YOUR GO-TO INTELLIGENCE RESOURCE FOR THE SCHOLARLY PUBLISHING INDUSTRY

Online research content and search technologies provider EBSCO Information Services (EBSCO) is embracing the potential of generative artificial intelligence (AI). Recognizing the transformative possibilities of generative AI in the academic research and library domain, EBSCO is proactively integrating AI into its product portfolio. The company is embarking on AI pilot projects within specific environments, with the aim of bolstering research effectiveness. This strategic move combines EBSCO's core strength — offering access to high-quality, pertinent content — with AI's remarkable capabilities, with the aspiration to redefine information discovery and comprehension for students, faculty, and researchers.

EBSCO has been investing in modernizing its user interface (UI) for research platforms, including EBSCO Discovery Service^™ (EDS) and EBSCOhost^®. The AI pilot projects will seamlessly integrate with this modern UI and encompass features like natural language search, improved search result assessment through summaries and key insights, interactive engagement with source documents (including multilingual support), search alerts, and content recommendations. All AI-generated insights will be clearly marked, and library administrators will have the option to disable the feature if desired.

Having incorporated AI operationally for several years, EBSCO is expanding its utilization to enhance metadata, categorize content, and establish linked data relationships. Linked data allows generative AI models to tap into an extensive network of interconnected information, encompassing vocabularies and knowledge graphs. This approach not only aids in managing computational costs and minimizing the environmental footprint of generative AI but also ensures that the adoption process is efficient for libraries.

While generative AI holds immense promise, it's important to acknowledge its susceptibility to potential misinformation and factual inaccuracies. Given EBSCO's commitment to serving researchers with credible information, the company is adopting a meticulous, quality-oriented approach. The goal is to seamlessly integrate AI capabilities into the framework of meticulously curated content. This approach not only safeguards against the spread of misinformation but also guarantees that EBSCO's AI technology provides users with highly relevant and accurate scholarly resources, elevating the reliability and precision of research across the academic landscape. Pilot projects will focus on identifying affordable, high-impact solutions for libraries.

Tim Lull, Senior Vice President of Discovery, User Experience, and Platform Services at EBSCO, emphasizes the company's commitment to safeguarding content creators' and publishers' rights while leveraging licensed and proprietary high-quality metadata and full-text resources. He underlines EBSCO's dedication to superior relevancy ranking and thesauri mapping. Lull acknowledges and seeks to address the valid concerns and potential challenges linked to generative AI, including its potential misuse by students to complete their work. The company is taking a measured approach to integrate AI into its products, mindful of the environmental implications and substantial investment costs associated with AI models. EBSCO's focused efforts revolve around developing innovative solutions that harness AI to enhance customer and user experiences, while actively mitigating undesired AI-related behaviors, reducing hallucinations (AI-generated inaccuracies and misinformation), and preserving research integrity.