Artificial Intelligence is changing rapidly. The technology is simultaneously a consumer product to sell to us, to convince us to want to have for better or worse, as well as one to foist upon us whether we want it or not. It's likely moving to something we supposedly cannot do without. What now might be free, require a login or a personal subscription, might later be something libraries must purchase. That, or the AI technology will be quietly folded into our existing databases with or without fanfare. AI generated summaries already pop up in google searches, so why not. ProQuest, in February 2024, launched a beta test of an AI product. Web of Science and Scopus.ai, too. EBSCO has their Concept Map. And the people behind our system, the one we call OneSearch, now have one, which is there to use. Yes, it's a product they are imploring us to want.
Northern Michigan University's Center for Teaching and Learning has a webpage on Generative Artificial Intelligence (AI) at NMU, which goes beyond the scope of the resources listed here, although a few academic search engines are listed there.
These AI scholarly search tools remind me of the early days (late 1990s) of the Internet and the proliferation of search engines. Google, now the search engine, was one of many at the time vying for your business. As you look through these tools, you can't help but notice the hard sell from them, their self promotion. I can't emphasize that enough, they are marketing themselves as the one--the best ones will survive, certainly. Remember the "I got it off the Internet so it must be true?"--AI has the same air about it, so you'll want to be careful and look at the results.... If a total stranger tells you something with assured confidence, do you immediately believe them?
These AI search tools listed on this guide will probably come and go, I suspect (researchrabbit, an early player, seemed to have gone away but has now returned; Yewno is no longer). Some are free, some require you to register and have an account, and some you must pay for (what do the user agreements and privacy policies say?). I unabashedly found some of these resources from Aaron Tay's blog, in particular, two posts: List of academic search engines that use Large Language models for generative answers and List of Literature mapping tools. He is a librarian in Singapore and has delved into the nuances of these resources, into the back-end. I invite you to enjoy his commentary. I am just a user and a comparer of the results.
From the standpoint of library research, these resources again underline the "there is no one source". And some are very well done, providing a nice twist to research--I'll make another pass and pick out favorites. Some have deeper pools of works from which to draw; some will have more of works in your field. What personally identifiable information is taken from you when you sign up for them or use them? Does that matter to you?
Many or most of these are variations on a theme. They visually present direct or indirect connections from your resource with other resources, and speaking as a librarian, the results from these resources here come off as visual displays (greatly enhanced) of what a database provides you, but they come from or are enhanced by AI, which gives them an air of absolute authority if the hype about AI is to be believed. Some are overwhelming with the results. I have no favorite (yet). I invite you to systematically compare them yourself with a known paper or subject and draw your own conclusions. You will have your preferred ones tools here. Most, but not all, seem to draw from Semantic Scholar, which has partnered with many scholarly publishers and paper sources (should you create a free account there, here is their terms of service and here is their privacy policy). You do not need to have one, though.
CiteSpace
A visual analytic tool for analyzing trends and patterns in the scholarly literature of a field of research. You can select different data sources. There is a free version (recommended for new users) and an advanced, subscription version.
Connected Papers
Get a visual overview of a new academic field (enter a typical paper [keywords, title, DOI] and we'll build you a graph of similar papers in the field). Make sure you haven't missed an important paper (with Connected Papers you can just search and visually discover important recent papers). Use to create a bibliography for your thesis. Discover the most relevant prior and derivative works. Use their Prior Works view to find important ancestor works in your field of interest. Its source is the Semantic Scholar corpus.
Citrus-Search
Drop in a paper, search, add another paper if you wish, see what's related to either by content or citing network.
Inciteful
Over 240 million papers from which to draw, like the other resources on this page, similar topics and relevant literature about the subject.
PaperMap
Requires registration. Inputs papers from Semantic Scholar.
VOSViewer
Download (free) this software tool for constructing and visualizing bibliometric networks. These networks may for instance include journals, researchers, or individual publications, and they can be constructed based on citation, bibliographic coupling, co-citation, or co-authorship relations. VOSviewer also offers text mining functionality that can be used to construct and visualize co-occurrence networks of important terms extracted from a body of scientific literature. This is a different resource than appears on this page.
PaperFetcher
Requires you to stop and think about your search first. Input article or journal source, then your subject, and select output format. No visual results, just a bibliography of DOIs. The next step is up to you--find them, and most likely plug a resource into one of the others on this page.
CitationTree
Drop in the DOI of an article and see the citation connections. As this resource suggests, "Citationtree searches for papers in the citation environment of the input then displays only the most central papers...This is useful to have an overview of a new field, to spot important papers and accelerate your bibliography." Imports the bibliography into a citation manager such as zotero.
CitationChaser
Requires a token from lens.org (which sources PubMed, PubMed Central, CrossRef, Microsoft Academic Graph and CORE) for the results.
Open Knowledge Maps
Type in your concepts/subject and the most 100 relevant resources will be displayed.
PureSuggest
Add single or multiple DOIs and concepts and relevant/connected results will be mapped out.
LitMaps
Requires registration, which I haven't done yet. But reading their "About Us" echoes what all these other tools on this page do.
Citation Gecko
Frankly, this one is a little confusing to me. I am still trying to figure it out. Maybe I am missing something. My selected papers, used elsewhere in my testing these resources on this page, show nothing....
Local Citation Network
Visual citation network to aid in your literature review. Somewhat confusing--are these connections about the topic or more about the number of citations?
The resources below come mostly from Aaron Tay's blog, in particular, his list of academic search engines that use Large Language models for generative answers.
If you are unfamiliar with these, it might be a good idea to have a topic in mind that you are familiar enough with already to use as a benchmark for the results and are experienced enough with library databases to understand the searching, and the way the results are presented and discussed. In other words, if you are an experienced researcher, these tools are likely to help you a lot: If you haven't tinkered with a library database, these are likely to confuse you. I link to the policies and terms of these resources because whatever personal or proprietary information you share, these resources now have it.
Elicit.com
Reviewed here in the Journal of the Canadian Health Libraries Association. Free, but requires a login. Their Privacy Policy is here and their Terms of Service is here.
Consensus
Free, at first, but one must sign in and pony up for the AI aspects. The results are drawn from Semantic Scholar--nearly all the AI literature map tools draw from this, so your results aren't unlimited. In fact, you might already have found them elsewhere. Here is their Privacy Policy and their Terms of Service. Does anyone remember the old search engine AskJeeves, circa 1997 or 1998? I get the impression it wanted to be like Consensus.
scite.ai assistant
Privacy Policy, which pops right up, is here. 7 day free trial, otherwise. Prompting this with a question can all but write your paper, I have to say. The result is much better than ChatGPT if you are that kind of person--pretty sobering. You still have to do your own research. Their "assistant" here is really clever. Their sources come from many scholarly publishers.
Ask R Discovery
Draws from over 250 million papers across the scholarly spectrum. Privacy policy and terms of use. I ran a search, which was free initially, and at the bottom my my results: "Disclaimer: ‘Ask R Discovery’ aims to provide a snapshot of related research but is not a substitute for your own due diligence. Please exercise care when relying on these results, especially as this is an early version of the service. Your constructive feedback is crucial for its development."
LitMaps
Sourced from over 200 million articles. The more I look at these tools, the more I see they pull from the same places. Terms and conditions, privacy policy. Free searching? I was prompted to register for a free account.
SciSpace
Prices start at free to about 12.00 a month, which gives you more options. Similar to scite.ai above in that it can write you a nice sounding paper. Their privacy policy is here and their terms of use are here. SciSpace draws from a large pool of scholarly work.
Dimensions AI Assistant
Free registration. Privacy policy and website terms of use. About 370 million total records from which to draw from.
Lumina
Requires login to start. Their data sources are here.
Epsilon.ai
Free, at first, then you must subscribe. Terms of service and privacy policy. Searches Semantic Scholar.
ATLAS.ti
Free trial with login. Their splash page--you'll see why I believe these come off as a new consumer product wanting your business at the moment. Their policies are listed here.
ORKG Ask
Over 76 million sources. Maybe you can get it to work. Showed me nothing.
Wonders.ai
Free, but then you have to pay after 25 searches. Pulls from Semantic Scholar. Privacy policy, terms.
Undermind.ai
Requires you to have an account. Searches Semantic Scholar.
System Pro
Searches only the PubMed corpus, which is mostly biomedical and life sciences literature (but many other topics turn up there--you might be surprised. And AI might create additional associations). Terms of use and privacy policy.
The Literature
Another PubMed search tool. Terms of use.
Paper Digest
Offers a variety of choices (research copilot, literature review, conference paper summaries, etc). Sources come from a variety places (under Data & Compute Resource). Terms of use and privacy policy.
OpenRead
No idea the source. 300 million papers from somewhere.... Privacy policy and terms of use. You must sign in to see their AI enhancements, apparently.
ScienceOS.ai
Another with Semantic Scholar as a base. Free search, then you have to sign in to get the results. Terms of use and privacy policy.
Waldo
Requires sign in to use. Their privacy & sharing information. Scroll down to sourcing--Waldo sounds like a search engine. Nothing scholarly is mentioned.
Both APA and MLA, the two major styles used (based on experience) have webpages dedicated to citing AI. To reproduce them here is not necessary. They both cite ChatGPT as the sources, but the academic search engines here are not ChatGPT (even though their AI produces text for you after being prompted with a search).
The Chicago Manual of Style has this to say.