About this Research Topic
The rise of the Internet of Things (IoT) and Artificial Intelligence (AI) leads to the emergence of edge computing systems that push the training and deployment of AI to the edge of the network, so as to achieve reduced bandwidth cost, improved responsiveness, and better privacy protection. Motivated by such trends, this Research Topic focuses on a new paradigm, the Social Edge Computing (SEC), that empowers applications at the edge by revolutionizing the computing, intelligence, and learning of the human-centric systems. The integration of edge computing, humans, and AI in SEC allows machines and humans to make collaborative and optimized decisions in edge-based intelligent applications. The SEC paradigm generalizes the current machine-to-machine interactions in edge computing (e.g., mobile edge computing literature), and machine-to-AI interactions (e.g., edge intelligence literature) into a holistic human-machine-AI ecosystem.
The SEC paradigm presents a set of interesting research problems. Examples of such problems include the rational nature of device owners, pronounced heterogeneity of edge devices, real-time and efficient AI at the edge, generative AI and large language models (LLM) at the edge, trustworthy AI and responsible machine learning at the edge, human-like AI agent, and human-AI interaction in edge computing, and the privacy/security concerns and fairness issues of the edge users. By investigating these research questions, SEC targets to provide a set of human-centered AI solutions that jointly explores the collective intelligence of humans, machines, and AI in a resource-constrained computing environment. SEC also motivates novel AI for social good applications at the edge, such as personalized health monitoring, automatic disaster damage assessment, real-time language translation, crowd abnormal event detection, smart home automation, and vehicle-based criminal tracking.
The SEC paradigm’s specific connections to the Data Science section in Frontiers in Big Data address some of the novel human-centric data science questions at the edge of the network. Examples of such questions include: 1) how to efficiently acquire, clean, process, integrate and analyze human-centric data generated at the edge (e.g., video, audio, image, and text data generated by humans or smart devices on their behalf)? 2) How can novel distributed data-driven models and systems that can effectively explore the tradeoff between application performance (e.g., accuracy, delay) and limited resources (e.g., bandwidth, storage, processing capability) at the edge of the network be developed? 3) How can human-centric issues (e.g., privacy, fairness, trustworthiness) for data-driven applications at the edge be addressed?
This Research Topic looks for articles that have not been published in the prior literature or are under review/consideration of another publication venue. If the submission is an extension of previously published conference or workshop papers, it needs to have at least 30% new content and the difference between the submission and the previously published version needs to be explicitly cited in the submission. The themes of interest to this research topic include but are not limited to:
- Generative AI and Large Language Model (LLM) at the edge
- Resource allocation and management in social edge computing
- Human-centric data science (e.g., modeling, integration, processing, visualization) at the edge
- Heterogeneity (e.g., data, model, device heterogeneity) in social edge computing
- Robustness and Generality in social edge computing
- Scalability and Efficiency in social edge computing
- Federated learning and distributed modeling in social edge computing
- Real-time and efficient AI in social edge computing
- Human-Computer Interaction (CHI) and Human-AI interaction in edge computing
- Human-like AI agents and anthropomorphism at the edge
- Trustworthy AI and responsible machine learning in edge computing
- Crowdsourcing, social sensing, and crowdsensing in edge computing
- Multimedia, social media and multimodal applications in edge computing
- Bias and Fairness issues at the edge
- Privacy, Security and Ethical issues at the edge
- AI for Social Good applications in edge computing
- Novel and emerging applications in social edge computing
The SEC paradigm presents a set of interesting research problems. Examples of such problems include the rational nature of device owners, pronounced heterogeneity of edge devices, real-time and efficient AI at the edge, generative AI and large language models (LLM) at the edge, trustworthy AI and responsible machine learning at the edge, human-like AI agent, and human-AI interaction in edge computing, and the privacy/security concerns and fairness issues of the edge users. By investigating these research questions, SEC targets to provide a set of human-centered AI solutions that jointly explores the collective intelligence of humans, machines, and AI in a resource-constrained computing environment. SEC also motivates novel AI for social good applications at the edge, such as personalized health monitoring, automatic disaster damage assessment, real-time language translation, crowd abnormal event detection, smart home automation, and vehicle-based criminal tracking.
The SEC paradigm’s specific connections to the Data Science section in Frontiers in Big Data address some of the novel human-centric data science questions at the edge of the network. Examples of such questions include: 1) how to efficiently acquire, clean, process, integrate and analyze human-centric data generated at the edge (e.g., video, audio, image, and text data generated by humans or smart devices on their behalf)? 2) How can novel distributed data-driven models and systems that can effectively explore the tradeoff between application performance (e.g., accuracy, delay) and limited resources (e.g., bandwidth, storage, processing capability) at the edge of the network be developed? 3) How can human-centric issues (e.g., privacy, fairness, trustworthiness) for data-driven applications at the edge be addressed?
This Research Topic looks for articles that have not been published in the prior literature or are under review/consideration of another publication venue. If the submission is an extension of previously published conference or workshop papers, it needs to have at least 30% new content and the difference between the submission and the previously published version needs to be explicitly cited in the submission. The themes of interest to this research topic include but are not limited to:
- Generative AI and Large Language Model (LLM) at the edge
- Resource allocation and management in social edge computing
- Human-centric data science (e.g., modeling, integration, processing, visualization) at the edge
- Heterogeneity (e.g., data, model, device heterogeneity) in social edge computing
- Robustness and Generality in social edge computing
- Scalability and Efficiency in social edge computing
- Federated learning and distributed modeling in social edge computing
- Real-time and efficient AI in social edge computing
- Human-Computer Interaction (CHI) and Human-AI interaction in edge computing
- Human-like AI agents and anthropomorphism at the edge
- Trustworthy AI and responsible machine learning in edge computing
- Crowdsourcing, social sensing, and crowdsensing in edge computing
- Multimedia, social media and multimodal applications in edge computing
- Bias and Fairness issues at the edge
- Privacy, Security and Ethical issues at the edge
- AI for Social Good applications in edge computing
- Novel and emerging applications in social edge computing
Keywords: Edge Computing, Artificial Intelligence, Social Computing, Human-centered AI, Trustworthy AI, Efficient AI, Internet of Things (IoT), Distributed Systems, Privacy, Security, Social Good, Human Machine Interaction, Resource Management, Fairness and Ethics
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
