Bank of Italy

Communicating and Understanding a Knowledge Graph: From Search to a Visualization and Navigation Tool

  • Digital & Comms
industry
FSI
know how
  • App Modernization
  • Machine Learning

1 Starting Point

1 Need

Banca d’Italia ci ha presentato le seguenti sfide:

  • implement an internally designed IT methodology (KG-Model)
  • create an environment for modeling and visualizing Knowledge Graphs
  • simplify the Knowledge Graph design flow
  • communicate the project results in a way that is understandable even to non-experts
  • develop knowledge representation models suited to business needs

2 Discovery

2 Direction

Abbiamo sviluppato Clickable Graph Schema, un’applicazione su misura, che ha prodotto questi risultati:

  • improved modeling and visualization of Knowledge Graphs
  • standardization of data representation
  • clear communication of results that can be understood even by non-experts
  • incremento dell’efficienza nei progetti in termini di qualità, scalabilità e tempistiche
  • improved collaboration between research and business

3 How

3 The challenge

Researchers at ART, the research and development unit of the Bank of Italy , approached us to implement an IT methodology they had developed (KG-Model). Specifically, they wanted to create an environment for conceptual modeling and visualization of Knowledge Graph (KG) schemas. A KG is a data model particularly suited to representing knowledge in domains characterized by the presence of many interconnected entities.

Un KG è composto da dati di business (cosiddetta componente “estensionale”) e da conoscenza definita sotto forma di modelli logici, probabilistici, matematici in genere (la cosiddetta componente “intensionale”).

ART ha una notevole esperienza nella progettazione e implementazione di Knowledge Graph, per il supporto al Decision Making basato su dati nell’ambito di applicazioni economiche e finanziarie.

Dinova is the ideal partner to develop and make the methodology usable, and to open up new business scenarios, thanks to a strong expertise in data and machine learning combined with a knowledge of UX methodologies and tools and its communication style that speaks to people.

"They were building several different Knowledge Graphs for a variety of business objectives. The same knowledge representation technology was therefore being used in many complex and diverse contexts. So we created the KG-Model methodology to standardize our work," said Luigi Bellomarini, Deputy Director of ART (Applied Research Team) at the Bank of Italy.

ART, in consultation with Dinova, realized that KG-Model could best be activated through a visual tool to simplify the design flow of a KG , better understand the work done on each project and communicate the results.

The need was twofold: consistency of design and communication. Visual knowledge representation models serve to fully understand a certain domain of interest, but also to communicate its real value to non-experts and to respond to business requests.

4 What

4 Solution by Dinova

L’obiettivo era chiaro: serviva un’interfaccia per rendere visibile e attuabile un framework che altrimenti avrebbe rischiato di rimanere teorico.

Thanks to the collaboration with Dinova, an application called Clickable Graph Schema was developed, a visualization and modeling environment for Knowledge Graph schemas. A visual way to reveal not only the data, but also the knowledge generated.

We created a completely custom solution for the Bank of Italy, tailored to the client's specific needs. The developed tool offers consultation, interactivity, and navigation features and is part of a strategy that promotes continuous improvement and development.

Una volta ottenuta la visualizzazione dei dati è possibile derivare, in automatico, la relativa documentazione.

“Attraverso una metodologia visuale abbiamo potuto creare agilmente degli schemi per parlare con gli utenti, creare applicazioni più versatili e documentazioni utili”

Luigi Bellomarini, Vicedirettore di ART (Applied Research Team) di Banca d’Italia.

The application has allowed the bank's users to see how technologies like this can enable highly complex business activities, saving valuable time. It has also facilitated the work of interdisciplinary teams.

5 Why

5 Why Dinova?

Dati, persone e innovazione: un’alchimia di successo

Siamo stati scelti per la nostra solida esperienza nei progetti di gestione e modellazione di dati in contesti AI-intensive, ma anche per l’attenzione verso gli aspetti di User Interface, System Integration e comunicazione.

Our approach has always been people-focused. To understand the most appropriate solution for each client, we map the business needs at an informational level, collecting, interpreting, and qualifying data from the company's knowledge base.

The project, among other things, brought together the worlds of research and business. Creating a visual representation around a methodology allowed people with different skills to connect with a typical researcher's way of working

Bank of Italy has already begun applying the new technology to banking and financial supervision projects. The collaboration with Dinova continues, and future developments are already being planned to make this project even more impactful for the R&D department and the business.

Other success stories