# Temporal Heterogeneous Graph **Entity class:** Graph data model ## Definition A **temporal heterogeneous graph** contains multiple node and edge types whose relationships change over time. In a counterterrorism or investigative setting, nodes may represent people, devices, accounts, vehicles, places, organizations, transactions, identifiers, and events; edges carry type, time, direction, provenance, and confidence. ## Relationships - **analyzed through:** [[wiki/Identity and Relationship Analysis|Identity and Relationship Analysis]], [[wiki/Social Network Analysis|Social Network Analysis]], and graph analytics. - **dynamic inference:** [[wiki/State Estimation|State Estimation]] and [[wiki/Data Assimilation|Data Assimilation]]. - **model:** [[wiki/Predictive Graph Digital Twin|Predictive Graph Digital Twin]]. - **risk boundary:** ambiguous entity resolution and missing observations must remain explicit. ## Sources / Provenance - [ODNI — Data Mining Report CY2021–2023](https://www.odni.gov/files/documents/CLPO/CY2021-2023_Data_Mining_Report_FINAL.pdf) - [IARPA — AGILE program material](https://www.iarpa.gov/images/research-programs/AGILE/AGILE_ModSim_-_Final.pdf)