Behavior Change as a Signal for Identifying Social Media Manipulation
2026-03-04T19:56:39Z•29d25b0b4535e62ae3f606e7752b832c5b8b4621f2093d03f7928bb917fbc34c
LLM-detectionbehavioral-change-detectioncommunity-detectioncontent-moderationcoordinated-inauthentic-behaviorcredibility-governancedata-integrationdynamic-graphsexplainabilityhate-speech-datasetidentity-ecosystemmisinformationnetwork-analysisopen-sourceprivacy-riskprompt-engineeringscholarly-datasetssocial-botssocial-media-manipulation
What happened
Collection of recent academic papers focused on online information integrity and tooling: (1) a method using Behavioral Languages for Online Characterization (BLOC) to detect social bots and coordinated inauthentic behaviour by measuring behavioral change; (2) Credibility Governance, a dynamic credibility-scoring mechanism to mitigate manipulation of collective judgments; (3) HateMirage, a multi-dimensional explainable dataset for subtle/faux hate and benchmarking explainability in hate detection; (4) analyses of content moderation distortions and an LLM-based rephrasing mitigation; (5) a case
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_cs_si
- Record identifier
- 29d25b0b4535e62ae3f606e7752b832c5b8b4621f2093d03f7928bb917fbc34c
- Enrichment time
- 2026-03-04T19:56:39Z
- 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.