When large language models meet personalization: perspectives of challenges and opportunities

World Wide Web (2024) 27:42 https://doi.org/10.1007/s11280-024-01276-1 When large language models meet personalization: perspectives of challenges and opportunities Jin Chen1 · Zheng Liu2 · Xu Huang3 · Chenwang Wu3 · Qi Liu3 · Gangwei Jiang3 · Yuanhao Pu3 · Yuxuan Lei3 · Xiaolong Chen3 · Xingmei Wang3 · Kai Zheng4 · Defu Lian3 · Enhong Chen3 Received: 22 November 2023 / Revised: 15 March 2024 / Accepted: 14 May 2024 © The Author(s) 2024 Abstract The advent of large language models marks a revolutionary breakthrough in artificial intelligence. With the unprecedented scale of training and model parameters, the capability of large language models has been dramatically improved, leading to human-like performances in understanding, language synthesizing, common-sense reasoning, etc. Such a major leap forward in general AI capacity will fundamentally change the pattern of how personalization is conducted. For one thing, it will reform the way of interaction between humans and personalization systems. Instead of being a passive medium of information filtering, like conventional recommender systems and search engines, large language models present the foundation for active user engagement. On top of such a new foundation, users’ requests can be proactively explored, and users’ required information can be delivered in a natural, interactable, and explainable way. For another thing, it will also considerably expand the scope of personalization, making it grow from the sole function of collecting personalized information to the compound function of providing personalized services. By leveraging large language models as a general-purpose interface, the personalization systems may compile user’s requests into plans, calls the functions of external tools (e.g., search engines, calculators, service APIs, etc.) to execute the plans, and integrate the tools’ outputs to complete the end-to-end personalization tasks. Today, large language models are still being rapidly developed, whereas the application in personalization is largely unexplored. Therefore, we consider it to be right the time to review the challenges in personalization and the opportunities to address them with large language models. In particular, we dedicate this perspective paper to the discussion of the following aspects: the development and challenges for the existing personalization system, the newly emerged capabilities of large language models, and the potential ways of making use of large language models for personalization. Keywords Large language models · Personalization systems · Recommender systems · Tool-learning · AIGC Jin Chen and Zheng Liu contributed equally to this work. Extended author information available on the last page of the article 0123456789().: V,-vol 123 42 Page 2 of 45 World Wide Web (2024) 27:42 1 Introduction The emergence of large language models [1], which have demonstrated remarkable progress in understanding human expression, is profoundly impacting the AI community. These models, equipped with vast amounts of data and large-scale neural networks, exhibit impressive capabilities in comprehending human language and generating text that closely resembles our own. Among these abilities are reasoning [2], few-shot learning [3], and the incorporation of extensive world knowledge within pre-trained models [1]. This marks a significant breakthrough in the field of artificial intelligence, leading to a revolution in our interactions with machines. Consequently, large language models have become indispensable across various applications, ranging from natural language processing and machine translation to creative content generation and chatbot development. The introduction of ChatGPT, in particular, has gained significant attention from the human community, prompting reflections on the transformative power of large language models and their potential to push the boundaries of what artificial intelligence (AI) can achieve. This disruptive technology holds the promise of transforming how we interact with and leverage AI in countless domains, opening up new possibilities and opportunities for innovation. As these language models continue to advance and evolve, they are likely to shape the future of artificial intelligence, empowering us to explore uncharted territories and unlock even greater potential in human-machine collaboration. Personalization, the art of tailoring experiences to individual preferences, stands as an essential and dynamic connection that bridges the gap between humans and machines. In today’s technologically driven world, personalization plays a pivotal role in enhancing user interactions and engagements with a diverse array of digital platforms and services. By adapting to individual preferences, personalization systems empower machines to cater to each user’s unique needs, leading to more efficient and enjoyable interactions. Moreover, personalization goes beyond mere content recommendations; it encompasses various facets of user experiences, encompassing user interfaces, communication styles, and more. As artificial intelligence continues to advance, personalization becomes increasingly sophisticated in handling large volumes of interactions and diverse user intents. This calls for the development of more advanced techniques to tackle complex scenarios and provide even more enjoyable and satisfying experiences. The pursuit of improved personalization is driven by the desire to better understand users and cater to their ever-evolving needs. As technology evolves, personalization systems will likely continue to evolve, ultimately creating a future where human-machine interactions are seamlessly integrated into every aspect of our lives, offering personalized and tailored experiences that enrich daily routines. Large language models, with their deep and broad capabilities, have the potential to revolutionize personalization systems, transforming the way humans interact and expanding the scope of personalization. The interaction between humans and machines can no longer be simply classified as active and passive, just like traditional search engines and recommendation systems. However, these large language models go beyond simple information filtering and they offer a diverse array of additional functionalities. Specifically, user intent will be actively and comprehensively explored, allowing for more direct and seamless communication between users and systems through natural language. Unlike traditional technologies that rely on abstract and less interpretable ID-based information representation, large language models enable a more profound understanding of users’ accurate demands and interests. This deeper comprehension paves the way for higher-quality personalized services, meeting users’ needs and preferences in a more refined and effective manner. Moreover, the integration of 123 World Wide Web (2024) 27:42 Page 3 of 45 42 various tools is greatly enhan (...truncated)