Name: Bhavika Padidala
Course: User Experience Design
Topic: Architectures for interaction.
The goal is to establish the settled outline standards for building complex visual analytics (VA) web applications that could productively incorporate visualizations with data management, computational change, hypothesis testing, and knowledge discovery.
Clairvoyant work by Bender et was the first to suggest framework architectures that implement web visualization pipeline as containing 3 separate steps: data preparation, representation and rendering
The authors examined how old practices of large-scale web application improvement, together with present-day visualization tools can be adjusted to fabricate integrative VA applications. They contributed the fundamental outline and usage of a platform for Statistics Online Computational Resource Analytical Toolbox (SOCRAT).
The author obtained the results by presenting a prototype of SOCRAT, a multilevel platform to build VA applications. We use AngularJS, as it provides useful featres that suites in the context of proposed architecture. Define Modularity at the system level, modules at platform level AngularJs components at module level. They used npm to include system dependencies . Webpack is used to develop application and to separate code in to small parts that can be loaded dynamically to represent the highest modularity level.
In the experiment they combined ngHandsontable a data grid editor for raw data input and display. Datavore to store and retrieve the data and a lowlevel D3 library for visualization. ngHandsontable ,D3, and jstat are infused straight in to Data input,Charting, and clustering module. This application provides an interface to upload custom CSV/TSV file using K-Means algorithm.
Apart from all the libraries and tools used in this paper, the developers can choose any visualization library in combination with any analytical methods that can be further used in developing the applications.