Cognigo, provider of an AI-driven enterprise platform for data protection and compliance, has introduced a data protection capability that it says can differentiate between sensitive and non-sensitive data based on language context. The NLP Contextualization feature is included with the Cognigo platform to help organizations take control of their data and ensure regulatory compliance.
New data privacy and governance laws (GDPR, CCPA) are forcing companies to become more aware of the personal customer info they are holding, but existing solutions cannot identify most forms of sensitive information. Cognigo’s NLP Contextualization uses AI and proprietary deep learning models to filter through both structured and unstructured data and find sensitive personal information such as ethnicity or religion that may be used to identify specific individuals.
NLP Contextualization collects, extracts, and labels data across all data storage silos whether on-prem or in the cloud. It then analyzes the context of terms that may relate to specific persons in the sentence to determine whether or not sensitive data is present. For example, it can spot the difference between “the customer likes Italian food” and “the customer is an Italian citizen”; the former statement is permitted, but the latter is sensitive data that is subject to privacy laws, such as in Article 9 of the GDPR.
The new capability identifies all types of personal information, including sexual preferences; ethnicity; religious and political views; health issues, and criminal background.
“The most challenging aspect of compliance regulations is that they specify a wide range of sensitive data definitions that only human logic can understand,” said Guy Leibovitz, CEO and founder of Cognigo. “Rule-based and statistical models can match terms and categorize text at a very high level, but miss out on important sentence-level information. NLP Contextualization removes the need for that human logic element by automatically locating and identifying sensitive data regardless of its location or context.”
More information is available from Cognigo.