Discover the power of AI in our Library catalog
Primo Research Assistant (Intelligent Assistant). It's a generative AI-based tool designed to optimize time-consuming tasks. It allows users to query scholarly content in natural language and uses an extensive library to pinpoint five articles that can help answer the question posed. The tool provides the most relevant information from the abstracts of each article to build a description of the answer, including inline references to the sources to provide transparency into how each source contributed to the answer.
How does it work?
Primo Research Assistant uses a Retrieval Augmented Generation (RAG) architecture to combine the linguistic capabilities of the LLM (Large Language Model) with knowledge of the material indexed in the Central Discovery Index (CDI). Answers are generated as follows:
Query conversion: the user's query is sent to the LLM, where it is converted into a Boolean query containing a number of variations of the query, connected with an OR. If the query is not in English, some of the variations will be in the query language and the other variations will be in English.
Result retrieval: the Boolean query is sent to CDI to retrieve the results.
Reclassification: the top results (up to 30) are reclassified using embeddings to identify five sources that best address the user's query.
Overview creation: the top five results are sent to LLM with instructions to create the overview with inline references, based on the abstracts.
Response delivery: the overview and sources are returned to the user in the response.
Deploy research questions related to the topic; this can significantly help in the research process.
Examples:
As examples of how it works, it is suggested that they ask any of the following questions:
How does the sun influence landscape architecture?
What is magical realism?
How did Leonardo Da Vinci influence the Renaissance?
Result:
You can use this option provided by our discoverer to explore the topic of your interest and the benefits and potential of this tool.