Integrated value chain risk-based profiling and optimization

A computer system for integrated value chain risk-based profiling and optimization, comprising:

• a hardware processor configured to execute software instructions stored on non-transitory machine-readable storage media that cause the computing device to:
• analyze an integrated value chain ontological database comprising aggregated data from a plurality of heterogeneous sources, the database being ontologically structured, for information related to a risk query, wherein the information related to the risk query comprises entities, individuals, and locations;
• create or update a weighted and directed temporospatial knowledge graph based on the database, the weighted and directed temporospatial knowledge graph comprising a first set of nodes representing entities, a second set of nodes representing entity descriptors, and weighted edges representing relationships among the nodes, wherein the weighted and directed temporospatial knowledge graph spans a plurality of time periods and geographical locations, and incorporates provenance and contextual metadata, and wherein the weighted and directed temporospatial knowledge graph represents a digital version of the integrated value chain;
• identify one or more paths or clusters of a subset of the weighted and directed temporospatial knowledge graph that meet a pre-determined risk threshold, wherein each of the one or more paths or clusters represents a risk category;
• perform one or more discrete event simulations using data from at least part of the integrated value chain ontological database, the data used in performing the one or more discrete event simulations comprising real-world and partial synthetic data, wherein:
• each discrete event simulation models propagation of disruption events through entity relationships in the weighted and directed temporospatial knowledge graph to determine a probability and disruption impact associated with the disruption event;
• and
• a subset of the plurality of discrete event simulations models alternative actions in response to disruptive events to identify potential mitigation strategies for disruption events;
• and
• assign, using a trained machine-learning model that processes outputs of the one or more discrete event simulations, a risk score to each risk category represented by the one or more paths or clusters, based on the probability and the disruption impact associated with the risk query;
• wherein the analysis of the aggregated data from the plurality of heterogeneous sources, the updating of the weighted and directed temporospatial knowledge graph, the identification of one or more paths or clusters, the performance of the one or more discrete event simulations, and the assigning of the risk score are performed iteratively such that updates occur in response to new data from the plurality of heterogeneous sources or new outputs of the discrete event simulations.