Selected work

Representative Papers

Beyond Topicality: A Conceptual Analysis of Societal Relevance and Its Application to Search Results and AI Responses

Jul 10, 2026

This work proposes the concept of “social relevance” to address the limitations of existing search relevance models, which primarily focus on topical matching and often fail to identify or mitigate harmful content such as misinformation and discriminatory material, thereby neglecting broader societal interests. Through conceptual analysis and a three-dimensional modeling framework encompassing system, user, and societal perspectives, the study clarifies the definition, boundaries, and distinction of social relevance from traditional information quality metrics. By transcending conventional relevance paradigms, this framework provides a theoretical foundation for designing retrieval systems that integrate ethical values and social welfare, ultimately advancing search engines toward a value-driven paradigm.

0 citationsRead paper

Secrets Best Not Shared: DNS Privacy Enhancements for the Constrained IoT

Jun 08, 2026

This study addresses the vulnerability of resource-constrained IoT devices to identification and privacy leakage in conventional DNS communications, which also exposes them to service disruption attacks. The work presents the first systematic evaluation of the privacy-preserving potential of DNS over CoAP and introduces a traffic obfuscation strategy integrating fixed-length packetization, block-wise transfer, and header-payload compression. To further enhance anonymity, the approach incorporates a variant of onion routing. Experimental results demonstrate that, across diverse link conditions, the proposed method reduces DNS frame identification accuracy to 77%, substantially outperforming DNS over HTTPS, which remains trivially identifiable via IP addresses. The authors also release a public dataset to establish a benchmark for future research in this domain.

0 citationsRead paper

Contrastive Learning and Correlation Clustering for Sequences of Network Telescope Data

Jun 03, 2026

This work addresses the challenge of effectively associating network scanning sources in the absence of semantic annotations. The authors propose an unsupervised contrastive learning approach based on the Transformer architecture that learns semantic embeddings directly from raw network flow sequences, without requiring pretraining or manual labeling. By modeling inter-sequence similarity and integrating correlation-aware clustering, the method enables automatic association analysis of scanning sources. Experimental results demonstrate that the learned embeddings exhibit significantly higher similarity for sequences originating from the same source compared to those from different sources, indicating strong generalization capability. Furthermore, the clustering outcomes align closely with ground-truth scanning labels, confirming both the effectiveness and novelty of the proposed approach.

0 citationsRead paper

LZn : Robust LoRa Frame Synchronization Under Frame Collisions and Ultra-Low SNR Conditions

Apr 30, 2026

This work addresses the vulnerability of LoRa to frame collisions under identical spreading factors and the inability of conventional receivers to achieve reliable synchronization and decoding in ultra-low signal-to-noise ratio (SNR) environments. To overcome these limitations, the authors propose LZn, a low-complexity synchronization scheme that introduces, for the first time, a spectral intersection mechanism into the LoRa synchronization stage. LZn significantly enhances robustness under high collision rates and extremely low SNR while maintaining low computational overhead and full compatibility with the existing physical layer. Experimental results demonstrate that LZn improves detection sensitivity by up to 10 dB and increases detection probability by 1.54× across simulations and three real-world datasets. Furthermore, it achieves a 3.46× improvement in decoding performance in the worst-case single-user scenario and a 1.22× gain under collision conditions.

0 citationsRead paper

StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning

Apr 09, 2026

This work addresses the often-overlooked dynamic programming–like information propagation structure in reinforcement learning. By analyzing the temporal evolution of return distributions in distributional reinforcement learning, the study identifies—within a model-free setting—an implicit dynamic programming structure for the first time. Leveraging this insight, the authors define a state learning order and introduce a temporal learning metric \( t^*(s) \) that characterizes when value information becomes reliably estimable at each state. Building upon this structure, they propose StructRL, a structure-aware sampling strategy that guides the agent to learn states in accordance with the intrinsic information propagation sequence. Empirical results demonstrate that the derived learning signal effectively captures the diffusion process of value information, leading to substantial improvements in both learning efficiency and stability.

0 citationsRead paper

Recent publications

Latest Papers

Beyond Topicality: A Conceptual Analysis of Societal Relevance and Its Application to Search Results and AI Responses

Jul 10, 2026

This work proposes the concept of “social relevance” to address the limitations of existing search relevance models, which primarily focus on topical matching and often fail to identify or mitigate harmful content such as misinformation and discriminatory material, thereby neglecting broader societal interests. Through conceptual analysis and a three-dimensional modeling framework encompassing system, user, and societal perspectives, the study clarifies the definition, boundaries, and distinction of social relevance from traditional information quality metrics. By transcending conventional relevance paradigms, this framework provides a theoretical foundation for designing retrieval systems that integrate ethical values and social welfare, ultimately advancing search engines toward a value-driven paradigm.

0 citationsRead paper

Secrets Best Not Shared: DNS Privacy Enhancements for the Constrained IoT

Jun 08, 2026

This study addresses the vulnerability of resource-constrained IoT devices to identification and privacy leakage in conventional DNS communications, which also exposes them to service disruption attacks. The work presents the first systematic evaluation of the privacy-preserving potential of DNS over CoAP and introduces a traffic obfuscation strategy integrating fixed-length packetization, block-wise transfer, and header-payload compression. To further enhance anonymity, the approach incorporates a variant of onion routing. Experimental results demonstrate that, across diverse link conditions, the proposed method reduces DNS frame identification accuracy to 77%, substantially outperforming DNS over HTTPS, which remains trivially identifiable via IP addresses. The authors also release a public dataset to establish a benchmark for future research in this domain.

0 citationsRead paper

Contrastive Learning and Correlation Clustering for Sequences of Network Telescope Data

Jun 03, 2026

This work addresses the challenge of effectively associating network scanning sources in the absence of semantic annotations. The authors propose an unsupervised contrastive learning approach based on the Transformer architecture that learns semantic embeddings directly from raw network flow sequences, without requiring pretraining or manual labeling. By modeling inter-sequence similarity and integrating correlation-aware clustering, the method enables automatic association analysis of scanning sources. Experimental results demonstrate that the learned embeddings exhibit significantly higher similarity for sequences originating from the same source compared to those from different sources, indicating strong generalization capability. Furthermore, the clustering outcomes align closely with ground-truth scanning labels, confirming both the effectiveness and novelty of the proposed approach.

0 citationsRead paper

LZn : Robust LoRa Frame Synchronization Under Frame Collisions and Ultra-Low SNR Conditions

Apr 30, 2026

This work addresses the vulnerability of LoRa to frame collisions under identical spreading factors and the inability of conventional receivers to achieve reliable synchronization and decoding in ultra-low signal-to-noise ratio (SNR) environments. To overcome these limitations, the authors propose LZn, a low-complexity synchronization scheme that introduces, for the first time, a spectral intersection mechanism into the LoRa synchronization stage. LZn significantly enhances robustness under high collision rates and extremely low SNR while maintaining low computational overhead and full compatibility with the existing physical layer. Experimental results demonstrate that LZn improves detection sensitivity by up to 10 dB and increases detection probability by 1.54× across simulations and three real-world datasets. Furthermore, it achieves a 3.46× improvement in decoding performance in the worst-case single-user scenario and a 1.22× gain under collision conditions.

0 citationsRead paper

StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning

Apr 09, 2026

This work addresses the often-overlooked dynamic programming–like information propagation structure in reinforcement learning. By analyzing the temporal evolution of return distributions in distributional reinforcement learning, the study identifies—within a model-free setting—an implicit dynamic programming structure for the first time. Leveraging this insight, the authors define a state learning order and introduce a temporal learning metric \( t^*(s) \) that characterizes when value information becomes reliably estimable at each state. Building upon this structure, they propose StructRL, a structure-aware sampling strategy that guides the agent to learn states in accordance with the intrinsic information propagation sequence. Empirical results demonstrate that the derived learning signal effectively captures the diffusion process of value information, leading to substantial improvements in both learning efficiency and stability.

0 citationsRead paper