Dynamic exploration of large dimension graphs
for community detection
Graph is a powerful data representation for BI people and analysts, but visualizing a graph can be complexe as nodes and vertices number grows. Community detection algorithms solve this problem by gathering nodes into communities in a hierarchical way. But it is computationally intensive and limits come when the graph is dynamic.
LumenAI brought to life a new algorithm capable of refreshing communities in real time within a time evolving graph.
This has applications in social networks analysis, intelligence services, fraud detection, etc.
We support actors on these main problematics :
Detect fraud while analyzing dynamic data in realtime.
Business fraud can be dismantled through modeling this business and finding irregularities within the model. LumenAI can help you practice your business rules and streams to opportunely identify fraud prone actions or behaviors.
LumenAI deploys to its clients new architectures and methods for operating realtime analysis on this dynamic data. The discovery of new, previously unseen behavior is instantaneous.
Ease the search and improve the analyst performances.
What is the problem : on dynamic data, as it is often the case today with continuous information streams to monitor, analyse and interpret, lack of dynamic analysis and visualization tools result in an incomplete perception of the phenomena to be observed.
Indeed, existing commercial tools exhibit poor features for dynamic data stream exploration and visualization. The analyst must iterate its search in a constantly interrupted and troubled manner. This result in a loss of time, and loss of clear perception of the phenomena to be identify.
LumenAI brings its new tools for dynamic analytics and visualization to handle to most complex and rich data stream exploration problems. LumenAI's tools assist the analysts in his search for singularities, new patterns, or patterns knowns to be worth finding. Its continuous analytics and visualization features make it particularly friendly to keep the analyst connected to the phenomena under operation.
Reveal sensitive information within a lake of data