#gpt3chat
LightNews LightNews
v-i-o-l-a.openbiblio.social.ap.brid.gy
"Leveraging LLMs in Library Publishing: JATS XML Encoding with ChatGPT" https://doi.org/10.31274/jlsc.18048
#librarypublishing #jats #xml #gpt3chat
Leveraging LLMs in Library Publishing: JATS XML Encoding with ChatGPT
Introduction: Reliable and lightweight conversions of Microsoft Word documents to HTML have long eluded library publishers. We demonstrate how off-the-shelf large language models (LLMs) like ChatGPT offer a lean pathway forward for generating JATS XML, which current platforms are equipped to render into user-friendly HTML publications. Methods: With careful prompting, ChatGPT can turn a plain text typescript into valid JATS. Leveraging a one- and few-shot approach for the <front> part of an XML file ensures that boilerplate data included in example(s) prompts the LLM to populate the correct data in its output. In <body> and <back/references> parts, zero-shot prompts with only the name and version of our JATS specification produce valid XML in ChatGPT 4.0. Results: One- and few-shot prompting proved effective in directing ChatGPT 3.5 to consistently encode discrete, sequential sections of article typescripts. In retesting with ChatGPT 4.0, zero-shot approaches demonstrated that <body> and <back/references> parts need only the JATS specification name and version to convert typescript into valid XML. The <front> parts still benefit from a one- and few-shot approach. Discussion: The primary bottleneck is token or source size limitations. Content must be broken up into separate sections for input and the output manually “stitched” together to form a complete XML file. Conclusion: LLMs may offer a solution for publishers without the resources to encode JATS files by other means. As LLMs increase in scale, we expect workflows for encoding research articles in JATS to become even more accurate, with fewer restrictions on capacity.
www.iastatedigitalpress.com
January 17, 2025 at 12:49 PM
mariannabolog.bsky.social
Wait! I figured why #gpt3chat makes my skin crawl.
It reminds me of those students during exams who know that they dont know the answer to a question, but nevertheless keep talking authoritatively, trying to sound coherent. 🙄😂😤
November 21, 2024 at 8:56 PM
lamrongol.bsky.social
百週ぐらい回ってこんな創作ができる #gpt3chat はすごいのではないかという気がしてきた。
November 20, 2024 at 6:11 AM
mariannabolog.bsky.social
#ChatGPT texts screenshotted all around #Twitter these days.
I wonder whether anyone ran these texts into a #plagiarism detection software out of curiosity to see % of overlap with human-written texts on the internet.
#GPT3 #gpt3chat
November 21, 2024 at 9:02 PM
russwilcox.bsky.social
#gpt3chat learns from what humans write on the Internet, which means it is likely to be an expert resource on niche topics but often misinformed about topics of general interest.
November 17, 2024 at 1:14 AM