SCALING, MODELING AND REASONING ON COMPLEX TYPES OF DATA FOR HIGH-LEVEL ANALYSIS APPLICATIONS

摘要

High-level scalable fusion of structured and unstructured data involves the ingestion and processing of unstructured data to produce a statistical model stored as extracted entities and then mapped to a collection of Resource Description Structure (RDF) triples ) and applying semantic analysis to a set of structured data to produce a logical model stored as a collection of triples. Reasoners are applied to the two models generating an extended knowledge graph of basic and inferred knowledge which is broken down into a large database, each row storing a corresponding triple, and a reasoner converting RDF triples to associated triplets by adding a new column to the database in response to the detection of a new attribute for a subject already present in one of the rows of the database so that the new attribute is stored in the new column in a new row created for the subject already present.