Manuel Peña Muñoz, Alejandro Llaves and Terunobu Kume.
Abstract: In Fujitsu Laboratories of Europe (FLE), we are developing a platform to get insights in the financial sector from the analysis of multiple heterogeneous data sources. At the core of the platform, there is a knowledge graph that is populated with new nodes and relationships as new data are ingested. The benefit of using a knowledge graph in our platform is threefold. First, it reduces the time spent on simple and repetitive data integration tasks. Second, the acquired knowledge is used by a virtual assistant to facilitate the data reconciliation process, thus users do not need to be experts to merge a new dataset. And third, it adds extra value to the customer information by linking knowledge to open and private data sources.
Keywords: knowledge graph; dynamic data loading; financial domain