Becoming JILDA

The difficulty in finding useful dialogic data to train a conversational agent is an open issue even nowadays, when chatbots and spoken dialogue systems are widely used. For this reason we decided to build JILDA, a novel data collection of chat-based dialogues, produced by Italian native speakers and related to the job-offer domain. JILDA is the first dialogue collection related to this domain for the Italian language. Because of its collection modalities, we believe that JILDA can be a useful resource not only for the Italian research community, but also for the international one.