This paper investigates the problem of automatic humour recognition, and provides and in-depth analysis of two of the most frequently observed features of humorous text: human-centeredness and negative polarity. Through experiments performed on two collections of humorous texts, we show that these properties of verbal humour are consistent across different data sets.

In a familiar parable, a group of blind men try to discern the shape of an elephant, but each man’s perspective differs depending on whether he’s touching the trunk, tusk, or tail (Fig. 1). After 2,500 years of studying humor, scientists similarly have differing perspectives on what makes things funny (1). We present three common perspectives on humor. Although each is insightful, no one perspe...

The purpose of this study was to investigate the neuropsychological process and timing of appreciating humor upon reading a set of Japanese riddles which were made up of 4 distinct phases: “Given A (1st phase), I'd say B (2nd phase). Do you know why? (3rd phase) It is because X (4th phase).” To investigate how the brain responds when an individual finds the answer to the riddle (“It is because ...

Humans use humour to ease communication problems in human-human interaction and in a similar way humour can be used to solve communication problems that arise with humancomputer interaction. We discuss the role of embodied conversational agents in human-computer interaction and we have observations on the generation of humorous acts and on the appropriateness of displaying them by embodied conv...

The humor effect refers to a robust finding in memory research that humorous information is easily recalled, at the expense of recall of nonhumorous information that was encoded in close temporal proximity. Previous research suggests that memory retrieval processes underlie this effect. That is, free recall is biased toward humorous information, which interferes with the retrieval of nonhumorou...

The success of a humour-generation program is usually assessed by having human judges rate texts. However, there has been little consideration of the patterns shown by such judgements, particularly in terms of consistency. We present two small studies which attempt to gauge the consistency of human judgements about humorous aspects of texts, and discuss some of the methodological issues involved.