Skip to content
Home » The Effortless Blog » SciSpace: An all-in-one AI tool for literature reviews

SciSpace: An all-in-one AI tool for literature reviews

Literature Review is a complex and tedious process. Luckily, modern tools make it much more manageable. Just like the internet search was a paradigm shift from a library, AI is a paradigm shift from a simple search. In this deep dive, we are looking at one of the most versatile tools that combines many features and saves you a lot of time: SciSpace.


Join the Effortless Newsletter. Receive free tips on note-taking, literature review, AI tools, and other productivity topics specifically tailored for academics and students.


The primary aim of SciSpace is to use AI to understand and draw connections between research papers, which is essential for anyone undertaking a literature review. Scispace can help you during every step, from finding papers relevant to your research question to helping you understand critical findings. With SciSpace, you can see the source of all information: all answers are cited with a quick link. Here are six use-cases of how this single tool can enhance your literature review:

  1. Searching literature from a question rather than keywords
  2. Scanning your existing collection of PDFs for specific criteria using AI
  3. Interacting with a single PDF to find relevant literature and explain concepts
  4. Using SciSpace as a browser extension to interact with websites
  5. Scientific Writing, paraphrasing and AI detection
  6. Conversation on SciSpace using ChatGPT Plus

Let’s get started:

1. Semantic Search for Literature: Starting From Scratch

SciSpace is an effective tool to search for literature. With SciSpace, you don’t need to rely on keyword searches, which can potentially cause you to miss out on finding related literature. Instead, you can search your papers using a question or phrase. This is called a semantic search. Internally, SciSpace utilizes AI first to understand what you are looking for and then cross-reference it with the content of the papers in its database (282M at the time of writing).

Your search result will be a paragraph of text answering your question. It will generally contain 5-10 citations and use academic jargon. Below, you will find the papers listed individually and can add columns that summarize a specific aspect of the study. In the example above, we are looking to summarize the model organism and location of the study for a set of ecological papers.

To refine your search you can then select the papers that seem relevant and click the “Show more like selected” button. SciSpace will learn a little bit better what you are looking for and add similar papers to the list of selected papers.

Save relevant papers to “Collections” for later analysis. A collection is essentially a folder containing a number of papers.

Can I use AI search for a systematic literature review?

Generally, no. A systematic literature review systematically reviews all literature that fulfills specific criteria (e.g., employs a particular methodology). This is done so others can reproduce your search (and findings). Using an AI algorithm, however, often leads to (1) slightly deviating results every time you run it and (2) may be incomplete. It is not reproducible, which makes your review not systematic. Therefore, do not use AI to search in the context of a systematic literature review.

But once you have found the papers using a reproducible keyword search in a database like Pubmed, you can use SciSpace to analyze these for criteria. This will save you a lot of time.

2. Analyzing your existing PDF collection using AI

When you already have a collection of PDFs, you can use SciSpace to search through the content of these PDFs. This is useful if you are dealing with more niche literature that might not be in their database or when you are working on a systematic literature review and selecting papers with a reproducible keyword search. The first step is to upload your papers:

After uploading the papers, they will be displayed as a table in your library. You can add columns to this table. Each column answers a question or summarizes an aspect of each paper, respectively. The most powerful feature is that you can create custom columns. For example, I have a giant collection on range shifts in climate change. It is generally assumed that the changing climate makes plants shift their locations/ranges. However, a growing body of literature suggests that competition between plant species also plays a significant role. To find all studies that deal with these so-called “biotic interactions,” I made a column titled “Relevance to biotic interactions.” Note how, in the screenshot below, some studies have an N/A in this column. This means the paper is irrelevant to this particular topic and we didn’t even have to open the PDF to find this out!

3. SciSpace CoPilot: Detailed insights for a single paper

CoPilot is a PDF reader feature of SciSpace. With Copilot, you can have an AI conversation with a paper. Upload a paper or select one from your library and start asking questions. You can even tell CoPilot to recommend questions, helping you think more like a scientist. Additionally, CoPilot speaks many languages, so while the paper may be in English, you can converse in any language you choose.

Some of the best use cases for using CoPilot are less summarizing the content and more utilization of it as an AI-powered search engine within the paper. For example, when you are writing a paper, you will make statements that need citations. To double-check your cite correctly, you can upload the cited PDF to SciSpace and ask, “Does this paper support the statement…”. You can find more information on this in the guide to academic writing with AI.

Conversational AI agents work best if you give them a single task at a time. To solve more complex tasks, SciSpace offers the possibility to ask follow-up questions. For example, ask it to identify the core problem in a paper and then the proposed solution.

4. CoPilot: as a browser extension

CoPilot also works as a Chrome extension, too. When browsing for papers, use this tool to quickly skim papers to decide whether the paper is worth correctly reading. Click on the CoPilot Chrome extension while you have a paper open and start asking the paper questions. Try questions like “Does this paper cover X?” or “Does this paper use X method?”. For a quicker understanding, try to “summarise the results.”

Another use case for the CoPilot extension is to translate and summarize insights. Some domains have a large body of work not translated into English. Some Wikipedia pages also vary significantly in detail depending on their language. The extension allows you to extract information from and to any language.

5. SciSpace for Academic Writing

The paraphraser tool, as the name suggests, can paraphrase text. The tone can be changed, which is helpful when converting highly technical information into more casual language or vice versa. Your text can then be shorter or longer while adjusting the magnitude of these changes.

Lastly, the paraphraser tool can also translate your texts into any language. This is useful if you write a “dirty” paragraph in your language and use AI to translate it to academic English.

SciSpace has an AI detector for academic writing as well. It is currently debated whether AI-generated texts can be detected. Still, if you have worked with ChatGPT for a while, you will notice that it tends to recreate predictable structures across its replies (e.g., “In summary…”). Here is an example of a text I wrote with a lot of assistance from AI (see this workflow), but never by just copying and pasting sentences. Instead, AI was used to inspire the writing structure and paraphrase some stubborn sentence parts. The result is pretty on point:

6. Using SciSpace within ChatGPT

ChatGPT is a potent and versatile tool that lacks precise understanding of scientific topics. Connecting SciSpace’s giant publication database with the conversational abilities of ChatGPT makes it a pretty powerful tool. This works through a custom GPT assistant. You can read what GPT assistants are and how to build your own in this tutorial.

The result is a table with citations to your question. Each entry in the table is a link to SciSpace, where you can use the CoPilot to read this article in-depth (See Section 3.)

There are currently 100s of GPTs available in ChatGPT; here are my 5 top AI assistant picks for academics.

7. Saving Notes to Notebooks

SciSpace has a promising new feature called notebooks it allows you to combine highlights in papers, conversations with the Copilot (step 3) and your own writing into a single document. Think of a notebook just as a single document where you can add content in a semi-automatic way. In itself, this is not remarkable, but SciSpace allows you to treat these notebooks just like uploaded PDFs. This means you can analyze notebooks (whether you uploaded them manually or collected in SciSpace) and bulk analyze them with AI (see step 2). This can be really handy if your notes form a big part of your knowledge base.

Pricing and 20-40% discount codes

You will get an incredible amount of work done for free with SciSpace. However, just like with ChatGPT, you will need a premium subscriptions to leverage AI fully. AI models are usually differentiated on their size. Larger models are exponentially more complex but allow for larger inputs and more sophisticated outputs.

A notable feature missing in the free version is the export to Zotero (via CSV or BibTeX). While you can import papers from your Zotero library, exporting is only available to premium users.

If you are considering getting SciSpace premium, you can use the codes Effortless20 and Effortless40 to get 20% off the monthly subscription or 40% off the yearly subscription, respectively.

Learn a complete AI literature Review workflow

SciSpace is a perfect first step in an in-depth literature review process. If you want to learn more about what comes next, check out this course:

Effortless AI literature Review preview image.

Leverage AI to find the most impactful literature quickly, cut your reading volume by 75%, and uncover hidden reference gaps combining various tools into a tested workflow.

Summary

SciSpace offers features that assist in all stages of the literature review process:

When looking for papers: use the literature review feature to find papers that match your question and add them to your library.

Once you’ve created your library: Use the AI columns to compare and contrast papers, while narrowing in on the key literature.

When the Key Literature is found: Use CoPilot to ask questions alongside your thorough reading and note-taking process.

Keep in mind that a semantic search is not a great idea for a systematic literature review as the search results will not always be reproducible.

SciSpace is entirely free. However, they do offer a premium subscription that includes a more advanced and unlimited use of the copilot feature. SciSpace is a versatile tool you don’t want to miss in your literature review tool belt.