A Crowd-Annotated Spanish Corpus for Humor Analysis

Castro, Santiago; Chiruzzo, Luis; Rosá, Aiala; Garat, Diego; Moncecchi, Guillermo

doi:10.48550/arxiv.1710.00477

Cited by 2 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Typically, these datasets come with a binary labelling, i. e., texts are annotated on whether they are humorous or not. This is different in, e. g., #HashtagWars [24] and the dataset by Castro et al [25], where texts are annotated with different degrees of humour. For task 7 of SemEval-2017 [31], English puns have been labelled for the location and interpretation of the word being responsible for the pun.…”

Section: Textual Humour Recognitionmentioning

confidence: 93%

“…Another dataset often used is the Pun of the Day dataset created by Yang et al [23], consisting of puns scraped from a website and non-humorous data acquired from, amongst others, news websites. Recently, social media data became a popular source for textual humour datasets, with datasets based on tweets [24,25,26] and Reddit posts [11] being proposed. While the majority of the available datasets are in English, corpora in Italian [27], Spanish [25], Chinese [28], and Russian [29,30] have been created.…”

Section: Textual Humour Recognitionmentioning

confidence: 99%

“…Recently, social media data became a popular source for textual humour datasets, with datasets based on tweets [24,25,26] and Reddit posts [11] being proposed. While the majority of the available datasets are in English, corpora in Italian [27], Spanish [25], Chinese [28], and Russian [29,30] have been created. Typically, these datasets come with a binary labelling, i. e., texts are annotated on whether they are humorous or not.…”

Section: Textual Humour Recognitionmentioning

confidence: 99%

See 2 more Smart Citations

Multimodal Prediction of Spontaneous Humour: A Novel Dataset and First Results

Christ¹,

Amiriparian²,

Kathan³

et al. 2022

Preprint

View full text Add to dashboard Cite

Humour is a substantial element of human affect and cognition. Its automatic understanding can facilitate a more naturalistic human-device interaction and the humanisation of artificial intelligence. Current methods of humour detection are solely based on staged data making them inadequate for 'real-world' applications. We address this deficiency by introducing the novel Passau-Spontaneous Football Coach Humour (Passau-SFCH) dataset, comprising of about 11 hours of recordings. The Passau-SFCH dataset is annotated for the presence of humour and its dimensions (sentiment and direction) as proposed in Martin's Humor Style Questionnaire. We conduct a series of experiments, employing pretrained Transformers, convolutional neural networks, and expert-designed features. The performance of each modality (text, audio, video) for spontaneous humour recognition is analysed and their complementarity is investigated. Our findings suggest that for the automatic analysis of humour and its sentiment, facial expressions are most promising, while humour direction can be best modelled via text-based features. The results reveal considerable differences among various subjects, highlighting the individuality of humour usage and style. Further, we observe that a decision-level fusion yields the best recognition result. Finally, we make our code publicly available at https://www.github.com/EIHW/passau-sfch. The Passau-SFCH dataset is available upon request.

show abstract

Section: Textual Humour Recognitionmentioning

confidence: 93%

Section: Textual Humour Recognitionmentioning

confidence: 99%

Section: Textual Humour Recognitionmentioning

confidence: 99%

See 1 more Smart Citation

Multimodal Prediction of Spontaneous Humour: A Novel Dataset and First Results

Christ¹,

Amiriparian²,

Kathan³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Textual Content-related features Key concepts (e.g. situation) on which the humorous story is built [8], [9], [24], [39] Incongruity Disconnection Semantic disconnection (e.g., contrast) between two content words in a sentence [8], [9], [14], [45] Intra-sentence repetition Repeating similar objects in a sentence [8], [9], [14], [39], [45] Human-centeredness Polarity Positive/negative orientation of emotion [8], [9], [12], [45], [46] Subjectivity Subjective/objective orientation [9], [45], [47], [48] Phonetic style Alliteration Occurrences of the same letter or sound at the beginning of a group or words [8], [9], [12], [20], [45] Rhyme Repetition of similar sounds in the final stressed syllables of a group of words [9], [12], [20], [45] Build-ups Inter-sentence repetition Concepts (e.g., a person) that have been previously told before a punchline [4], [6], [49] Audio Volume Volume variation Variation in volume: softer and louder [16], [25], [26], [50] Pitch Stress Vocal stress by raising pitch [15], [16], [25], [26],…”

Section: Humor-related Features Subcategory Description Referencesmentioning

confidence: 99%

DeHumor: Visual Analytics for Decomposing Humor

Wang

Yao

et al. 2022

IEEE Trans. Visual. Comput. Graphics

View full text Add to dashboard Cite

Despite being a critical communication skill, grasping humor is challenging-a successful use of humor requires a mixture of both engaging content build-up and an appropriate vocal delivery (e.g., pause). Prior studies on computational humor emphasize the textual and audio features immediately next to the punchline, yet overlooking longer-term context setup. Moreover, the theories are usually too abstract for understanding each concrete humor snippet. To fill in the gap, we develop DeHumor , a visual analytical system for analyzing humorous behaviors in public speaking. To intuitively reveal the building blocks of each concrete example, DeHumor decomposes each humorous video into multimodal features and provides inline annotations of them on the video script. In particular, to better capture the build-ups, we introduce content repetition as a complement to features introduced in theories of computational humor and visualize them in a context linking graph. To help users locate the punchlines that have the desired features to learn, we summarize the content (with keywords) and humor feature statistics on an augmented time matrix. With case studies on stand-up comedy shows and TED talks, we show that DeHumor is able to highlight various building blocks of humor examples. In addition, expert interviews with communication coaches and humor researchers demonstrate the effectiveness of DeHumor for multimodal humor analysis of speech content and vocal delivery.

show abstract

A Crowd-Annotated Spanish Corpus for Humor Analysis

Cited by 2 publications

References 0 publications

Multimodal Prediction of Spontaneous Humour: A Novel Dataset and First Results

Multimodal Prediction of Spontaneous Humour: A Novel Dataset and First Results

DeHumor: Visual Analytics for Decomposing Humor

Contact Info

Product

Resources

About