TopoBudget: Persistent-Connectivity-Preserving Web Graph Sparsification for Reusable Community Analytics

2026-08-11T08:52:08Z9b2d0bbbe327edeef438274ba0f05bfc2315a9308c9f4507d2563ad1f0b00280
IoT securityLLMTikTok data collectionacademic researchadversarial machine learningcontent moderationdisinformationevasion attacksgraph analyticsopinion dynamicssocial media

What happened

An arXiv RSS collection containing research on graph sparsification, TikTok data acquisition, network centrality, IoT power control, agent-based social dynamics, SocialFi simulation, perceptions of AI financial advice, adversarial testing of LLM-generated disinformation detectors, and opinion polarization control. The security-relevant item studies iterative adversarial rewriting—including back-translation and LLM persona-based transformations—that evades social-media disinformation classifiers, achieving up to a 95% label-flip rate while often preserving the original meaning. No direct system

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
arxiv_cs_si
Record identifier
9b2d0bbbe327edeef438274ba0f05bfc2315a9308c9f4507d2563ad1f0b00280
Enrichment time
2026-08-11T08:52:08Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.