Dashboard: A Tool for Integration, Validation, and Visualization of Distributed NLP Systems on Heterogeneous Platforms

Dashboard is a tool for integration, validation, and visualization of Natural Language Processing (NLP) systems. It provides infrastructural facilities using which individual NLP modules may be evaluated and refined, and multiple NLP modules may be combined to build a large end-user NLP system. It helps system integration team to integrate and validate NLP systems. The tool provides a visualization interface that helps developers to profile (time and memory) for each module. It helps researchers to evaluate and compare their module with the earlier versions of same module. The tool promotes reuse of existing modules to build new NLP systems. Dashboard supports execution of modules that are distributed on heterogeneous platforms. It provides a powerful notation to specify runtime properties of NLP modules. It provides an easy-touse graphical interface that is developed using Eclipse RCP. Users can choose an I/O perspective (view) that allows him better visualization of intermediate outputs. Additionally, Eclipse RCP provides plugin architecture; hence customized end-user specific functionality can be easily added to the tool.