Skip to Main Content

AI (Artificial Intelligence): AI for Literature searching

Responsible and Effective Use of Artificial Intelligence in Academic Research and Writing

AI in Academic Research

The literature review is an important part of any research. It gives a summary of research already published about a specific topic, identifying knowledge gaps, and areas for further research.

AI tools can complement academic databases to enhance the efficiency and depth of your literature search. It is important to check with your lecturer if the use of AI tools is permitted in your academic work and the Responsible and Ethical Use of AI and Citing AI Usage pages of this guide.

AI is not a substitute for your own critical analysis and thinking. Utilize AI as a tool to bolster, rather than replace, your research and preparatory work. This approach ensures that your unique insights and interpretations shine through clearly in your finished project.

Key points to remember

Brainstorming with AI:

While general AI tools can be incredibly helpful for brainstorming search terms or suggesting potentially relevant databases, students should approach them as starting points, not comprehensive solutions. AI can assist in clarifying concepts and expanding keyword ideas, but should not be used as a substitute for a thorough literature search. 

The importance of human expertise in database searching:

To obtain comprehensive and high-quality results, AI-generated suggestions should be refined and tested using the native search functionalities of trusted academic databases and guided by consultations with librarians [insert link to Library page post-grad student services], who can offer invaluable expertise in developing robust search strategies and navigating complex resources.

How is AI prompting different from keyword searching in a database?

While AI prompting and keyword/search string creation are related in the context of research, they serve different purposes and operate in different environments. 

Purpose:

Retrieval of the best academic sources from academic databases on your topic

How It works:
  • Uses Boolean logic (AND, OR, NOT)
  • Involves controlled vocabulary specific to the database
  • Requires precision to filter results effectively

Visit our How to Search guide for more detailed instructions.

Components:
  • Keywords: Core concepts of your research question
  • Synonyms/related terms: To broaden or narrow the search
  • Operators: AND (narrow), OR (broaden), NOT (exclude)
  • Field tags: e.g., title:, abstract:, author:

Example:

Topic: What are the effects of social media on mental health in teenagers?

Keywords: social media, mental health, teenagers

Search string: (social media OR Instagram OR TikTok) AND (mental health OR wellbeing OR depression) AND (teenagers OR adolescents OR youth)

Use in:

Academic databases (like ProQuest, Scopus, EBSCOhost, etc.).

Key skill:

Knowing your topic well enough to include synonyms, broader/narrower terms, and combine them correctly for the database.  

Limitation:

- Requires knowledge of database-specific syntax (Boolean, truncation, field tags).

- May miss relevant results if synonyms or controlled vocabulary are not used.

- Limited to what is indexed in the database.

 

While AI prompting and keyword/search string creation are related in the context of research, they serve different purposes and operate in different environments. 

AI prompt summary

Purpose:

Give an AI tool instructions to help you generate ideas or summaries for you.

How It Works:
  • Uses natural language
  • Can be open-ended or specific
  • AI interprets intent and generates a response based on patterns

Components:
  • Context: What are you researching
  • Task: What do you want the AI to do
  • Constraints: Optional limits (e.g., peer-reviewed only, last 5 years)

Example:

Topic: What are the effects of social media on mental health in teenagers?

Prompt: "What are the psychological effects of social media use on teenagers? Please include evidence-based insights, mention specific mental health, and refer to studies published in the last 5 years."

Use in:

AI Tools (ChatGPT, Copilot etc.)

Key skill:

Knowing how to write specific enough prompts for the AI tool. 

Learn how to write better prompts with the Prompt Engineering section of this guide.

Limitation:
  • May generate inaccurate or fabricated information (“hallucinations”).
  • Lacks access to subscription-based academic content.
  • May not understand nuanced academic tasks without clear prompting.

 

  • Use AI prompts to brainstorm alternative keywords and search strings for your topic.
  • Refine AI’s suggestions by manually testing them in your chosen database and then refine them until you find the best results.
  •  AI cannot replace the precision and reliability of structured database searching.
Key Differences between Keyword/Search String Creation and AI prompts

Keyword/Search String Creation

AI Prompting
  •  Databases are reliable but require technical skill and domain knowledge.
  • You must create the correct keywords to search databases
  • Use to retrieve relevant academic sources
  • Needs logical structure (Boolean, nesting, truncation)
  •  You can trust the results from the database to be true and accurate
  • You must tell AI what you want it to do
  • Use it to generate ideas or drafts
  • It needs clear instructions
  • AI is creative but unreliable for factual accuracy as it can hallucinate or make up sources and facts
  •  You must check the accuracy of the information
Example: ("remote work" OR telecommuting) AND ("employee wellbeing" OR "job satisfaction") Example: “Suggest search terms for remote work wellbeing”

 

How to acknowledge using AI prompts

Always check with your course outline or lecturer whether AI tools are permitted and to what degree they may be used. If AI use is permitted following the instructions on its use, document and acknowledge the type of AI tool used and how it contributed to the research and cite and reference the use.

Why should you acknowledge the use of AI in your research?

To uphold academic integrity and transparency it is very important to acknowledge when, and how, you have used AI tools in your research. This allows readers to understand the methodology you have used and upholds academic honesty.

Monash University's recommendations for how to acknowledge when and how you've used generated material as part of an assignment or project is very helpful. It suggests that your acknowledgement includes:

  • That you have used generative AI
  • Specify the AI tool used
  • Detailed descriptions of how the information was generated
  • Identify the prompts used
  • Explain how you used the outputs was used in your own work

 

An example of an appropriate acknowledgement format they suggest is as follows:

I used [ADD AI tool (ADD link if needed)] to [ADD how used] and [ADD number of iterations/drafts]. I modified the outputs in [ADD ways].

For detailed examples please visit their Acknowledging the use of generative artificial intelligence - Student Academic Success

If you used a chat tool you can also use the chat session transcripts to document AI use.