Scite_ is an AI based citation/reference service that can assist in managing a large body of literature.
It allows for e.g. searching citation statements for facts and insights, asking questions to your reference list, check what others have published in support of a paper (or if any have reported lack of reproducibility), track trends and insights, check references, etc.
The DTU Library have secured a license and it can be used/access via the DTU network. It is not required to create a user to access the tool, but doing so allow additional features to be unlocked.
The library details how it works:
The tool uses artificial intelligence to ‘get behind the citations’ - and provide a qualitative (and quantitative) entry to how scholarly publications are cited, based on ‘supporting citations’, ‘mentioned citations’, ‘contracting citations’, as well as undefined, all showing how papers are used and in which sections of a paper the citations appear.
Scite.AI can create ‘custom dashboards’ both based on searches from Scite.ai, but also with results from searches done in Scopus, Web of Science or more. These dashboards can be ‘analysed using the features in Scite.ai’ for data extraction, and all can be exported as a table for later use.
SciteAI also incorporate ‘Editorial notifications’ like Errors, Retractions, Errata, etc. It is possible to use SciteAI as an LLM engine, actually searching in scholarly papers - with real references. So ‘ask your research question’ in natural language, and get a result with scientific papers you can trust. SciteAI collects their information/data from several agreements with publishing houses, as well as Open Repositories and sources like Semantic Scholar, OpenAlex and CrossRef.
You can read more about Scite_ and other tools at the library (DTU findit -> more tools).