6 Multimodal humor in human-robot interaction
-
Thomas Kiderle
Abstract
Over the last decade, humor has been of increasing interest in humanrobot interaction research. Most research focuses on entertainment, whereas the use of humor in human-robot dialog, such as small talk, is still scarce. Meanwhile, robots are envisioned to be part of our everyday work and domestic environments of the future. Thus, robot humor is a thankful opportunity for making everyday humanrobot interactions more pleasant and equipping robots with social intelligence. The process of creating and presenting robot humor encompasses different building blocks. First, the humor must be scripted in advance or generated dynamically based on the interaction context. Next, the presentation of humor with a social robot requires synchronized verbal and non-verbal behaviors, such as vocal gestures, adjusted prosody, facial expression, gaze and gestures. Since humor is a subject of taste, finally, the robot should also be able to consider the individual user’s preferences and personalize its presentation based on human feedback. This chapter provides a structured overview of the building blocks involved in creating and communicating robot humor. Since robot humor is inspired by human communication, we (1) give a short overview of the most commonly used humor types (canned humor, irony) and how they are communicated by humans. The humor signals reported in the literature are the foundation for a structured implemention of multimodal humor for social robots. Next, we (2) give an overview of state-of-the-art robots performing humor, including scenarios, contents and adaptation approaches. Afterwards, we (3) present selected computational approaches from the literature for generating verbal humor. They serve as basis for a (4) step by step guide for augmenting verbal humor with verbal and non-verbal humor signals, which is applicable to humanoid robots. Finally, we (5) illustrate challenges for robot humor using the example of domestic human-robot interaction. This includes recognizing the dialog context, appropriateness and predictability of the humorous effect, limitations resulting from robot embodiment, salience of different modalities and adaptation of humor to individual users.
Abstract
Over the last decade, humor has been of increasing interest in humanrobot interaction research. Most research focuses on entertainment, whereas the use of humor in human-robot dialog, such as small talk, is still scarce. Meanwhile, robots are envisioned to be part of our everyday work and domestic environments of the future. Thus, robot humor is a thankful opportunity for making everyday humanrobot interactions more pleasant and equipping robots with social intelligence. The process of creating and presenting robot humor encompasses different building blocks. First, the humor must be scripted in advance or generated dynamically based on the interaction context. Next, the presentation of humor with a social robot requires synchronized verbal and non-verbal behaviors, such as vocal gestures, adjusted prosody, facial expression, gaze and gestures. Since humor is a subject of taste, finally, the robot should also be able to consider the individual user’s preferences and personalize its presentation based on human feedback. This chapter provides a structured overview of the building blocks involved in creating and communicating robot humor. Since robot humor is inspired by human communication, we (1) give a short overview of the most commonly used humor types (canned humor, irony) and how they are communicated by humans. The humor signals reported in the literature are the foundation for a structured implemention of multimodal humor for social robots. Next, we (2) give an overview of state-of-the-art robots performing humor, including scenarios, contents and adaptation approaches. Afterwards, we (3) present selected computational approaches from the literature for generating verbal humor. They serve as basis for a (4) step by step guide for augmenting verbal humor with verbal and non-verbal humor signals, which is applicable to humanoid robots. Finally, we (5) illustrate challenges for robot humor using the example of domestic human-robot interaction. This includes recognizing the dialog context, appropriateness and predictability of the humorous effect, limitations resulting from robot embodiment, salience of different modalities and adaptation of humor to individual users.
Chapters in this book
- Frontmatter I
- Contents V
- Introduction 1
-
Part 1: Face-to-face interactions
- 1 A multimodal approach to children’s development of humor in family life 15
- 2 On target. On the role of eye-gaze during teases in face-to-face multiparty interaction 53
- 3 Humorous Smiling: A Reverse Cross-Validation of the Smiling Intensity Scale for the Identification of Conversational Humor 87
- 4 Alternative conceptualizations of the Smiling Intensity Scale (SIS) and their applications to the identification of humor 109
- 5 Facial gestures and laughter as a resource for negotiating humor in conversation 131
- 6 Multimodal humor in human-robot interaction 169
-
Part 2: Mediated interactions
- 7 Facial expressions as multimodal markers of humor: More evidence from scripted and non-scripted interactions 209
- 8 Emojis and jocular flattery in Chinese instant messaging interactions 231
- 9 More than laughter: Multimodal humour and the negotiation of ingroup identities in mobile instant messaging interactions 263
- 10 Humour and creativity in a family of strangers on Facebook 289
- 11 “Loanword translation and corrective acts are incongruous”: Debating metapragmatic stereotypes through humorous memes 319
- Index 355
Chapters in this book
- Frontmatter I
- Contents V
- Introduction 1
-
Part 1: Face-to-face interactions
- 1 A multimodal approach to children’s development of humor in family life 15
- 2 On target. On the role of eye-gaze during teases in face-to-face multiparty interaction 53
- 3 Humorous Smiling: A Reverse Cross-Validation of the Smiling Intensity Scale for the Identification of Conversational Humor 87
- 4 Alternative conceptualizations of the Smiling Intensity Scale (SIS) and their applications to the identification of humor 109
- 5 Facial gestures and laughter as a resource for negotiating humor in conversation 131
- 6 Multimodal humor in human-robot interaction 169
-
Part 2: Mediated interactions
- 7 Facial expressions as multimodal markers of humor: More evidence from scripted and non-scripted interactions 209
- 8 Emojis and jocular flattery in Chinese instant messaging interactions 231
- 9 More than laughter: Multimodal humour and the negotiation of ingroup identities in mobile instant messaging interactions 263
- 10 Humour and creativity in a family of strangers on Facebook 289
- 11 “Loanword translation and corrective acts are incongruous”: Debating metapragmatic stereotypes through humorous memes 319
- Index 355
