Using AI Tools for Research | Built-in Database Tools

In this Guide, you'll find:

• Information on AI-powered tools to help with your research, including:
  • Built-in library database tools: JSTOR | EBSCO | Library Catalog
  • Tips for using generative AI in the research process

What to know:

• JSTOR's research tool helps you...
  • Find relevant results that may not be at the top of your search results
  • Decide whether articles are relevant to your research question
  • Explore related topics and articles
• You must be logged in to a JSTOR account to use it (registering is free and accounts aren't managed by UJ; we recommend using your UJ email but it isn't required)
• The tool only answers direct questions about scholarly content within JSTOR--it won't generate essays or provide you with sources outside of JSTOR

How to use it:

• Semantic Results: Enter search terms a search results page, toggle over to "Semantic results" to work with the top 25 articles and book chapters for your search
  • Ask the Research Tool questions about the set of results! Questions about the content of those 25 articles seem to work best, but always remember use the articles themselves to verify the tool's answers to your questions
  • Some limitations: The tool will only answer questions about that set of 25 articles; it won't suggest additional sources

• Article Analysis: When you select a result from your search results and view an article details page, the Research Tool appears on the right side of the screen
  • Note that some publishers have opted out, so it may not appear for all articles or book chapters
  • Use the suggested prompts or ask your own questions about the article
  • Click the three dots next to the chat box to cite or download your interaction with the Research Tool

What to know:

• EBSCO's AI options allow you to
  • Search using natural language, like Google does
  • View plain-language summaries of articles
• It's not a "conversational" AI tool like ChatGPT--it only affects search and provides article summaries

How to use it:

• Natural Language Searching: Toggle on "Natural language searching" from a new search or from your search results, then type in your research question or topic without reducing it to keywords
  • Librarian Amanda's tip: try the same search with and without natural language searching turned on, and try a keyword version of your natural language search with each option, too! You will get different results each time you search and find more possibly relevant articles

• AI Article Insights: In your search results, click on the "AI Insights" icon below a result
  • Coming soon!

What to know:

• The library catalog (or Primo) Research Assistant can help you
  • Search using natural language, like Google does
  • Identify relevant sources
  • Get a brief, summary-based answer to a research question
• Only uses article content that is available through our library catalog (and mostly scholarly article content)
• Automatically saves your questions and the responses the tool provides

How to use it:

• Click on "Research Assistant" in the menu options along the top of any catalog page and log in with your usual UJ email address and password. 

• Type in a research question and don't worry about reducing it to keywords - the Research Assistant will return 5 relevant results and a summary answer to your question based on those 5 results, plus an option to search for more articles from your library and related research questions down below the summary
• Some tips:
  • Questions that require evidence to answer (ie. empirical research questions) work best, so try to avoid questions with yes or no answers
  • You can include different limits in your question such as publication dates or article types
  • Right now, the tool doesn't support follow-up questions; you have to ask separate questions one at a time