Linked Data

Semantic approach

Semantic approach incorporates the use of technology and methods of technical knowledge which derive from both academic and industrial strains of research. On one side, there is Natural Language Processing (NLP), which to date has developed various applications over the years and has recently been infused by technology that integrates the vision of Semantic Web – a bundle of technologies and best practices supported by W3C (World Wide Web Consortium), designed by Tim Berners-Lee, who aims not only to make the web more navigable but also to query through the use of software applications. This gave rise to the Linked Data concept.

Recently NLP (tokenisation based on statistics and machine learning) and Semantic Web (reasoning, representations based on ontologies and controlled wording) techniques have been working well side by side.

NLP, in particular, can be seen as a tool for the analysis of unstructured data which can bring out the syntactic and reasoning structures (parts of conversations, sentences, dependency features). On the basis of this analysis, a representation can be inserted at a higher level, which will highlight the existing related texts, concepts and entities.

The shared vision of the two disciplines is ruled by the intention to represent, process and analyse structured or unstructured data emphasising as much as possible the data’s meaning.


Advantages of the semantic approach

The advantages of a semantic approach for search engines can be summarised as follows:

Semantic Approach offers moreover a distinctive advantage in terms of a framework for Information, allowing the consistent use of annotation resources (taxonomies, synonyms, word lists) with an enhanced vision by reusing documents uninterruptedly.

Discover how to get the best results by using Sophia Semantic Search

[cml_media_alt id='1222']celi-prodotti-2[/cml_media_alt]