Esperienza di creazione di entrate lessicografiche combinatorie: metodi e dati dal progetto CombiNet

The present dissertation aims at simulating the construction of lexicographic layouts for an Italian combinatory dictionary based on real linguistic data, extracted from corpora by using computational methods. This work is based on the assumption that the intuition of the native speaker, or the lexicographer, who manually extracts and classifies all the relevant data, are not adequate to provide sufficient information on the meaning and use of words. Therefore, a study of the real use of language is required and this is particularly true for dictionaries that collect the combinatory behaviour of words, where the task of the lexicographer is to identify typical combinations where a word occurs. This study is conducted in the framework of the CombiNet project aimed at studying Italian Word Combinationsand and at building an online, corpus-based combinatory lexicographic resource for the Italian language. This work is divided into three chapters. Chapter 1 describes the criteria considered for the classification of word combinations according to the work of Ježek (2011). Chapter 1 also contains a brief comparison between the most important Italian combinatory dictionaries and the BBI Dictionary of Word Combinations in order to describe how word combinations are considered in these lexicographic resources. Chapter 2 describes the main computational methods used for the extraction of word combinations from corpora, taking into account the advantages and disadvantages of the two methods. Chapter 3 mainly focuses on the practical word carried out in the framework of the CombiNet project, with reference to the tools and resources used (EXTra, LexIt and "La Repubblica" corpus). Finally, the data extracted and the lexicographic layout of the lemmas to be included in the combinatory dictionary are commented, namely the words "acqua" (water), "braccio" (arm) and "colpo" (blow, shot, stroke).

La voce per gli altri. Il Libro Parlato e la voce dell'interprete al servizio della comunità non vedente e ipovedente

L’obiettivo del presente elaborato è esplorare un ulteriore possibile impiego delle competenze dell’interprete di conferenza in un ambito ancora poco conosciuto ai più: la produzione di audiolibri e l’attività, a sfondo sociale, dei donatori di voce. Partendo da una panoramica generale sull’oralità e la valenza della voce in termini metafisico-antropologici, seguita dall’iniziazione del lettore al mondo degli audiolibri per non vedenti e ipovedenti e all’attività del Centro Internazionale del Libro Parlato di Feltre (oggetto del secondo capitolo), nel terzo capitolo il lavoro si concentra sul valore del gesto vocale nella comunicazione orale e, in modo specifico, per l’interprete di conferenza, riflessione che assume particolare interesse alla luce del percorso di studi di chi scrive. Quindi, si giunge, nel quarto capitolo, all’affermazione dell’importanza della pedagogia vocale e dei suoi strumenti regolatori del gesto vocale in relazione agli obiettivi espressivo-comunicativi di un “interprete”, sia esso attore, cantante, speaker radiofonico, audiodescrittore, donatore di voce o interprete interlinguistico e interculturale. Nel quinto capitolo il gesto vocale è descritto in una prospettiva anatomo-fisiologica, con un’analisi dei vari sistemi che costituiscono l’apparato pneumofonoarticolatorio preposto all’atto della fonazione. Infine, l’elaborato si chiude con i punti di vista di un’ex donatrice di voce, della professoressa di Tecniche di Presentazione Orale della Scuola di Lingue e Letterature, Traduzione e Interpretazione di Forlì, e di un utente non vedente di audiolibri, raccolti dalla loro viva voce attraverso delle interviste che hanno l’obiettivo di dare un riscontro reale a quanto affermato teoricamente nel corso della disamina.

Emotional contagion in consecutive interpreting: An empirical study with novice interpreters

It has recently been noticed that interpreters tend to converge with their speakers’ emotions under a process known as emotional contagion. Emotional contagion still represents an underinvestigated aspect of interpreting and the few studies on this topic have tended to focus more on simultaneous interpreting rather than consecutive interpreting. Korpal & Jasielska (2019) compared the emotional effects of one emotional and one neutral text on interpreters in simultaneous interpreting and found that interpreters tended to converge emotionally with the speaker more when interpreting the emotional text. This exploratory study follows their procedures to study the emotional contagion potentially caused by two texts among interpreters in consecutive interpreting: one emotionally neutral text and one negatively-valenced text, this last containing 44 negative words as triggers. Several measures were triangulated to determine whether the triggers in the negatively-valenced text could prompt a stronger emotional contagion in the consecutive interpreting of that text as compared to the consecutive interpreting of the emotionally neutral text, which contained no triggers—namely, the quality of the interpreters’ delivery; their heart rate variability values as collected with EMPATICA E4 wristbands; the analysis of their acoustic variations (i.e., disfluencies and rhetorical strategies); their linguistic and emotional management of the triggers; and their answers to the Italian version of the Positive and Negative Affect Schedule (PANAS) self-report questionnaire. Results showed no statistically significant evidence of an emotional contagion evoked by the triggers in the consecutive interpreting of the negative text as opposed to the consecutive interpreting of the neutral text. On the contrary, interpreters seemed to be more at ease while interpreting the negative text. This surprising result, together with other results of this project, suggests venues for further research.

Voice conversion with pre-trained representations for audio anonymization

The recording and processing of voice data raises increasing privacy concerns for users and service providers. One way to address these issues is to move processing on the edge device closer to the recording so that potentially identifiable information is not transmitted over the internet. However, this is often not possible due to hardware limitations. An interesting alternative is the development of voice anonymization techniques that remove individual speakers characteristics while preserving linguistic and acoustic information in the data. In this work, a state-of-the-art approach to sequence-to-sequence speech conversion, ini- tially based on x-vectors and bottleneck features for automatic speech recognition, is explored to disentangle the two acoustic information using different pre-trained speech and speakers representation. Furthermore, different strategies for selecting target speech representations are analyzed. Results on public datasets in terms of equal error rate and word error rate show that good privacy is achieved with limited impact on converted speech quality relative to the original method.