Skip to Main Content

Today's hours:

Main & Art Libraries 8:30am – 4:30pm

See all library hours »

  • Ask a Librarian
  • FAQ

Learn More About Generative AI

Resources and questions to consider related to using generative AI for research

Considerations Before Using AI

 

  • What does your professor say?
  • What are you trying to accomplish?
  • When/where in the process might you use an AI tool? Is it free or “freemium”?
  • What’s the right mix of tools (not just AI)? 

AI for Research Discovery

Selected AI research discovery tools

Tools

Semantic Scholar

Research Rabbit

Elicit 

Description

Search engine that indexes over 214 million papers, with filters such as journals and conferences, authors, publication types, and date range; suite of tools in beta.

Citation-based research literature and author mapping tool with over 100s of millions of academic articles. “Spotify for papers.”

Research discovery, summary, and data/ content extraction platform, searching across over 125 million papers.

Data Sources

Web indexing, content providers, publisher partners.

OpenAlex, Semantic Scholar, PubMed

Semantic Scholar

Business Model

Free

Free, with account 

Freemium, with account 

Owner

Ai2

Independently owned

Public benefit corporation

Similar Tools

OpenAlex

Inciteful.xyz

Consensus

Semantic Scholar

Can help you: 

  • Discover research by searching a large, multidisciplinary index; similar to Google Scholar but with better filtering and content transparency

  • Quickly grasp the overview of an article through AI-generated TLDR summaries 

  • Identify potentially relevant and important sources through filters and sort options

  • Understand how a reference has been used by other researchers with the article section filter

Pros:

  • Research dashboard similar to personal database accounts

  • Commitment to remaining free

  • Active development of ancillary tools including an article reader 

  • Easy to use, intuitive interface

Cons:

  • No advanced search options; not compatible with Boolean and wildcard strategies 

  • No natural language searching; searches by keyword only

When in the research process might you use Semantic Scholar? 

  • Similar to when a researcher might use Google Scholar at the beginning or end of the research process, or when struggling to locate research publications

  • To help inform your selection and use of sources

Research Rabbit 

Can help you: 

  • Discover new research related to a seed paper or small collection of publications 

  • Visualize and explore networks of references and cited by publications

  • Visualize and explore networks of authors 

  • Find key researchers within an area, their works, and their connections with other researchers

Pros:

  • The horizontally scrolling interface fosters backtracking as you explore networks

  • Provides paper and author recommendations

  • Can create shareable collections

  • Commitment to remaining free

  • Compatible with Zotero

Cons:

  • Works best when you start with a set of key papers or researchers (so should have this first) 

  • Has some issues with author name ambiguity or changes

  • Best to begin with a publication rather than a keyword search 

  • The metadata, including abstracts and author info, displays in a very messy way, so it can be difficult to see what an article is actually about at a glance

When in the research process might you use Research Rabbit? 

  • When starting from a seed paper or set of papers and would like to more explore across citation and author networks 

  • When you would like to map the scholarly conversation to see the “bigger picture”

  • When interested in finding key researchers

  • When a traditional literature search only yields a single or a few results 

Elicit

Can help you: 

  • Find research that might address your specific research question

  • Quickly review publications through AI-generated summaries that note relevancy to your research question

  • Compare and contrast publications discovered in Elicit, or your own collection of pdfs

  • Extract common elements of a paper, such as hypothesis, theoretical framework, variables, future research directions

  • Synthesize research findings

  • Identify key concepts within an area of research

Pros:

  • Natural language searching

  • Integration with Semantic Scholar 

  • Transparency around Elicit’s potential limitations

Cons:

  • Prompts require a research question versus a keyword or topic search 

  • Data/content extraction does not appear to be available for every publication 

  • Basic publication data (journal title, format, etc.)  is not consistently available

  • AI integration offers the opportunity to fail or provide misleading information more so than other tools

When in the research process might you use Elicit? 

  • Jumping off point in the beginning of your research process 

  • During the literature review process 

  • Towards the end of your literature search to compare and contrast publications or quickly extract data