Event Website
https://docs.google.com/presentation/d/1rCMoc130VzmjTUvnaasJgOYVOWb5F-ZWWOYoUysMJio/edit?usp=sharing
Start Date
15-12-2020 3:05 PM
Description
Recent work in Natural Language Processing (NLP) has created machine learning models that can generate language very similarly to humans using deep learning models called Transformers. A Transformer-based model called GPT-2 may improve a creative writing tool that generates a continuation of a story for a writer. This proposal explores research to determine whether the application of GPT-2 along with suggestions from prior studies will improve coherence and creativity of suggestions in creative writing tools. To determine the success of this tool, I will collect user feedback to comparatively determine if this implementation improves upon previous systems, and where it still has room to improve.
Click below to download individual papers.
So what happens next? Improving Story Continuation in Creative Writing Tools
Recent work in Natural Language Processing (NLP) has created machine learning models that can generate language very similarly to humans using deep learning models called Transformers. A Transformer-based model called GPT-2 may improve a creative writing tool that generates a continuation of a story for a writer. This proposal explores research to determine whether the application of GPT-2 along with suggestions from prior studies will improve coherence and creativity of suggestions in creative writing tools. To determine the success of this tool, I will collect user feedback to comparatively determine if this implementation improves upon previous systems, and where it still has room to improve.
https://digitalcommons.lmu.edu/honors-research-and-exhibition/2020fall/section-03/4
Comments
Mentor: Mandy Korpusik