Vector Label Propagation (VLP) is a machine learning technique used for tasks like semi-supervised learning and graph-based analysis. It is valuable in various fields, including social network analysis and recommendation systems, where each data point has multiple attributes. This paper discusses the development of a web application aimed at demonstrating the Agent-based vector-label propagation algorithm's (AVPRA) operation. Prior to this application, AVPRA execution and visualization were challenging for non-experts, relying on scripts. The web app offers a user-friendly, graphical interface, enabling real-time tracking and transparent visualization of the algorithm's iterations.
A web application to demonstrate the properties of Label Propagation algorithms / P. Ceravolo, S. Maghool, A.G. Talpalaru (CEUR WORKSHOP PROCEEDINGS). - In: CoopIS-D 2023 : CoopIS Demonstration Track 2023 / [a cura di] F. Mannhardt, N. Assy. - [s.l] : CEUR-WS, 2023. - pp. 16-20 (( convegno CoopIS tenutosi a Groningen nel 2023.
A web application to demonstrate the properties of Label Propagation algorithms
P. Ceravolo;S. Maghool
;
2023
Abstract
Vector Label Propagation (VLP) is a machine learning technique used for tasks like semi-supervised learning and graph-based analysis. It is valuable in various fields, including social network analysis and recommendation systems, where each data point has multiple attributes. This paper discusses the development of a web application aimed at demonstrating the Agent-based vector-label propagation algorithm's (AVPRA) operation. Prior to this application, AVPRA execution and visualization were challenging for non-experts, relying on scripts. The web app offers a user-friendly, graphical interface, enabling real-time tracking and transparent visualization of the algorithm's iterations.| File | Dimensione | Formato | |
|---|---|---|---|
|
paper-4.pdf
accesso aperto
Tipologia:
Publisher's version/PDF
Licenza:
Creative commons
Dimensione
1.25 MB
Formato
Adobe PDF
|
1.25 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.




