Towards User-Centric Text-to-Text Generation: A Survey

Yang, Diyi; Flek, Lucie

doi:10.1007/978-3-030-83527-9_1

Cited by 8 publications

(5 citation statements)

References 105 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While our past as well as ongoing experiments, employing natural language inference, line masking, and character pseudonymization, have shown promising results, they only seem to constitute partial superficial remedies for a deep and complex issue. In the future, we intend to approach the problem by adapting and employing current NLG personalization techniques (Yang and Flek, 2021 Another serious problem, identified by the theatre professionals while working with our hierarchical setup, is the fact that the script generation often strays away from the synopsis. So far, we have been only operating with flat textual representations of script parts in the hierarchical setup, aligning parts of the script to parts of its synopsis.…”

Section: Discussion and Limitationsmentioning

confidence: 99%

“…13 In future work, we plan to incorporate natural language inference checks (Welleck et al, 2019) or experiment with dialogue act semantic representations (Kumar et al, 2018) in order to increase coherence. To improve character consistency, we plan to follow percharacter personalization approaches (Yang and Flek, 2021). AI Benjamin, Oscar Sharp, and Ross Goodwin.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Proceedings of the 4th Workshop of Narrative Understanding (WNU2022)

2022

View full text Add to dashboard Cite

Welcome to the 4th Workshop on Narrative Understanding! This is the 4th iteration of the workshop, which brings together an interdisciplinary group of researchers from AI, ML, NLP, Computer Vision and other related fields, as well as scholars from the humanities to discuss methods to improve automatic narrative understanding capabilities. We are happy to present 8 papers on this topic (along with 7 non-archival papers to be presented only at the workshop). These papers take on the complex challenges presented by diverse texts in areas of film, dialogue and literature as they look to improve methods for event extraction, gender and representation bias, controllable generation, quality assessment, and other tasks related to the workshop theme. We would like to thank everyone who submitted their work to this workshop and the program committee for their helpful feedback.

show abstract

Section: Discussion and Limitationsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Proceedings of the 4th Workshop of Narrative Understanding (WNU2022)

2022

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

GPT-2-based Human-in-the-loop Theatre Play Script Generation

Rosa¹,

Schmidtová²,

Dušek³

et al. 2022

Proceedings of the 4th Workshop of Narrative Understanding (WNU2022)

View full text Add to dashboard Cite

We experiment with adapting generative language models for the generation of long coherent narratives in the form of theatre plays. Since fully automatic generation of whole plays is not currently feasible, we created an interactive tool that allows a human user to steer the generation somewhat while minimizing intervention. We pursue two approaches to long-text generation: a flat generation with summarization of context, and a hierarchical text-to-text two-stage approach, where a synopsis is generated first and then used to condition generation of the final script. Our preliminary results and discussions with theatre professionals show improvements over vanilla language model generation, but also identify important limitations of our approach.

show abstract

“…In recent years, various generative models have emerged for tasks such as text-to-text [18] and text-to-image 3 generation, music composition based on prompts 4 , and text simplification [110]. Other models generate images [89], poetry [101],…”

Section: Generative Models and Personalizationmentioning

confidence: 99%

“…The same idea can also be used for personalization: the output of the system is conditioned by the information available about the user [110] and their context [34] to better match generated texts to a writer's expectations and needs. According to Wang et al [106], personalization can happen at two levels: it can include factual knowledge such as personal data, user attributes, and user preferences.…”

Section: Generative Models and Personalizationmentioning

confidence: 99%

The AI Ghostwriter Effect: Users Do Not Perceive Ownership of AI-Generated Text But Self-Declare as Authors

Draxler¹,

Werner²,

Lehmann³

et al. 2023

Preprint

View full text Add to dashboard Cite

Human-AI interaction in text production increases complexity in authorship. In two empirical studies (n1 = 30 & n2 = 96), we investigate authorship and ownership in human-AI collaboration for personalized language generation models. We show an AI Ghostwriter Effect:Users do not consider themselves the owners and authors of AI-generated text but refrain from publicly declaring AI authorship.The degree of personalization did not impact the AI Ghostwriter Effect, and control over the model increased participants' sense of ownership. We also found that the discrepancy between the sense of ownership and the authorship declaration is stronger in interactions with a human ghostwriter and that people use similar rationalizations for authorship in AI ghostwriters and human ghostwriters. We discuss how our findings relate to psychological ownership and human-AI interaction to lay the foundations for adapting authorship frameworks and user interfaces in AI in text-generation tasks.

show abstract

Towards User-Centric Text-to-Text Generation: A Survey

Cited by 8 publications

References 105 publications

Proceedings of the 4th Workshop of Narrative Understanding (WNU2022)

Proceedings of the 4th Workshop of Narrative Understanding (WNU2022)

GPT-2-based Human-in-the-loop Theatre Play Script Generation

The AI Ghostwriter Effect: Users Do Not Perceive Ownership of AI-Generated Text But Self-Declare as Authors

Contact Info

Product

Resources

About