CARS: Workshop on Context-Aware Recommender Systems

Contextual information has long been recognized as a fundamental modeling dimension in various social science and technological disciplines and is becoming increasingly important for enhancing recommendation quality and retrieval performance. Over the years, several CARS workshops have explored how contextual information can be incorporated into traditional recommender systems.

While a substantial amount of research has been conducted, many existing approaches to context-aware recommender systems (CARS) still follow the representational view, where context is modeled using predefined and static factors such as time, location, or device.

However, recent advances in generative AI and foundation models are fundamentally reshaping how contextual information can be modeled and utilized. Large Language Models (LLMs) and multimodal foundation models enable the extraction of context directly from unstructured and heterogeneous signals, including text, images, interaction sequences, and environmental data. As a result, recommender systems are increasingly moving toward latent, inferred, and dynamic contextual representations, allowing for richer modeling of user intent and evolving preferences.

Recent developments in the field include generative context-aware recommender systems, intent-aware and sequence-aware models, latent context modeling, and multimodal context understanding. These approaches allow systems to capture both short- and long-term user behavior and adapt to complex, partially observable environments. At the same time, the growing deployment of recommender systems in mobile applications, IoT ecosystems, conversational systems, and multi-stakeholder platforms introduces new challenges. These include handling dynamic and uncertain contexts, while also addressing critical concerns such as privacy, security, fairness, and transparency. Emerging solutions such as federated learning, privacy-preserving modeling, and explainable AI are essential to ensure trustworthy and responsible recommendation systems.

The CARS 2026 workshop aims to rethink the role of contextual information in recommender systems in the era of generative AI and foundation models. The workshop will focus on how contextual signals and user intent can be inferred from multimodal data, how they can be integrated into next-generation recommender architectures, and how systems can remain adaptive, explainable, and privacy-aware. 

The workshop seeks to foster cross-disciplinary discussions, identify key research challenges, and advance innovative approaches for building effective and trustworthy context-aware recommender systems.

Topics of interest

We invite contributions to the workshop about topics related to CARS (but are not limited to):

• Context in generative recommender systems 
• Foundation Models and LLMs for contextual recommendation 
• Multimodal context understanding (text, image, speech, interaction signals) 
• Understanding user intent and dynamic preferences 
• Sequence-aware and time-aware recommender systems 
• Latent context models for recommender systems 
• Explainable context-aware recommender systems 
• Privacy and security modeling in CARS 
• Federated Learning for CARS 
• Ethics, fairness, and bias mitigation in CARS 
• Context in Augmented and Virtual Reality 
• Context processing in IoT environments 
• Human context recognition (e.g., health applications) 
• GUI and interaction strategies for increasing context awareness 
• Datasets and benchmarks for context-aware recommendation

Submission Types and Guidelines

CARS submissions should be prepared in PDF format according to the ACM single-column format (Microsoft Word or Latex formats). If you are using Overleaf, you can use the following code (\documentclass[manuscript]{acmart}). The peer-review process is single-blind and handled electronically through EasyChair. Accepted papers will be included in the workshop proceedings and at least one author of each accepted contribution must attend the workshop. Accepted papers are given an oral or a poster presentation slot at the workshop. 

The ideal length of a paper for the CARS workshop is between 4-8 pages (excluding references). Submitted work should be original. However, technical reports or ArXiv disclosure prior to or simultaneous with the workshop submission is allowed, provided they are not peer-reviewed. The organizers also encourage authors to make their code and datasets publicly available.