Text-based communication👇 doesn’t have that extra layer of vocal tone🗣 which tells you if sarcasm👻 is being used on you. Portuguese🇵🇹 researchers have built a neural network🤖 which determines whether someone is being sarcastic👻 or not😐. Being able to tell the difference between “great!” which can be both sarcastic😑 or literal😧 is vital for language processing for the future. Silvio Amir from the University of Lisbon🇵🇹 writes in a paper💬, “Lexical cues alone are insufficient to discern ironic intent. Appreciating the context of utterances is critical for this, even for humans.”
The paper📄 talks about a method wherein the neural network locates the user’s “embeddings,”🔎 or contextual cues🗣 from previously written content. It uses these factors to plot🕵 the 👤user with others & probably finds that they form 👥groups. Amir’s paper is due to be presented at CoNLL, a natural language learning conference👥 taking place next month.