At the moment we do sentiment on a project by project basis using sentiment kernels based on a knowledge-based approach. The goal of intents is that on the analysis side we don’t rely on the subjectivity of sentiment but on actionable categories that may have sentiment features. For example, the intent “customer+service” may have features like “good” or “horrible” or there may be intents like “never+fly” or “worst+ever” about bad experiences with companies that can be used to trigger alerts.
 
Our partner company Séntisis provides sentiment for the Spanish language through an API that is sold together with lang.ai’s intent induction process and other products they offer.

Did this answer your question?