Map the field, separate evidence from opinion, and make the gap unavoidable.
A literature review is not a list of papers. It is an argument about what the field already knows, where methods disagree, and why your question still matters.
gumu can keep source PDFs, notes, citation checks, and draft sections in one paper workspace.
Write the topic, population or system, method family, outcome, and time range. A clear question keeps the review from becoming an annotated bibliography.
Start with a rough idea and ask gumu to propose review scopes, inclusion criteria, and likely keyword families.
Begin with supervisor recommendations, recent survey papers, top venue papers, and highly cited works. Record why each seed is relevant.
Upload PDFs or paste BibTeX and ask for a source map by method, dataset, claim, and limitation.
Use backward references, forward citations, venue searches, and author clusters. Keep inclusion and exclusion decisions explicit.
Ask gumu to turn a seed list into search strings and a screening rubric you can apply consistently.
For each paper, capture question, method, data, result, limitation, and relevance to your gap. The matrix is the review's real backbone.
Use a literature-review helper to summarize each source into the same fields and flag missing metadata.
Organize papers by competing explanations, datasets, methods, threat models, or evaluation style. Chronology is useful only when the field's logic is historical.
Ask for thematic outlines that explain how clusters disagree and where your work fits.
End the review by naming the unresolved problem, why previous approaches do not solve it, and what kind of contribution would close it.
Run a gap-strengthening pass to remove vague phrases and connect the gap to your paper, thesis, or grant aim.
Start from a rough prompt, paper draft, PDF, review, call text, or supervisor note. gumu keeps the chat, source, references, and PDF regions together so every answer can become a concrete edit.
Enough to support the review question and venue expectations. A thesis chapter may need dozens to hundreds; a paper section may need a focused set of the most relevant work.
A literature review can be a standalone analytical chapter or study. Related work is usually shorter and positions one paper against prior work.
AI can help organize, summarize, and draft, but you still need to verify sources, decide inclusion criteria, and own the final argument.