Important Dates
- Paper submission deadline: 1 April 2024
Publication: TBA
We are continuously witnessing the big data era. In research communities of big data management and analytics, the problem of how to deal with the growing amount of spatiotemporal data has been attracting much attention. Due to the proliferation of a variety of location-based social media, people find it increasingly challenging to pick out useful and interesting context-aware information from the overwhelming volumes of spatiotemporal data. As such, various applications aiming to enhance users’ experience have been developed. Among these applications, location-based recommender systems have been becoming an indispensable part of spatiotemporal data analytics. They help users find their preferred context-aware news articles, desired items to purchase, routes to travel, points of interest, potential nearby friends on social networks, etc.
In related research communities, it is of great importance to develop effective recommender systems that are capable of producing personalized high-quality and timely recommendation results based on spatiotemporal data. Although a host of recommender systems have been developed by existing studies, how to address the following three limitations regarding existing recommender systems remains an open problem to research communities. First, the result quality of existing recommender systems is vulnerable to rapidly changing spatiotemporal data inputs (e.g., spatiotemporal data streams with high arrival rates). Second, traditional recommendation algorithms are designed based on input data from a single source or a small number of similar sources. It is of great interest to take advantage of many-source heterogeneous spatiotemporal data to improve the quality of recommendations. Third, it is challenging to balance the trade-off between recommendation result quality and the efficiency of recommendations. It is important to develop recommendation algorithms under parallel and distributed environments. Such a requirement is particularly crucial in dealing with big spatiotemporal data.
This special issue will focus on developing high-performance recommender systems on the basis of spatiotemporal data with the ability to output personalized high-quality and context-aware recommendation results. For this purpose, we need to investigate the following challenges. (1) How to develop effective context-aware recommendation algorithms based on many-source heterogeneous spatiotemporal data streams? (2) How to develop high-performance recommendation algorithms by leveraging parallel and distributed data processing mechanisms? (3) How to balance the trade-off between recommendation result quality and the efficiency of recommendations? (4)In recent years, large-scale pre-trained models (e.g. ChatGPT, GPT4) have shown their great potential in various fields. Could it be applied to the field of recommendation to improve model performance?
To address the aforementioned challenges, this special issue targets to pioneer novel spatiotemporal recommendation algorithms, context-aware recommender system model training frameworks, parallel and distributed data processing algorithms, and heterogeneous streaming data analytics to improve existing location-based recommender systems regarding both effectiveness and efficiency.
The list of possible topics includes, but are not limited to:
- Big spatiotemporal data preprocessing for context-aware recommender systems, including data cleaning, feature selection and extraction, data clustering, etc
- Distributed and parallel data processing for location-based recommender systems
- Efficient trajectory and route recommender systems
- Deep learning/reinforcement learning/federated learning techniques for spatiotemporal data analytics
- Crowdsourcing techniques for recommender systems
- Big spatiotemporal data management and analytic platforms for recommender systems
- Measurements and characterization of innovative context-aware recommender-system applications
- Individual, group behavior analysis and social activity discovery
- Data-driven solutions for location-based recommender system
- Privacy preserving in spatiotemporal recommender systems
- Fairness, debias, robust, and trustworthy spatiotemporal recommender systems
- User modeling and user behavior studies in the context of spatiotemporal data
- Large pretrained model and its applications for location-based recommender systems.
Submission Guidelines
For author information and guidelines on submission criteria, please visit the TBD’s Author Information page. Please submit papers through the ScholarOne system, and be sure to select the special-issue name. Manuscripts should not be published or currently submitted for publication elsewhere. Please submit only full papers intended for review, not abstracts, to the ScholarOne portal.
Questions?
Contact the guest editors:
- Shuo Shang, Professor, UESTC, China
- Qi Liu, Professor, USTC, China
- Renhe Jiang, Assistant Professor, The University of Tokyo, Japan
- Ryosuke Shibasaki, Professor, The University of Tokyo, Japan
- Panos Kalnis, Professor, KAUST, Saudi Arabia
- Christian S. Jensen, Professor, Aalborg University, Denmark