=Paper=
{{Paper
|id=Vol-2362/paper16
|storemode=property
|title=Frequency Dictionaries to the Instructions to Medical Products
|pdfUrl=https://ceur-ws.org/Vol-2362/paper16.pdf
|volume=Vol-2362
|authors=Roksolana-Yustyna Perkhach,Yuliia Shyika
|dblpUrl=https://dblp.org/rec/conf/colins/PerkhachS19
}}
==Frequency Dictionaries to the Instructions to Medical Products==
https://ceur-ws.org/Vol-2362/paper16.pdf
Frequency Dictionaries to the Instructions to
Medical Products
Roksolana-Yustyna Perkhach[0000-0002-6466-6122] and Yuliia Shyika[0000-0003-2474-0479]
Lviv Polytechnic University, 79005 Lviv , Ukraine
roksoliana-iustyna.t.perkhach@lpnu.ua,
yuliia.i.shyjka@lpnu.ua
Abstract. The actuality of the research stems from the necessity to investigate
the linguistic cognitive and linguistic cultural characteristics of the terms in the
instructions to medical products. Since the number of imported drugs has grown
rapidly in recent years and the need for the correct translation of the instructions
has also increased. The methodological basis for the study has been formed
from researches of contemporary scientists. Thus, the compilation of frequency
dictionaries of modern medical terminology will facilitate the study of the terms
used in the instructions to medical products. The tokenization of the instructions
texts has been made using the KWIC Concordance corpus manager. 195
instructions to medical products have been analyzed and sorted according to
Anatomical Therapeutic Chemical Classification System. As a result, the
frequency dictionaries have been created: Ukrainian language dictionary
includes 1000 words, Polish language dictionary includes 1000 words and the
frequency dictionary of German language includes 1000 words.
Keywords: research corpus, frequency dictionary, medical terminology, KWIC
Concordance, text corpora.
1 Introduction
Rapid development of medicine has resulted in the growth of the number of medical
terms which are difficult to process without appropriate technical means. One of the
most common forms of linguistic information systematization is the linguistic
corpora, which helps to create a text resource. Afterwards such data can be used in
various scientific researches.
Nowadays there are numerous medical dictionaries of high quality [1, 2, 3],
however there is no frequency dictionary for medical terminology.
The purpose of the study is to create frequency dictionaries using corpus-based
technologies as well as to introduce corpus technologies of representation of various
texts of contemporary Ukrainian, Polish and German languages.
The set goal requires the solution of the following tasks:
1. to describe and systematize existing scientific researches of Ukrainian and foreign
scholars on the texts corpus and the role of corpus based researches in modern
science;
2. to study classifications of linguistic text corpora most popular among scholars;
3. to create and describe the factual bases of the research (for Ukrainian, Polish and
German instructions to medical products);
4. to describe the method of frequency dictionary compilation on the basis of the text
corpus of the instructions to medical products.
5. to identify and describe the benefits of frequency dictionaries regarding the
specific features of modern medical terminology.
From a cognitive point of view, such a study is of high importance because it will
enable us to identify the peculiarities of Ukrainian, Polish and German terminology,
and to standardize Ukrainian terminology.
The terminological and methodological basis for the study has been formed from
researches of O. Baranov [4], M. Baker [5], A. Hardie, P. Baker, T. McEnery [6], S.
Buk., Y. Levus, Y. Yavorskyi, [7], Y. Karpilovska [8] O. Maksymiv [9] et al.
2 Problem and proposed method
2.1 Related works and researches
To compile a frequency dictionary for medical terminology, it is necessary to create a
text resource of the linguistic information storage – a linguistic corpus which consists
of three subcorpora: Ukrainian, Polish, and German instructions to medical products.
The first attempts to create an electronic text resource for conducting linguistic
research were made by «Centro Automazione Analisi Linguistica» (CAAL) (Italy,
1956) [10, p. 11].
Later many studies in the field of automated data processing have been conducted,
which resulted in the formation of several meanings of the notion «linguistic corpus»
in linguistics. These meanings are the following: electronic archive, electronic
libraries, collection of texts, array of texts, full text database, etc.
O. Demska offers the following definition of the corpus: «Organized electronic
collection of written and oral texts of any natural language, which has obligatory
characteristics and which is assigned for scientific study of a language» [11, p. 89].
However, M. Svidzinskyi believes that a corpus is structured and philologically
competent linguistic data which are represented in an automated form, intended to
solve specific linguistic problems [12, p. 27].
A. Baranov, P. Baker, Ye. Karpilovska, G. Sinclair and others suggest classifying
linguistic corpora according to:
type of linguistic data (written, oral, mixed);
«parallelism» (monolingual, bilingual, multilingual);
«literacy» (literary, dialect, colloquial, terminological, mixed);
the purpose of creation (multi-purpose and specialized);
genre (literary, folk, drama, journalistic);
accessibility (available for free, commercial and private);
purpose (research and illustrative);
dynamism (dynamic and static);
markup (marked and not marked);
markup type (morphological, semantic, syntactic, prosodic, and others);
text volume (full text and "fragmented") [4; 8; 13].
Some researchers assume that it is necessary to reduce the number of types of
linguistic corpora and to distinguish the following ones: specialized, reference,
multilingual, parallel, educational, diachronic, and instructive. O. Demska-Kulchytska
claims that there are the following corpora types:
full text (texts in the corpus are full) and fragmentary (there are fragments of texts
in corpus);
research (are used in linguistic studies to formulate new theories and concepts);
illustrative (are used to prove theorems or hypotheses about the language);
monitoring / dynamic (provide the possibility to monitor changes in language,
taking into account the aspect of diachrony; statistic corpora verify the state of
language in a certain synchronous period);
diachronic (represent the language over time period, and the synchronous corpora
represent language or type of the text in a fixed time period);
general language (represent the general, national language, and specialized
corpora, are aimed to solve specific and research tasks) [11, p. 156-157].
Multilingual and parallel corpora as well as comparative corpora form a separate field
of linguistics, which is highly important for the theory and practice of translation [5,
p. 234].
The main features of the corpus are:
representativeness: the ability of the corpus to reflect all the properties of the
subject area, that is the level of implementation of the language system, which
contains linguistic descriptive phenomena;
authenticity: involves the selection of the written or oral text(s) or passage(s) of the
text(s) created by the native speaker(s) in the process of real communication.
Meeting the requirement of authenticity is one of the components of empirical
research of the actual corpus material;
selection demands the limitation of the actual material by selecting certain
fragments of the language;
balance: the proportional amount of text resources in the corpus [11, p. 102].
Scientists use four basic parameters to characterize linguistic corpora: firstly, it
should be sufficiently large; secondly, the corpus should be structured or tagged;
thirdly, the texts, components of a particular corpus, should be in e-form; fourthly,
«electronic corpus» should include special software for further corpus processing [4;
5; 6].
V. Shyrokov notes that during the creation of the Ukrainian National Corpus the
following criteria for the texts selection have been used:
diachronic aspect (which texts and of which time period should be selected);
stylistic (texts should also represent substyles of the national language);
territorial (texts should represent the specificity of the literary language depending
on the region of Ukraine, as well as represent examples of oral or written texts
created abroad);
quantitative (clearly specifies the number of words in the text or passage, which are
included in the corpus, the number of texts and / or passages). Generally, in
corpora studies the criteria for selecting text fragments are separate problem,
within which linguistic and technical criteria are developed, and in the scientific
literature on this problem they are generalized as the theory of criteria for text
material selection for corpora of different types [10, p. 11].
2.2 Experimental results
One of the tasks of our research is to create a research corpus of instructions for
medical products. We have considered the linguistic corpus as a text collection stored
in electronic databases, which consists of three subcorpora: Ukrainian, Polish, and
German [14, 15, 16].
There are a lot of programs for text processing:
1. programs of analysis and linguistic processing of texts;
2. text conversion programs;
3. psycholinguistic programs;
4. texts generators and "talking programs";
5. systems of natural language processing and others [13, 18].
The program for analysis and linguistic processing of texts KWIC Concordance for
Windows [17] was used in our research.
The program possesses the following features:
building up the concordance;
searching both separate words and combinations;
sorting the list of words according to several criteria set by a user;
the ability to display the found word forms in the context;
saving the results;
quick processing of the request;
supporting text file format (txt).
Computer program for linguistic data processing KWIC Concordance for Windows
is attached to the Windows operating system, has lists of combinations to the
reference element specified by the token set (Collocation tool).
Fragments of factorial databases of research corpora on the material of the
languages under research are given in Fig. 1, 2, 3. [21, p. 27].
Frequency dictionary is «a type of a dictionary, which shows the number of uses,
that is, the frequency of a certain part of speech in the texts under survey. As a rule, it
consists of several lists: lists of words, usually in a root form, listed according to the
frequencies reduction and in alphabetical order and the similar lists for word forms»
[19, p. 292].
The importance of frequency dictionaries data has been repeatedly emphasized by
modern scholars: «Statistics also indicates that the first 100 of the most frequent words
covers about half the average text. However, the frequency dictionary data objectively
reflect vocabulary of the language only to the extent of the objective selection of
sources for this vocabulary. Their reliability depends on the quantitative and qualitative
characteristics of the lexical array for counting» [9, p. 109].
Fig. 1. Ukrainian database
Fig. 2. Polish database
Fig. 3. German database
The process of frequency dictionary compilation for the instructions to medical
products included the following steps:
1. The absolute frequency of each lemma is calculated by the semi-automatic method.
2. Relative frequency of the words is calculated automatically.
3. Words are listed according to their frequency.
4. The ranks in the dictionary are singled out.
5. The lexical minimum of the vocabulary of words for the instructions to medical
preparations is distinguished on the basis of separated ranks [20, p. 71].
So, the frequency dictionaries not only reflect the modern languages, but also
contribute to linguistic research of terminology and its standardization.
2.3 Our methods
For a comparative analysis of the instructions to medical products, it is necessary to
create a text resource of the linguistic information storage – a linguistic corpus which
consists of three subcorpora: Ukrainian, Polish, and German instructions to medical
products. [21, p.27]
The word forms are reduced to the base form. It indicates to which part of speech
belongs each of the word form. In addition, the traditional approach to lemmatization,
which is particularly important «for the quantitative study of lexemes in texts written
in the synthetic languages (to which the Ukrainian belongs)» [18, p. 90] is used:
Noun – all grammatical cases are reduced to the nominative case singular, for
example: medytsynoiy – medytsyni – medytsyna. Plural nouns are reduced to the
nominative case plural.
Adjective – case singular and plural forms are given as the root form of the
masculine gender. Adjectives in the comparative and the superlative degrees as
well as suppletive adjectives also follow this rule.
Verb – the synthetic forms of tense (present, past, future) are reduced to the
infinitive form of the verb. According to S. Buk [19, p. 298] analytical tense forms
are reviewed as syntactic forms and each component is considered as a separately
registered word.
Participle – case forms of singular and plural are given as the root form of the
masculine gender.
Numerals, pronouns, prepositions, particles and interjections are not given in the
dictionary, because they are not terminological units. Instead, meaningful commonly
used words, such as nouns, adjectives, verbs, adjectives, adverbs and participles, are
listed.
Our research has been based on the instructions sorted according to Anatomical
Therapeutic Chemical Classification System; 195 instructions have been analyzed.
Thus, the frequency dictionary of Ukrainian language includes 1000 words (see
Tab.1), but we shared in this article their English equivalents, the frequency dictionary
of Polish language includes 1000 words and the frequency dictionary of German
language includes 1000 words [21, p. 26].
Table 1. Ukrainian frequency dictionary.
Rank Word Absolute frequency Relative frequency
1 patient 1604 1,879012
2 preparation 1514 1,773581
3 application 1462 1,712666
4 dose 1349 1,580291
5 treatment 1076 1,260485
6 mg 1062 1,244084
7 to be 1043 1,221827
8 to be able 752 0,880933
9 research 718 0,841104
10 trace 665 0,779017
11 reaction 564 0,6607
12 therapy 552 0,646643
13 impairment 525 0,615013
14 patient 510 0,597442
15 ml 494 0,578698
16 level 492 0,576355
17 introduction 488 0,57167
18 clinical 483 0,565812
19 concentration 481 0,563469
20 to apply 430 0,503725
It is important that the dictionaries based on the corpora bring the users facts about the
real functioning of the language, since the corpus-based studies of the language
through the volume of the analyzed material and the technical capabilities of the
corpus toolkit can reveal such linguistic realia that scientists have not even suspected
to exist.
Since terms are coined in a professional environment and are mainly used in
professional texts, in particular in the instructions to medical products, it is possible to
introduce special meaningful components which have specific meanings and thus can
perform the classification function.
Undoubtedly, in addition to one-component terms, the systematic nature is given to
medical terminology by complex terms, which are formed by borrowed components
(which can be located both in the preposition and in the postposition) with the
nominal basis.
The main trends in term coining, found in the frequency dictionaries to the
instructions to medical products, are one-component terms and composite terms.
Composite terms have an advantage over term combinations, since they are more
economical and they serve as a word-building base for the derivative words: ukr.
hinekolohiia – hinekoloh – hinekolohichnui. Components, used in the language as
independent words (akusher-hinekoloh), are used to coin complex and compound
terms, as well as components that are not used independently, such as: lat. -algia,
auto-, bio-, gastro-. Thereby, it is necessary to describe not only independent words in
the dictionary, but also elements of the terms. Since, focusing on international
elements of term and morphemes, derived from Latin words, the user of the dictionary
can better remember, or even understand the name, still unknown to him. For
example, lat. rhil - 'inclination', lat. -lyt- 'destroying', lat. -haem- 'blood', lat. -troph-
'nutrition', lat. -gen- 'producing', lat.'- oid-' similar'.
It is obvious that international term elements and morphemes function only in the
Ukrainian and German instructions. This is especially noticeable after comparing
terms given in the frequency dictionaries to the instructions to medical products. For
example: a suffix of Latin origin lat. itis – ukr. -it-(-yt-) represents the meaning
‘inflammation’: ukr. apendytsyt, pol. zapalenie wyrostka robaczkowego, germ. die
Appendizitis 'inflammation of the appendix'. While compilers of Ukrainian and
German instructions and dictionaries use the international Ukrainian suffix -it-(- yt-)
or German – itіs, in Polish sources there is a term derived on a national basis.
A certain problem is caused by the transfer of terms-eponyms in the Ukrainian
language, since under the influence of extralinguistic factors, the compilers of
Ukrainian-language instructions use Russian transliteration.
As for the German umlauts, we follow the instructions of the Ukrainian-Latin-
English Medical Encyclopedic Dictionary compilers [1], which emphasize on the
necessity to reproduce the sound image, since the German umlauts do not have exact
equivalent in the Ukrainian phonetic system. Thus, the most similar sounds must be
chosen: o - labial [e] – ukr., [ü], u – labial – [i] – ukr.: Gete, Shenbakh, Shreder,
Minkhen (not Miunkhen). When original sound image is chosen during translation,
we get the most accurate pronunciation of the word.
The terms-eponyms, for example: ukr. – Bazedova khvoroba - pol. choroba
Basedowa – germ. Basedowkrankheit / Morbus Basedow (in German-language
instructions, terms-eponyms are represented not only by term combinations, but also
by composite terms). Obviously, the use of terms-eponyms is practical for specialists
in the medical and pharmaceutical industries, but embarrasses the non-professionals.
That is why it is necessary to indicate a synonym for such a term, or to explain it.
Due to the quantitative indicators in frequency dictionaries, it was discovered that
the terms-eponyms are used differently by the compilers of the Ukrainian-language
instructions: S. Stivensa or S. Styvensa (according to the system of Russian
transliteration). Existence of two variants of the writing of one term-eponym indicates
the need to standardize the Ukrainian rules of transliteration.
The surname components have two- or more - stems and also include a root word
that is the general name: the syndrome and proper name (surname or surnames) of
scientists, which semantically specifies the term. Dual transliteration, both on the
basis of graphical (visual) and sound images in the text of instructions to medical
products, lead to inaccurate reproduction of terms.
Thus, frequency dictionaries provide information on the frequency of the use of
terminological synonyms (doublets) and variating forms. Data on the frequency and
consistency of term combination as well as on the functioning of international term
elements are important. To identify lexical minimum and to compile terminology
dictionaries, information about the kernel and the periphery is highly important,
because such data indicate the 'archaization' / 'neologization' of the system.
Compiling terminological dictionaries the peculiarities of the term formation of
each particular terminology should be taken into account, therefore the compilation of
frequency dictionaries is important because they represent not only professional
terminology but also reflect the present state of language.
3 Conclusions
The source of the research is based on the instructions sorted according to
Anatomical Therapeutic Chemical Classification System. 195 instructions to medical
products have been analyzed.
Linguistic research corpus contains three sub-corpora: Ukrainian, Polish and
German. Frequency dictionaries contain 3000 words that are specific to medical
products.
The results of the study confirm the need for further research on medical
terminology. The processed material allows not only to reveal the peculiarities of
Ukrainian, Polish and German terminology, but also to understand better the
differences between Ukrainian, Polish and German cultures. It is necessary to
continue improving the standardization of medical terminology and compiling
dictionaries of various types on the basis of instructions to medical products.
The current state of intercultural communication and the world globalization
requires such studies on the material of other languages.
References
1. Petrukh L., Holovko I.: Ukrainian-Latin-English Medical Encyclopedic Dictionary, VSV
"Medicine", Kyiv (2012).
2. Badziński A.: Medyczny słownik kolokacji polsko-angielski-angielsko-polski, Warszawa
(2011).
3. Nechai S.: Russian-Ukrainian Medical Dictionary with foreign names, Fund "Third
Millennium", Kyiv (2000).
4. Baranov A.: Automation of linguistic research: the body of texts as a lingual-caustic
problem. In: Russian studies today, vol. 1-2., pp. 179-191. (1998).
5. Baker M.: Corpora in translation studies: An overview and some suggestions for future
research, pp. 223-243. Target (1995).
6. Hardie A, Baker M, McEnery T.: A Glossary of Corpus Linguistics, Edinburgh Univ
Press, (2006).
7. Buk S., Levus Ye., Yavorskyi Ye.: Algorithm for reflecting changes in the lexical
saturation of the text. In: Scientific journal of Lviv Polytechnic National University, vol.
771, pp. 349-353. Lviv (2013).
8. Karpilovska Ye: Introduction to Applied Linguistics: Computer Linguistics, LLC
"Southeast", Donetsk (2006).
9. Maksymiv O.: Corpus of the Persian language as a source of material for the frequency
dictionary, Eastern world,vol. 4, pp. 109-114. (2008).
10. Shyrokov V., Buhakov O., Hriaznukhina T., Kostyshyn O., Kryhin M.: Corpus linguistics,
Dovira, Kyiv (2005).
11. Demska O.: Text Corpus: the idea of another form, VPTS NaUKMA, Kyiv (2011).
12. Świdziński M.: Lingwistyka korpusowa w Polsce – źródła, stan, perspektywy. In:
LingVaria, vol. 1 pp. 23-32. (2006).
13. Sinclair J.: Corpus, Concordance, Collocation, Oxford University Press, Oxford (1991).
14. Regulatiry and directive documents of the Ministry of Healthcare of Ukraine,
www.mozdocs.kiev.ua/liki.php, last accessed 2018/03/21.
15. Rejestr Produktów Leczniczych, http://pub.rejestrymedyczne.csioz.gov.pl.375, last
accessed 2018/03/21.
16. European Medicines Agency, www.ema.europa.eu/drugs, last accessed 2018/03/21.
17. Kwic, http://www.chs.nihon- u.ac.jp/eng_dpt/tukamoto/kwic_e.html, last accessed
2018/04/02.
18. Buk S.: Statistical characteristics of the novel “Basis of the Society” by Ivan Franko (based
on the frequency dictionary of the novel). In: Scientific journal of Lviv Polytechnic
National University, vol. 676, pp. 90-93. Lviv (2010).
19. Buk S.: Quantitative parametrization of Ivan Franko’s texts: project and its realization. IIn:
Scientific journal of Lviv Polytechnic National University, vol. 58, pp. 290-307. Lviv
(2013).
20. Perkhach R.-Yu.: Terminology in patient information leaflets: cognitive linguistic and
cultural linguistic aspects (based on Ukrainian, Polish, German): dys. kand. filol. n.:
10.02.15, Odesa (2017) .
21. Perkhach R.-Yu, Shyika Y.: The methology of frequency dictionaries to the instructions to
medical products. (based on trilingual corpus) In: Materials of 13th International Scientific
and Technical Conference on Computer Sciences and Information Technologies (CSIT
2018), vol. 2, pp. 26-30. Lviv (2018).