Scientists create a neural network that can spot sarcasm on Twitter

We have all received a tweet that we're not sure is sarcastic or just plain mean. Now scientists have built a neural network to detect'the lowest form of wit' on Twitter. It does this by putting a social media message in context and if a message seems to disagree with the majority of messages about a topic, the writer is probably being sarcastic. Scientists have built a neural network to detect'the lowest form of wit' on Twitter. The neural network seeks to identify sarcastic comments by first finding the user's'embeddings' which include contextual cues such as the content of previous tweets and their interests.