Middle-Out Graph Analytics

Pacific Northwest National Laboratory, Richland, Washington

Graph analytics is a way of facilitating guided graph exploration through visual and interactive means. Unlike many graph visualization research efforts that focus predominantly on layout algorithms and rendering techniques, graph analytics research strives to provide an engaging interactive journey that bridges the gap from data to information to knowledge. Graph visualization still plays an important role in building this analytical journey, as do database querying, graph mining, interactive interrogation, human judgment, and senses.

Graph analytics tools can be used to study a variety of situations, from geographical locations to levels of response.

Two primary schools of thought have developed when designing graph analytics tools: top-down and bottom-up. The top-down approach often provides an initial full view of the entire dataset, and then gradually reaches out to the local details. The bottom-up approach frequently starts with seed nodes or a subset of nodes, and then builds the rest of the graph through associations. These graph analytics tools are suited to different tasks or goals.

Whether in corporate or government organizations, analysts today are bombarded with massive amounts of information from a multitude of sources. This vast amount of information may easily overwhelm an analyst’s cognitive capacity. A visual analytics approach was developed that transforms data into a graph representation consisting of nodes and links. Using this graph analytics software, investigators can query, organize, and link information about individuals, facts, locations, events, objects, and data to discover key trends, patterns, and insights.

The approach exploits the rich middle-ground information typically overlooked by most traditional top-down and bottom-up analysis tools. The software employs a multi-resolution, middle-out, cross-zooming technique that allows users to interactively explore their graphs on a common desktop computer or handheld device, interactively analyzing graphs with up to one million nodes.

The graph analytics tools can be tailored to a variety of areas, including social networks, cybersecurity, electric power grids, forensic analysis for law enforcement, critical infrastructure, bio-informatics, and earth sciences.

For more information, contact Kannan Krishnaswami at This email address is being protected from spambots. You need JavaScript enabled to view it.; 509-375-4597.