AIDAinformazioni Anno 42 – N. 3-4 – luglio-dicembre 2024

AIDAinformazioni

RIVISTA SEMESTRALE DI SCIENZE DELL’INFORMAZIONE

NUMERO 34

ANNO 42

LUGLIODICEMBRE 2024

cacucci

editore

bari

Editrice: Cacucci Editore S.a.s.

Via D. Nicolai, 39 – 70122 Bari (BA)

www.cacuccieditore.it

e-mail: riviste@cacuccieditore.it

Telefono 080/5214220

Proprietario della rivista:

Università della Calabria

Direttore Scientico:

Roberto Guarasci, Università della Calabria

Direttore Responsabile:

Fabrizia Flavia Sernia

Comitato scientico:

Anna Rovella, Università della Calabria;

Maria Guercio, Sapienza Università di Roma;

Giovanni Adamo, Consiglio Nazionale delle Ricerche †;

Claudio Gnoli, Università degli Studi di Pavia;

Ferruccio Diozzi, Centro Italiano Ricerche Aerospaziali;

Gino Roncaglia, Università della Tuscia;

Laurence Favier, Université Charles-de-Gaulle Lille 3;

Madjid Ihadjadene, Université Vincennes-Saint-Dénis Paris 8;

Maria Mirabelli, Università della Calabria;

Agustín Vivas Moreno, Universidad de Extremadura;

Douglas Tudhope, University of South Wales;

Christian Galinski, International Information Centre for Terminology;

Béatrice Daille, Université de Nantes;

Alexander Murzaku, College of Saint Elizabeth, USA;

Federico Valacchi, Università di Macerata.

Comitato di redazione:

Antonietta Folino, Università della Calabria;

Erika Pasceri, Università della Calabria;

Maria Taverniti, Consiglio Nazionale delle Ricerche;

Maria Teresa Chiaravalloti, Consiglio Nazionale delle Ricerche;

Assunta Caruso, Università della Calabria;

Claudia Lanza, Università della Calabria.

Segreteria di Redazione:

Valeria Rovella, Università della Calabria

AIDAinformazioni

R   S ’I

Fondata nel 1983 da Paolo Bisogno

AIDAinformazioni

R   S ’I

«AIDAinformazioni» è una rivista scientiﬁca che pubblica articoli inerenti alle Scienze dell’In-

formazione, alla Documentazione, all’Archivistica, alla Gestione Documentale e all’Organiz-

zazione della Conoscenza ma amplia i suoi conni in ulteriori campi di ricerca ani quali

la Terminologia, la Linguistica Computazionale, la Statistica Testuale, ecc. È stata fondata

nel 1983 quale rivista uciale dell’Associazione Italiana di Documentazione Avanzata e nel

febbraio 2014 è stata acquisita dal Laboratorio di Documentazione dell’Università della Ca-

labria. La rivista si propone di promuovere studi interdisciplinari oltre che la cooperazione e

il dialogo tra proﬁli professionali aventi competenze diverse, ma interdipendenti. I contributi

pubblicati arontano questioni teoriche, metodologie adottate e risultati ottenuti in attività di

ricerca o progettuali, denizione di approcci metodologici originali e innovativi, analisi dello

stato dell’arte, ecc.

«AIDAinformazioni» è riconosciuta dall’ANVUR come rivista di Classe A per l’Area 11 –

Gruppo Scientico Disciplinare 11/HIST-04 – Scienze del libro, del documento e storico-re-

ligiose e come rivista scientica per le Aree 10 – Scienze dell’antichità, ﬁlologico-letterarie e

storico-artistiche; 11 – Scienze storiche, ﬁlosoﬁche, pedagogiche e psicologiche; 12 – Scienze

giuridiche; 14 – Scienze politiche e sociali. È anche annoverata dall’ARES (Agence d’éval-

uation de la recherche et de l’enseignement supérieur) tra le riviste scientiﬁche dell’ambito

delle Scienze dell’Informazione e della Comunicazione. La rivista è, inoltre, indicizzata in:

ACNP – Catalogo Italiano dei Periodici; BASE –Bielefeld Academic Search Engine; ERIH

PLUS – European Reference Index for the Humanities and Social Sciences – EZB – Elektro-

nische Zeitschriftenbibliothek – Universitätsbibliothek Regensburg; Gateway Bayern; KVK

– Karlsruhe Virtual Catalog; e Library Catalog of Georgetown University; SBN – Italian

union catalogue; Ulrich’s; Union Catalog of Canada; LIBRIS – Union Catalogue of Swedish

Libraries; Worldcat.

I contributi sono valutati seguendo il sistema del double blind peer review: gli articoli ricevuti

sono inviati in forma anonima a due referee, selezionati sulla base della loro comprovata espe-

rienza nei topics speciﬁci del contributo in valutazione.

AIDAinformazioni

Anno 42

N. 3-4 – luglio-dicembre 2024

cacucci

editore

bari

  

2024 Cacucci Editore – Bari

Via Nicolai, 39 – 70122 Bari – Tel. 080/5214220

http://www.cacuccieditore.it e-mail: info@cacucci.it

Ai sensi della legge sui diritti d’Autore e del codice civile è vietata la

riproduzione di questo libro o di parte di esso con qualsiasi mezzo,

elettronico, meccanico, per mezzo di fotocopie, microfilms, registra-

zioni o altro, senza il consenso dell’autore e dell’editore.

Sommario

Contributi

A A, Il nuovo regolamento eIDAS e alcune “quisquilie”

archivistiche 9

F B, M T, Exploration du réseau numérique

YouTube autour de la santé des militaires: quelles sont les thématiques des

discours, les sources d’informations et les acteurs de la communication? 29

E C, L F, Assisted morbidity coding: the

SISCO.web use case for identifying the main diagnosis in Hospital

Discharge Records 51

V F, A humanistic approach to datafication 79

R P, Testimonianze di un impegno culturale per

l’Università di Salerno. Le carte di Alfonso Menna 101

F S, A B, E G,

S M, CompL-it: a Computational Lexicon of Italian 119

Rubriche

C G, Non solo libri 151

Contributi

AIDAinformazioni

ISSN 1121–0095

ISBN 979-12-5965-456-4

DOI 10.57574/596545646

pag. 119-148 (luglio-dicembre 2024)

CompL-it: a Computational Lexicon of Italian

Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Abstract: This paper describes CompL-it, a new open computational lexicon for contem-

porary Italian. The resource was constructed from three sources: an already available Italian

lexicon, a lemmatized list of inflected forms obtained from a morphological analyser, and a set

of treebanks. Integrating these resources required a standardisation process in accordance with

the standards of the Linguistic Linked Open Data community, which was necessary for the

subsequent conversion into the OntoLex-Lemon model. The resulting computational lexicon

comprises approximately 100,000 lexical entries, 790,000 forms, 57,000 senses, and 86,000

semantic relations. The lexicon, thanks to its rich and articulated linguistic structure, can be

used, as shown, to enhance information retrieval in the context of full-text search tasks.

Keywords: Computational Lexicon, Linguistic Resources, Linguistic Linked Open Data, On-

toLex-Lemon, Information Retrieval.

1. Introduction

While a significant number of digital lexical resources are available for many

languages (such as various multilingual WordNets) (Princeton University n.d.;

MultiWordNet n.d.; Global WordNet Association n.d.), only a few integrate

different layers of linguistic information, such as morphology, semantics, and

syntax. This is not surprising, as the construction of a computational lexicon

that conveys linguistic information across different layers can be an extremely

time-consuming task that requires advanced linguistic expertise.

CNR-Istituto di Linguistica Computazionale (ILC) “A. Zampolli”, Pisa, Italy. flavia.scio-

lette@ilc.cnr.it; andrea.bellandi@ilc.cnr.it; emiliano.giovannetti@ilc.cnr.it; simone.marchi@

ilc.cnr.it.

In this context, a computational lexicon can be defined as a resource that contains in-

formation about words, their meanings, and linguistic properties, designed to be used by

computer systems for tasks like natural language processing (NLP), machine translation, or

text analysis. It typically includes details such as word categories (e.g., noun, verb), syntactic

information, and semantic relationships.

120 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

On the other hand, the need for resources of this kind is long-standing.

Italian linguistics, for example, has always shown an interest for lexical data

(Sabatini 2006), which has been encouraged by the increasing availability of

many corpus-based resources. As documented in Chiari (2012) many projects

involving corpora (monolingual, parallel, domain-specific) have flourished

and both digitised traditional dictionaries and computational dictionaries

have taken advantage of them, for example to calculate the frequency of words

or to increase their lexical coverage (for example by adding neologisms). In

terms of exploitation, a number of applications are meant to take advantage of

lexical resources, such as sentiment analysis (Prakash and Aloysius 2021), and

«semantic role labelling, verb sense disambiguation, and ontology mapping»

(Brown et al. 2022, 2).

In the context of archival science and document management, the availabi-

lity of linguistic resources to support the organisation and retrieval of informa-

tion has been considered crucial for many years (Chen et al. 1995; Smith 1997;

Thompson et al. 2011). In these fields, the development of computational

lexicons can provide a fundamental contribution to the community, expan-

ding the potential for knowledge analysis and management. It is believed that

a resource capable of formalising a language’s lexical and semantic structures

in a complex way can improve the efficiency of archiving, classification, and

information retrieval activities, within a document management paradigm in-

creasingly supported by IT tools (Bamman and Crane 2010; Hmeidi et al.

2016; Passarotti and Mambrini 2021). The creation of increasingly efficient

tools for automating archival and document practices can greatly simplify the

management of large volumes of unstructured data, enhancing precision in

indexing and retrieving information. Furthermore, a computational lexicon

can serve as a key linguistic resource for building Knowledge Organization Sy-

stems (KOSs), such as ontologies and thesauri, crucial elements for knowledge

organisation (Hodge 2000; Shiri 2015).

In this work, we illustrate CompL-it, an Italian computational lexicon

built by leveraging existing resources, whose data have been thoroughly an-

alysed, extracted, converted, and interconnected. CompL-it has been made

freely available as Linguistic Linked Open Data (LLOD) on the CLARIN

repository (CLARIN-IT n.d.a).

2. State of the art

In order to define the state of the art regarding computational lexical re-

sources for the Italian language we first conducted a search on the Virtual

Language Observatory (CLARIN VLO n.d.) (VLO) of the European infra-

structure CLARIN (Common Language Resources and Technology Infra-

structure). This database, which contains hundreds of thousands of references

CompL-it: a Computational Lexicon of Italian 121

to language resources and tools, was browsed using the available filters. In

particular, a search was carried out by type of resource, selecting lexicalRe-

source, and specifying the Italian language. In addition, from the obtained list

we excluded automatically produced resources (i.e, not revised by hand), lists

of idiomatic expressions, lists of terms with no linguistic information at all,

multilingual named entities, sets of embeddings, parallel corpora, metadata,

resources that only appear by virtue of some references to the Italian language

and, finally, all the resources whose data are not open and freely available.

Among this last type of resources, however, it is worth mentioning BabelNet

(Navigli and Ponzetto 2012; BabelNet n.d.), Senso Comune (Vetere et al.

2011), and Italian FrameNet (Basili et al. 2017), particularly for the richness

of data they offer.

The resources identified on CLARIN VLO that met the above criteria were

14, and can be classified into two categories: 11 lexical resources and 3 termi-

nological resources.

The available lexical resource that, for this work, has been taken as the

main reference (and used as one of the sources) is LexicO (Sciolette, Giovan-

netti, and Marchi 2023), a multi-layered computational lexicon developed at

CNR-ILC and built from Parole-Simple-Clips (PSC) (Bel et al. 2000; Ruimy

et al. 2002; ILC4CLARIN CNR 2016). More details on the nature of LexicO

will be provided later in section 3.1.

Another very rich lexical resource, also developed at CNR-ILC, is ItalWor-

dNet (Roventini et al. 2003), available as an SQL dump on the VLO in its se-

cond version (Roventini, Marinelli, and Bertagna 2016). The VLO also men-

tions MultiWordNet (Pianta, Bentivogli, and Girardi 2002; MultiWordNet

n.d.), realised as an extension of Princeton’s WordNet (Miller 1995; Princeton

University n.d.), and which also includes data for the Italian language. Multi-

lingual language resources include OmegaWiki (Meijssen 2014), an open and

collaborative resource whose aim is «to describe all words of all languages with

definitions in all languages» and includes lexical, terminological and ontologi-

cal information (WikiMedia 2022).

In addition to the resources listed so far, which are structured as lexicons,

other resources are available for Italian that convey individual layers of lin-

guistic information. Building on the aforementioned PSC and ItalWordNet,

a resource that provides semantic data called the Italian Sense Inventory was

created. This resource was developed within the ELEXIS project (ELEXIS

n.d.) to support Word Sense Disambiguation tasks.

On the VLO, there are also two resources developed by the same author

that complement each other: Italian Function Words (Grella 2018a) and Ital-

ian Content Words (Grella 2018b). The former, as the name suggests, contains

Italian function words and is designed to support tasks such as POS tagging

and syntactic parsing. The second constitutes a morphological dictionary of

122 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

over 2 million inflected forms that, however, includes hundreds of thousands

of forms that, although morphologically correct, are not represented in linguis-

tic usage. The last two lexical resources we include in this review are Universal

Derivations (Kyjánek et al. 2021) and Universal Segmentations (Žabokrtský

et al. 2022), both multilingual, in which about 10 thousand Italian lemmas

are linked to their respective segmentations, derived forms and compounds.

With regard to the three terminological resources identified in the VLO we

first mention Geodomain WordNet, a collection of geographical terms linked

to the English and Italian WordNets (Frontini, Del Gratta, and Monachini

2016). The other two resources, developed within the Pan-Latin Terminology

Network (Realiter n.d.), are the Pan-Latin Lexicon of Collars and Sleeves in

Fashion and Costume (Zanola et al. 2023) and the Pan-Latin Textile Fibres

Vocabulary (Dankova, Zanola, and Calvi 2022).

Although not available on CLARIN VLO, Morph-it! (Zanchetta and Ba-

roni 2005) is a freely accessible and rich morphological resource for Italian,

consisting of 504,906 inflected forms and 34,968 lemmas. However, as the

authors note (Morph-it! 2018), because it is derived from an Italian newspaper

corpus, the resource has «many gaps in basic, every-day vocabulary».

Another interesting resource worth mentioning is SimpleLEX-IT, as it was

built similarly to CompL-it, i.e., by combining together different existing re-

sources (Mazzei 2016; SimpleLEX-IT n.d.). In particular, SIMPLELex-it was

developed by integrating morphological data from the previously cited Mor-

ph-it!, the Vocabolario di base della lingua italiana by Tullio De Mauro (De

Mauro 1980; 2016), two entries of the Italian Wikipedia concerning verbs

(Wikipedia 2024a; 2024b) and, finally, data from the Italian Universal De-

pendencies (UD) treebanks (Universal Dependencies n.d.a).

In the context of Linked Open Data – or, more precisely, LLOD, un-

derstood as the reference community for the creation and sharing of resour-

ces according to LOD principles (Cimiano et al. 2020; LLOD n.d.) – the

linguistic resources currently available for Italian include RDF datasets for

the previously mentioned PSC (Del Gratta et al. 2015) and IWN (Bartolini

2016) resources. The LLOD landscape, however, offers resources for different

languages, both contemporary and historical varieties; as an illustrative and

non-exhaustive example in a constantly expanding field, it is worth mentio-

ning Dbnary (Sérasset 2015), the multilingual resource based on Wiktionary,

made available according to LLOD principles. For historical varieties, we cite

LiLa – Linking Latin, a knowledge base for Latin that now includes several

resources (Mambrini and Passarotti 2023), and the DigitAnt project for an-

cient language varieties in Italy (Mallia et al. 2024). On the terminological

front, CHAMUÇA is noted, a resource for Portuguese loanwords in Asian

languages (Khan et al. 2024), and initial studies for a resource related to terms

in the Babylonian Talmud (Sciolette 2024), with the formalisation of contex-

CompL-it: a Computational Lexicon of Italian 123

ts through the OntoLex module FrAC (Frequency, Attestation and Corpus

Information), which is currently under development (Chiarcos et al. 2022;

Github n.d.a).

As a source of reference data for the construction of the CompL-it lexicon,

and as already mentioned, we chose LexicO, one of the freely available lexical

resources for the Italian language. The motivation for this choice is twofold,

and is partly also evident from the data reported in section 4.3, where LexicO

has been quantitatively compared to five other resources. First of all, Lexi-

cO (see section 3.1) is a multi-layered linguistic resource, in which informa-

tion of various kinds (phonological, morphological, syntactic and semantic)

is encoded: in this sense, it constitutes a unicum of its kind, given that all

the others limit themselves to representing, essentially, lexemes linked to each

other through semantic relations. Moreover, from a more quantitative point of

view, LexicO with its dense network of relations constitutes an extremely rich

resource of linguistic data.

However, LexicO also has limitations, both in terms of coverage in the

number of lexical entries, in terms of specific content (such as inflected forms

or missing lexical senses) and, finally, in terms of the data format in which it

is currently represented.

The idea of building CompL-it arose precisely due to these limitations of

LexicO: to ensure maximum lexical coverage, the linguistic data from LexicO

was integrated with data from two additional sources. Furthermore, standards

defined by the LLOD community were adopted for the model and represen-

tation format.

3. The sources

As stated in the previous section, LexicO was selected as the foundational

resource for constructing CompL-it. Additionally, we considered two other

sources: M-GLF (MAGIC-Generated Lemmatized Forms), a list of lemma-

tized forms with morphological information generated by the MAGIC tool

(Battista and Pirrelli 1999; Pirrelli and Battista 2000), and a set of Italian

language treebanks available through the UD repository (Universal Depen-

dencies n.d.b). All these resources have been chosen both for the richness of

the data they provide and because they have been manually constructed or

validated.

These three resources are very different from each other in terms of for-

mats, models and purposes, and therefore their integration required a process

of standardisation, as described in Section 4.1. In the following sections, we

describe the three resources together with some specific pre-processing inter-

ventions carried out prior to the data standardisation and conversion steps

necessary to create CompL-it.

124 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

3.1. LexicO

LexicO is a computational lexicon of Italian, available on CLARIN as a re-

lational database (CLARIN-IT n.d.b). This resource is derived from the above

mentioned PSC, with which it shares the same model based on the theory of

Generative Lexicon by James Pustejovsky (Pustejovsky 1995).

LexicO contains four layers of linguistic information: a morphological

layer, which describes lemmas, parts of speech (POS), and inflectional rules;

a semantic layer, which includes information about senses and their relation-

ships; a syntactic layer, detailing the syntactic behaviour of units and their

phrase structure; and finally, a phonological layer, which involves inflected for-

ms generated from the inflectional rules in the morphological layer. Althou-

gh each layer operates independently, there are connections between different

units, such as between syntactic and semantic entries. Since its initial use in

tasks such as full-text search (Giovannetti et al. 2022), it has become evident

that there is a need to convert all the data into a format compliant with current

standards.

Morphological units form the basis of lexical entries in LexicO. Each unit

is associated with a POS value and a set of morphological rules used to gene-

rate grammatically correct forms. These forms are defined as a type of entry

called phonological units.

Each association between a lemma and a form is described with a POS and

a certain number of morphological traits, as shown in Table 1.

Lemma Form POS MorphFeat

abbandonare abbandonai V 1|singular|indicative|past

Table 1: abbandonare (to abandon) with its form abbandonai, its POS (verb), and its morphological

features (first person singular, indicative, past).

As already mentioned in the previous section, LexicO is directly derived

from the PSC computational lexicon. This initial resource, while already quite

comprehensive, contained redundant or duplicated data and some entries

lacking in information: an emblematic example is the absence of the form

vado (I go) of the verb andare (to go). Although these issues did not diminish

the intrinsic value of the source, they required interventions to address the

gaps where possible. All interventions are documented in Sciolette, Giovan-

netti, and Marchi (2023).

3.2. M-GLF

The second lexical source we used to build up CompL-it was M-GLF, a

list of lemmatised forms generated by MAGIC, a morphological analyser for

CompL-it: a Computational Lexicon of Italian 125

Italian (Battista and Pirrelli 1999). The tool includes three modules: a lexicon

compiler, the morphological analyser itself, and a morphological generator.

We used this latter to generate the M-GLF list of forms (CLARIN-IT n.d.c)

by starting from a list of morphological rules for lemmas, endings and idiosyn-

cratic entries, contained in a morphological database.

An example of a M-GLF entry follows which is relative to a form of the

Italian verb abbaiare (to bark):

[1] MACRO:word[l_abbaiare,abbaiera’,v_fin,3,!,s,-

fut,ind,!,!]

In this example, l_abbaiare is the lemma of the form abbaierà, which is the

third person singular of the finite verb abbaiare in the indicative mood and

future tense. These morphological traits are indicated in the line, separated

by commas. Exclamation points represent null values for unspecified features,

such as degree, which is only relevant for adjectives.

The MAGIC generation tool is based on rules that constitute an extremely

rigorous model. First of all, the tool was unable to generate certain forms, such

as the absolute superlative. Moreover, the generation of entries produced some

inconsistencies. In particular, we found entries having multiple POS, such as

noun and adjective (e.g., svedese can indicate both the noun for a resident of

Sweden and the adjective for denoting the quality of being Swedish). In these

cases, we decided to intervene by splitting the entries with double POS into

distinct entries, each of which having its own POS with the correct morpho-

logical traits.

3.3. Treebanks

To further enrich the morphological layer of CompL-it, we also decided to

consider lemmas, forms, and morphological information obtained from the

available treebanks for Italian. The treebanks are collected in the UD repos-

itory, according to a common annotation scheme (Universal Dependencies

n.d.c), used for resources in different languages.

We only included treebanks that have been manually revised: three based

on balanced corpora of general-purpose texts (such as newspapers, legal doc-

uments, etc.) and one from a specific domain. We considered the following

treebanks:

• ISDT (Italian Stanford Dependency Treebank) (Universal Dependen-

cies n.d.d): this resource was obtained through a semi-automatic con-

version process starting from MIDT (the Merged Italian Dependency

Treebank). It is the result of merging pre-existing dependency-based

resources, aimed at improving the interoperability of available data.

126 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

The schema was partially adapted to account for the specific features of

the Italian language (Simi, Bosco, and Montemagni 2014);

• VIT (Universal Dependencies n.d.e): it is a conversion of VIT (Veni-

ce Italian Treebank), developed at the Laboratory of Computational

Linguistics at Università Ca’ Foscari in Venice. Originally a constituen-

cy-based treebank, VIT includes linguistic materials of various types,

extracted from five text typologies and spoken dialogues. The data un-

derwent conversion to the CoNLL-U format (Universal Dependencies

n.d.f), along with several stages of data harmonisation;

• ParTUT (ParallelTut) (Universal Dependencies n.d.g): it is a conver-

sion of a multilingual parallel treebank developed at the University of

Turin consisting of a variety of text genres, including talks, legal texts

and Wikipedia articles (Sanguinetti and Bosco 2015);

• ParlaMint-It (Universal Dependencies n.d.h): it is a collection of tran-

scriptions of parliamentary sessions of the Italian Senate, annotated in

Universal Dependencies. The corpus is part of a larger multilingual col-

lection of parliamentary transcripts built during the ParlaMint project

(CLARIN n.d.).

Although the selected treebanks were chosen precisely because they un-

derwent manual revision, they are not entirely free of errors, including gaps

and inconsistencies, particularly in morphological features, which can intro-

duce noise.

For example, there are cases where a word appearing in the treebanks is

annotated with fewer morphological features than the same word appearing

in the other two resources. This is the case for the form abilitati (enabled, as

in “enabled users”), described in the treebanks only through lemma and POS,

while in LexicO and M-GLF, this word is also provided with number (plural)

and gender (masculine) features. In all these cases, in CompL-it, the words

from the resource richer in linguistic information have been added.

4. The nature of CompL-it

This section illustrates the resource CompL-it, by starting with the neces-

sary standardisation process that had to be carried out, described in Section

4.1. The conversion in RDF format is briefly described in Section 4.2, while

a quantitative analysis of the resource is carried out in Section 4.3. To ensu-

re data interoperability we chose Ontolex-Lemon as the backbone model of

Compl-It, as it is the de facto standard for representing lexical resources in the

Linked Open Data community.

CompL-it: a Computational Lexicon of Italian 127

In order to proceed to the standardisation and conversion processes, it was

necessary to carry out some pre-processing steps. Some of these interventions

concerned all the resources and involved, in particular: i) the conversion of

superscripts into accented letters and distinction of high and low accents; ii)

the removal of proper nouns, such as named entities (e.g. Petrarca) and trade

names (e.g. Xerox); iii) the exclusion of abbreviations (e.g. Dott. instead of

doctor); iv) the exclusion of multiword expressions, which included nouns,

adjectives, adverbs and prepositions (e.g. the expression a ferro e fuoco); v) the

removal of unadapted loanwords (e.g. word processor).

Loanwords, multi words and proper nouns require different treatment and

will therefore be the subject of future work.

4.1. Standardisation

The following paragraphs describe the interventions undertaken to make

the models of the considered resources homogeneous in terms of morphology

and semantic relations.

4.1.1 Morphology

As can be seen from Section 3, the models and reference vocabularies of

LexicO, M-GLF and treebanks differ from each other, often profoundly. This

divergence between linguistic information representation systems is also moti-

vated by the different approach used to represent linguistic data.

In fact, M-GLF and LexicO can be included in the category of lexico-

graphic resources, whereas the standard used for treebanks, based on the UD

paradigm, pertains to the annotation of linguistic corpora.

In order to standardise the vocabularies, we decided to use LexInfo, an

inventory of types, values and properties designed to describe linguistic data

categorie

s (LexInfo n.d.). LexInfo comprises morphological properties, such as

gender, number, mood, grammatical categories (POS), and semantic relations,

such as synonymy, hypernymy, and so on. This choice is justified primarily

by the alignment of this vocabulary with the OntoLex-Lemon model (W3C

2016) used to represent CompL-it, as LexInfo serves as the reference linguistic

ontology for resources created with this model. Additionally, LexInfo complies

with other standards related to the OntoLex-Lemon model, including OLiA

(Ontologies of Linguistic Annotation) (Chiarcos and Sukhareva 2015; OLiA

n.d.), a repository of linguistic categories specific to annotated corpora. Ulti-

mately, the selection of the LexInfo vocabulary was largely driven by the need

to produce a lexical resource that is as interoperable as possible with other re-

128 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

sources based on the OntoLex-Lemon model, in accordance with the Linked

Data paradigm.

This difference in the representation of linguistic information in the mod-

els of the three sources occurs mainly at the level of the association between

POS and morphological traits. In fact, LexicO and M-GLF are based on a

model and specific vocabularies of labels, which are characterised by a very

fine-grained categorisation of POS. In the case of the treebanks, the UPOS

has a coarser grain: for example, the combination of UPOS and trait “pos-

sessive” with the value “yes” in the treebanks has been mapped to a specific

LexInfo POS. For example, the entry mio (mine) appears in treebanks with

the following annotation: PRON for the UPOS, with the feature “Poss=Yes”.

In CompL-it, this entry has been described with POS “possessivePronoun”,

according to the LexInfo vocabulary. More in general, the vocabularies of Lex-

icO, M-GLF and the treebanks were mapped into LexInfo, according to the

following scenarios: i) direct mapping between POS, if available (as was often

the case for LexicO and M-GLF); ii) conversion of POS and trait combina-

tions present in the treebanks into a LexInfo POS; iii) conversion into OLiA

or proposal of an ad hoc label, if the trait was not present in LexInfo. The

conversion tables have been made available on (Github n.d.b).

4.1.2 Semantics

In CompL-it, 137 types of semantic relations derived from LexicO have

been included. These relations are categorised into eight classes, listed below.

• Four classes are related to the four qualia roles taken from the Genera-

tive Lexicon theory

, namely:

– Formal: the role that describes the entity conveyed by the sense in

relation to other entities. An example of a relation associated with

the formal role is hyponymy, e.g., gatto-mammifero (cat-mammal);

– Agentive: the role that provides information about the origin of an

entity. An example of a relation associated with the agentive role is

caused by, e.g., infezione-batterio (infection-bacterium);

– Telic: the role that specifies a function of an entity. An example of a

relation associated with the telic role is Object of activity, linking an

object to a certain event, such as libro-leggere (book-to read);

OLiA has been used in the conversion of two traits in M-GLF for “Diminutive” and

“Augmentative”.

For an overview of the theory and the relationship between qualia roles and relations, see

(Sciolette, Giovannetti, and Marchi 2023). Following the terminology in the PSC documen-

tation, entities refer to the concept expressed by the sense, conveyed by a specific entry. These

entities can be connected to each other through semantic relations. Semantic relations are also

classified according to qualia roles.

CompL-it: a Computational Lexicon of Italian 129

– Constitutive: the role that describes the composition of an entity;

an example of a relation associated with the constitutive role is me-

ronymy, as senatore-senato (senator-senate).

• A derivational class, reserved for relations concerning senses that un-

dergo a change in grammatical category, e.g., from adjective to noun,

as triste-tristezza (sad-sadness).

• A class related to polysemy relations, as listed in Malmgren (1988). An

example of a regular polysemy class is Substance-Colour, as seen in the

sense of turchese (turquoise), which can refer to both the gemstone and

the colour.

• Two classes not documented in the original PSC model, namely synon-

ymy, e.g. ciclone-uragano (cyclone-hurricane), and metaphor, e.g. leone

(lion) to relate the sense of a brave man to the sense of the animal.

These few examples convey the image of a system of relations aimed at

defining meaning according to very fine-grained categories, as exemplified in

the case of meronymy, which distinguishes senses related to parts of a set, com-

ponents of a group, and, as a subclass, followers of a certain movement, as for

example Marxista (Marxist).

At present, it has proved particularly complex to find exact correspondences

between the semantic relations described in the reference ontologies (LexInfo

and OLiA) for the OntoLex-Lemon model. In some cases, it was necessary to

define relations from scratch.

The need to update lexical resources in Linked Data formats was also felt

in the past and led to the creation of some resources conforming to previous

versions of the OntoLex-Lemon model (Del Gratta et al. 2015; Villegas and

Bel 2015). However, it was not possible to reuse these resources either because

they did not include updates to the OntoLex-Lemon model, or because they

did not include a mapping with other reference models, such as LexInfo.

To ensure maximum interoperability, where possible, relations formalised

from scratch were linked to the corresponding reconstructed resources in pre-

vious versions of the OntoLex-Lemon model, with the seeAlso relation (W3C

2005).

For the construction of a vocabulary of CompL-it relations, the following

were also considered: i) equivalences, where possible, with LexInfo, such as

in the case of synonymy; ii) additional properties, not previously defined by

other resources, but reconstructed through documentation and analysis of the

resource.

The vocabulary definition phase also had a direct effect on the enrichment

interventions of the resource. For example, the mapping with LexInfo made it

possible to define an additional relation, hypernymy, as the inverse of hypon-

ymy (which translates the isA relation present in the LexicO model); since hy-

130 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

pernymy is the inverse of hyponymy, even if the relation is not described in the

source resource, it was still possible to infer a number of instances. This hap-

pened for all relations of which we could formalise additional properties from

the study of the documentation (as in the case of causes, inverse of causedBy).

4.2. Conversion to Linked Data

Once the data from the three lexical sources had been standardised, they

were converted into the Linked Data format. After introducing the reference

model adopted, an example of converted data is provided.

4.2.1 The OntoLex-Lemon model

In the context of the representation and publication of lexical data as

knowledge graphs and/or as Linguistic Linked Open Data, the OntoLex-Le-

mon model has become a de facto standard. This model was created with the

aim of supporting the linguistic foundation of a given ontology by adding

information on how ontological entities are lexicalised in different langua-

ges. However, OntoLex-Lemon can also be used as a lexicographic model to

represent linguistic entities without any concept they denote being defined.

OntoLex-Lemon is inspired by many other models, in particular the Lexical

Markup Framework (LMF) (Francopoulo et al. 2006), LexInfo (Cimiano et

al. 2011) — aligned with DatCatInfo (DatCatInfo n.d.) — and LIR (Lingui-

stic Information Repository) (Montiel-Ponsoda et al. 2008).

Figure 1 represents the core of the model, called ontolex. The rectangles

represent the classes of the model, the arrows with full heads represent the

properties of the objects, and the arrows with empty heads represent the sub-

class relationships.

CompL-it: a Computational Lexicon of Italian 131

Figure 1: The core model of OntoLex-Lemon (ontolex). Picture taken from the W3C OntoLex Final

Community Report at (W3C 2016).

ontolex is based on the definition of three fundamental classes: i) LexicalEntry,

that «represents a unit of analysis of the lexicon that consists of a set of forms

that are grammatically related and a set of base meanings that are associated

with all of these forms»; ii) Form, that «represents one grammatical realisation

of a lexical entry»; iii) LexicalSense, that «represents the lexical meaning of

a lexical entry when interpreted as referring to the corresponding ontology

element, if it is given». With reference to Figure 1, it is necessary to emphasise

that ontolex also allows us to express the fact that a given lexical entry evokes

a certain mental concept or refers to an entity with a formal interpretation

defined in an ontology. Therefore, OntoLex-Lemon introduces a fourth ele-

ment, the

LexicalConcept class, which represents a mental abstraction, concept

or unit of thought that can be lexicalised by a given collection of meanings.

The rest of the architecture of OntoLex-Lemon is divided into 4 modules,

each representing a different linguistic aspect, namely: i) decomposition (de-

comp), i.e. the process of describing which elements constitute a multiword

or compound; ii) the lexical and semantic relations between lexical entries

and lexical senses respectively (vartrans); iii) the syntactic behaviour of lexical

entries (synsem); iv) the description of the metadata of the lexical resource

(lime). However, in this paper, our conversion work will mainly use ontolex,

dealing with neither composition nor syntactic aspects in particular.

132 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

It is important to emphasise that the model abstracts from specific lin-

guistic theories or category systems used to describe the properties of lexical

entries and their syntactic behaviour. The re-use of existing category systems

or linguistic ontologies is therefore strongly encouraged. In our case, as rec-

ommended by the community that developed the model and as described in

section 4.1.1, the LexInfo model was used, which offers a rich vocabulary of

linguistic categories and relationships for morphology, syntax and semantics.

4.2.2 Representing data in RDF OntoLex-Lemon

The conversion of the standardised data coming from the three sources

into OntoLex-Lemon was performed by an algorithm in two steps: i) conver-

sion of the linguistic information according to the formalisation described in

the core ontolex module of the model; ii) serialisation of the data into Turtle

The obtained lexicon was then loaded into Ontotext GraphDB (Ontotext

n.d.), a semantic repository compliant with RDF and SPARQL (W3C 2013).

Below is an example of an RDF OntoLex-Lemon representation of a Com-

pL-it lexical entry in Turtle format.

:coniglio_entry a ontolex:Word;

lexinfo:partOfSpeech lexinfo:noun;

ontolex:canonicalForm coniglio_lemma;

ontolex:otherForm coniglio_form_1;

ontolex:sense coniglio_sense_1, coniglio_sense_2, co-

niglio_sense_3.

:coniglio_lemma a ontolex:Form;

lexinfo:gender lexinfo:masculine;

lexinfo:number lexinfo:singular;

ontolex:writtenRep “coniglio”@it, “rabbit”@en.

:coniglio_form_1 a ontolex:Form;

lexinfo:gender lexinfo:masculine;

lexinfo:number lexinfo:plural;

ontolex:writtenRep “conigli”@it, “rabbits”@en.

:coniglio_sense_1 a ontolex:LexicalSense;

skos:denition “mammifero della famiglia dei Lepori-

di, con pelame di vario colore, lunghe orecchie, occhi

Turtle is a serialisation format for RDF data types (W3C 2014).

CompL-it: a Computational Lexicon of Italian 133

grandi e sporgenti e grossi incisivi”@it, “Mammal of

the Leporidae family, with variously colored fur, long

ears, large, protruding eyes and large incisors”@en;

lexinfo:hyponym mammifero_sense;

simple:polysemyAnimalFood coniglio_sense_3.

:coniglio_sense_2 a ontolex:LexicalSense;

skos:denition “persona timida e molto paurosa”@it,

“shy and very

fearful person”@en;

lexinfo:hyponym persona_sense;

simple:metaphor coniglio_sense_1.

:coniglio_sense_3 a ontolex:LexicalSense;

skos:denition “carne dell’omonimo animale”@it, “meat

of the animal”@en.

In this example, the lexical entry coniglio (rabbit) is associated with two

forms, one of which is defined as the canonical form (the lemma) and the

other suitable for representing the plural form conigli (rabbits), both of which

are equipped with the appropriate morphological traits. The lexical entry is

also associated, via the ontolex:sense relation, with three lexical senses, each of

which has a natural language definition. Furthermore, the first two senses are

also endowed with semantic relations that link them to other lexical senses.

For example, rabbit_sense_2 is defined as a hyponym of mammal_sense.

4.3. CompL-it in numbers

In this section, the CompL-it lexicon is described from a quantitative per-

spective, both by enumerating the entities and relations it comprises and by

comparing it with lexicographic resources available for the Italian language of

a similar nature.

From a morphological standpoint, the resource is composed of 101,795

lexical entries (comprising a total of 791,541 word forms), classified with 36

POS categories and described with morphological traits. Figure 2 depicts a

Venn diagram representing the different dimensions, in terms of lexical en-

tries, of the three source resources and their intersections. As observed, the

most significant contribution of words comes from M-GLF. However, both

LexicO and the treebanks contribute significantly with a total of 47,069 forms

and 9,028 additional lexical entries.

134 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Figure 2: Representation, by the number of lexical entries, of the three source resources and their

various intersections.

Going more into the specific linguistic content, Table 2 shows the distri-

bution of word forms by POS: as expected, most are verbal forms (69% of the

total). The other class encompasses the 32 POS not explicitly specified, such

as, for example, number, conjunction, determiner, and preposition.

POS LexicO M-GLF TBs CompL-it

verb 345,307 526,635 10,741 545,342

noun 76,396 115,562 11,723 136,744

adj. 45,712 98,603 7,083 103,869

adv. 742 2,818 845 3223

other

935 670 926 2,364

total 469,092 744,288 31,318 791,542

Table 2: Distribution of word forms by POS

This category includes the following POS: adposition, article, auxiliary, cardinalNumeral,

conjunction, coordinatingConjunction, definiteArticle, demonstrativeDeterminer, demon-

strativePronoun, determiner, exclamativeDeterminer, exclamativePronoun, fusedPreposition,

indefiniteArticle, indefiniteDeterminer, indefinitePronoun, interjection, interrogativeAdverb,

interrogativeDeterminer, interrogativePronoun, numeral, numeralDeterminer, numeralPro-

noun, particle, personalPronoun, possessiveAdjective, possessiveDeterminer, possessivePro-

noun, pronoun, relativeDeterminer, relativePronoun, subordinatingConjunction.

CompL-it: a Computational Lexicon of Italian 135

As for the data related to the semantic layer, CompL-it describes 55,713

word senses connected to each other through 137 types of semantic relations,

totalling 86,577 instances. Table 3 shows a distribution of the 10 most nume-

rous types of semantic relation instances.

Semantic relation # instances an example

hyponym 43,069 medicina, scienza (medicine, science)

approximateSynonym 5,666 sciocco, stupido (foolish, stupid)

usedFor 3,291 matita, scrivere (pencil, to write)

partMeronym 3,159 giorno, settimana (day, week)

partHolonym 3,159 cinghiale, grugno (boar, snout)

createdBy 2,857 quadro, dipingere (painting, to paint)

ObjectOfeActivity 1,366 bistecca, mangiare (steak, to eat)

memberMeronym 1,318 segretario, partito (secretary, party)

ResultingState 1,063 bruciare, bruciato (to burn, burnt)

memberHolonym 979 stormo, uccello (ock, bird)

other 20,255 -

total 86,577

Table 3: Distribution of semantic relations instances.

To provide an overview of the dimensions and richness of linguistic infor-

mation conveyed by CompL-it, we finally present, in Table 4, a comparison

with other lexical resources available for Italian

Data updated at the time of writing.

136 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

entries forms senses/synsets

semantic

relations

instances

semantic

relations

types

LexicO 71,021 469,708 56,870 senses 89,340 137

IWN 48,416 - 49,350 synsets 138,385 83

MWN 41,491 - 32,673 synsets 45,593 14

OmegaWiki 30,258

- 23,417 senses 66,005

SIMPLELex-IT 7,022 26,560 - - -

Morph-it! 34,968 504,906 - - -

CompL-it 101,795 791,541 56,870 senses 86,577

137

Table 4: Concise comparison of some of the main freely available resources containing lexical data for

the Italian language.

5. CompL-it: access and use

The resource, in addition to being available for download, can be queried

through a dedicated web interface (KLAB n.d.). This interface, shown in Figu-

re 3, allows the user to select a series of precompiled SPARQL queries (visible

on the left), modify one of them using the right panel, or formulate a new

query from scratch.

As an example, the figure includes a precompiled query that allows for di-

splaying all meanings of the verb fare (to do). If selected, the interface queries

the resource and returns 7 senses of that verb, displaying their definitions and

some examples. Using the corresponding SPARQL query shown in the right

panel, it is possible to modify the label fare (highlighted in the figure) to insert

another Italian verb, click the execute query button at the top right, and view

the meanings of that verb in CompL-it.

In addition to its presentation as a linguistic resource in itself, as it is freely

distributed, numerically rich and conforms to Linked Open Data standards,

CompL-it can also be described in relation to the uses that can be made of it

for information management and retrieval tasks.

For instance, a computational lexicon can be used in full-text search ap-

proaches in which queries can fully exploit the morphological information

contained therein and the complex and articulated system of semantic rela-

This number includes both dictionary entries and encyclopaedic data (such as named

entities).

Semantic relations are defined between “concepts” and not between senses of a specific

language.

There are fewer instances of semantic relations in CompL-it than in LexicO because prop-

er nouns were not extracted from the latter resource (as specified at the beginning of section

4) and the associated semantic relations were excluded with them.

CompL-it: a Computational Lexicon of Italian 137

tions between the words of the lexicon (especially synonymy and hyponymy).

Better results are naturally obtained if searches are carried out on linguistical-

ly pre-analysed texts (at least with POS tagging) to reduce ambiguity in the

results. In the following section we provide an example of a full-text search

supported by a computational lexicon. For more details, see (Giovannetti et

al. 2022).

Figure 3: The query interface with the example of search for the senses of the Italian verb fare (to do).

Finally, we would also like to emphasise that representing data according

to the LOD paradigm can bring some advantages, namely: (i) the federation

mechanism with other datasets potentially allows the integration and im-

provement of queries results for more enriched searches, e.g., linking with

etymological datasets (see Section 6); (ii) the addition of a semantic layer to

the data through ontologies, allows the implicit knowledge in the dataset to

be inferred and exploited in queries to the text, e.g., exploiting the transitivity

of synonymy or hypernymy.

5.1. An example: lexicon-based search of the Babylonian Talmud

In this example, we show a query of the Italian translation of the Babylo-

nian Talmud. This translation is being carried out by expert translators in the

context of the Babylonian Talmud Translation Project (PTTB n.d.) using Tra-

duco, a computer-aided translation tool developed at the CNR-ILC (Giovan-

netti et al. 2016). The search function of the tool allows the text to be accessed

with complex queries, which can include information conveyed by the Italian

lexicon related to morphology and semantics.

138 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Figure 4: An example of query and relative results with the verb iniziare.

Considering the lemma iniziare (to start) it is possible to insert both mor-

phological and semantic traits into the query (Fig. 4). By adding, for example,

a restriction on the indicative mood and the third person, it is possible to

extract all the contexts of the Talmud with the forms of the verb iniziare char-

acterised by these traits. On the semantic side, we can also expand the query to

include all lemmas having at least one sense as synonym to one of the (three)

senses available in the lexicon for iniziare: the lexicon returns introdurre (to

introduce) and cominciare (to begin). With the aforementioned morpholog-

ical and semantic restrictions, the system is able to return Talmudic contexts

containing, for example, the form inizia (singular, present indicative), but

also comincia, cominciano (plural, present indicative), and iniziò (singular, past

indicative).

6. Conclusions and perspectives

In this article, we presented CompL-it, a new computational lexicon for

contemporary Italian. In the first part, we described the state of the art for

this type of resource and outlined a landscape that, although rich, presen-

ts some challenges, such as the heterogeneity of data formats and linguistic

models. Next, we described the three resources from which the lexicon was

constructed: a computational lexicon (LexicO), a lemmatised form list obtai-

ned from a morphological analyser (M-GLF) and a set of treebanks. The three

sources were based on different formats and models, which made it necessary

CompL-it: a Computational Lexicon of Italian 139

to standardise the data, also in accordance with the standards used by the

Linguistic Linked Open Data community. Standardisation was followed by a

conversion phase, which led to the final version of the lexicon in the form of

Linguistic Linked Open Data according to the OntoLex-Lemon model.

We have also described the resource on a quantitative basis, also comparing

CompL-it with some of the lexicographic resources available for Italian. The

lexicon has been released as an open resource, is freely downloadable and can

be consulted through a SPARQL interface. Finally, it was shown, through an

example of a search on the Italian text of the Babylonian Talmud, how such

a resource can be usefully exploited to provide linguistic-semantic access to

textual corpora.

By its very nature, the editing of a lexicon can never be called a finished

work. In the immediate future, CompL-it will first be further enriched in the

semantic layer from the data that have not yet been extracted from LexicO

(including templates and semantic traits). Subsequently, a merging method-

ology similar to the one adopted for morphology will also be applied to the

semantic layer, in particular by considering available semantic resources such

as ItalWordNet. A version of CompL-it will also be released, albeit limited to

the morphological layer, conforming to the Universal Dependencies model.

The resource will also be further extended to include cliticised forms, multi-

word forms and forms generated with suffixes, as in the case of papà-papino

(dad, daddy). Finally, other linguistic layers will be considered, such as syntax

(including data related to syntax-semantics interface) and phonetics, starting

by leveraging on such data already available in LexicO.

Acknowledgement

Scientific publication produced thanks to the agreement between the Na-

tional Research Council – Institute of Computational Linguistics and the

PTTB S.c.a r.l. – Babylonian Talmud Translation Project.

References

BabelNet. n.d. “BabelNet | Il Più Grande Dizionario Enciclopedico e Rete Se-

mantica Multilingue.” Accessed December 3, 2024. https://babelnet.org/.

Bamman, David, and Gregory Crane. 2010. “Computational Linguis-

tics and Classical Lexicography.” In Changing the Center of Gravity, ed-

ited by Melissa Terras and Gregory Crane, 297-322. Gorgias Press.

https://doi.org/10.31826/9781463219222-015.

Bartolini, Roberto. 2016. “IWN-LOD.” http://hdl.handle.net/20.500.11752/ILC-66.

140 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Basili, Roberto, Silvia Brambilla, Danilo Croce, and Fabio Tamburini.

2017. “Developing a Large Scale FrameNet for Italian: The IFrameN-

et Experience.” In Proceedings of the Fourth Italian Conference on Com-

putational Linguistics CLiC-It 2017, edited by Roberto Basili, Malvina

Nissim and Giorgio Satta, 59-64. Torino: Accademia University Press.

https://doi.org/10.4000/books.aaccademia.2364.

Battista, Marco, and Vito Pirrelli. 1999. “Una Piattaforma di Morfologia

Computazionale per l’analisi e la Generazione delle Parole Italiane.”

ILC-CNR Technical Report.

Bel, Nuria, Federica Busa, Nicoletta Calzolari, Elisabetta Gola, et al. 2000.

“SIMPLE: A General Framework for the Development of Multilin-

gual Lexicons.” In Proceedings of the Second International Conference

on Language Resources and Evaluation (LREC’00), edited by M. Gavri-

lidou, G. Carayannis, S. Markantonatou, S. Piperidis and G. Stainhau-

er. Athens, Greece: European Language Resources Association (ELRA).

http://www.lrec-conf.org/proceedings/lrec2000/pdf/61.pdf.

Brown, Susan Windisch, Julia Bonn, Ghazaleh Kazeminejad, Annie Zaenen,

James Pustejovsky, and Martha Palmer. 2022. “Semantic Representations

for NLP Using VerbNet and the Generative Lexicon.” Frontiers in Artificial

Intelligence 5 (April):821697. https://doi.org/10.3389/frai.2022.821697.

Chen, Hsinchun, Tak Yim, David Fye, and Bruce Schatz. 1995. “Automatic

Thesaurus Generation for an Electronic Community System.” Journal of

the American Society for Information Science 46 (3): 175-93.

Chiarcos, Christian, and Maria Sukhareva. 2015. “OLiA – Ontologies of

Linguistic Annotation.” Edited by Sebastian Hellmann, Steven Moran,

Martin Brümmer, and John P. McCrae. Semantic Web 6 (4): 379-86.

https://doi.org/10.3233/SW-140167.

Chiarcos, Christian, Elena-Simona Apostol, Besim Kabashi, and Cipri-

an-Octavian Truică. 2022. “Modelling Frequency, Attestation, and Cor-

pus-Based Information with OntoLex-FrAC.” In Proceedings of the 29th

International Conference on Computational Linguistics, edited by Nicoletta

Calzolari, Chu-Ren Huang, Hansaem Kim, James Pustejovsky, Leo Wan-

ner, Key-Sun Choi, Pum-Mo Ryu, et al., 4018-27. Gyeongju, Repub-

lic of Korea: International Committee on Computational Linguistics.

https://aclanthology.org/2022.coling-1.353.

Chiari, Isabella. 2012. “Il Dato Empirico in Lessicografia: Dizionari Tradizio-

nali e Collaborativi a Confronto.” Bollettino Di Italianistica II (January):

94-125.

CompL-it: a Computational Lexicon of Italian 141

Cimiano, Philipp, Christian Chiarcos, John P. McCrae, and Jorge Gracia. 2020. Lin-

guistic Linked Data: Representation, Generation and Applications. Cham: Spring-

er International Publishing. https://doi.org/10.1007/978-3-030-30225-2.

Cimiano, Philipp, Paul Buitelaar, John McCrae, and Michael Sintek. 2011. “Lex-

Info: A Declarative Model for the Lexicon-Ontology Interface.” Journal of Web

Semantics 9 (1): 29-51. https://doi.org/10.1016/j.websem.2010.11.001.

CLARIN. n.d. “ParlaMint: Comparable and Interoperable Parliamen-

tary Corpora | CLARIN ERIC.” Accessed December 3, 2024.

https://www.clarin.eu/parlamint.

CLARIN-IT. n.d.a. “CompL-It.” Accessed December 3, 2024. https://dspace-

clarin-it.ilc.cnr.it/repository/xmlui/handle/20.500.11752/ILC-1007.

CLARIN-IT. n.d.b. “LexicO.” Accessed December 3, 2024. https://dspace-

clarin-it.ilc.cnr.it/repository/xmlui/handle/20.500.11752/ILC-977.

CLARIN-IT. n.d.c. “MAGIC - Generated Lemmatized Forms.” Accessed De-

cember 3, 2024. https://dspace-clarin-it.ilc.cnr.it/repository/xmlui/han-

dle/20.500.11752/ILC-1002.

CLARIN VLO. n.d. “Virtual Language Observatory.” Accessed September 30,

2024. https://www.clarin.eu/content/virtual-language-observatory-vlo.

Dankova, Klara, Maria Teresa Zanola, and Silvia Calvi. 2022. “Pan-Latin Tex-

tile Fibres Vocabulary.” http://hdl.handle.net/20.500.11752/OPEN-975.

DatCatInfo. n.d. “Welcome to DatCatInfo.” Accessed December 3, 2024.

https://datcatinfo.net/.

De Mauro, Tullio. 1980. Guida all’uso delle parole: parlare e scrivere semplice e

preciso per capire e farsi capire. Libri di base. Roma: Editori Riuniti.

De Mauro, Tullio, a cura di. 2016. Il Nuovo Vocabolario Di Base Della Lingua

Italiana. December 23, 2016. https://www.dropbox.com/scl/fi/zg2y99x-

qik4k11nj19fgi/nuovovocabolariodibase.pdf?rlkey=s0uf8ggv11kf44i-

p6a2ldz16n&e=1&dl=0.

Del Gratta, Riccardo, Francesca Frontini, Anas Fahad Khan, and Mo-

nica Monachini. 2015. “Converting the PAROLE SIMPLE CLIPS

Lexicon into RDF with Lemon.” Semantic Web 6 (4): 387-92.

https://doi.org/10.3233/SW-140168.

ELEXIS. n.d. “ELEXIS European Lexicographic Infrastructure.” Accessed

September 30, 2024. https://elex.is/.

142 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Francopoulo, Gil, Monte George, Nicoletta Calzolari, et al. 2006. “Lexical Mark-

up Framework (LMF).” In Proceedings of the Fifth International Conference on

Language Resources and Evaluation (LREC’06), edited by Nicoletta Calzolari,

Khalid Choukri, Aldo Gangemi, Bente Maegaard, Joseph Mariani, Jan Odijk,

and Daniel Tapias. Genoa, Italy: European Language Resources Association

(ELRA). http://www.lrec-conf.org/proceedings/lrec2006/pdf/577_pdf.pdf.

Frontini, Francesca, Riccardo Del Gratta, and Monica Monachini. 2016. “Geodomain

WordNet ITA ENG V 1.0.” http://hdl.handle.net/20.500.11752/ILC-68.

Giovannetti, Emiliano, Davide Albanesi, Andrea Bellandi, and Giulia Benot-

to. 2016. “Traduco: A Collaborative Web-Based CAT Environment for

the Interpretation and Translation of Texts.” Digital Scholarship in the Hu-

manities 32 (suppl_1): i47-62. https://doi.org/10.1093/llc/fqw054.

Giovannetti, Emiliano, Davide Albanesi, Andrea Bellandi, Simone Marchi,

Mafalda Papini, and Flavia Sciolette. 2022. “The Role of a Computa-

tional Lexicon for Query Expansion in Full-Text Search.” In Proceedings

of the Eighth Italian Conference on Computational Linguistics CliC-It 2021,

edited by Elisabetta Fersini, Marco Passarotti, and Viviana Patti, 162-

68. Accademia University Press. https://doi.org/10.4000/books.aacca-

demia.10638.

Github. n.d.a. “The Ontolex Module for Frequency, Attestation and Corpus

Information.” Accessed December 3, 2024. https://github.com/acoli-re-

po/frac-addenda/blob/master/index.md.

Github. n.d.b. “CompL-It Mapping Tables.” Accessed December 3, 2024. https://

github.com/klab-ilc-cnr/Tables-for-mapping-of-Italian-lexicon-CompL-it.

Global WordNet Association. n.d. “Main Page.” Accessed December 3, 2024.

http://globalwordnet.org/.

Grella, Matteo. 2018a. “Italian Content Words V3.” http://hdl.handle.

net/11372/LRT-2894.

Grella, Matteo. 2018b. “Italian Function Words V3.” http://hdl.handle.

net/11372/LRT-2893.

Hmeidi, Ismail, Mahmoud Al-Ayyoub, Nizar A. Mahyoub, and Moham-

med A. Shehab. 2016. “A Lexicon Based Approach for Classifying Arabic

Multi-Labeled Text.” International Journal of Web Information Systems 12

(4): 504-32. https://doi.org/10.1108/IJWIS-01-2016-0002.

Hodge, Gail. 2000. Systems of Knowledge Organization for Digital Libraries:

Beyond Traditional Authority Files. Washington, DC: Digital Library Fed-

eration, Council on Library and Information Resources.

ILC4CLARIN CNR. 2016. “PAROLE-SIMPLE-CLIPS.” http://hdl.handle.

net/20.500.11752/ILC-88.

CompL-it: a Computational Lexicon of Italian 143

Khan, Fahad, Ana Salgado, Isuri Anuradha, et al. 2024. “CHAMUÇA: To-

wards a Linked Data Language Resource of Portuguese Borrowings in

Asian Languages.” In Proceedings of the 9th Workshop on Linked Data in

Linguistics @ LREC-COLING 2024, edited by Christian Chiarcos, Kater-

ina Gkirtzou, Maxim Ionov, Fahad Khan, John P. McCrae, Elena Montiel

Ponsoda, and Patricia Martín Chozas, 44-48. Torino, Italia: ELRA and

ICCL. https://aclanthology.org/2024.ldl-1.6.

KLAB. n.d. “CompL-It SPARQL Search Interface.” Accessed December 3,

2024. https://klab.ilc.cnr.it/CompL-it-SPARQL-interface/.

Kyjánek, Lukáš, Zdeněk Žabokrtský, Jonáš Vidra, and Magda Ševčíková. 2021.

“Universal Derivations v1.1.” http://hdl.handle.net/11234/1-3247.

LexInfo. n.d. “About the Ontology.” Accessed September 30, 2024. https://

lexinfo.net/.

LLOD. n.d. “Linguistic Linked Open Data.” Accessed December 3, 2024.

https://linguistic-lod.org/.

Mallia, Michele, Michela Bandini, Andrea Bellandi, et al. 2024. “DigItAnt: A

Platform for Creating, Linking and Exploiting LOD Lexica with Hetero-

geneous Resources.” In Proceedings of the 9th Workshop on Linked Data in

Linguistics @ LREC-COLING 2024, edited by Christian Chiarcos, Kater-

ina Gkirtzou, Maxim Ionov, Fahad Khan, John P. McCrae, Elena Montiel

Ponsoda, and Patricia Martín Chozas, 55-65. Torino, Italia: ELRA and

ICCL. https://aclanthology.org/2024.ldl-1.8.

Malmgren, Sven-Göran. 1988. “On Regular Polysemy in Swedish.” In Studies

in Computer-Aided Lexicology, 179-200. Data Linguistica 18. Stockholm:

Almqvist & Wiksell.

Mambrini, Francesco, and Marco Carlo Passarotti. 2023. “The LiLa Lemma

Bank: A Knowledge Base of Latin Canonical Forms.” Journal of Open

Humanities Data 9 (November):1-5. https://doi.org/10.5334/johd.145.

Mazzei, Alessandro. 2016. “Building a Computational Lexicon by Using

SQL.” In Proceedings of the Third Italian Conference on Computational

Linguistics CLiC-It 2016, 200-04. Napoli: Accademia University Press.

https://doi.org/10.4000/books.aaccademia.1808.

Meijssen, Gerard. 2014. “OmegaWiki.” http://hdl.handle.net/11372/LRT-853.

Miller, George A. 1995. “WordNet: A Lexical Database for English.” Communica-

tions of the ACM 38 (11): 39-41. https://doi.org/10.1145/219717.219748.

Montiel-Ponsoda, Elena, Wim Peters, Mauricio Espinoza, Asunción Gó-

mez-Pérez, and Margherita Sini. 2008. “Multilingual and Localization

Support for Ontologies.” Technical Report 2.4.2. http://neon-project.

org/deliverables/WP2/NeOn_2008_D242.pdf.

144 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Morph-it! 2018. “Resources:Morph-It.” Last Modified May 03. https://docs.

sslmit.unibo.it/doku.php?id=resources:morph-it.

MultiWordNet. n.d. “NLP Research Group - MultiWordNet.” Accessed

December 3, 2024. https://nlplab.fbk.eu/tools-and-resources/lexical-re-

sources-and-corpora/multiwordnet.

Navigli, Roberto, and Simone Paolo Ponzetto. 2012. “BabelNet: The Au-

tomatic Construction, Evaluation and Application of a Wide-Coverage

Multilingual Semantic Network.” Artificial Intelligence 193 (December):

217-50. https://doi.org/10.1016/j.artint.2012.07.001.

OLiA. n.d. “Ontologies of Linguistic Annotation (OLiA) | Olia.” Accessed

December 3, 2024. https://acoli-repo.github.io/olia/.

Ontotext. n.d. “Ontotext GraphDB.” Accessed December 3, 2024. https://

www.ontotext.com/products/graphdb/.

Passarotti, Marco Carlo, and Francesco Mambrini. 2021. “Linking Latin: In-

teroperable Lexical Resources in the LiLa Project.” In Building New Re-

sources for Historical Linguistics, edited by Erica Biagetti, Chiara Zanchi

and Silvia Luraghi, 103-24. https://doi.org/10.5281/zenodo.5994271.

Pianta, Emanuele, Luisa Bentivogli, and Christian Girardi. 2002. “Multi-

WordNet: Developing an Aligned Multilingual Database.” In Proceedings

of the First International Conference on Global WordNet.

Pirrelli, Vito, and Marco Battista. 2000. “The Paradigmatic Dimension of

Stem Allomorphy in Italian Verb Inflection.” Italian Journal of Linguistics

12 (2): 307-80.

Prakash, T. Nikil, and Amalanathan Aloysius. 2021. “Textual Sentiment Anal-

ysis Using Lexicon Based Approaches.” Annals of the Romanian Society for

Cell Biology 25 (4): 9878–85.

Princeton University. n.d. “WordNet.” Accessed December 3, 2024. https://

wordnet.princeton.edu/.

PTTB. n.d. “Progetto Traduzione Talmud Babilonese.” Accessed December 3,

2024. https://www.talmud.it/it/.

Pustejovsky, James. 1995. The Generative Lexicon. The MIT Press. https://doi.

org/10.7551/mitpress/3225.001.0001.

Realiter. n.d. “Home | Realiter.” Accessed December 3, 2024. https://www.realiter.net/.

Roventini, Adriana, Antonietta Alonge, Francesca Bertagna, et al. 2003. “‘Ital-

WordNet’: Building a Large Semantic Database for the Automatic Treat-

ment of Italian.” Linguistica computazionale: XVIII/XIX, 1998/1999, 745-

91. https://doi.org/10.1400/18178.

Roventini, Adriana, Rita Marinelli, and Francesca Bertagna. 2016. “ItalWord-

Net v.2.” http://hdl.handle.net/20.500.11752/ILC-62.

CompL-it: a Computational Lexicon of Italian 145

Ruimy, Nilda, Monica Monachini, Raffaella Distante, et al. 2002. “CLIPS,

a Multi-Level Italian Computational Lexicon: A Glimpse to Data.” In

Proceedings of the Third International Conference on Language Resources and

Evaluation, LREC 2002, May 29-31, 2002, Las Palmas, Canary Islands,

Spain. European Language Resources Association. http://www.lrec-conf.

org/proceedings/lrec2002/sumarios/197.htm.

Sabatini, Francesco. 2006. “La Storia dell’Italiano nella Prospettiva della Cor-

pus Linguistics.” In Proceedings of the 12th EURALEX International Con-

gress, edited by Cristina Onesti, Elisa Corino and Carla Marello, 31-37.

Torino: Edizioni dell’Orso.

Sanguinetti, Manuela, and Cristina Bosco. 2015. “PartTUT: The Turin Uni-

versity Parallel Treebank.” In Harmonization and Development of Resources

and Tools for Italian Natural Language Processing within the PARLI Project,

edited by Roberto Basili, Cristina Bosco, Rodolfo Delmonte, Alessandro

Moschitti, and Maria Simi, 589: 51-69. Studies in Computational Intelli-

gence. Springer. https://doi.org/10.1007/978-3-319-14206-7_3.

Sciolette, Flavia, Emiliano Giovannetti, and Simone Marchi. 2023. “LexicO: An

Italian Computational Lexicon Derived from Parole-Simple-Clips.” Umanis

tica Digitale 7 (15): 169-93. https://doi.org/10.6092/issn.2532-8816/15176.

Sciolette, Flavia. 2024. “Modeling Linking between Text and Lexicon with

OntoLex-Lemon: A Case Study of Computational Terminology for the

Babylonian Talmud.” In Proceedings of the 9th Workshop on Linked Data in

Linguistics @ LREC-COLING 2024, edited by Christian Chiarcos, Kater-

ina Gkirtzou, Maxim Ionov, Fahad Khan, John P. McCrae, Elena Montiel

Ponsoda, and Patricia Martín Chozas, 103-7. Torino, Italia: ELRA and

ICCL. https://aclanthology.org/2024.ldl-1.13.

Sérasset, Gilles. 2015. “DBnary: Wiktionary as a Lemon-Based Multilin-

gual Lexical Resource in RDF.” Semantic Web 6 (4): 355-61. https://doi.

org/10.3233/SW-140147.

Shiri, Ali. 2015. “Semantic Access and Exploration in Cultural Heritage Dig-

ital Libraries.” In Cultural Heritage Information: Access and Management,

edited by Ian Ruthven and Gobinda G. Chowdhury, 177-96. Facet Pub-

lishing.

146 Flavia Sciolette, Andrea Bellandi, Emiliano Giovannetti, Simone Marchi

Simi, Maria, Cristina Bosco, and Simonetta Montemagni. 2014. “Less Is

More? Towards a Reduced Inventory of Categories for Training a Parser

for the Italian Stanford Dependencies.” In Proceedings of the Ninth In-

ternational Conference on Language Resources and Evaluation (LREC’14),

edited by Nicoletta Calzolari, Khalid Choukri, Thierry Declerck, Hrafn

Loftsson, Bente Maegaard, Joseph Mariani, Asuncion Moreno, Jan Odi-

jk, and Stelios Piperidis, 83-90. Reykjavik, Iceland: European Language

Resources Association (ELRA). http://www.lrec-conf.org/proceedings/

lrec2014/pdf/818_Paper.pdf.

SimpleLEX-IT. n.d. “Alexmazzei/SimpleLEX-IT: SimpleLEX-IT Is the Com-

putational Lexicon Employed into SimpleNLG-IT.” Accessed December

3, 2024. https://github.com/alexmazzei/SimpleLEX-IT.

Smith, Jocelyn C. 1997. “The Use of Lexicons in Information Retrieval in

Legal Databases.” In Proceedings of the Sixth International Conference on

Artificial Intelligence and Law - ICAIL ’97, 29-38. Melbourne, Australia:

ACM Press. https://doi.org/10.1145/261618.261625.

Thompson, Paul, John McNaught, Simonetta Montemagni, et al. 2011. “The

BioLexicon: A Large-Scale Terminological Resource for Biomedical Text

Mining.” BMC Bioinformatics 12 (1): 397. https://doi.org/10.1186/1471-

2105-12-397.

Universal Dependencies. n.d.a. “Introduction.” Accessed December 3, 2024.

https://universaldependencies.org/it/overview/introduction.html.

Universal Dependencies. n.d.b. “Universal Dependencies.” Accessed Decem-

ber 3, 2024. https://universaldependencies.org/.

Universal Dependencies. n.d.c. “UD Guidelines.” Accessed December 3,

2024. https://universaldependencies.org/guidelines.html.

Universal Dependencies. n.d.d. “UD_Italian-ISDT.” Accessed December 3,

2024. https://github.com/UniversalDependencies/UD_Italian-ISDT/

tree/master.

Universal Dependencies. n.d.e. “UD_Italian-VIT.” Accessed December 3,

2024. https://github.com/UniversalDependencies/UD_Italian-VIT/.

Universal Dependencies. n.d.f. “CoNLL-U Format.” Accessed December 3,

2024. https://universaldependencies.org/format.html.

Universal Dependencies. n.d.g. “UD_Italian-ParTUT.” Accessed December

3, 2024. https://github.com/UniversalDependencies/UD_Italian-Par-

TUT/.

Universal Dependencies. n.d.h. “UD_Italian-ParlaMint.” Accessed December

3, 2024. https://github.com/UniversalDependencies/UD_Italian-Parla-

Mint/.

CompL-it: a Computational Lexicon of Italian 147

Vetere, Guido, Alessandro Oltramari, Isabella Chiari, Elisabetta Jezek, Laure

Vieu, and Fabio Massimo Zanzotto. 2011. “Senso Comune, an Open

Knowledge Base of Italian Language.” Traitement Automatique Des Lan-

gues 52 (3): 217-43.

Villegas, Marta, and Núria Bel. 2015. “PAROLE/SIMPLE ‘Lemon’ Ontolo-

gy and Lexicons.” Edited by Sebastian Hellmann, Steven Moran, Martin

Brümmer, and John McCrae. Semantic Web 6 (4): 363-69. https://doi.

org/10.3233/SW-140148.

W3C. 2005. “UsingSeeAlso - W3C Wiki.” Last Modified January 13. https://

www.w3.org/wiki/UsingSeeAlso.

W3C. 2013. “SPARQL 1.1 Overview.” https://www.w3.org/TR/spar-

ql11-overview/.

W3C. 2014. “RDF 1.1 Turtle.” https://www.w3.org/TR/turtle/.

W3C. 2016. “Lexicon Model for Ontologies: Community Report, 10 May

2016.” https://www.w3.org/2016/05/ontolex/.

WikiMedia. 2022. “OmegaWiki - Meta.” Last Modified December 03. ht-

tps://meta.wikimedia.org/wiki/OmegaWiki.

Wikipedia. 2024a. “Verbi Incoativi - Wikipedia.” Last Modified January 26.

https://it.wikipedia.org/wiki/Verbi_incoativi.

Wikipedia. 2024b. “Verbi Irregolari Italiani - Wikipedia.” Last Modified No-

vember 6. https://it.wikipedia.org/wiki/Verbi_irregolari_italiani.

Žabokrtský, Zdeněk, Nyati Bafna, Jan Bodnár, et al. 2022. “Universal Segmen-

tations 1.0 (UniSegments 1.0).” http://hdl.handle.net/11234/1-4629.

Zanchetta, Eros, and Marco Baroni. 2005. “Morph-It! A Free Corpus-Based

Morphological Resource for the Italian Language.” In Proceedings of Cor-

pus Linguistics Conference Series 2005 (ISSN 1747-9398), 1: 1-12. Univer-

sity of Birmingham.

Zanola, Maria Teresa, Klara Dankova, Claudio Grimaldi, and Anna Serpente.

2023. “Pan-Latin Lexicon of Collars and Sleeves in Fashion and Cos-

tume.” http://hdl.handle.net/20.500.11752/OPEN-987.

ISBN 979-12-5965-456-4 ISSN 1121-0095

AIDAinformazioni Anno 42 – N. 3-4 – luglio-dicembre 2024

AIDAinformazioni

Rivista semestrale di Scienze dell’Informazione

Anno 42

N. 3-4 – luglio-dicembre 2024

Contributi

A A

Il nuovo regolamento eIDAS e alcune “quisquilie”

archivistiche

F B, M T

Exploration du réseau numérique YouTube

autour de la santé des militaires: quelles sont les

thématiques des discours, les sources d’informations

et les acteurs de la communication?

E C, L F

Assisted morbidity coding: the SISCO.web

use case for identifying the main diagnosis in

Hospital Discharge Records

V F

A humanistic approach to datafication

R P

Testimonianze di un impegno culturale per

l’Università di Salerno. Le carte di Alfonso

Menna

F S, A B,

E G, S M

CompL-it: a Computational Lexicon of Italian

Rubriche

C G

Non solo libri

In copertina

Disegno di Paul Otlet, Collections Mundaneum, centre d’Archives, Mons (Belgique).