This repository stores the relevant PubMed search query (pubmed_query.txt), resource PMIDs (pmid-v3_2011.txt), and LLM prompts (topicModeling_LLM_assist.ipynb) mentioned in the submission of:
Shu Yang, PhD1,#, Elizabeth Mamourian, MS1,#, Lingyao Li, PhD2, Zixuan Wen, MA1, Tianqi Shang, MS1, Bojian Hou, PhD1, Jiayu Gao1, Yanbo Feng, MS1, Weiqing He, MS1, George Demiris, PhD1, Ryan Urbanowicz, PhD3, Li Shen, PhD1,† LLM-Enhanced Topic-Driven Literature Curation for Resource Discovery to Support Caregiving in Aging. †: Correspondance to li.shen@pennmedicine.upenn.edu.
In addition to the initial entries, the files here will be continuously updated so that more papers/resources related to caregiving in aging (esp., AI-focused) will be collected and updated on the website
As part of PennAITech’s efforts to identify AI and technology resources that support caregiving for the aging population, we developed an automated pipeline that integrates classic probabilistic topic modeling with LLM-based methods to surface relevant resources from 2,000+ PubMed entries published since 2011. We also built a website to host these resources and their topic hierarchy, marking the first systematic attempt to assist literature curation in this domain.
This work is supported in part by NIH P30 AG073105.