RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

OntosMiner

Product
Developers: AviComp Services
Date of the premiere of the system: 2001/06/15

OntosMiner is an end-to-end system of recognition in texts in a natural language of references of objects of attention (named entities) and semantic links between them (facts) and also defining the general tonality[1].

The core of technology is the linguistic OntosMiner processor analyzing the text using ontologies and dictionaries, editable the user, special evristik.

Setup of the linguistic processor gives to the user opportunities:

  • Editing model of the interesting data domain;
  • Editing dictionaries of objects of attention;
  • Task of the list of phrases on which an object can be found in the text;
  • Refining of an obligatory context of an object;
  • Setup of options of recognition of an object from the text;
  • Task of a method of identification of synonymous objects;
  • Replenishment of dictionaries the new objects selected by heuristic rules

At the heart of the OntosMiner mechanism - the patent technology belonging to the class NLP (Natural Language Processing - natural languag processing).


Operation of the semantic processor, 2014

Ontosminer can be used for the analysis and systematization of arrays of correspondence, corporate documents and news publications for the benefit of state agencies and intelligence agencies. In one of federal agencies where a system is implemented, it keeps the automated analysis and classification of incoming correspondence (more than 1 thousand addresses a day on 1 thousand categories)[2].

According to developers, one of the most important options of a system - a possibility of its setup by the customer: unlike the traditional systems of search and monitoring, Ontosminer allows thin setup of the linguistic processor according to the purposes and the customer's tasks, by means of machine learning technologies (machine learning). On the basis of the text examples entered by the user a system studies understanding of sense of separate words and texts on the set subject.

For April 15, 2015 a system is capable to analyze semantic structure of any kind of data, declared in the company.

History

In the late nineties the group of the Swiss businessmen decided to invest means in advanced program development. Selected the direction from an analysis result of the market - knowledge management, namely - the project of difficult analytical processing of texts[3].

For its implementation the Ontos company was created. Studying offers, foreign investors came into contacts with the Russian company Avicomp Services having experience of creation of the analytical systems and a ten-year experience of software developer and system integrator.

Avicomp Services acquired the rights to use Ontos technology in the solutions and transferred a part of the developments for inclusion in Ontos core.

On the next stage of cooperation the Russian company suggested the Swiss partner to develop functionality of products towards analytical processing of text information using NLP technology what involved group of linguists of computer center of the Russian Academy of Sciences and DFKI company to.

At the beginning of 2003 the market was entered by the first products Ontos Series, in the summer of 2003 the first project on their basis in the Ministry of Internal Affairs of Kyrgyzstan is implemented.

The product for semantic information processing is delivered not only to the Russian state departments - to neighboring countries, in RIA Novosti. Abroad semantic solutions are implemented by means of related company "Eventos".

Notes