Bridging the semantic gap: Visualizing transition graphs with user-defined diagrams

Users are often faced with large data sets in which the underlying meaning of the data is clear, but sense-making is difficult. State transition graphs fit this description well. Users associate precise semantics with multivariate attributes that describe nodes, but current visualization techniques do not exploit this. Consequently, users have to invest significant effort to interpret and understand visualizations. We present a novel technique that bridges this semantic gap by enabling the user to define custom diagrams that closely reflect the associated semantics. User-defined diagrams are incorporated in a number of correlated visualizations that support different user needs. To show the merit of our approach, the authors discuss two real-world applications. © 2007 IEEE.