Call for Papers
Sixth Knowledge-aware and Conversational Recommender Systems Workshop (KaRS 2024)
In conjunction with the ACM International Conference on Recommender Systems (RecSys 2024)
October 14-18, 2024
Bari, Italy
https://kars-workshop.github.io/2024/
Submission deadline: August 30, 2024
We are pleased to invite you to contribute to the Sixth Knowledge-aware and Conversational Recommender Systems Workshop held in conjunction with the ACM International Conference on Recommender Systems (RecSys 2024), Bari, Italy, from October the 14th to October the 18th, 2024.
SCOPE
Recommender systems have achieved ubiquity across various domains, ranging from e-commerce to media content suggestions, playing pivotal roles in facilitating user online experiences. However, despite their prevalence, these systems often encounter challenges in effectively engaging with human users. While data-driven algorithms have demonstrated success in uncovering latent connections within user-item interactions, they frequently overlook the central actor in this loop: the end-user.
A prevalent behavior among human users, which is rarely encoded in recommendation engines, is the utilization of domain-specific knowledge. Fortunately, Knowledge-aware Recommender Systems are garnering increasing attention within the recommendation community. By leveraging explicit domain knowledge represented via ontologies or knowledge graphs, these approaches can understand the semantic relationships between users, items, and other entities, thus offering tailored recommendations to users and addressing inherent limitations of purely data-driven systems. Despite their existence for over two decades, their significance has been revitalized due to the Linked Open Data initiative and the availability of large knowledge-graphs such as DBpedia and Wikidata. Linked data and their ontologies underpin many recommendation approaches and challenges proposed in recent years, such as Knowledge Graph embeddings, hybrid recommendation, link prediction, knowledge transfer, interpretable recommendation, and user modeling. Moreover, a new wave in this domain is marked by the emergence of neuro-symbolic systems, integrating data-driven methodologies with symbolic reasoning. This fusion of machine learning systems, adept at harnessing data, with symbolic approaches, adept at leveraging knowledge, holds promise in enhancing recommendation quality, particularly in scenarios with sparse training data.
In parallel, the rise of Conversational Recommender Systems (CRSs) highlights the crucial role of content features in facilitating user interactions, particularly in multi-turn dialogues between users and the system, which bring about novel challenges, such as the incorporation of both short- and long-term preferences, prompt adaptation to user feedback, the limited availability of datasets, and the evaluation beyond simple accuracy metrics. In this context, the emergence of Large Language Models (LLMs) is pivotal and has breathed new life into CRSs. LLMs significantly impact CRSs by leveraging their advanced capabilities in understanding user queries and generating relevant recommendations in natural language. They excel in processing complex and nuanced user inputs, allowing for more seamless and engaging interactions between users and the system. Moreover, LLMs contribute to the adaptability and responsiveness of CRSs by continuously learning from user interactions, thus refining their recommendations and improving the user experience over time.
The Sixth Knowledge-aware and Conversational Recommender Systems (KaRS) Workshop aims to spark a new generation of research that prioritizes user experience, engagement, and satisfaction over mere accuracy. Drawing upon a multifaceted set of expertise including Machine Learning, Deep Learning, Human-Computer Interaction, Information Retrieval, and Information Systems, this workshop endeavors to catalyze a fresh wave of research.
KaRS serves as a dynamic forum where researchers and practitioners, scholars and industry professionals, converge to not only disseminate their latest findings but also to identify the emerging topics, delineate the next key challenges, and forecast the future opportunities for research and development. Through fostering active participation and facilitating the exchange of ideas, KaRS aims to cultivate an interdisciplinary community that collaborates on the topics of knowledge-aware and conversational recommenders, alongside the emerging topics of LLMs and neuro-symbolic methodologies, which further extend the scope of this new edition of KaRS.
TOPICS
This workshop aims at establishing an interdisciplinary community with a focus on the exploitation of (semi-)structured knowledge and conversational approaches for recommender systems and promoting collaboration opportunities between researchers and practitioners.
Topics of interest include, but are not limited to:
- KNOWLEDGE-AWARE RECOMMENDER SYSTEMS.
- Models and Feature Engineering:
- Knowledge-aware data models based on structured knowledge sources (e.g., Linked Open Data, BabelNet, Wikidata, etc.)
- Semantics-aware approaches exploiting the analysis of textual sources (e.g., Wikipedia, Social Web, etc.)
- Knowledge-aware user modeling
- Methodological aspects (evaluation protocols, metrics, and data sets)
- Logic-based modeling of a recommendation process
- Knowledge Representation and Automated Reasoning for recommendation engines
- Deep learning methods to model semantic features
- Large language models (LLMs) for Knowledge-aware Recommender Systems
- Beyond-Accuracy Recommendation Quality:
- Using knowledge bases and knowledge graphs to increase recommendation quality(e.g., in terms of novelty, diversity, serendipity, or explainability)
- Explainable Recommender Systems
- Knowledge-aware explanations to recommendations (compliant with the General Data Protection Regulation)
- Online Studies:
- Using knowledge sources for cross-lingual recommendations
- Applications of knowledge-aware recommenders (e.g., music or news recommendation, off-mainstream application areas)
- User studies (e.g., on the user’s perception of knowledge-based recommendations), field studies, in-depth experimental offline evaluations
- Models and Feature Engineering:
- CONVERSATIONAL RECOMMENDER SYSTEMS.
- Design of a Conversational Agent:
- Design and implementation methodologies
- Dialogue management (end-to-end, dialog-state-tracker models)
- UX design
- Dialog protocols design
- Large language models (LLMs) for Conversational Recommender Systems
- User Modeling and interfaces:
- Critiquing and user feedback exploitation
- Short- and Long-term user profiling and modeling
- Preference elicitation
- Natural language-, multi-modal-, and voice-based interfaces
- Next-question problem
- Methodological and Theoretical aspects:
- Evaluation and metrics
- Datasets
- Theoretical aspects of conversational recommender systems
- Design of a Conversational Agent:
SUBMISSIONS
We invite three kinds of submissions, which address novel issues in Knowledge-aware and Conversational Recommender Systems:
- LONG PAPERS should report on substantial contributions of lasting value. The Long papers must have a length of a minimum of 6 and a maximum of 8 pages (plus an unlimited number of pages for references).
- SHORT/DEMO PAPERS typically discuss exciting new work that is not yet mature enough for a long paper. In particular, novel but significant proposals will be considered for acceptance in this category despite not having gone through sufficient experimental validation or lacking a strong theoretical foundation. Applications of recommender systems to novel areas are especially welcome. The Short/Demo papers must have a length of a minimum of 3 and a maximum of 5 pages (plus an unlimited number of pages for references).
- POSITION/DISCUSSION PAPERS describe novel and innovative ideas. Position papers may also comprise an analysis of currently unsolved problems, or review these problems from a new perspective, in order to contribute to a better understanding of these problems in the research community. We expect that such papers will guide future research by highlighting critical assumptions, motivating the difficulty of a certain problem, or explaining why current techniques are not sufficient, possibly corroborated by quantitative and qualitative arguments. The Position/Discussion papers must have a length of a minimum of 2 and a maximum of 3 pages (plus an unlimited number of pages for references).
Papers may range from theoretical works to system descriptions. We particularly encourage Ph.D. students or Early-Stage Researchers to submit their research. We also welcome contributions from the industry and papers describing ongoing funded projects which may result useful to the Knowledge-aware and Conversational Recommender Systems community. Submission will be peer-reviewed and accepted papers will appear in the workshop proceedings (CEUR workshop series). The review process is single-blind. Submitted papers will be evaluated according to their originality, technical content, style, clarity, and relevance to the workshop.
Long and short/demo paper submissions must be original work and may not be under submission to another venue at the time of review. Each accepted long or short paper will be included in the CEUR online Workshop proceedings and presented in a plenary session as part of the Workshop program. Original position/discussion accepted papers will be included in the CEUR online Workshop proceedings. Selected position/discussion papers will be invited as oral presentations.
Submissions of full research papers must be in English, in PDF format in the CEUR-WS two-column conference format available as compressed archive or as Overleaf template. Papers must be submitted through EasyChair by selecting the track “KaRS: Sixth Knowledge-aware and Conversational Recommender Systems Workshop”.
IMPORTANT DATES
- Paper submissions deadline: August 30th, 2024
- Paper acceptance notification: September 16th, 2024
- Camera-ready deadline: September 30th, 2024
- Workshop day: October 14th-18th, 2024
Deadlines refer to 23:59 (11:59 pm) in the AoE (Anywhere on Earth) time zone.
Leave a Reply