The Impact of Semantic and Morphosyntactic Ambiguity on Automatic Humour Recognition

Humour is one of the most amazing characteristics that defines us as human beings and social entities. Its study supposes a deep insight into several areas such as linguistics, psychology or philosophy. From the Natural Language Processing (NLP) perspective, recent researches have shown that humour can be automatically generated and recognized with some success. In this work we present a study carried out on a collection of English texts in order to investigate whether or not semantic and morphosyntactic ambiguities may be employed as features in the automatic humour recognition task. The results we have obtained show that it is possible to discriminate humorous from non humorous sentences through features like perplexity or sense dispersion.