Data modelling su database non relazionali: confronto di prestazioni su operazioni CRUD

Gli obiettivi di questi tesi sono lo studio comparativo di alcuni DBMS non relazionali e il confronto di diverse soluzioni di modellazione logica e fisica per database non relazionali. Utilizzando come sistemi di gestione due DBMS Document-based non relazionali, MongoDB e CouchDB, ed un DBMS relazionale, Oracle, sarà effettuata un’analisi di diverse soluzione di modellazione logica dei dati in database documentali e uno studio mirato alla scelta degli attributi sui quali costruire indici. In primo luogo verrà definito un semplice caso di studio su cui effettuare i confronto, basato su due entità in relazione 1:N, sulle quali sarà costruito un opportuno carico di lavoro. Idatabase non relazionali sono schema-less, senza schema fisso, ed esiste una libertà maggiore di modellazione. In questo lavoro di tesi i dati verranno modellati secondo le tecniche del Referencing ed Embedding che consistono rispettivamente nell’inserimento di una chiave (riferimento) oppure di un intero sotto-documento (embedding) all’interno di un documento per poter esprimere il concetto di relazione tra diverse entità. Per studiare l’opportunità di indicizzare un attributo, ciascuna entità sarà poi composta da due triplette uguali di attributi definiti con differenti livelli di selettività, con la differenza che su ciascun attributo della seconda sarà costruito un indice. Il carico di lavoro sarà costituito da query definite in modo da poter testare le diverse modellazioni includendo anche predicati di join che non sono solitamente contemplati in modelli documentali. Per ogni tipo di database verranno eseguite le query e registrati i tempi, in modo da poter confrontare le performance dei diversi DBMS sulla base delle operazioni CRUD.

Hierarchical Multi-Label Text Classification in a Low-Resource Setting

In this thesis we address a multi-label hierarchical text classification problem in a low-resource setting and explore different approaches to identify the best one for our case. The goal is to train a model that classifies English school exercises according to a hierarchical taxonomy with few labeled data. The experiments made in this work employ different machine learning models and text representation techniques: CatBoost with tf-idf features, classifiers based on pre-trained models (mBERT, LASER), and SetFit, a framework for few-shot text classification. SetFit proved to be the most promising approach, achieving better performance when during training only a few labeled examples per class are available. However, this thesis does not consider all the hierarchical taxonomy, but only the first two levels: to address classification with the classes at the third level further experiments should be carried out, exploring methods for zero-shot text classification, data augmentation, and strategies to exploit the hierarchical structure of the taxonomy during training.

Knowledge-based Chord Embeddings

This thesis develops AI methods as a contribution to computational musicology, an interdisciplinary field that studies music with computers. In systematic musicology a composition is defined as the combination of harmony, melody and rhythm. According to de La Borde, harmony alone "merits the name of composition". This thesis focuses on analysing the harmony from a computational perspective. We concentrate on symbolic music representation and address the problem of formally representing chord progressions in western music compositions. Informally, chords are sets of pitches played simultaneously, and chord progressions constitute the harmony of a composition. Our approach combines ML techniques with knowledge-based techniques. We design and implement the Modal Harmony ontology (MHO), using OWL. It formalises one of the most important theories in western music: the Modal Harmony Theory. We propose and experiment with different types of embedding methods to encode chords, inspired by NLP and adapted to the music domain, using both statistical (extensional) knowledge by relying on a huge dataset of chord annotations (ChoCo), intensional knowledge by relying on MHO and a combination of the two. The methods are evaluated on two musicologically relevant tasks: chord classification and music structure segmentation. The former is verified by comparing the results of the Odd One Out algorithm to the classification obtained with MHO. Good performances (accuracy: 0.86) are achieved. We feed a RNN for the latter, using our embeddings. Results show that the best performance (F1: 0.6) is achieved with embeddings that combine both approaches. Our method outpeforms the state of the art (F1 = 0.42) for symbolic music structure segmentation. It is worth noticing that embeddings based only on MHO almost equal the best performance (F1 = 0.58). We remark that those embeddings only require the ontology as an input as opposed to other approaches that rely on large datasets.