We use essential cookies for the proper functioning of the website and additional ones to make interaction with the site as convenient as possible. It helps us personalize your user experience as well as obtain analytical information to improve the service.
If you agree to accept all cookies, click "Accept all"; if not, click "Only essential". To learn more, view the Cookie Policy.
New advances in AI: scientists develop a sarcasm detector on social networks
American scientists have managed to develop a detector that allows you to determine the presence of sarcastic statements on social networks. The principle of operation of "devices" is based on the analysis of user sentiments.
Artificial intelligence allows you to analyze emotions in posts posted on social networks. Positive, negative or neutral, which are related to the text posted on the web. The technique developed by American researchers allows reading this information with an accuracy of 60-70%.
Huge databases have been loaded to find patterns in publications that indicate sarcasm. Then their accuracy was checked. Through AI, the technique not only analyzes, but also learns - learns to recognize sarcastic statements.
It is noted that even a person cannot always understand where and when sarcasm is used. For software, this is even more difficult. But the development of American researchers helps to find sarcastic words and expressions, determine the relationship between them and sort the text.
Early recognition of sarcasm caused difficulties due to the nature of human use. After all, as a rule, he depended on intonations, facial expressions, which are difficult to convey in the text. But in practice, it turned out that AI can cope with this.
Scientists want to use their development to improve the work of applications that work online. As well as devices for identifying the identity of users, other automated systems such as "smart" devices.
Recently it was reported that scientists know how to turn used grain into protein and fiber.
Artificial intelligence allows you to analyze emotions in posts posted on social networks. Positive, negative or neutral, which are related to the text posted on the web. The technique developed by American researchers allows reading this information with an accuracy of 60-70%.
Huge databases have been loaded to find patterns in publications that indicate sarcasm. Then their accuracy was checked. Through AI, the technique not only analyzes, but also learns - learns to recognize sarcastic statements.
It is noted that even a person cannot always understand where and when sarcasm is used. For software, this is even more difficult. But the development of American researchers helps to find sarcastic words and expressions, determine the relationship between them and sort the text.
Early recognition of sarcasm caused difficulties due to the nature of human use. After all, as a rule, he depended on intonations, facial expressions, which are difficult to convey in the text. But in practice, it turned out that AI can cope with this.
Scientists want to use their development to improve the work of applications that work online. As well as devices for identifying the identity of users, other automated systems such as "smart" devices.
Recently it was reported that scientists know how to turn used grain into protein and fiber.