- Polished guard seeks to help polished chatbots and less subject to exploitation
- With NLP to its heart, it works by classifying the text on a four -point polite scale
- The data set and the source code are available on GitHub and Face Embraged
Intel unveiled Polite Guard, an open source AI tool aimed at assessing the politeness of a text and allowing AI chatbots to remain constantly polite to customers.
In an article on the Blog of the Intel Community, the latest addition to the Intel’s Intel portfolio hopes to provide a standardized framework to assess linguistic nuances in the communication led by AI.
Taking advantage of the natural language treatment (PNL), Intel says that polished guard, classifying the text in four different categories of policies, somewhat polished, neutral and impolies, helps to mitigate the Vulnerabilities of AI by “providing a Defense mechanism against contradictory attacks “.
The role of Guard Intel Poli for SMEs
According to Intel, the polished guard strengthens the resilience of the system by guaranteeing a coherent polished outing even when managing the potentially harmful text.
The company hopes that this approach will be “[improve] Customer satisfaction and loyalty “for the companies that implement it.
Released under the MIT license, Polite Guard grants developers the flexibility of modifying it and integrating it into their own projects.
Its data set and its source code are available on Github and Hugging Face, with other developments to publish via the Intel community blog.
