=Paper=
{{Paper
|id=Vol-2523/paper30
|storemode=property
|title=
Towards a Tatar Wordnet: a Methodology of Using Tatar Thesaurus (short paper)
|pdfUrl=https://ceur-ws.org/Vol-2523/paper30.pdf
|volume=Vol-2523
|authors=Alfiya Galieva,Alexander Kirillovich,Natalia Loukachevich,Olga Nevzorova
|dblpUrl=https://dblp.org/rec/conf/rcdl/GalievaKLN19
}}
==
Towards a Tatar Wordnet: a Methodology of Using Tatar Thesaurus (short paper)
==
https://ceur-ws.org/Vol-2523/paper30.pdf
Towards a Tatar Wordnet: a Methodology of Using
Tatar Thesaurus
Alfiya Galieva1, Alexander Kirillovich2,
Natalia Loukachevich3 and Olga Nevzorova1,2
1 Tatarstan Academy of Sciences, Kazan, Russia
2 Kazan (Volga region) Federal University, Kazan, Russia
3 Lomonosov Moscow State University, Moscow, Russia
amgalieva@gmail.com, alik.kirillovich@gmail.com,
louk_nat@mail.ru, onevzoro@gmail.com
Abstract. For wordnet developing for a new language, the key problem is to
find original resources that contain enough lexical data of the language in an
appropriate format. This article discusses the structure, methodology of compi-
lation and the current state of the bilingual Russian-Tatar Social-Political The-
saurus, which can serve as an initial resource for building the Tatar Wordnet.
This thesaurus reflects the logical-semantic organization of lexical elements
(synonymous, generic, and some other relationships) at the conceptual and lexi-
cal levels. Mainly, we focus on building synsets for nouns (single nouns and
noun phrases).
Keywords: Tatar language, WordNet, Thesaurus, Linguistic ontology, Socio-
political terminology.
1 Introduction
A great hindrance to develop linguistic ontologies for a new language and conceptual
modeling is the lack of original lexicographic resources containing full and relevant
linguistic data description.
Success of Princeton WordNet has determined emergence of wordnets and word-
net-like projects for different languages and multilingual wordnets. In wordnet build-
ing developers often use the Expand Model (Vossen 2002: 52) when available word-
nets that serve as mapped linguistic relations between the items and ready synsets of a
source language are translated using bilingual dictionaries into equivalent synsets in
the target language. Most of the wordnets existing for today are implemented by
translating Princeton English WordNet. The alternative approach is very laborious,
time-consuming and difficult to implement, based on compiling synsets and mapping
semantic relations between word senses directly on the data of the language for which
a wordnet is developed. In this case good dictionaries of synonyms and other seman-
tic dictionaries are required.
Copyright © 2019 for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
316
So the possibility of developing wordnets is largely determined by the presence of
bilingual dictionaries or fairly complete descriptions of the semantic system of the
language.
The absence of large English-Tatar dictionaries (the available ones are of very lim-
ited volume and may be used only for education purposes) makes it impossible to use
the Expand Model to Tatar wordnet development, as well as absence of Tatar seman-
tic dictionaries makes it almost impossible to develop original Tatar wordnet.
The objective of this article is to describe the methodology for constructing a Tatar
wordnet based on a lexical resource such as the Tatar social-political thesaurus. This
approach allows you to directly use the data of the thesaurus, primarily a set of
synsets and relationships between synsets.
The body of the paper is organized as follows. Section 2 outlines the basic theoret-
ical background of the study, and the main attention is paid to wordnet projects devel-
oped for the Turkic languages. Section 3 presents the methodology of compiling the
Russian-Tatar socio-political thesaurus and its current state. Section 4 describes the
most important aspects of implementing a wordnet-like resource using Tatar thesau-
rus synsets for Tatar nouns. Section 5 discusses the conclusions and outlines the pro-
spects of future work.
2 Related Works
At present time, there are various wordnets for some Turkic languages.
Two Turkish wordnet projects have been developed for the Turkish language. The
first one (Çetinoğlu, et al, 2018; Bilgin, et al, 2004) has been created at Sabancı Uni-
versity as part of the BalkaNet project (Tufis, et al, 2004). The BalkaNet project was
built on the basis of a combination of expand and merge approaches. All wordnets
contain many synonyms for Balkan common topics, as well as synsets typical for
each of the BalkaNet languages. The size of the Turkish Wordnet is about 15,000
synsets.
Another Turkish wordnet is the KeNet (Ehsani, 2018; Ehsani, et al, 2018). This
wordnet was built on the basis of modern Turkish dictionaries. A bottom-up approach
was used to build this resource. Based on dictionaries, words were selected and then
they were manually grouped into synsets. The relationships between words have been
automatically extracted from dictionary definitions and then these relationships have
been created between synsets. The size of this resource is about 113,000 synsets.
Unfortunately, the lack of large Turkish-Tatar dictionaries (as well as English-
Tatar ones) makes it impossible to translate Turkish resources into the Tatar language.
In this respect the Tatar language can be attributed to low-resource languages.
The Extended Open Multilingual Wordnet (Bond, 2013) resource is built from
Open Multilingual Wordnet by replenishing the WordNet data automatically extracted
from the Wiktionary and Unicode Common Locale Data Repository (CLDR). The
resource contains wordnets for 150 languages, including several Turkic: Azerbaijani,
Kazakh, Kirghiz, Tatar, Turkmen, Turkish, Uzbek. The Tatar wordnet contains a total
of 550 concepts, which cover 5% of the PWN core concepts.
317
The BabelNet (Navigli and Ponzetto, 2012) resource contains a common network
of concepts that have text inputs in many languages. The BabelNet contains 90,821
Tatar text entries that refer to 63,989 concepts. However, due to the fact that this re-
source was built automatically, it has quality problems.
Thus, the development of a qualitative Tatar wordnet with an emphasis on the spe-
cific features of the Tatar language based on the existing lexical resources is very
relevant.
3 Tatar Socio-Political Thesaurus: Methodological Issues
of Compiling and Its Current State
The conceptual model of the Tatar socio-political thesaurus (hereinafter referred to as
TatThes), the general principles of displaying linguistic data are taken from the
RuThes project (http://www.labinform.ru/pub/ruthes/) (Loukachevitch and Dobrov,
2014; Loukachevitch, Dobrov and Chetviorkin, 2014). The RuThes thesaurus build as
is a hierarchical network of concepts with attributed lexical entries for automatic text
processing.
In the RuThes each concept is linked with a set of language expressions (nouns,
adjectives, verbs or multiword expressions of different structures – noun phrases and
verb phrases) which refer to the concept in texts (lexical entries). The RuThes con-
cepts have no internal structure as attributes (frame elements), so concept properties
are described only by means of relations with other concepts.
Each of the RuThes concept is represented as a set of synonyms or near-synonyms
(plesionyms). The RuThes developers use a weaker term, ontological synonyms, to
designate words belonging to different parts of speech (like stabilization, to stabilize),
the items may be related to different styles and genres. Ontological synonyms are the
most appropriate means to represent cross-linguistic equivalents (correspondences),
because such approach allows us to fix units of the same meaning disregarding sur-
face grammatical differences between them. For example, Table 1 represents basic
ways of translating Russian adjective + noun phrases into Tatar.
Table 2. Examples of Russian Adj + Noun phrases and ways of translating them into Tatar
Russian unit Corresponding Tatar unit The structure of English translation
Tatar unit
Пенсионный Пенсия яше N + NPOSS_3 Retirement age
возраст
Рабочий класс Эшчеләр сыйныфы NPL + NPOSS_3 Working class
Консульская Консуллык хезмәте NNMLZ +NPOSS_3 Consular service
служба
Сексуальное Сексуаль азчылык ADJ + N Sexual minority
меньшинство
Именная стипендия Исемле стипендия NCOMIT + NPL Nominal scholarship
318
The TatThes is based on the list of concepts of the RuThes, i.e. the Tatar compo-
nent is based on the list of concepts of the RuThes thesaurus. The methodology of
compiling the Tatar part of the thesaurus includes the following steps:
1. Search for equivalents (corresponding words and multiword expressions) which
are actually used in Tatar as translations of Russian items.
2. Adding new concepts representing topics which are important for the sociopoliti-
cal and cultural life of the Tatar society and which are not presented in the original
RuThes (for example, Islam-related concepts, designations of Tatar culture specific
phenomena, etc.).
3. Revising relations between the concepts considering the place of each new con-
cept in the hierarchy of the existing ones and, if necessary, adding the new concepts
of the intermediate level. So an important step is to check up the parallelism of con-
ceptual structures between the languages.
The TatThes is mainly being compiled by manual translation of terms from the
RuThes into Tatar, besides the Tatar language specific concepts and their lexical en-
tries are added (about 250 new concepts). Search for equivalents in the Tatar language
in many cases became a time-consuming task, because available Russian-Tatar dic-
tionaries of general purpose contain obsolete lexical data (Galieva, Kirillovich, et al.,
2017). So when compiling the lists of concept names and lexical entries we manually
browsed large arrays of official documents and media texts in Tatar. In the process of
compiling the Thesaurus, data from the following available Tatar corpora is used:
1. Tatar National Corpus (http://tugantel.tatar/?lang=en);
2. Corpus of Written Tatar (http://www.corpus.tatar/en).
In the course of the project we found that distinguishing feature of the contempo-
rary Tatar lexicon is a great deal of absolute synonyms of different origin in and struc-
ture, the main cause of the phenomenon is language contacts (Galieva, Nevzorova, et
al., 2017; Galieva, 2018).
The TatThes is implemented as a web application and has a special site
(http://tattez.turklang.tatar/). Additionally, it has been published in the Linguistic
Linked Open Data cloud as part of RuThes Cloud project (Kirillovich, et al, 2017).
Currently the TatThes contains 10,000 concepts, and 6,000 of them provided with
lexical entries.
4 Tatar Thesaurus Data for Wordnet Implementation:
Case of Nouns
Previously, the RuThes thesaurus has been semi-automatically converted to the
WordNet-like structure, and Russian wordnet (RuWordNet) has been generated
(Loukachevitch, et al, 2016; Loukachevitch, et al., 2018). The conversion included
two main steps:
319
1. the automatic subdivision of the RuThes text entries into three nets of synsets ac-
cording to parts of speech;
2. the semi-automatic conversion of RuThes relations to WordNet-like relations.
The current version of RuWordNet (http://ruwordnet.ru/eng) contains 110 thou-
sand Russian unique words and expressions. The same approach can be used to trans-
form TatThes to Tatar wordnet.
The TatThes data may be serve as an initial basis for wordnet building by the fol-
lowing reasons:
1. The sociopolitical sphere covers a broad area of modern social relations. This ar-
ea comprises generally known terms of politics, international relations, economics
and finance, technology, industrial production, warfare, art, religion, sports, etc.
2. Currently the TatThes, in addition to terminology, comprises some general lexi-
con branches representing lexical items which can be found in various domain specif-
ic texts.
3. Semantic relations in the TatThes are necessary and sufficient to arrange the Tatar
nominal vocabulary (nouns and noun phrases) as a wordnet-like network of synsets.
Thesaurus concepts unite synonymous items, so we have ready sets of synonyms
as building blocks for wordnet. The concepts are linked by semantic relations with
each other. In the RuThes and in the TatThes there are four main types of relation-
ships between concepts, see Table 2. Semantic relations, mapped in wordnet, are not
all shared by all lexical categories, so thesaurus data converting into wordnet format
require dissimilar ways for different parts of speech.
Table 2. Semantic relations between nouns in thesaurus and in wordnets
Semantic relations in Thesaurus Semantic relations in wordnets
Hypernym — hyponyms Hypernym — hyponyms
Holonym — meronym Holonym — meronym
Symmetrical association (Asc)
Asymmetric association (Asc1/Asc2)
Asc and Asc1/Asc2 association relations need additional explanations. The Asc sym-
metrical association, distinguished in RuThes and inherited by Tatar Socio-Political
Thesaurus, connects very similar concepts, which the developers did not dare to com-
bine into the same concept (for example, cases of presynonymy of items).
The Asc1/Asc2 asymmetric association connects two concepts that cannot be de-
scribed by the relations mentioned above, but neither of them could not exist without
the existence of the other (for example, a concept SUMMIT MEETING needs exist-
ing the concept HEAD OF THE STATE). In studies of ontologies this relation may be
mapped as the ontological dependence relation.
Nevertheless, basic semantic relations which we need to group nouns concepts into
wordnet are presented in the TatThes.
320
The core of the TatThes is made up of nouns and noun phrases (see Table 3), so the
bulk of thesaurus data may be used for Tatar wordnet building without significant
changes (synonymous items are yet joined into synsets and the required relations
between them are selected).
Table 3. Number of noun concepts and noun phrase concepts in TatThes
(on data of the Russian part).
Structure of TatThes items Number items
Noun 3387
Adj + Noun 3135
Noun + NounGEN 352
Other 3126
Total 10000
An important issue is reflecting Tatar language specific word usage features in the
resource. Presence alone of the shared concepts in languages do not necessarily evi-
dences the same ways of usage of individual words or of usage words of individual
semantic classes. Consider this with an example. Specific feature of the Tatar lan-
guage is using of hypernyms before a corresponding hyponym, and such using is not
regarded as pleonasm in many cases (examples 1–3):
(1) Париж шәһәрендә ‘in the city of Paris’ (instead of ‘in Paris’);
(2) кыз кеше ‘girl human’ (instead of ‘a girl’);
(3) май аенда ‘in the month of May’ (instead of ‘in May’).
Table 4. Representing lexical entries of month names in Thesaurus.
Rus con- Russian Rus Tatar con- Tatar lexical entries Tat
cept name lexical entries POS cept name POS
ДЕКАБРЬ Декабрь ‘December’ N Декабрь Декабрь ‘December’ N
Декабрьский ‘of De- ADJ Декабрь ае ‘month of NP
cember’ December’
ЯНВАРЬ Январь ‘January’ N Гыйнвар Гыйнвар ‘January’ N
Январский ‘of Janu- ADJ Гыйнвар ае ‘month of NP
ary’ January’
Январь ‘January’ N
Январь ае ‘month of Janu- NP
ary’
ФЕВРАЛЬ Февраль ‘February’ N Февраль Февраль ‘February’ N
Февральский ‘of ADJ Февраль ае ‘month of NP
February’ February’
In cases when such a usage is conventionalized and corpus data evidences that the
usage has a high frequency, we include such hyponym-hypernym items into a list of
lexical entries of a concept. Such manner of designating is a feature of using topo-
321
nyms and some classes of general lexicon, so it should be considered in Tatar wordnet
building. For example, lexical entries of month names include such conventionalized
noun phrases, composed of the month name and the hyponym, designating month in
general, see Table 4.
Because the RuThes concepts assemble ontological synonyms, the RuThes lexical
entries bring together words of different part of speech. Therefore, in standard case a
Russian synset joins a noun (often we use it as a concept name) and a relative adjec-
tive derived from a noun (Table 5; only core items of synsets are represented). In
Tatar, like in other Turkic languages, there is no original relative adjectives (and ex-
isting ones are borrowed from European or Oriental languages), so in many cases the
TatThes synsets are composed of items of the same part of speech, mainly of nouns.
This circumstance greatly facilitates cleaning thesaurus synsets data for wordnet de-
veloping.
Table 5. Typical arrangement of Russian and Tatar Thesaurus synsets.
Basic lexical entries of a Russian Part of speech of Basic lexical entries of Part of speech of
concept Russian words a Tatar concept Tatar words
Река ‘river’ N Елга ‘river’ N
Речной ‘of river, fluvial’ ADJ
Факультет ‘faculty’ N Факультет ‘faculty’ N
Факультетский ‘of faculty’ ADJ
Преподаватель ‘teacher N Укытучы ‘teacher’ N
Преподавательский ‘of teacher’ ADJ
Больница ‘hospital’ N Хастаханә ‘hospital’ N
Больничный ‘of hospital’ ADJ Сырхауханә ‘hospital’ N
So the core of the TatThes is made up of nouns and noun phrases (69% of total
number of concepts). At the moment semantic relations between nouns mapped in
thesaurus, are necessary and sufficient to convert Tatar thesaurus data into the word-
net format.
4 Conclusion
When building a wordnet for a new language, in particular, for a low-resource one, a
crucial issue is searching for appropriate sources. We are planning to use data of the
TatThes as a base resource for developing Tatar wordnet.
The TatThes is being compiled by manual translation of terms from the RuThes in-
to Tatar, with searching Tatar equivalents used in real texts, so the thesaurus contains
relevant lexical data. In the TatThes each concept is linked with a set of language
expressions (single words or multiword expressions) which refer to the concept in
texts – lexical entries.
The analysis of thesaurus data shows that the bulk of the thesaurus synsets are
formed around nouns or noun phrases. A mapping semantic relations of nouns in the
thesaurus reproduces a mapping semantic relations in wordnets.
322
Future work includes adding material of verbs and other parts of speech. Also we
are planning to develop some automatic approaches to mining terms and to asses the
Tatar terminology coverage in Thesaurus on Tatar socio-political texts data.
Acknowledgements
This work was partially funded by the subsidy allocated to Kazan Federal University
for the state assignment in the sphere of scientific activities, grant agreement no.
1.2368.2017 and partially by a subsidy assigned to the Institute of Applied Semiotics
of the Tatarstan Academy of Sciences for the state assignment.
References
1. Orhan Bilgin, Özlem Çetinoğlu, and Kemal Oflazer: Building a Wordnet for Turkish. Ro-
manian Journal of Information Science and Technology 7 (1–2), 163–172 (2004).
2. Francis Bond and Ryan Foster: Linking and Extending an Open Multilingual Wordnet. In:
Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics
(ACL 2013), 1352–1362. ACL (2013).
3. Özlem Çetinoğlu, Orhan Bilgin, and Kemal Oflazer: Turkish Wordnet. In: K. Oflazer and
M. Saraçlar (eds). Turkish Natural Language Processing. Springer (2018).
doi:10.1007/978-3-319-90165-7_15
4. Razieh Ehsani. KeNet: A Comprehensive Turkish Wordnet and Using It in Text Cluster-
ing. PhD Thesis. Işık University (2018).
5. Razieh Ehsani, Ercan Solak, and Olcay Taner Yildiz: Constructing a WordNet for Turkish
Using Manual and Automatic Annotation. ACM Transactions on Asian and Low-Resource
Language Information Processing 17 (3), Article No. 24 (2018). doi:10.1145/3185664
6. Alfiya Galieva, Alexander Kirillovich, Bulat Khakimov, Natalia Loukachevitch, Olga
Nevzorova, and Dzhavdet Suleymanov: Toward Domain-Specific Russian-Tatar Thesau-
rus Construction. In: R. Bolgov, N. Borisov, et al. (eds.) Proceedings of the International
Conference on Internet and Modern Society (IMS-2017), pp. 120–124. ACM Press, New
York (2017). doi:10.1145/3143699.3143716
7. Alfiya Galieva, Olga Nevzorova, and Dilyara Yakubova: Russian-Tatar Socio-Political
Thesaurus: Methodology, Challenges, the Status of the Project. In: R. Mitkov and G. An-
gelova (eds.) Proceedings of the International Conference Recent Advances in Natural
Language Processing (RANLP 2017), pp. 245–252. INCOMA Ltd., Varn (2017).
doi:10.26615/978-954-452-049-6_034
8. Alfiia Galieva: Synonymy in Modern Tatar Reflected by the Tatar-Russian Socio-Political
Thesaurus. In: J. Čibej, V. Gorjanc, I. Kosem and S. Krek (eds.) Proceedings of the XVIII
EURALEX International Congress: Lexicography in Global Contexts (Euralex 2018),
pp. 585-994. Ljubljana University Press (2018).
9. Alexander Kirillovich, Olga Nevzorova, Emil Gimadiev, and Natalia Loukachevitch:
RuThes Cloud: Towards a Multilevel Linguistic Linked Open Data Resource for Russian.
In: P. Różewski and C. Lange (eds.) Proceedings of the 8th International Conference on
Knowledge Engineering and Semantic Web (KESW 2017). Communications in Computer
and Information Science 786, 38–52. Springer (2017). doi:10.1007/978-3-319-69548-8_4
10. Natalia Loukachevitch and Boris Dobrov: RuThes Linguistic Ontology vs. Russian Word-
nets. In: H. Orav, C. Fellbaum and P. Vossen (eds.) Proceedings of the 7th Conference on
Global WordNet (GWC 2014), pp. 154–162. University of Tartu Press (2014).
323
11. Loukachevitch, N.V., Dobrov, B.V., and Chetviorkin, I. I.: RuThes-Lite, a Publicly Avail-
able Version of Thesauru of Russian Language RuThes. In: Computational Linguistics and
Intellectual Technologies: Papers from the Annual International Conference “Dialogue”,
pp. 340–349. RGGU (2014)
12. Loukachevitch, N.V., Lashevich, G., Gerasimova, A.A., Ivanov, V.V., and Dobrov, B.V.:
Creating Russian wordnet by conversion. In: Computational Linguistics and Intellectual
Technologies: papers from the Annual Conference “Dialogue”, pp. 405-415. RGGU
(2016).
13. Natalia Loukachevitch, German Lashevich, and Boris Dobrov: Comparing Two Thesaurus
Representations for Russian. In: F. Bond, T. Kuribayashi, C. Fellbaum and P. Vossen
(eds.) Proceedings of the 9th Global WordNet Conference (GWC 2018), pp. 35–44. Global
Wordnet Association (2018).
14. Roberto Navigli and Simone Paolo Ponzetto: BabelNet: The automatic construction, eval-
uation and application of a wide-coverage multilingual semantic network. Artificial Intel-
ligence 193, December 2012, 217–250 (2012).
15. Tufis, D., Cristea, D., and Stamou, S.: BalkaNet: Aims, Methods, Results and Perspec-
tives. A General Overview. Romanian Journal of Information Science and Technology 7
(1–2), 9–43 (2004).
16. Piek Vossen (ed). EuroWordNet: General Document (2002).
324