The Cogito Graph Engine is the core of the Cogito Knowledge Center architecture. The engine represents information as entities (nodes) and relationships (arcs) in a scalable store and executes high performance traversal and retrieval operations. The Cogito Graph Engine provides modeling and query services in addition to fundamental graph services. Unlike others that use an 'in memory graph', the Cogito Knowledge Center provides a 'persistent graph' that spans the size of memory available.
Users can model the information to establish various contextual relationships, ontologies and views. Data points are identified as class-typed nodes in the graph overlay, and relationships are represented as arcs with definable arc types. Modeling flexibility allows analysts to change the graph model structure to easily see different perspectives on potential patterns or relationships.
Immediately after fusing the data, which includes the process of modeling, importing and linking, all of the data is ready and able to be queried and analyzed. Users can query or visualize data to see if relationships exists between seemingly unrelated data points and identify the shortest path between them or query for specific patterns within the data. Analysis allows users to see which connections are the most powerful or the weakest, how one data point affects other data points, or how information is interrelated.
The Cogito Graph Engine and resultant graph(s) can be included as part of a services oriented architecture (SOA) consisting of distributed graph databases, distributed Cogito engines, web services applications, and centralized management.
The Cogito Graph Engine handles massive data sets ingested from multiple sources. The engine can manage up to fifteen billion nodes with nodes capable of supporting one million arcs. GQL, the first commercially available graph query language, enables users to programmatically interact with the engine.
Key Benefits:
- Handles large databases of fully connected data
- Simplifies aggregation and fusion of disparate datasets
- Provides advanced tools for network relationship analysis
- Enables new ways to look at and evaluate patterns
Top Features:
- Powerful and fast assimilation
- Generates fully normalized datasets
- Rich APIs for custom applications
Requirements (minimum):
- Processor: Pentium Class, 1.2 GHz+
- RAM: 512 MB
- Disk Space: 50 MB plus data
- Operating System: Windows Server 2003 or Windows XP
|
