Threat and Leak Analysis

Try API for Free
Ask for Customization

Threat tracking

Intellexer tools can be used to improve the effectiveness of the security system by the recognition of predetermined entities (opinions, people, companies, locations…) and identification of semantic relations between them.

With the help of such Intellexer modules as Linguistic Processor, Sentiment Analyzer, and Named Entity Recognizer you can automatically extract semantic signs of potential threats from the information channels (emails, messengers, social networks, etc.):

  • Potentially dangerous subject-action-object patterns ( i.e., We - are - at war or you - will be - our targets )
  • Potentially dangerous sentiments and opinions ( The Federal Reserve warned that froth in the riskiest parts of debt markets may lead to more defaults. )
  • The mention of specific entities (persons, organizations, locations, etc.) ( Al-Qaeda, ISIS )

Extracted semantic signs are sought in the knowledge base of precedent threats. If some security triggers are detected, system will generate a warning message.

Our semantic solutions can be applied to the various subject fields, such as criminal actions, terrorist activities, economy monitoring, infrastructure security, mass media, etc. The results obtained are extremely accurate with the minimum information noise of the extracted objects and their parameters.

Leak tracking

Use our semantic tools for document search and comparison enables creation of a system, which can trace the appearance of certain documents on the intranet/internet, or improve existing security systems by adding natural language analysis feature.

Search for similar documents functionality is provided by Comparator module. It allows users to find similar documents on the basis of content analysis. Our tool looks for similar documents resting on the model document, hence, it doesn't require any training stage. Comparator is able to estimate proximity level leaning on meaning analysis, even if the document content has been deliberately changed.

Search for protected idea can be realized on the basis of Question Answering System. You'll be able to scan the document base (emails, network storage, messenger's history) for the presence of certain ideas and concepts, which are objects of security. Due to natural language core our tools can retrieve any messages or documents, where the required concept is mentioned, no matter which words were used to describe the information.