Identrics AD


: 14 déc. 2021


Recherche partenariat En recherche de financements
A propos

Identrics is an innovation hub enabling automatic analysis of information by employing machine learning and artificial intelligence. At Identrics we:

  • Offer solutions that turn data into patterns - semantic technologies for high automation of repeated processes for data processing and enrichment, and automated multilingual abstracting of texts which can read over 45 languages and produce abstracts in English.  
  • Can study information, analyze and offer patterns of events and behaviors online – statistics for volume, topics, and types of fake content per language, country or region, topic, or even named entity (victims of fake content or producers of fake content). To do this, we continuously research open-source frameworks and models, select and adapt them for our ongoing or ad-hoc purposes.
  • Utilise a media data collection solution, currently crawling more than 120 000 sources, aggregating about 1.3 mln. documents daily, also maintaining a historical archive going back to 2015. This solution and data warehouse can be used as an integral part of any solution which aims at monitoring fake news dissemination and origin detection.

Striving to stay at the top of the latest technological advances, we constantly conduct active research activities to develop and implement technologies that add value to the data-information-analysis chain and build award-winning AI solutions that transform unstructured data into digestible knowledge and actionable insights. In line with the team’s expertise in data science and semantic technologies we:

  • Are preparing a methodology for assessment of „trustworthy“ AI.
  • Have elaborated and implemented a methodology for identification of propaganda, disinformation or fake news. It relies on fact-checking, usage of extreme words and expressions, pics and visualizations, lack of authors, etc. This methodology is language agnostic and utilizes the human-in-the-loop model - a combination of machine learning models and people domain and language expertise.
  • We have created an abstractive summarization solution that generates near human-like shorter retellings of different types of texts. The algorithm “understands” the text and summarises it in “its own words”. Our product transforms multilingual content from all over the world and summarises it in English.

We are open to take an interest in ventures on distinctive subjects where we will be able to create innovations and to unfold our interests on topics like artificial intelligence and ML, digital media,fake news, human rights and democracy,  cybersecurity and others.

Mots clés et domaine d'activité:

 Transfert de technologie
 Agenda numérique pour l'Europe
 Innovation & Recherche
 Nouveau média
 Politique de cohésion
 Horizon Europe
 Open source
 Intelligence artificielle
 Défence et cyber-sécurité

Idées proposées

Appels à projets suivis
Pas d'appels à projets suivis!