Sasuke Semantic Classifier



Knowledge production is one of the main task of enterprises and administrations, but since resources are shared between different department, groups or agencies it’s often needed to retrieve those documents in order to plan and set up our own business. And when resources are defined by different schemas, store in multiple repositories or simply wrongly generated search for the very useful information could be an hopeless quest.

Ontology-based techniques use semantic to organize knowledge through a domain specific conceptual network, allowing user to refine content search. The shift from classic to semantic search is the shift of focus from syntactic analysis to semantic computing, by which knowledge manager are able to model relationship between entities that represent the domain’s information.


Conceptual modeling does not increase search improvement alone, but it offer a new way to update database information starting from concepts you’re already using in ontology in order to define and structure knowledge. So, semantic properties are extended to new resources along their consistency with already organized information in order to classifying it in an easy an semi-automatic way. Sasuke is able to:

  • classify documents;
  • classify internet/intranet website;
  • enhance semantic based search;
  • automatic metadata attribution to new resources;
  • find multilingual correlates;
  • canalize documents taking into consideration the need of different professional skills.




Sasuke can be easily employed to solve:

  • large data volume classification;
  • single document classification;
  • domain ontology modeling, enabling ad hoc modification;
  • multilingual classification based on the very same ontology;
  • enable single/multiple concept’s classification;
  • embedded web spider to whole website’s analysis;
  • use W3C standards such as OWL (ontology web language) SKOS (simple knowledge organization system) ADNOM (Public administration’s ontology standard committee) CEN/ISSS (European network for administrative nomenclature);
  • use Wordnet ontology in order to expand concept’s taxonomy;
  • make use of faceted browsing.

Sasuke could be easily integrated with other products and can be used remotely via SaaS (software as a service. It’s built on J2EE – JAVA 2 Enterprise Edition and Application Server (TomCat, Oracle, etc). It’s accessible via API Java or via Web Services retrievable through C, C++, PHP applications. EvoDevo support business partnership with the aim of:

  • personalizza Sasuke sulle applicazioni aziendali, interfacciandosi con le applicazioni già esistenti (sistemi documentali, CRM, siti web, Content Manager…);
  • helping companies to customize and integrate Sasuke on already existing applications (CRM, website, CMS etc.);
  • helping companies to develop their own domain ontology.

Evodevo Sasuke Classifier was integrated with two new a complementary modules:

  • FM module: it integrates with the Facility Management (FM) product ref2space by IdeaRe;
  • generic data: access to generic data, including Facebook ones.

The modules are funded by a CoResearch project by Lazio Innova.