AI applied to finding articles
Works a bit like ResearchScreener which is used at the end of the lit search process to seek relevant papers after a search has been done.
Both use ‘seed papers’ to help the AI determine other papers that are closely related. In ResearchRabbit’s case the linked citations are drawn from Microsoft Academia and papers must have a PMID or a DOI be found or linked. PMID = PubMed Identifier DOI = Digital Object Identifier.
Early days and validity and reliability would need to be checked. At minimum it would be a useful to double check nothing has been missed even if there is still a need for a replicable search strategy. Value will also depend on Microsoft Academic having sufficient coverage of the literature under review. It can only find from the papers harvested.
Use of such a search tool would need to be reported according to the new PRISMA-S standard, along with the more conventional search strategies.
Aaron Tay reports on developments with many of the open citation options linked to below in his blog Musings about Librarianship.
There are other options for tracking connections between papers or for identifying relevant search terms: