RTVE Play's contextual information interrogation, search and retrieval system.

A knowledge graph is not only a means to integrate heterogeneous and distributed information, to improve the interoperability of systems or to facilitate the representation of knowledge to documentalists and other expert personnel, it is also a means to build intelligent utilities for people. RTVE has implemented for this purpose and on the basis of the RTVE Grafo knowledge graph, a faceted semantic search engine that allows textual search, contextual search, advanced search by entities, as well as the generation of pages with enriched information. In short, it has built a more intuitive, simple, natural and contextual system of interrogation and retrieval of RTVE PLay information. This means, in practice, that searches are not only based on keywords, but on the meaning and context of the terms, which makes them more efficient and relevant.

The RTVE Grafo project aims to provide the best possible experience to its digital visitors, offering a search engine that interrogates a knowledge graph where audiovisual resources are linked to each other, making it possible to present results well organized by entities, as well as enriched and contextualized. In short, one of the aims of the project has been to provide the public with a more intuitive, intelligent, personalized, semantically meaningful and effective browsing and search experience.

And all this, ensuring that this new experience of knowledge discovery and navigation through the contents of the Spanish Radio and Television works equally on any kind of device, so that all users can access what interests them and act as they wish in the RTVE Grafo framework at any time and place.

RTVE Grafo is built on GNOSS technology, a company that for two decades has been working on the development of its own capabilities for the construction and exploitation of knowledge graphs, a mathematical structure, computable by machines and interrogable by people, that represents well, both the structure of reality and that of our cognition and that, therefore, are at the core of the Artificial Intelligence Symbolic Program. GNOSS is also a pioneer in the construction of the Semantic Web in Spain and in the wider Spanish language. In 2010, GNOSS became the first Spanish company in linking its GNOSS and DIDACTALIA projects with Freebase, New York Times and DBPedia and, as a consequence, in being part of the global project of building the semantic web with open and linkable data orLinked Open Data Cloud.

In order to make its purpose a reality, GNOSS has developed GNOSS Semantic AI Platform, development platform for the construction of projects Inteligencia Artificial Simbólica o Semántica, which includes an ecosystem of Artificial Intelligence Cognitive Services. The Artificial Intelligence family of services from GNOSS Semantic AI Platform consists of GNOSS Sherlock NLP-NLU, which encompasses a set of natural language processing services oriented to the recognition and extraction of entities and topics in texts and their subsequent consolidation in a knowledge graph, and GNOSS LOKUT, a natural language interrogation system that hybridizes knowledge graphs with LLMs in order to ensure an auditable, traceable and reproducible AI output, which are the three attributes of an Explainable AI.