Citing/analyzing entire books with Deep Research

Prompting
1

In many of the social sciences, researchers frequently read and cite entire books instead of the journal articles that are so common in STEM research. This poses a bit of a problem when working with Deep Research. Even when giving specific links to book-length texts (like in this chat ChatGPT - Shakespeare AI Prophecy Theory ) or providing an attached full-length text (like in this chat ChatGPT - Confucian Microbiota Analysis ), the linked/attached book level material appears ignored by Deep Research in favor of its standard web crawl approach.

Does anyone have any suggestions, tips, or workflows for dealing with book-level analysis? TIA!

1 Like
2

Here are a few ideas:

  • Break the book into sections. Splitting it by chapters or themes can help the tool concentrate on a smaller amount of text at once.

  • Prepare summaries. Create brief outlines of each section and use these summaries along with your query.

  • Point out key passages. Highlight excerpts or specific pages that are important for your research.

  • Use iterative queries. Start by asking for an overview of a section, then follow up with more detailed questions.

  • Reference connections. Mention recurring topics or themes in your questions so the tool can link the ideas.

The above might help Deep Research to focus on the parts of the book that matter most for your analysis.

2 Likes