There is a correlation between our increasingly digitalized daily life and the development of semantic technologies in the field of Security . Going beyond the dream, we realize how today Machine Learning systems can represent an indispensable tool for public and private security operators, not only to classify, but also to correlate images, sounds and words in an increasingly accurate way. Let’s talk about GeoCyber in the Security field, enabled by Semantic Recognition, Audio Recognition & Video Recognition technologies.
Semantics and NLP
Let’s start with two brief definitions that help us to better focus on what we are dealing with:
- Semantics is that branch of linguistics that deals with studying the meaning of the words and utterances of a language.
- The Natural Language Processing (NLP) which we have already talked about in the article “Natural Language Processing: Tasks, Applications and Sentiment Analysis” is the branch of Artificial Intelligence which is concerned with making human (or natural) language understandable to a computer. To take a trivial example: customer service chatbots .
So, to capture the meaning of semantic data and turn it into valuable information , we use Natural Language Processing. The recognition of the entities is given by a classification of the parts of the text through the assignment of tags. The categories can be varied such as places , companies , personal names , etc.
An increasingly common use is that of extracting information from governance documents, such as reports and procedures, to allow for quick consultation.
GeoCyber & Security
The GeoCyber solutions developed by Drive2Data in collaboration with Pragma Etimos in the Security & Investigation field , are the result of over twenty years of experience in the processing of semantic and territorial data, combined with the latest innovations in the field of Artificial Intelligence, Neuronal Models and NLP.
Security breach tracking: Drive2Data’s solutions
Thanks to Drive2Data’s cutting-edge technology, it is possible to track criminal activity and breaches and gather useful information for investigations.
The true innovative strength of GeoCyber lies in the union of different technologies which, starting from the analysis of large amounts of heterogeneous data, automatically create classifications and correlations between them, offering security operators an unprecedented added value.
Let’s see specifically the three integrable solutions that enable the GeoCyber :
Vulgaris : Semantic Recognition Engine
Vulgaris , allows the reusability of heterogeneous unstructured data (personal data, territorial data, etc ..) through the use of mixed NLP algorithms and Neuronal Models. Furthermore, the data are classified and correlated with each other, creating in a few moments and automatically a network / map of information that can include: names, surnames, company, company role, places, and other keywords of interest for the surveys. and / or analyzes.
Polyphonic : Audio Recognition Engine
Polyphonic is the solution that deals with audio analysis , classifying voices and places through Neuronal Models . This translates into the possibility of recognizing people and places of interest, correlating the two types of data, for example to find out if there are commonplaces to certain criminal subjects.
Furthermore, thanks to the “ Speech to Text ” tool, fruit of the important know – how acquired by Pragma Etimos in the Audio Recognition and NLP fields , it is possible to transform a vocal speech into a textual one to keep track of conversations, thus more easily analyzed.
Effigies : Video Recognition Engine
Effigies allows the automated intelligent analysis of large video streams , thanks to the use of Machine Learning and Deep Learning algorithms. It can identify and search in real time : vehicles, license plates, objects, people, animals , even by selecting color classes. The distinctive element of our solution is the creation of ad hoc neuronal models also for the recognition of scenarios . It can also be perfectly integrated into devices (such as surveillance cameras) already in use.
The GeoCyber innovation comes by integrating Vulgaris , Polyphonic and Effigies thanks to which it is possible to create a flow of information deriving from texts, audio and video able to geographically track security breaches at 360 °.