Requests for materials of a particular type (for example, "peer reviewed articles") or from a certain time or period (for example, "the newest research") are not currently supported.
Follow-up questions are not currently supported. Each question stands by itself.
Currently, there are no restrictions on the number of searches.
"Primo Research Assistant is a generative AI-powered tool designed to streamline time-intensive tasks. It enables users to query academic content in natural language and utilizes the breadth of your library to pinpoint five articles that can aid in answering your question" (Getting Started with Primo Research Assistant). The first five results are analyzed by a Large Language Model (LLM) which crafts an overview of the results with in-line citations. Clicking ‘view all results’ will take you to a full list of search results and the chance to filter or refine the results.
To make the most of Primo Research Assistant, it is essential that you ask clear and detailed questions about your topic. Be as specific as possible and phrase your query in the form of a question/request.
The following example questions can be found under the search box on Primo Research Assistant's Start Page:
For a more detailed tutorial visit: Primo Research Assistant Tutorial.
Traditional and AI-enabled approaches to literature reviews each have common pitfalls:
Traditional methods rely heavily on keywords. This makes it harder to capture flexible vocabulary, work across disciplines, or incorporate metadata effectively.
AI-enabled methods can capture related terms and synonyms more easily with vector search, but they introduce the problem of hallucination, which students may overlook without careful evaluation. They can also be limiting in output scope (e.g., only retrieving five results in Primo), and some users stop searching after a single trial prompt.
Neither approach guarantees complete or perfectly accurate outcomes-- important studies can still be missed. The best practice is to use both approaches together while staying critical of the results.
|
