In the contemporary music landscape, the process of networking and establishing strategic collaborations represents a significant challenge, especially for emerging artists and professionals who often rely on fragmented and data-unsupported methods. Existing digital tools, while effective at tracking explicit relationships, lack the ability to quantify and visualize implicit affinities between professionals. To address this need, this thesis introduces Synergos, a web application designed to map and analyze professional networks. The system builds a centralized knowledge base by enriching contact profiles with aggregated information from various public platforms Muso.ai, Spotify, Wikipedia. Through the application of machine learning models (VGGish for audio, Sentence-BERT for text), Synergos transforms this multimodal data into a combined affinity metric, quantifying the compatibility between each pair of professionals. The main result is an interactive platform that allows for the calculation and visualization of a realistic affinity score. Through a dynamic graph, Synergos not only represents direct connections but also models implicit ties, where the spatial arrangement of nodes and their interconnections reflect the calculated affinity score. The system's effectiveness demonstrates the validity of a data-driven approach for the analysis of professional music networks. However, it is noted that the precision of the affinity metric is highly dependent on the availability and richness of online data, proving more robust for established artists than for emerging ones. This work thus lays the groundwork for future developments in data integration and network analysis, while also highlighting the need for a more inclusive digital ecosystem for emerging talent.
Nel panorama musicale contemporaneo, il processo di networking e la creazione di collaborazioni strategiche rappresentano una sfida significativa, specialmente per artisti e professionisti emergenti, che spesso si affidano a metodi frammentati e non supportati da dati. Gli strumenti digitali esistenti, pur essendo efficaci nel tracciare relazioni esplicite, mancano della capacità di quantificare e visualizzare le affinità implicite tra professionisti. Per rispondere a questa esigenza, la presente tesi introduce Synergos, un'applicazione web progettata per mappare e analizzare reti professionali. Il sistema costruisce una knowledge base centralizzata arricchendo i profili dei contatti con informazioni aggregate da diverse piattaforme pubbliche (Muso.ai, Spotify, Wikipedia). Attraverso l'applicazione di modelli di machine learning (VGGish per l'audio, Sentence-BERT per il testo), Synergos trasforma questi dati multimodali in una metrica di affinità combinata, quantificando la compatibilità tra ogni coppia di professionisti. Il risultato principale è una piattaforma interattiva che permette di calcolare e visualizzare un valore di affinità realistico. Attraverso un grafo dinamico, Synergos non solo rappresenta le connessioni dirette, ma modella anche i legami impliciti, dove la disposizione spaziale dei nodi e le loro interconnessioni riflettono il punteggio di affinità calcolato. L'efficacia del sistema dimostra la validità di un approccio data-driven per l'analisi delle reti professionali musicali. Tuttavia, si evidenzia come la precisione della metrica di affinità sia strettamente dipendente dalla disponibilità e ricchezza dei dati online, risultando più robusta per artisti affermati rispetto a quelli emergenti. Questo lavoro pone quindi le basi per futuri sviluppi nell'integrazione di dati e nell'analisi di rete, evidenziando al contempo la necessità di un ecosistema digitale più inclusivo per i talenti emergenti.
Synergos: a knowledge graph-based platform for music professionals
PIRRELLO, ELIA
2024/2025
Abstract
In the contemporary music landscape, the process of networking and establishing strategic collaborations represents a significant challenge, especially for emerging artists and professionals who often rely on fragmented and data-unsupported methods. Existing digital tools, while effective at tracking explicit relationships, lack the ability to quantify and visualize implicit affinities between professionals. To address this need, this thesis introduces Synergos, a web application designed to map and analyze professional networks. The system builds a centralized knowledge base by enriching contact profiles with aggregated information from various public platforms Muso.ai, Spotify, Wikipedia. Through the application of machine learning models (VGGish for audio, Sentence-BERT for text), Synergos transforms this multimodal data into a combined affinity metric, quantifying the compatibility between each pair of professionals. The main result is an interactive platform that allows for the calculation and visualization of a realistic affinity score. Through a dynamic graph, Synergos not only represents direct connections but also models implicit ties, where the spatial arrangement of nodes and their interconnections reflect the calculated affinity score. The system's effectiveness demonstrates the validity of a data-driven approach for the analysis of professional music networks. However, it is noted that the precision of the affinity metric is highly dependent on the availability and richness of online data, proving more robust for established artists than for emerging ones. This work thus lays the groundwork for future developments in data integration and network analysis, while also highlighting the need for a more inclusive digital ecosystem for emerging talent.| File | Dimensione | Formato | |
|---|---|---|---|
|
Thesis 20250924.pdf
non accessibile
Dimensione
788.59 kB
Formato
Adobe PDF
|
788.59 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/243498