The paper titled "DeHumor: Visual Analytics for Decomposing Humor" addresses the challenge of understanding and analyzing humor in public speaking. The authors highlight that humor is a critical communication skill that requires engaging content build-up and appropriate vocal delivery. Existing studies on computational humor focus mainly on textual and audio features near the punchline, neglecting the longer-term context setup. To bridge this gap, the authors introduce DeHumor, a visual analytical system designed to analyze humorous behaviors in public speaking. DeHumor decomposes each humorous video into multimodal features and provides inline annotations on the video script to reveal the building blocks of each example. In addition to the features introduced in previous theories of computational humor, DeHumor incorporates content repetition as a complement to better capture the build-ups. These features are then visualized in a context linking graph. To assist users in locating punchlines with desired features for learning purposes, DeHumor summarizes the content using keywords and provides humor feature statistics on an augmented time matrix. The effectiveness of DeHumor is demonstrated through case studies on stand-up comedy shows and TED talks, where it successfully highlights various building blocks of humor examples. Expert interviews with communication coaches and humor researchers validate its effectiveness for multimodal analysis of speech content and vocal delivery in terms of humor. This research contributes to a deeper understanding of concrete humor snippets by providing a comprehensive framework for analyzing humorous behaviors in public speaking through visual analytics.
Humor is when something is funny and makes people laugh. It's important to have good jokes and say them in a funny way. Some people study how to make jokes using computers, but they usually only focus on the part where the joke is told, not the whole story leading up to it. DeHumor is a computer program that can analyze funny videos and show how they are made. It breaks down the video into different parts and explains what makes each part funny. It also uses graphs to show how everything fits together. DeHumor can summarize the video using important words and show statistics about what makes it funny. It has been tested on comedy shows and speeches and experts say it works well for understanding humor."
Understanding Humor in Public Speaking with DeHumor
Humor is a powerful communication tool that can be used to engage an audience and make a lasting impression. However, understanding the building blocks of humor in public speaking has been challenging for researchers due to its complexity. In order to bridge this gap, researchers from the University of California, Berkeley have developed DeHumor – a visual analytical system designed to analyze humorous behaviors in public speaking.
The Challenge of Analyzing Humor
Previous studies on computational humor have mainly focused on textual and audio features near the punchline, neglecting the longer-term context setup that is essential for understanding humor. This makes it difficult to identify the key elements that contribute to successful comedic delivery. Additionally, existing systems are limited by their reliance on manual annotation or pre-defined templates which lack flexibility when analyzing different types of humorous content.
Introducing DeHumor
DeHumor addresses these challenges by decomposing each humorous video into multimodal features and providing inline annotations on the video script which reveal the building blocks of each example. In addition to traditional features such as vocal delivery and timing, DeHumor incorporates content repetition as a complement for better capturing build-ups leading up to punchlines. These features are then visualized in a context linking graph which allows users to easily identify patterns within videos and locate desired examples for learning purposes. Furthermore, DeHumor summarizes content using keywords and provides statistics on an augmented time matrix so users can quickly assess feature distributions across multiple videos at once.
Demonstrating Effectiveness Through Case Studies
To demonstrate its effectiveness, DeHumor was tested through case studies on stand-up comedy shows and TED talks where it successfully highlighted various building blocks of humor examples including vocal delivery techniques, wordplay jokes, puns etc.. Expert interviews with communication coaches and humor researchers further validated its effectiveness for multimodal analysis of speech content in terms of humor.
Conclusion
This research contributes significantly towards our understanding of concrete humor snippets by providing a comprehensive framework for analyzing humorous behaviors in public speaking through visual analytics tools like DeHumor . It also highlights how important it is to consider both long-term context setup as well as short term audio/visual cues when analyzing humorous behavior – something that has been neglected by previous studies but is essential for accurately assessing comedic performance .
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.
