A new semantic frame forecast tool could help authors overcome writer's block by anticipating the future development of an ongoing novel-length story.
Pennsylvania State University (Penn State) researchers first define the story’s narrative into 1,000 or more text blocks or semantic frames, with each frame embodying clustered concepts and related knowledge.
Then a predictive algorithm analyzes the story so far, and predicts possible semantic frames occurring in the next 10, 100, or 1,000 sentences.
Penn State's Kenneth Huang said tests showed the technology is generalizable and applicable to novels and scientific articles, so "we could probably [use] it on news and on other genres."
Huang envisions the tool enhancing human creativity, suggesting "the machine's outputs could inspire something that the writer didn't think of before."
From Penn State News
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA