SciSpace Academy
Tutorial

Visualising Research with SciSpace Agent: Figures and Diagrams

SciSpace AcademyEnglishData analysis

TL;DR

1-min read

Complex research is often easier to grasp as a picture than as a paragraph. This tutorial shows how SciSpace Agent can turn papers, notes or early ideas into research-ready visuals, including figures, diagrams, workflows, conceptual frameworks and graphical-abstract-style summaries. It is intended for students, researchers and educators in any field who want to explain their work more clearly to reviewers, students or wider audiences. Participants practise each of these steps, so the tutorial moves from understanding the principles to producing visuals that match what the underlying research actually reports.

Key takeaways

  • The first step is converting research text into a visual outline that captures the key components, the flow between them and the labels each element needs.
  • Planning a figure's structure means deciding what to show, what to simplify and what to emphasise before any drawing begins.
  • For a graphical abstract, identify the single main finding and the few steps that lead to it, sketch a simple flow, keep text to short labels and check the target journal's size and format requirements.
  • A good caption states what the figure shows, defines abbreviations and symbols and explains any panels, so the figure can be understood on its own.
  • Alt-text should describe the key content and takeaway in plain language for readers who cannot see the image, and both caption and alt-text must reflect the data accurately rather than overstate it.
  • Prompting guidelines help you iterate on visuals faster while keeping the process responsible and transparent.
Apply it

To apply this, open SciSpace Agent with a paper or your notes, ask for a visual outline first, then refine the figure and draft its caption and alt-text.

How to do it

  1. Turn text into a visual outline

    Give SciSpace Agent your paper, notes or idea and convert the text into a visual outline with components, flow and labels. This is the foundation for any figure.

  2. Plan the figure structure

    Decide what to show, what to simplify and what to highlight before generating a figure. A clear plan keeps visuals focused and accurate.

  3. Create diagrams with SciSpace Agent

    Create research figures with AI by asking SciSpace Agent for diagrams, workflows, conceptual frameworks or graphical-abstract-style summaries. Pictures often explain complex research faster than paragraphs.

  4. Draft captions, legends and alt-text

    Write clear, accurate captions, legends and alt-text for each visual. These make figures understandable and accessible.

  5. Iterate prompts responsibly

    Refine your prompts to iterate faster, while staying responsible and transparent about AI use. Check every visual for accuracy before using it.

Try this on SciSpace

Frequently asked questions

How do I make a graphical abstract for my paper?

Start by identifying the single main finding and the few steps or components that lead to it. Sketch a simple left-to-right or top-to-bottom flow, keep text to short labels, and use consistent colours and icons. Check the target journal's size and format requirements, and make sure the visual matches what the paper actually reports.

What makes a good figure caption and alt-text?

A good caption states what the figure shows, defines abbreviations and symbols, and explains any panels so the figure can be understood on its own. Alt-text should describe the key content and takeaway for readers who cannot see the image, in plain language. Both must accurately reflect the data rather than overstate it.

Related tutorials