Big Semantic Data

These projects collect massive amounts of data that represent a certain field of knowledge (research, industry 4.0, health, insurance, security, etc.), with a descriptive interpretation that allows obtaining relevant information and anticipating emerging issues; identify frontiers of knowledge; understand how the entities in the graph condense, hybridize and feed back; locate relevant people and organizations and understand their connections and affinities. In short, speed up the process of consultation and acquisition of knowledge, making the information more meaningful, useful and pertinent.

Semantic Big Data projects are no different from projects with less massive data; They continue to need a technological solution that responds to the challenge of representing, asking questions in an expressive and unmanaged way, enriching, visualizing, anticipating future trends and scenarios and making personalized recommendations on the set of massive data that represent their field of knowledge. Our solution consolidates the data in a semantically interpreted massive knowledge graph, which allows intelligent software agents and people to consult these semantic databases and obtain enriched results, applying on them in a combined way interrogation technologies based on the possibilities to compute First Order Logic and Deep and Machine Learning (Machine and Deep Learning).

Making this kind of calculation possible runs into limitations relative to the usual computational strategies on knowledge graphs that include summarization, particularly if they are very massive. Counting can be too expensive a task when it is done on the relationships contained in a very expressive and massive knowledge graph. To overcome these limits, we apply analytical and predictive strategies based on the partition and fragmentation of the knowledge graph and on the probabilistic and logarithmic estimation of massive quantities, which allow us to develop rapid calculation processes on large amounts of structured data.