Make your literature search strategy visible
New AI Query Builder drafts an editable Boolean keyword query. You decide what the search requires before Scopus retrieves the records. This is just the first version. Stay tuned for further developments in advanced search queries.
Introducing AI Query Builder in Scopus
A battery scientist investigating polymer binders for silicon-based anodes must decide which studies belong in the search before comparing how those materials perform. Binder chemistry may define the topic; porosity, electrode loading and cycling conditions help determine whether individual studies can be compared.
AI Query Builder turns a natural-language question or topic into a proposed Boolean expression. Researchers can inspect and refine the strategy before retrieving papers, then save it to revisit as the literature develops.
Generate a query then run a document search
Sign in, open Documents and enable AI Query Builder if it is off. Enter the topic below, then select Generate query. Review which concepts the proposed expression requires, which terms it treats as alternatives and how those groups are combined.
Find studies of polymer binders for silicon-based lithium-ion battery anodes.
Enter the focused topic in Documents with AI Query Builder enabled. Select Generate query to propose a search expression.
That review sets the scope of retrieval. Requiring every experimental condition in titles, abstracts or keywords may exclude relevant studies whose methods are described elsewhere. Inspect those conditions in the retrieved records and papers you can access.
A language model helps propose the query. Selecting Search then runs a conventional document search. TITLE-ABS-KEY searches titles, abstracts and keywords, rather than full text. The returned publication records are not an AI-written answer.
The generated query remains editable. Review its complete terms and grouping before selecting Search.
Conventional document results after Search. Inspect abstracts, refine the strategy and assess which studies address the research question.
A literature search makes choices before the first result appears. AI Query Builder makes a proposed strategy visible and editable, so researchers can examine its concepts and logic before they search.Bissan Rafe-Duz
Senior Product Management, Scopus at Elsevier
Apply the method to your research question
For an antibody–drug conjugate researcher, the critical boundary may lie between linker stability in plasma and intracellular release. Reviewing the proposed query reveals whether the search reflects that distinction. The retrieved studies still need to be read for the relevant plasma species or cellular model, incubation conditions and the endpoint used to assess stability or release.
Force-feedback research exposes another part of the strategy: terminology. A device engineer needs to decide how haptics, instrument force sensing and tissue interaction relate to the task before accepting them as search alternatives. Bench and ex-vivo studies can then be screened for the setting and measurements relevant to the engineering question.
Inspect abstracts and refine the strategy in Advanced Search. Save the search or set an alert where available; rerunning the same logic can retrieve different records as the database changes.
Frequently asked questions
AI proposes the query. Search retrieves Scopus document records using the reviewed expression.
No. Start with a question or topic, then review and edit the proposed logic.
Signed-in users can save searches and set alerts where available. Results can change.
Semantic Search retrieves papers by meaning. Where enabled, its separate AI Summary can summarize up to 20 selected references using abstracts. AI Query Builder constructs search logic.
No. Researchers must assess strategy, coverage, screening and underlying studies.
Follow your account’s access and organization’s AI policy. Avoid confidential, personal or sensitive inputs.