article is to conduct the analysis of discourse markers found
in English Last Wills and Testaments. The classification of discourse markers by B. Fraser has
been used in the study. A structural method has been applied to single out groups of discourse
markers. Discourse markers of sequence as a subtype of discourse activity markers, parallel
discourse markers, contrastive discourse markers, elaborative discourse markers and
inferential discourse markers as subtypes of message relationship markers are common in the
texts of Last Wills and Testaments. These markers complement the content of a previous
statement, combine parts of a sentence, introduce new information, contrast events, actions
and even participants. The usage of discourse markers facilities communication and ensures
the compositional integrity of the text.
(2021-05-04) Liashenko, Oleksii; Kazmina, Darina; Rosinskiy, Dmytro; Dukh, Yana; Kharkiv National University of Radio-Electronics
Relevance and the problem setting: at present, vulnerabilities in the firmware of IoT-devices
pose a serious threat, as attackers, who at first have exploited the vulnerabilities, gain remote
access to devices which allows them to form botnets that are then used to capture new
devices or organize serious DDos attacks. Therefore, currently, there is an urgent need to
increase the effectiveness of vulnerability detection methods in the firmware. The purpose of
this work is to analyze and define the term “vulnerability”, to provide the classification of
vulnerabilities of IoT-devices, the causes of vulnerabilities of IoT-devices, to analyze the
stages of vulnerability detection, and to present the example of a search algorithm for
(2021-05-04) Petrenjuk, Volodymyr; Petrenjuk, Dmytro; Centralukrainian national technical university; V. M. Glushkov Institute of Cybernetics of the NAS of Ukraine
The graph is outer-projective-planar, if embeds on the projective-plane with all vertices on
the boundary of one distinguished cell, and non-outer-projective-planar in another case. The
main result: diagrams of graphs as a result of the algorithm are given and the numbers of
reachability of sets of vertices of minors of the projective plane and sets with points of
connection of a star to subgraphs of these minors are calculated. The list of non-outer
projective-planar graphs that was declared in  has presented here.
(2021-05-04) Kirichenko, Lyudmyla; Radivilova, Tamara; Stepanenko, Juliia; Kharkiv National University of Radio Electronics
The article describes a new approach to the classification of time series based on the
construction of their recurrence plots. After transforming the time series into recurrence plots,
two approaches are applied for classification. In the first case, quantitative recurrence
characteristics are used for classification as features. In the second case, the time series is
presented in the form of a black and white image of its recurrence plot. A convolutional
neural network is used as an image classifier. The data for the classification are the
electrocardiograms realizations of 100 values, which contained records of healthy people and
patients with a diagnosis of ischemia. Research results showed the advantages of classifying
images of recurrence plots, indicate a good classification accuracy in comparison with other
methods and the potential capabilities of this approach.
(2021-05-04) Malion, Vasyl; Hryhorovych, Viktor; Lviv Polytechnic National University
The problem of choosing a way to buy a ticket is relevant in the world. Therefore, buyers
often prefer unprofitable and problematic alternatives. In order to choose the best way to buy
a ticket, all the alternatives to the usual booking offices are analyzed. Based on these data, a
comparison of these methods is made at a price, time spent, range of tickets, convenience,
reliability and provision of additional services.
(2021-05-04) Didushok, Valeriia; Khairova, Nina; National Technical University “Kharkiv Polytechnic Institute”
These days, an increased prevalence of post-traumatic stress disorder (PTSD) and severe
depression has been reported in populations exposed to war. This paper introduces using
linguistic analysis of trauma narratives in the context of the study of post-traumatic stress
disorder of combatants. As a subject of the analysis, posts of people who participated in
combat, obtained from topic-related discussion boards were used. The approach utilizes
vocabulary adaptation in NLP using the pre-trained language BERT model in addition to
descriptive statistics obtained from text. The novelty of the research lies in the use of a
context-sensitive model, while most of the existing research in this area is based on statistical
models that use statistical inference to discover hidden patterns.
(2021-05-04) Sokoliuk, Vitalii; Hryhorovych, Viktor; Lviv Polytechnic National University
Education is one of the best ways to form one’s personality so he is able to think and act
independently. It has always been given a great importance. Education systems are constantly
changing in order to adapt to new conditions and get the most out of them.
Recently, self-education aimed at gaining knowledge related to cognitive interests or
professional development has become increasingly popular. For this purpose, the relevant
literature is studied, different lectures and research centers are attended, somebody`s own
experiments and researches in the chosen field are conducted. Various methods on the
Internet are rapidly developing. One of the most popular is learning on educational platforms.
(2021-05-04) Khramtsov, Vladyslav; Orobinska, Olena; National Technical University "Kharkiv Polytechnic Institute"
The subject of the research "Special aspects of translation of medical instructions" is due to
the fact that in our time the assortment of new types of technologies in medicine which fulfils
the market of Europe and Ukraine is growing. The aim of the study is to convey the content
accurately, as much as possible to preserve the features of the style. In order to achieve it, we
need to know the subject and the related terminology (in our case, the medical one) and to
achieve the adequacy of the translation of the text of this industry.
(2021-05-04) Khluieva, Anastasiia; Kochuieva, Zoia; Borysova, Natalia; National Technical University “Kharkiv Polytechnic Institute”
This article describes implementation of the removing homonymy by collocation system
method in Ukrainian language, which can be implemented on Python programming language.
It also includes the relevance of removing homonymy phenomenon and difficulties
associated with it. A study of various methods of removing homonymy was carried out and
conclusions are mentioned in this work. In article there is an algorithm of the removing of
homonymy, particularly homoforms, by collocation system method.
(2021-05-04) Holshtein, Maiia; Babkova, Nadiia; National Technical University "Kharkiv Polytechnic Institute"
Nowadays a lot of descriptions of pieces of musical art can be found in Internet or in
specialized collections. There is no recommendation system that offers certain composition
for performance according to its difficulty level. This paper suggests the approach to creating
the recommendation system of piano pieces. The approach is based on checking for
collocations in descriptions of each composition. This paper shows the statistical method
PMI used for searching the collocations indicating on certain difficulty level. In addition it
also discusses the main problems during creating own recommendation system.
(2021-05-04) Bieliaiev, Oleksandr; Selivorstova, Yuliia; Liutenko, Irina; National Technical University "Kharkiv Polytechnic Institute"
In the process of creating software, the task of choosing the best framework in this case often
arises, for which it is necessary to evaluate the frameworks. The work considered the features
of the frameworks that are used in web development. Criteria for the rationality of using
frameworks for developing web applications are given. The components of front-end
development are considered and the classification of frameworks is given. It is proposed to
evaluate frameworks using the ISO / IEC 25010 quality model. The main functionality of
front-end frameworks, which are used for evaluation, are formulated. Such methods as the
synthesis of a formal decision-making model, qualimetry, SACS, analysis of variance, and
factor analysis can be used to evaluate frameworks. Subsequently, to evaluate the
frameworks, the SACS method was chosen as the best methodology for multi-criteria
(2021-05-04) Tarasenko, Yaroslav; Petrasova, Svitlana; National Technical University “Kharkiv Polytechnic Institute”
The paper describes an approach to open relation extraction based on unsupervised machine
learning. The state-of-the-art methods for extracting semantic relations are analyzed. The
algorithm of automatic open relation extraction using statistical, syntactic and contextual
information is proposed. The results of the study can be used in information retrieval,
summarization, machine translation, question-answering systems, etc.
(2021-05-04) Budko, Daria; Sharonova, Nataliia; National Technical University "Kharkiv Polytechnic Institute"
The article provides an overview of existing methods of processing textual data in order to
improve the site's performance and quickly find information about the product that the user
needs: substantiated the relevance of the research topic, analyze the classification of
problems, search and processing of textual information.
(2021-05-04) Shleiko, Anna; Borysova, Natalia; Kochuieva, Zoia; Melnyk, Karina; National Technical University "Kharkiv Polytechnic Institute"
The paper presents an overview of the existing machine learning methods for solving the
problem of gender classification of the authors of the written texts by names: substantiates
the relevance of the research topic, analyzes the existing methods of solving the task and
selects the direction of further research.
(2021-05-04) Romanova, Uliana; Petrasova, Svitlana; National Technical University “Kharkiv Polytechnic Institute”
The paper describes an approach to extraction of Verb-Noun patterns from news data stream.
The linguistic tagging, namely algorithms for parsing, and methods for extracting
collocations are analyzed. The algorithm for the automatic extraction of Verb collocations
from the designed corpus of news texts is proposed. The Stanford Universal Dependencies
parser is applied to identify Verb-Noun patterns. Then t-score is implemented for extracting
(2021-05-04) Shapovalova, Anastasiia; Petrasova, Svitlana; National Technical University “Kharkiv Polytechnic Institute”
The paper provides an algorithm of designing a test system for automated knowledge
assessment through open-ended questions. The relevance of the use of open-ended tasks and
problems of processing natural language answers are analyzed. The application of WordNet
and regular expressions is proposed for designing samples of correct answers in the