SciSpace Academy
TutorialLiterature review › Step-by-step AI literature review

Complete Literature Review Workflow Using SciSpace

Dr. Joanne DonoghueEnglishLiterature reviewStep-by-step AI literature review

TL;DR

1-min read

A literature review asks for much more than a list of papers: you also have to read closely, compare studies and shape everything into coherent writing. This hands-on tutorial walks through a complete review workflow in SciSpace organised into four stages: discover, analyze, understand and synthesize. It is aimed at PhD scholars, master's students, faculty and research professionals at any point in a review, from early scoping to the final write-up, and attendees were able to try each step live on their own topics.

Key takeaways

  • Semantic, AI-powered search helps you surface studies that address your question even when they use different wording, so relevant work is less likely to slip past a strict keyword query.
  • Pairing semantic search with conventional database searching is a sensible way to explore a topic broadly while still checking that coverage is complete.
  • Side-by-side comparison tables let you line up the methods, results and limitations of many papers at once, which makes patterns and disagreements across the literature easier to see.
  • Chat with PDF can explain dense passages, equations and unfamiliar terminology, helping you understand a difficult paper before deciding how it fits your argument.
  • The synthesize stage turns your reading into structured notes, summaries and drafts that keep their citations attached, giving you material you can build directly into a thesis, dissertation or journal article.
  • Mapping the familiar review steps of searching, screening, extracting and comparing onto these four stages gives you a repeatable routine rather than an ad hoc process.
Apply it

To apply this, start your next review in SciSpace with a semantic search, move promising papers into a comparison table, and draft your synthesis from the resulting notes.

How to do it

  1. Discover papers with semantic search

    Start your literature review workflow by running a semantic AI search in SciSpace. It surfaces relevant papers beyond exact keyword matches, so you miss less important work.

  2. Compare papers in structured tables

    Add multiple papers to a table and pull out their methods, results and limitations. Side-by-side comparison makes patterns across studies easy to see.

  3. Chat with PDFs to understand them

    Open difficult papers with Chat with PDF and ask about complex sections, equations or technical jargon. This helps you read closely without getting stuck.

  4. Synthesise notes into drafts

    Turn your reading into structured notes and summaries, then shape them into citation-ready drafts. This connects analysis directly to writing.

Watch the tutorial37 min

Chapters

  1. Welcome: why use SciSpace instead of ChatGPT for literature searches
  2. How an AI literature review replaces one-paper-at-a-time PubMed searching
  3. Zotero vs Mendeley: which free reference manager to use
  4. How to set up Zotero: connector, collections and adding papers by DOI
  5. SciSpace browser extension: chat with any journal page and save to Zotero
  6. Standard vs High Quality vs Deep Review in SciSpace literature review
  7. What a SciSpace Deep Review report includes: methods table, gaps and PDF export
  8. High Quality mode and filtering for full-text and open-access papers
  9. Chat with PDF in SciSpace: explain text, podcast and save to Zotero
  10. How to sync your Zotero library with SciSpace
  11. Can you chat with multiple PDFs at once? Chat with files in SciSpace
  12. How to build a PubMed Boolean search string with an AI agent
  13. How to import multiple PubMed papers into Zotero at once
  14. Using SciSpace and Zotero inside Google Scholar
  15. Reproducible literature search in SciSpace across PubMed and Zotero
  16. How to find research gaps with AI using a custom prompt
  17. How to cite while you write: Zotero citations in Word and Google Docs
  18. How to use SciSpace inside ChatGPT to find research papers
  19. Recap: a literature review workflow that feeds back into SciSpace

Try this on SciSpace

Frequently asked questions

What are the main steps of a literature review?

Most reviews follow a similar path: define the question, search for relevant studies, screen and select them, read and extract key information, compare findings, and synthesise them into themes or arguments. This tutorial maps those steps onto four practical stages in SciSpace: discover, analyze, understand and synthesize, ending with notes and drafts that are ready for citation.

How does semantic search differ from keyword search for papers?

Keyword search matches the exact terms you type, so papers using different wording can be missed. Semantic search interprets the meaning of your question and returns studies that address the same concept even if they use other terminology. It is useful early in a review, though combining it with database searches helps ensure coverage.