Exploring Terminology Transfer: Greek-French Perspectives in Employment and Economy

Abstract

The study of specialised language is enabled through the analysis of lexical units at the morphological, syntactic, and semantic levels. Various theories of terminology approach the fundamental questions of the field from different perspectives [15], [25]. This paper presents a study based on the Theory of Textual Terminology, focusing on the specialised language and terminology in Public Administration documents from Greece and France, specifically within the domains of Employment and Economy. The research adopts a corpus-driven approach, utilising an original, static, synchronic, bilingual, and comparable corpus. The paper discusses the study's findings, including the methodology for terminology extraction and the potential for reusing the resulting resources.

Keywords

• eol>Corpus Linguistics
• Textual Terminology
• LSP
• Language of the Public Administration
• Comparative study
• Semi-automatic Term extraction
• FAIR principles
• 1

1. Introduction

4th International Conference on "Multilingual digital terminology today. Design, representation formats and management systems" (MDTT) 2025, June 19–20, 2025, Thessaloniki, Greece Email: [e-mail] 0000-0002-7231-0863 (A. Christopoulou) © 2025 Copyright for this paper by its authors.

Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). text. As L’Homme explains, “the term is realised within texts2,” even though it is understood in relation to an extralinguistic element: the thematic field [ 15 ]. Lexical elements that may initially appear to belong to general language can, upon closer analysis, be shown to function as specialised terms within a specific domain, carrying meanings unique to that field. The meaning and terminological properties of lexical elements are shaped by the thematic field, their level of usage, and the communicative context. Understanding specialised texts requires a contextual approach to language, considering text production and setting. Thus, terminology must move beyond solely norm-setting, onomasiological methods. According to L’Homme, once terms are identified as lexical units (single-word terms), the focus of terminology and terminography shifts to examining word combinations where these terms appear together –referred to as lexical collocations [ 15 ]. Collocations are defined as word pairs or groups that frequently co-occur and are characterised by three key features: their frequency of co-occurrence, their proximity in the text, and their collocational strength. Corpus pattern analysis is used to explore how the surrounding words (syntagmatic criteria) reveal the varied meanings of a term with multiple interpretations [ 9 ] or of a word which obtains a specialized meaning in relation to its context and the register.

2. Research questions

This research focuses on the specialised language3 and terminology used in Public Administration (PA) documents issued in Greece and France between 2010 and 2018 within two thematic domains: a) Employment and b) Economy, which reflect the political identity of any administration. The study specifically examined: a) the morphosyntactic structure of recurring word patterns and grammatical constructions in (PA) documents from both administrations, b) the consistency in the use of PA language compared to the morphological variation of terms, and c) the functional equivalence of multi-word complex terms (composed of a head and an expansion [ 11 ]) in the two languages, based on comparable corpora.

3. Methodological framework

The principles of authenticity, representativeness, and sampling were applied to compile the corpus [ 24 ]. A specialised bilingual corpus was constructed, designed to be comparable, static, comprehensive, and synchronic. It consists of 679 PA documents, encompassing a total of 4,119,530 words The textual material was divided into four sub-corpora –two for each thematic field– across the two languages. The criteria for selecting textual data from Greek and French sources included: a) documents published by Public Administration entities, b) available in both printed and/or electronic formats, c) specialised texts, d) relevant to the two thematic fields (Employment & Economy), and e) covering the period from January 2010 to December 2018. The widely used and highly dynamic software Sketch Engine, known for its advanced technology and regularly updated tools and features, was selected for text analysis and processing in this research.

4. Term identification and retrieval

Α semi-automated approach was used to examine a) the formation of multi-word complex terms (MWCTs) and the most frequent collocations in Greek and French, as found in the bilingual corpus, 2 «(…) le terme se matérialise dans les textes» ([ 15 ], p. 81). 3 The terms special language ([ 21 ], [ 5 ]) and specialised language ([ 22 ], [ 4 ]) are used interchangeably. They both refer to language used in specific domains or professions. A more prevalent term in academic circles is Languages for Specific Purposes (LSP) ([ 12 ], [ 19 ], [ 2 ], [ 3 ], [ 26 ]), which encompasses both special and specialised languages. LSP is defined as the “natural language used in communication between experts in a domain and characterised by the use of specific linguistic means of expression” [ 13 ]. and b) the most frequent and statistically significant terms and collocations as well as broader clusters, regardless of their frequency.

Initially, a word list was extracted, followed by the generation of a keyword list using the relevant tools in Sketch Engine. The GreekWeb 2014 corpus (elTenTen14) was chosen as the reference corpus for Greek, while the FrenchWeb 2012 corpus (frTenTen12) was selected for French4. The research focused on nouns, as they constitute the primary building blocks of terms (both single-word and multi-word) and play a crucial role in forming collocations and larger lexical clusters. To ensure a comprehensive analysis of the wordlist, the inclusion of a keyword list was essential. To this end, the keyword tool was utilised with the single-words tab selected for identifying single-word terms. The data was analysed based on relative frequency and score to extract keywords from each subcorpus. The most frequent lexical elements were then identified in comparable language pairs. Specifically, the study aimed to determine whether semantically and translationally equivalent words –referred to as node terms (NTs)–could be found a) in the wordlist and/or b) in the keyword list of the comparable corpora. An additional statistical factor considered was the collocational strength of the co-occurring words that form terms and collocations, which could serve as a criterion for automatic term extraction. As noted by Ananiadou and Zervanou [ 27 ], the measurement of mutual information reflects the degree to which a word provides insight into other words that tend to co-occur with it.

Once common NTs were identified across the comparable corpora, the research proceeded with a contrastive analysis of the findings. To select the NTs for further examination, the following cumulative criteria were applied: a) The NT must appear in the wordlist, indicating its frequency within the corpus; b) It should be semantically and pragmatically significant; c) It must be representative of the thematic field relevant to the specific text type, and while not a strict requirement, it is preferable that the NT also appears in the keyword list; d) It should facilitate the formation of complex terms and collocations. Based on the assumption that Greek and French terms would correspond, this contrastive study used quantitative and qualitative analysis to determine the extent and type of correspondences for the selected NTs within the two comparable corpora. Fourteen terms were selected from each language. Of these, six belong to the thematic field of Employment in Greek, six to Employment in French, six to the thematic field of Economy in Greek, and six to Economy in French. Additionally, two NTs are consistently present across all four subcorpora and they appear in both thematic fields and in both languages. The terms in question appear within the 350 most frequent words in the wordlist. Given that the frequency list comprises over 10,000 entries for each sub-corpus, it becomes evident that there is a correspondence and equivalence across the four sub-corpora concerning the occurrence frequency of the selected NTs and the hypothesis is confirmed by the experiment [ 6 ].

5. Highlights of the analysis

The findings of the study include: 1) An examination of the methods used to form MWCTs in Greek and French, along with the ways these terms are expanded. MWCTs can generally be analysed into two parts, particularly when bilexical: the immediate constituents, comprising a head constituent and a modifier constituent. As term element is defined, any morphological part of a term that carries meaning or has semantically differentiating value [ 28 ]. Tri-lexical terms and broader clusters, comprised of more than two term elements, are formed by adding new terms to the core of the bi-lexical term, thereby augmenting the 4 The Greek corpus Greek Web 2019 (elTenTen19), containing 2,342,091,029 words, became available in Sketch Engine as a reference corpus in 2022. However, the research for this study had already been completed, and the new corpus was not used. The French corpus French Web 2023 (frTenTen23), containing 23,874,070,858 billion words, became available in Sketch Engine as a reference corpus in 2023. However, the research for this dissertation had already been completed, and the new corpus was not used. terminological structure [ 29 ]. According to L'Homme [ 14 ], MWCTs in French are formed using a finite number of grammatical category sequences. “Essentially, most complex French terms are formed by combining a noun modified by an adjective (...), by another noun (...), or by a prepositional phrase”. In fact, there are no 4-word terms and co-occurring word patterns that are more than 3 words are considered as lexical clusters. Apart from the co-occurrence frequency the collocational strength of the term elements is one of the elements under study, therefore the examination of the MWCTs must be limited to the tri-word terms. The prepositions do not count as term elements. According to Riegel, Pellat & Rioul [ 20 ] any preposition or prepositional phrase can introduce a nominal complement. The semantic interpretation of this complement is contingent upon both the inherent meaning of the preposition and the contextual relationship between the elements it connects. For the measurement of the collocational strength the log Dice score was deemed as the most suitable for the analysis of MWCTs, as this index allows for the comparison of collocations across corpora, in contrast to other statistical measurements of collocational strength. The value of log Dice is independent of the corpus size, which allows researchers to compare textual data consistently and stably, and for this reason, it is preferred in the qualitative analysis of very large corpora. Usually, the values of statistical indices are dependent on the reference corpus, thus rendering it less practical for comparing collocations from comparable corpora or for data interpretation. 2) A structural and morphological analysis of MWCTs derived from the selected NTs in the comparable corpora. 3) A semantic analysis of structures forming bi-lexical terms in both Greek and French. 4) A contrastive comparison of the functional equivalence of MWCTs. 5) A quantitative and qualitative analysis of MWCTs based on the corpora. 6) An exploration of the characteristics of specialised language in administrative documents in Greece and France. Among the characteristics of the specialised language of PA are the long-period sentences, high frequency of conjunctions (indicative of numerous subordinate clauses), word repetition in lieu of pronoun usage (to ensure referential precision), adverbial clauses, nominal groups, repetitive collocations, elaborate syntax, extensive nominalisation, use of the passive voice, use of transitive verbs.

The language of the PA constitutes a register, a term that refers to the functional variety of language and related to a given type of context or situation type. The study of the textual material from the four comparable sub-corpora yielded the following results: a) Statistical Data on MWCTs from the Corpora: A consolidated table presenting term structures from both corpora, along with a visualisation of the NTs that most frequently participate in term formation. The same NTs in both languages, μέτρο and δαπάνη (from the Greek corpus) and dépense and mesure (from the French corpus), were found to form the most numerous MWCTs. b) Lemmatological Lists of MWCTs: The Greek corpus contains 745 MWCTs, while the French corpus includes 641, resulting in a total of 1,386 MWCTs. c) Lemmatological Lists of Acronyms and Abbreviations: Using the keyword lists from each of the four sub-corpora, shortened forms were manually extracted into four distinct lemmatological lists. A notable number of acronyms and abbreviations were identified in the French sub-corpora. d) Lemmatological List of Named Entities: The list includes only those named entities that are associated with the NTs being analysed. From the analysis, it follows that the Greek Corpus contains a greater number of named entities compared to the French Corpus. Specifically, 157 named entities from the Greek Corpus, and 69 named entities from the French Corpus are included in the lemmatological list from the selected NTs. e) Bilingual Glossary of Term Equivalences: A contrastive analysis of the terms at the pragmatic level led to the compilation of a bilingual glossary of common terms from both corpora with semantic and translational equivalence –or at least pragmatic correspondence. The glossary comprises 90 term equivalences. f) Correspondence in the Morphosyntactic Analysis of Terms: Correspondence was observed in the morphosyntactic structure of terms in both Greek and French, with symmetry found in the formation of MWCTs from specific parts of speech in both languages.

The comparison of the lexical data reveals that the 447 Greek bi-lexical terms adhere to three syntactic structures, with the most common being: i) Noun + Noun in genitive (N+Ngen), example εισφορά αλληλεγγύης (/isforá allilengíis/, in English: solidarity levy). ii) Adjective + Noun (A+N), example, αδήλωτη εργασία (/adíloti ergasía/, in English: undeclared work). In the French corpus, 425 bi-lexical terms were identified, formed using four syntactic structures, with the most frequent structures being: i) Noun + Preposition + Noun (N+PRE+N), example mesures de protection (mə.zyʁ də pʁɔ.tɛk.sjɔ̃/, in Eglish: protective measures). ii) Noun + Adjective (N+A), example travail illegal (/tʁa.vaj i.le.ɡal/, in English: illegal work). It is observed that the French structure Noun + Preposition + Noun (N+PRE+N) aligns with the most frequent structure of bi-lexical terms in Greek (N+Ngen), as the preposition de in French primarily functions to express the genitive case. This demonstrates perfect symmetry between the two languages in the formation of bi-lexical terms.

The comparative analysis of the syntactic structures of three-word terms from the corpora revealed a total of 318 Greek three-word terms and 223 French three-word terms. In total, fifteen distinct syntactic structures were identified for Greek three-word terms, compared to ten for their French counterparts.

In the Greek corpus, the most frequent three-word terms adhere to the following structures: i) Noun + Adjective in genitive + Noun in genitive (N+Adjgen+Ngen), example σύστημα κοινωνικής ασφάλισης (/sístima kinonikís asfálisis/, in English: social security system). ii) Noun + Noun in genitive + Noun in genitive (N+Ngen+Ngen), example Ημερολόγιο Μέτρων Ασφάλειας (/imerolóyio métron asfálias/, in English: Safety Measures Log Book).

In the French corpus, the most frequent syntactic structure for the three-word clauses of the node terms (NTs) is: i) Noun + Preposition + Noun + Adjective (N+PRE+N+Adj), example, ratio de dette publique (/ra.si.o də dɛt py.blik/, in English: public debt ratio) followed in frequency by: ii) Noun + Preposition + Noun + Preposition + Noun (N+PRE+N+PRE+N), example rupture du contrat de travail (/ʁyp.tyʁ dy kɔ̃.tʁa də tʁa.vaj/, in English: termination of employment contract).

The analysis of the syntactic structures of three-word terms in the French corpus reveals a notably high frequency of prepositions, particularly the preposition de, in the formation of MWCTs. Both the quantitative and qualitative analysis of the three-word terms in the corpus demonstrate a clear correspondence and symmetry in the predominant formation patterns across both languages.

The metadata for all documents in the bilingual corpus were recorded across four tables, one for each sub-corpus. Each table contains six columns detailing the document's serial number, title, issuing authority, issue date, ADA (Document Internet Posting Number), and registration number. The documents were organised chronologically by year, ranging from 2010 to 2018.

6. Applications

The research has applications in: a) Lexicography: Using the methodology applied, specialised vocabulary from PA documents can be systematically extracted for purposes such as compiling a dictionary, preparing a terminology list, and/or creating a thesaurus. In its electronic format, the thesaurus can be linked through electronic references to the relevant legislation and institutional framework governing the terms or phrases being searched. Additionally, it can include contextual examples illustrating the usage of these MWCTs or clusters. b) Translation of PA Documents: Identifying established MWCTs, collocations, and broader clusters associated with specific PA concepts improves translation quality and addresses challenges in translating regulations and directives issued by supranational institutions. c) Terminology Teaching: The methodology applied can be effectively used in teaching specialised terminology. By examining collocations and terms in authentic texts while considering the context in which they appear, comprehension is enhanced, contributing to the training of learners in specific scientific disciplines or professional fields [30]. Learners can leverage parallel or comparable texts and corpora to identify recurring patterns and understand the usage of terms or collocations in realworld contexts. Analysing the concordancer provides valuable insights into the full semantic range of a term, even without the need for a dictionary. d) Applications for Natural Language Processing: PA entities can develop a comprehensive software system integrating various language technology applications, such as an enhanced spell checker with domain-specific terminology and collocations, a grammar checker to identify syntactic and orthographic errors, a dictionary of acronyms and named entities, a dedicated website spell checker, and other language validation tools.

7. Critical analysis

Linguistic techniques for extracting candidate terms primarily aim at identifying recurring patterns grounded in the morphological and syntactic structures specific to term formation within a given cognitive field or specialised domain. Since conventional sources of terminology and lexicography (such as specialised dictionaries, glossaries, and thesauri) no longer meet the needs of their users [ 16 ], [ 17 ] suggest broadening the search for terms to include parallel texts, electronic corpora, and the World Wide Web, to access a rich variety of contexts that better reflect the behavior of terminological units. Frérot and Pecman [ 10 ] emphasise that with the rise of specialised corpora providing textual resources in digital formats, alongside computational tools for searching and analysing these materials, specialised languages have become more accessible, and their analysis has become more practical. Terminological research utilising specialised corpora offers reliable evidence for creating specialised dictionaries, standardising terminology, defining terms, and ensuring consistency across documents exchanged within professional sectors, such as the PA documents. As Pearson [ 18 ] observes, analysing a term within a specialised corpus is particularly useful for newly introduced terms that may not yet appear in a dictionary or when learners lack access to suitable specialised dictionaries. This is very common for the Public Administration vocabulary.

A key application of the proposed methodology for terminology extraction from PA documents is the creation of databases that enable users to search and retrieve information from administrative documents. This can be achieved either by formulating natural language queries for individual users or through automatic extraction processes in Natural Language Processing (NLP) systems. The integration of linguistics and computer science has significantly influenced the field of terminology, leading to the development of high-quality and reliable terminology technology applications [ 31 ].

Specialised corpora provide data and methods for analysing large volumes of text, which are necessary for cognitive engineering. These corpora help develop tools that support reasoning and knowledge representation, such as semantic networks or conceptual diagrams. By integrating realworld text data from specialised corpora, scholars can enhance and validate the representations used in cognitive engineering [ 7 ], [ 10 ]. The inclusion of metalinguistic data and concordancers, where terms frequently occur, helps confirm the terminological nature of collocations. Additionally, the conceptual organisation and classification of terms play a crucial role in the development of thesauri and terminologies. This approach forms an open scientific framework in which the principles of FAIR can be applied to manage scientific data, ultimately promoting the reuse and dissemination of scientific knowledge. The FAIR principles constitute a set of guidelines that ensure data is findable, accessible, interoperable, and reusable [ 25 ], all the above-mentioned very important and useful to the management of the PA register.

Declaration on Generative AI

During the preparation of this work, the authors used ChatGPT-4 for grammar and spelling checks. The authors have subsequently reviewed and edited the content and take full responsibility for the publication’s final version.

8. References

References

[1] Baker , P. ( 2006 ). Using Corpora in Discourse Analysis . London, New York: Continuum.

[2] Beaugrande , R. ( 1987 ). Special Purpose Language and linguistic theory . The Copenhagen School of Economics , 10 .

[3] Bergenholtz , H. , & Tarp , S. ( 1995 ). Manual of Specialised Lexicoraphy: the preparation of specialised dictionaries (Vol. 12 ) . (H. Bergenholtz , & S. Tarp, Eds.) Amsterdam/Philadelphia: John Benjamins B.V.

[4] Bowker , L. , & Pearson , J. ( 2002 ). Working with Specialized Language: A Practical Guide to Using Corpora . London & New York: Routledge.

[5] Cabré , M. ( 1998 ). La terminologie: théorie, méthode et applications . ( M. Cormier , & J. Humbley , Μεταφρ.) Ottawa: Les Presses de l' Université d' Ottawa.

[6] Christopoulou , A. ( 2024 ). A corpus-driven contrastive study of Greek-French Public Administration specialised language and Terminology in the fields of Employment and Economy . Doctoral Thesis , French Language and Literature , Athens.

[7] Condamines , A. ( 2005 ). Linguistique de corpus et terminologie . Langages , 157 ( 1 ), p. 36 - 47 . https://doi.org/10.3917/lang.157.0036.

[8] Condamines , Α. , & Picton , Α. ( 2022 ). Textual Terminology: Origins, Principles and New Challenges . In M.- C. L'Homme , & P. Faber, Theoretical Perspectives on Terminology. Explaining terms, concepts and specialized knowledge (p. 219 - 236 ). John Benjamins Publishing Company.

[9] Faber , P. , & L'Homme , M.- C. ( 2014 ). Lexical Semantic approaches to reminology: An introduction . Terminology. International Journal of Theoretical and Applied Issues in Specialized Communication , 20 ( 2 ), p. 143 - 150 . https://doi.org/10.1075/term.20.2.01int.

[10] Frérot , C. , & Pecman , M. (Eds.). ( 2021 ). Des corpus numériques à l'analyse linguistique en langues de spécialité . Grenoble: UGA Éditions. https://doi.org/10.4000/books.ugaeditions. 24195 .

[11] Hamon , T. , & Nazarenko , A. ( 2001 ). Detection of synonymy links between terms: Experiment and Results . In D. Bourigault, C. Jacquemin , & M.-C. L'Homme (Eds.), Recent Advances in Computational Terminology (Vol. 2 , p. 185 - 208 ). Amsterdam/Philadelphia: John Benjamins Publishing Company.

[12] Hoffmann , L. ( 1979 ). Towards a Theory of LSP. Elements of a Methodology of LSP Analysis . Fachsprache , 1 ( 1-2 ), p. 12 - 17 .

[13] ISO 1087 : 2019 . Terminology work and terminology science - Vocabulary

[14] L 'Homme , M.- C. ( 2004 ). La terminologie: principes et techniques . Québec: Les Presses de l' Université de Montréal.

[15] L 'Homme , M.- C. ( 2005 ). Sur la notion de «terme». Meta. Journal des traducteurs , 50 ( 4 ), p. 1112 - 1132 . https://doi.org/10.7202/012064ar.

[16] Liczner , A. ( 2016 ). Une contribution à l'amélioration des ressources terminographiques : Étude terminologique fondée sur un corpus de textes de spécialité du domaine du droit de l'Internet . Thèse de doctorat, Université Lumière Lyon 2 , École doctorale 484 3LA - Lettres , Langues, Linguistique et Arts, Lyon.

[17] Pantazara , M. , & Tziafa , E. ( 2018 ). Les termes de la crise économique grecque dans les corpus . Meta , 63 ( 3 ), p. 739 - 765 . https://doi.org/10.7202/1060171ar.

[18] Pearson , J. ( 1996 ). Strategies for identifying terms in specialised texts . TEANGA: The Irish Yearbook of Applied Linguistics(16) , p. 33 - 42 . Retrieved 03 17, 2025 , from https://files.eric.ed.gov/fulltext/ED414748.pdf.

[19] Picht , H. ( 1987 , 06). Terms and their LSP Environment - LSP Phraseology. Vers l' an 2000 . La terminotique , bilan et prospectives , 32 ( 2 ). https://doi:10.7202/003836ar.

[20] Riegel , M. , Pellat , J. , & Rioul , R. ( 2009 ). Grammaire méthodique du français . Paris: Quatridge - PUF.

[21] Sager , J. C. ( 1990 ). A Practical Course in terminology Processing . Amsterdam/Philadelphia: John Benjamins B.V.

[22] Sinclair , J. , & Ball , J. ( 1996 ). EAGLES Preliminary Recommendations on Text Typology. Internal Working Document. Retrieved 03 17 , 2025 , from http://www.ilc.cnr.it/EAGLES96/texttyp/texttyp.html.

[23] Teubert , W. ( 2004 ). Lexicology and Corpus Linguistics . An Introduction . In M. Halliday, W. Teubert , C. Yallop , & A. Čermáková . London & New York: Continuum.

[24] Tognini-Bonelli , E. ( 2001 ). Corpus Linguistics at Work. Amsterdam& Philadelphia: John Benjamins Publishing Company.

[25] Vezzani , F. ( 2022 ). Terminologie numérique : conception, représentation et gestion (Vol. 290 ). Bern, Berlin, Bruxelles, New York, Oxford: Peter Langa. https://doi:10.3726/b19407.

[26] Wright , S. , & Budin , G. ( 1997 ). Handbook of terminology management . Basic Aspects of Terminology Management (Τόμ. 1) . Amsterdam, Philadelphia: John BenjaminsPublishing Company.

[27] Ανανιάδου , Σ. , & Ζερβάνου , Κ. ( 2004 ). Αναγνώριση όρων . In Μ. Κατσογιάννου, & Ε. Ευθυμίου (Eds.) , Ελληνική Ορολογία: Έρευνα και Εφαρμογές (p. 283 - 311 ). Αθήνα: Εκδόσεις Καστανιώτη ΑΕ.

[28] Βαλεοντής , Κ. , Ζερίτη , Κ., & Νικολάκη , Α. ( 1999 ). Το προσδιοριστικό σύνθετο του ελληνικού σύμπλοκου όρου και η χρήση της γενικής . ΕΛΕΤΟ - 2ο Συνέδριο Ελληνική Γλώσσα και Ορολογία , (p. 283 - 316 ). Αθήνα.

[29] Γαβριηλίδου , Μ. , & Λαμπροπούλου , Π. ( 2004 ). Εξαγωγή όρων από κείμενα. Μια υβριδική μέθοδος . In Μ. Κατσογιάννου, & Ε. Ευθυμίου (Eds.) , Ελληνική Ορολογία: Έρευνα και Εφαρμογές (p. 313 - 326 ). Αθήνα: Ινστιτούτο Επεξεργασίας του Λόγου - Εκδόσεις Καστανιώτη Α .Ε.

[31] Κατσογιάννου , Μ. , & Ευθυμίου , Ε. ( 2004 ). Θεωρία, μέθοδοι και πρακτικές της ορολογίας . In Μ. Κατσογιάννου, & Ε. Ευθυμίου (Eds.) , Ελληνική Ορολογία: Έρευνα και Εφαρμογές (p. 25 - 60 ). Αθήνα: Εκδόσεις Καστανιώτη ΑΕ.