Multilingual Multimodal Detection of Humour in Stand-Up Comedy
| Název práce v češtině: | Vícejazyčná, multimodální detekce humoru ve stand-up komedii |
|---|---|
| Název v anglickém jazyce: | Multilingual Multimodal Detection of Humour in Stand-Up Comedy |
| Klíčová slova: | humor|automatická detekce|stand-up comedie|vícejazyčný|multimodální |
| Klíčová slova anglicky: | humour|automatic detection|stand-up comedy|multilingual|multimodal |
| Akademický rok vypsání: | 2022/2023 |
| Typ práce: | diplomová práce |
| Jazyk práce: | angličtina |
| Ústav: | Ústav formální a aplikované lingvistiky (32-UFAL) |
| Vedoucí / školitel: | RNDr. Martin Holub, Ph.D. |
| Řešitel: | skrytý - zadáno a potvrzeno stud. odd. |
| Datum přihlášení: | 29.03.2023 |
| Datum zadání: | 03.04.2023 |
| Datum potvrzení stud. oddělením: | 11.04.2023 |
| Datum a čas obhajoby: | 06.09.2023 09:00 |
| Datum odevzdání elektronické podoby: | 21.07.2023 |
| Datum odevzdání tištěné podoby: | 23.07.2023 |
| Datum proběhlé obhajoby: | 06.09.2023 |
| Oponenti: | Mgr. Mateusz Krubiński |
| Zásady pro vypracování |
| Humour is conveyed through a range of verbal and nonverbal cues (Aguert, 2022), but most research on humour detection has focused on textual data (Peyrard et al., 2021). However, there is growing interest in multimodal humour detection, with previous studies focusing on humour in specific domains such as sitcoms (Z. Liu et al., 2022; Kayatani et al., 2021) and TED talks (Hasan et al. 2019), primarily in English.
This thesis aims to investigate two new settings. First, it will explore humour detection in the distinct domain of stand-up comedy, characterized by a direct interaction between the comedian and the audience, and a higher density of jokes compared to TED talks. Second, it will experiment with multilingual and multicultural settings (American and British English, Russian and Italian), to investigate potential differences in humour modalities across cultures. To achieve these goals, student will collect a new dataset based on the stand-up routines of different comedians, evaluate laughter detection algorithms (e.g. Gillick et al., 2021), and automatically annotate the dataset on humorous messages using laughter from the audience. Then, experiments with multimodal and multilingual models will be conducted, including ablation studies to investigate how different modalities contribute to humour detection in different languages. |
| Seznam odborné literatury |
| Aguert, M. (2022). Paraverbal Expression of Verbal Irony: Vocal Cues Matter and Facial Cues Even More. Journal of Nonverbal Behavior, 46, 1–26. https://doi.org/10.1007/s10919-021-00385-z
Gillick, J., Deng, W., Ryokai, K., & Bamman, D. (2021). Robust Laughter Detection in Noisy Environments. Interspeech 2021, 2481–2485. https://doi.org/10.21437/Interspeech.2021-353 Hasan, M. K., Rahman, W., Zadeh, A., Zhong, J., Tanveer, M. I., Morency, L.-P., Mohammed, & Hoque. (2019). UR-FUNNY: A Multimodal Language Dataset for Understanding Humor. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2046–2056. https://doi.org/10.18653/v1/D19-1211 Kayatani, Y., Yang, Z., Otani, M., Garcia, N., Chu, C., Nakashima, Y., & Takemura, H. (2021). The Laughing Machine: Predicting Humor in Video. 2021 IEEE Winter Conference on Applications of Computer Vision (WACV), 2072–2081. https://doi.org/10.1109/WACV48630.2021.00212 Liu, Z.-S., Courant, R., & Kalogeiton, V. (2022, December 4). FunnyNet: Audiovisual Learning of Funny Moments in Videos. 16th Asian Conference on Computer Vision (ACCV2022). https://hal.science/hal-03839553 Peyrard, M., Borges, B., Gligorić, K., & West, R. (2021). Laughing Heads: Can Transformers Detect What Makes a Sentence Funny? (arXiv:2105.09142). arXiv. https://doi.org/10.48550/arXiv.2105.09142 |