Review of Article Galaxy Scholar

Abstract

really necessary? I believe that scite offers some compelling features that can make it competitive and attractive for both libraries and end users. The browser extension, Zotero plugin, and API make it easier than some tools to integrate into the research environment. It is highly customizable for individual users (some functionality requires a personal account, even with an institutional subscription), and accessible end-user pricing puts the tool in the reach of users for whom Web of Science or Scopus is not a viable option. It is intuitive and easy to use, where Web of Science is often complex and fussy. And its open (unpaywalled) interface makes it a prime candidate to be widely integrated into the online research landscape in way that other tools might never be. I also find scite promising in terms of being able to deliver on its chief claim: that it can save researchers time by helping them summarize the literature and quickly identify publications both valuable and discredited for their research. It is possible that Clarivate sees this as well, because, in the spring of 2022, Web of Science added a feature similar to scite’s “smart citations” to its interface (though it is not yet available for every item). Called “Citing Items Breakdown,” it classifies citations as background, basis, support, differ, and discuss and will show snippets from the paper that include the text of the citation. Some might worry that a tool like scite could oversimplify the research process, exacerbate blind spots and encourage students to use shortcuts to doing their own research. While all of this may be true, there are any number of other tools or practices that can contribute to these outcomes already. Textbooks summarize previously conducted research and present it in a form most likely to be digestible to a particular audience. Discovery layers collapse indexing and deliver broader search results with less precision. Faculty at my institution ranked Google Scholar, the ultimate research shortcut, as the most-popular place to start their research in a recent survey here. In discussing their citation classification model, scite’s creators emphasize that “scite supplements machine learning with human interventions.” This is a metaphor for what is necessary for conducting quality research in order to produce new knowledge, no matter what tool is being used. I think scite offers attractive features for students and researchers and can augment the discovery experience for academic libraries. I have already heard of one instance where a user asked for it by name at our reference desk; I would not be surprised to hear of this happening more in the future.