Web of Science Research Assistant: Functional analysis and usage recommendations

174 Web of Science Research Assistant: Functional analysis and usage recommendations Carlos Lopezosa Universitat de Barcelona, Spain https://orcid.org/0000-0001-8619-2194 Elisenda Aguilera-Cora Universitat Pompeu Fabra, Spain https://orcid.org/0000-0003-0923-9192 Lluís Codina Universitat Pompeu Fabra, Spain https://orcid.org/0000-0001-7020-1631 Juan-José Boté-Vericad Universitat de Barcelona, Spain https://orcid.org/0000-0001-9815-6190 Lopezosa, C., Aguilera-Cora, E., Codina, L., & Boté-Vericad, J. J. (2025). Web of Science Research Assistant: Functional analysis and usage recommendations. In J. Guallar, M. Vállez, & A. Ventura-Cisquella (Coords). Digital communication. Trends and good practices (pp. 174-189). Ediciones Profesionales de la Información. https://doi. org/10.3145/cuvicom.13.eng 175 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices Abstract This chapter provides a comprehensive functional analysis of the Web of Science Research Assistant, a generative AI tool integrated into Clarivate’s academic database. Designed to enhance research workflows, the tool supports tasks such as literature reviews, topic exploration, expert identification, and journal selection. Through a detailed examination of its interface and functionalities, including thematic summarization, co-citation mapping, and trend visualization, this chapter highlights the tool’s potential to streamline scientific discovery. Several practical use cases are presented, demonstrating its capabilities in optimizing searches, identifying seminal papers, and suggesting relevant publication venues. Despite its innovative features, the tool’s use must be guided by critical thinking, transparency, and ethical considerations. Researchers are encouraged to view the assistant not as a substitute, but as a complement to scholarly inquiry. The chapter concludes by emphasizing the value of reflective, layered interaction with AI to responsibly integrate these technologies into academic practice. Keywords Generative artificial intelligence; Academic search tools; Web of Science; Web of Science Research Assistant; Literature review automation; Research ethics and transparency. 1. Introduction Web of Science is one of the most important academic databases in the world, recognized for its multidisciplinary nature and the quality of its selection criteria. It is a database that covers everything from the experimental sciences to the humanities, including the social sciences. Web of Science covers documents published from 1900 to the present day in nearly 23,000 scientific journals, totalling around 180 million articles, as well as about 150,000 books (Clarivate, 2025a). At the same time as databases like Web of Science have grown in importance, so too have generative artificial intelligence tools, which have rapidly expanded and now permeate all areas of society — especially the scientific and academic spheres. This chapter introduces Web of Science Research Assistant, the artificial intelligence integrated into the Web of Science database. The following sections describe the tool’s interface and analyze its functionalities. As will be shown, Web of Science Research Assistant has the potential to transform access to, exploration of, analysis of, and synthesis of scientific information. However, its use must always be accompanied by critical thinking and an ethical perspective, which includes transparency. 2. What Is Web of Science AI Research Assistant and how does It work? Web of Science AI Research Assistant is a tool developed by Clarivate that uses generative AI models trained on the Web of Science Core Collection database to support researchers and students seeking to advance their academic work (Clarivate, 2025b). 176 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices This service goes beyond a traditional academic database, functioning instead as an assistant that can help (1) find seminal and relevant articles on a discipline or topic in seconds, (2) streamline advanced tasks such as literature reviews, expert identification, or journal selection for publication, (3) explore connections between academic concepts, authors, and scientific articles through interactive visualizations like trend maps and co-citation networks, and even (4) perform scientific document searches using natural language and multiple languages, thus facilitating access to scientific information without the need for advanced search operators. When a user submits a query, the tool retrieves the most relevant documents using semantic similarity algorithms and keyword searches. It then organizes the results based on relevance and generates responses or summaries using the content of the selected articles. Web of Science Research Assistant also allows both document-based searches and synthesis-type questions, tailoring the response format to the user’s needs. Like other tools such as Scopus AI (Aguilera-Cora et al., 2024a; 2024b), Elicit (Arroyo-Machado, 2024), Epsilon (Ren et al., 2025), Perplexity (Torres-Salinas & Arroyo-Machado, 2025), ChatGPT (Torres-Salinas et al., 2024; Boté-Vericad et al., 2024), Copilot (Lopezosa, 2023a; Boté-Vericad, 2024), or Scite (Codina, 2024), Web of Science Research Assistant reduces the manual effort required to construct complex searches and facilitates the identification of trends, research gaps, and non-obvious connections between scientific works. In the following section, we present a functional analysis of the tool to illustrate, with examples, the possibilities offered by Web of Science Research Assistant for research. 3. Practical use cases of Web of Science Research Assistant The Web of Science AI Research Assistant offers a variety of reports and visualizations designed to support academic research. These features allow users to efficiently analyze information and gain visual insights into relevant data and results. First, when accessing the main Web of Science platform (Figure 1), the user is presented with the primary interface, focused on its academic search function. At the top, there is a menu that provides access to features such as “Search” and “Research Assistant,” along with options to sign in or register. In our case, to activate the Web of Science AI service, we must select the “Research Assistant” tab. 177 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices Figure 1 Main Web of Science search interface with advanced options and research assistance. Once we have accessed the “Research Assistant” resource, a new interface appears in the form of an interactive chat (Figure 2). In the centre of the screen, the virtual assistant initiates the conversation with the message: “I am going to walk you through understanding a topic. What is your topic of interest or research question?,” inviting the user to enter their topic of interest or research question. Figure 2 Web of Science Research Assistant interface, guiding the user in understanding a topic based on an initial question. 178 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices On the left, there is a sidebar with the menu and the “Chat history” section, where the current conversation appears under the title “Understand a topic.” Navigation icons and system options are also visible. In the lower-left corner, there are links to send feedback, take a tour, or access information about the research assistant. Once the “Research Assistant” interface has been described, we test it with a prompt: “What is the relationship between SEO and the digital media?,” indicating that the topic to explore is the relationship between search engine optimization (SEO) and digital media. The result of this prompt (Figure 3) shows the response generated by the tool. At the top, there is a section titled “Overview,” where the main topics identified in the abstracts of the consulted articles are explained. In our case, these topics include search engine optimization (SEO) techniques, content and metadata features, and the impact of search engines on digital media. The overview also highlights the importance of SEO for improving web visibility, the role of content in search rankings, and how these dynamics affect journalism and academic research. Figure 3 Thematic summary generated by the Web of Science Assistant on the relationship between SEO and digital media, accompanied by key articles on optimization techniques and web visibility. 179 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices Below the summary, a list of fundamental articles (“Seminal papers”) related to the topic is provided. Among them are key works such as Brin and Page (1998), with more than 6,900 citations, and other studies on Google rankings, content characteristics, and techniques to increase online visibility (Figures 3 and 4). Each entry displays the article title, authors, source, publication date, number of citations, and a button to view more details. Additionally, below this list, there is a button to view more documents and an interactive section titled “What would you like to see next?,” offering complementary exploration options. These include analyzing the evolution of publications over time, viewing related concepts through a topic map, checking the most cited authors, and accessing new specific prompts that provide a guided and exploratory approach to the assistant. Figure 4 Second part of the thematic summary generated by the Web of Science Assistant on the relationship between SEO and digital media. In addition to these options, the Research Assistant suggests other prompts such as “Documents over time graph for search engine optimization” (Figure 5). This feature displays a graphical visualization. At the top, Clarivate’s chatbot notes that this type of chart allows users to observe publication trends on a specific topic over time, offering historical context, identifying emerging subtopics, and revealing shifts in academic interest. Just beside the explanation, users can adjust the chart parameters using a dropdown menu for the number of years (in this case, 25) and another for visualization settings. In the example shown, the bar chart illustrates the evolution of the number of documents published on “search engine optimization” from 1999 to 2024. A steady increase is observed starting in 2000, with a notable rise beginning around 2008. The peak in publications occurs in 2022 and 2023, with over 125 documents per year, with a slight decline in 2024. This type of analysis helps identify key moments of heightened research interest in SEO, possibly linked to technological advancements, changes in search engine algorithms, or the growth of digital communication. 180 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices Figure 5 Graph of publications on search engine optimization (SEO) over time, showing the sustained growth of academic interest from 1999 to 2024. Figure 6 Topic map on SEO visualizing related concepts and their interconnection, based on the co-occurrence of themes in the analyzed documents. 181 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices Additionally, we can create an interactive topic map. In the example presented, we once again use the topic of search engine optimization (Figure 6). At the top, it is explained that this type of visualization allows users to see the relationship between concepts within a research field, based on the documents covering those topics. The bubbles represent different themes, and their size indicates the number of documents related to each one, while the arrows show how the concepts reference one another. In our example, the central node of the map is search engine optimization, which is connected to terms such as data breach, collective intelligence, web scraping, knowledge integration, open source intelligence, and personal data. Other related and relevant topics are also displayed, such as cybercrime, general data protection regulation, automated decision making, fairness, social network analysis, security awareness, and intelligence analysis. Figure 7 Profiles of the six most cited authors in Web of Science related to the topic of search engine optimization (SEO). 182 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices This feature is particularly valuable, as it offers a form of representation that facilitates the exploration of subtopics, the identification of interdisciplinary connections, and the discovery of emerging trends in the field of SEO. Another valuable feature offered by the Research Assistant is the most cited authors panel (Figure 7). In our example, for the topic of search engine optimization, 84 relevant documents were identified, and the profiles of the six authors whose works have been most cited within that set are displayed. Each author is represented in an individual card that includes their name, institutional affiliation, Web of Science researcher ID, the most frequent topics in their publications (up to 10), and a button to view their full profile. This panel provides a clear overview of the leading researchers in the field and allows users to explore their contributions in greater depth. Figure 8 Thematic and bibliographic analysis on SEO and digital media, including a summary of key techniques, impact on journalism, and references to seminal articles. 189 Web of Science Research Assistent: Functional analysis and usage recommendations Carlos Lopezosa; Elisenda Aguilera-Cora; Lluís Codina; Juan-José Boté-Vericad Digital communication. Trends and good practices Boté-Vericad, J.-J, Fabeiro, R., & Anglada, L. (2024). Creando un chatbot con ChatGPT como soporte a la catalogación en bibliotecas, archivos y centros de documentación. Comparación de modelos de lenguaje en versión gratuita y premium. https://hdl.handle.net/2445/213600 Boté-Vericad, J.-J. (2024). Copilot Microsoft y ChatGPT: Análisis de datos, dispersión, relación y correlaciones de variables [vídeo]. https://hdl.handle.net/2445/213581 Clarivate. (2025a). Resources for Librarians and Administrators. https://clarivate.libguides. com/librarianresources/coverage Clarivate (2025b). Web of Science Research Assistant. https://clarivate.com/academia-government/scientific-and-academic-research/research-discovery-and-referencing/web-of-science/web-of-science-research-assistant Codina, L. (2024). Inteligencia artificial generative para trabajos académicos: método de análisis y comparación de Elicit, Epsilon, Perplexity, SciSpace y Scite. Lluís Codina. https:// www.lluiscodina.com/inteligencia-artificial-generativa-academia Codina, L. (2025). Interacción crítica escalonada con inteligencia artificial: explorando caminos para usar la IA en la academia. Lluís Codina. https://www.lluiscodina.com/interaccion-criticaia Font-Julián, C. I., Orduña-Malea, E., & Codina, L. (2024). ChatGPT Search as a tool for scholarly tasks: evolution or devolution?. Infonomy, 2(5). https://doi.org/10.3145/infonomy.24.059 Lopezosa, C. (2023a). Bing chat: hacia una nueva forma de entender las búsquedas. Anuario ThinkEPI, 17. https://doi.org/10.3145/thinkepi.2023.e17a04 Lopezosa, C. (2023b). La Inteligencia artificial generativa en la comunicación científica: retos y oportunidades. Revista de Investigación e Innovación en Ciencias de la Salud, 5(1), 1-5. https://doi.org/10.46634/riics.211 Orduña-Malea, E., & Cabezas-Clavijo, Á. (2023). ChatGPT and the potential growing of ghost bibliographic references. Scientometrics, 128(9), 5351-5355. https://doi.org/10.1007/s11192023-04804-4 Ren, Y., Mohammadi, L., & Codina, L. (2025). Aplicación de sistemas de IA en un trabajo de final de máster universitario: presentación de un caso. Barcelona: Universitat Pompeu Fabra. Departamento de Comunicación. http://hdl.handle.net/10230/70069 Torres-Salinas, D., & Arroyo Machado, W. (2025). Flujos de trabajo inteligentes con ChatGPT, Perplexity y NotebookLM. https://doi.org/10.5281/zenodo.14981646 Torres-Salinas, D., Thelwall, M., & Arroyo Machado, W. (2024). ChatGPT for Bibliometrics: Potential applications and limitations. https://hdl.handle.net/10481/92547