Skip to Main Content

MultiSearch

Provides an overview on using MultiSearch

MultiSearch has a new Research Assistant

Research Assistant (Beta) is an AI-enabled tool in MultiSearch that allows you to search the Macquarie University Library catalogue using natural language queries. You can use Research Assistant to: 

  • Start your search and get a generated overview of a new topic, drawn from academic sources.
  • Discover five key sources on your topic and browse additional related academic content.
  • Do a search in a language other than English to help you understand a complex topic – Research Assistant supports multiple languages.

How to access Research Assistant

  1. Navigate to the MultiSearch page.
  2. Sign in with your OneID by clicking on Sign in on the toolbar at the top.
  3. Click on the Research Assistant icon on the toolbar at the top. 

Screenshot showing Sign in and Research Assistant options on MultiSearch toolbar

  1. You must be logged in to MultiSearch to use Research Assistant.    

  1. Ask Research Assistant a question as if you were talking to another person, you can use full sentences and, if you wish to, your first language.   

  1. Research Assistant will answer the question and provide:  

  • Citations for these top five sources with the option to view more results. 

  • An overview based on its top five ranked academic resources, from their abstracts, to answer the question, including in-text references.  

  • Suggestions for related research questions that can enable a deep dive into the topic. 

  1. The results will be drawn from a selection of academic resources available via MultiSearch. 

  1. You can see your previous search results – they are saved between sessions but can be deleted if you no longer require them. Research Assistant keeps a history of your last 200 searches 

Research Assistant is built based on a Retrieval Augmented Generation (RAG) architecture, uses a Large Language Model (LLM) to convert your questions into keyword searches and is grounded in the Ex Libris Central Discovery Index (CDI), which contains over 5 billion records. Find more information about how the tool works by visiting the Getting Started with Primo Research Assistant webpage.

As with all information searching, it's important to approach results critically.

Here are some things to consider when using Research Assistant:

  • The five top ranked results are determined to be relevant by Research Assistant, but may not be the most useful results for your topic.  
  • The summary responses are only based on abstracts and descriptions and may leave out important details or misinterpret the resource.
  • Research Assistant only searches selected online content providers and excludes print resources, so it will miss important sources for your topic.
  • Searches may return articles or resources that aren't available in our library and may recommend resources with full-text restrictions.
  • Each search is handled independently. Research Assistant cannot build on previous search queries like other AI tools you may use.

For more comprehensive and in-depth searches, use the advanced search capabiltites of MultiSearch and consult the Library's discipline based Subject and Research Guides.

Chat with a Librarian if you need a hand.

  1. Make sure you are allowed to use Research Assistant for your assignments – check with your lecturer or tutor about assessment requirements in relation to the use of generative artificial intelligence.

  1. Do not paste the overview results into your assignment – this will not help you do well / this does not develop your understanding of the topic and may be a breach of academic integrity. For more information on using generative AI ethically at MQ see AI Tools at Macquarie University and Generative AI in Research.

  1. Always verify the accuracy and relevance of information in/from the results and the citation details of the sources.

 

Further information

Visit the Getting Started with Primo Research Assistant page for more information.

Learn how the tool works with this short tutorial video.