Audiovisual materials and second-language acquisition: a comparison between native language subtitles and target language captions.

Over the last few years, the massive popularity of video streaming platforms has managed to impact our daily habits by making the watching of movies and TV shows one of the main activities of our free time. By providing a wide range of foreign language audiovisual content, these entertainment services may represent a powerful resource for language learners, as they provide them with the possibility to be exposed to authentic input. Moreover, research has shown the beneficial role of audiovisual textual aids such as native language subtitles and target language captions in enhancing language skills such as vocabulary and listening comprehension. The aim of this thesis is to analyze the existing literature on the subject of subtitled and captioned audiovisual materials used as a pedagogical tool for informal language learning.

Bio-QA-GNN: Reasoning with Language Models and Knowledge Graphs for Interpretable Biomedical Question Answering

Nowadays the idea of injecting world or domain-specific structured knowledge into pre-trained language models (PLMs) is becoming an increasingly popular approach for solving problems such as biases, hallucinations, huge architectural sizes, and explainability lack—critical for real-world natural language processing applications in sensitive fields like bioinformatics. One recent work that has garnered much attention in Neuro-symbolic AI is QA-GNN, an end-to-end model for multiple-choice open-domain question answering (MCOQA) tasks via interpretable text-graph reasoning. Unlike previous publications, QA-GNN mutually informs PLMs and graph neural networks (GNNs) on top of relevant facts retrieved from knowledge graphs (KGs). However, taking a more holistic view, existing PLM+KG contributions mainly consider commonsense benchmarks and ignore or shallowly analyze performances on biomedical datasets. This thesis start from a propose of a deep investigation of QA-GNN for biomedicine, comparing existing or brand-new PLMs, KGs, edge-aware GNNs, preprocessing techniques, and initialization strategies. By combining the insights emerged in DISI's research, we introduce Bio-QA-GNN that include a KG. Working with this part has led to an improvement in state-of-the-art of MCOQA model on biomedical/clinical text, largely outperforming the original one (+3.63\% accuracy on MedQA). Our findings also contribute to a better understanding of the explanation degree allowed by joint text-graph reasoning architectures and their effectiveness on different medical subjects and reasoning types. Codes, models, datasets, and demos to reproduce the results are freely available at: \url{https://github.com/disi-unibo-nlp/bio-qagnn}.

Proof of Location through a Blockchain Agnostic Smart Contract Language: Design and Evaluation over Algorand and Ethereum

Le applicazioni che offrono servizi sulla base della posizione degli utenti sono sempre più utilizzate, a partire dal navigatore fino ad arrivare ai sistemi di trasporto intelligenti (ITS) i quali permetteranno ai veicoli di comunicare tra loro. Alcune di questi servizi permettono perfino di ottenere qualche incentivo se l'utente visita o passa per determinate zone. Per esempio un negozio potrebbe offrire dei coupon alle persone che si trovano nei paraggi. Tuttavia, la posizione degli utenti è facilmente falsificabile, ed in quest'ultima tipologia di servizi, essi potrebbero ottenere gli incentivi in modo illecito, raggirando il sistema. Diviene quindi necessario implementare un'architettura in grado di impedire alle persone di falsificare la loro posizione. A tal fine, numerosi lavori sono stati proposti, i quali delegherebbero la realizzazione di "prove di luogo" a dei server centralizzati oppure collocherebbero degli access point in grado di rilasciare prove o certificati a quegli utenti che si trovano vicino. In questo lavoro di tesi abbiamo ideato un'architettura diversa da quelle dei lavori correlati, cercando di utilizzare le funzionalità offerte dalla tecnologia blockchain e dalla memorizzazione distribuita. In questo modo è stato possibile progettare una soluzione che fosse decentralizzata e trasparente, assicurando l'immutabilità dei dati mediante l'utilizzo della blockchain. Inoltre, verrà dettagliato un'idea di caso d'uso da realizzare utilizzando l'architettura da noi proposta, andando ad evidenziare i vantaggi che, potenzialmente, si potrebbero trarre da essa. Infine, abbiamo implementato parte del sistema in questione, misurando i tempi ed i costi richiesti dalle transazioni su alcune delle blockchain disponibili al giorno d'oggi, utilizzando le infrastrutture messe a disposizione da Ethereum, Polygon e Algorand.

Le collocazioni in traduzione e interpretazione tra italiano e francese: Uno studio su eptic_01_2011

This dissertation aims at investigating differences in phraseological patterns in translated and interpreted language, on the basis of the intermodal corpus EPTIC_01_2011 and focusing on Italian and French. First of all, an overview is offered of the main studies and theories about corpus linguistics and collocations: the notion of corpus is defined and a typology (focusing on intermodal corpora) is presented, before moving on to the linguistic phenomenon of collocation and its investigation through corpus linguistics methods. Second, the general structure of EPTIC_01_2011 is presented, including the ways in which its texts have been assembled, edited through ad hoc conventions and enriched with metadata. The methodology proposed by Durrant and Schmitt (2009), slightly edited to fit the present study, has been used to extract and compare noun+adjective/adjective+noun bigrams from a quantitative point of view. A subset of these data have then been extracted and analysed manually. The results of the study are presented through graphs and examples, with an in-depth discussion of the bigrams considered. Lastly, the data collected are analysed and categorised in terms of shifts occurring in translation and in interpreting, potential causes are discussed and ideas for further research and for the development of the EPTIC corpus are sketched.