Cluster Research Papers using API or some other way?

I want to do the following:

Two things I need your help with:

  1. I’d like to cluster research papers by topic.
  2. I’m also interested in determining the number and frequency of co-authors across the papers.

What do I have?

  • txt file with all the info from the research papers, title, journal, coauthors, date, etc
  • this txt file does not include any of the research papers or abstract or anything, just the title, and other information

How can I use the API to help me with this? The TXT file is about 4000 words, which is slightly close to 8191 token limit for GPT 4.

I would appreciate any thoughts or guidance on how to do this. Thank you!!!