One of the basic tools of data analysis is visualization. Good, flexible visualizations make it easier to explore and understand the data, and this is useful at all stages of the data analysis process. At the beginning, visualizations make it easier to find errors and inconsistencies and to get to know your data and developing an intuition for it. It continues to drive insights throughout the process. In the end, visualizations make great supporting evidence and explanations in reports and presentations.

Visualizations will be an important part of this chapter and in understanding the results of topic modeling. To create and interact with the graphs, we're going to use some software that's recently become an important part of many data scientists' toolkits: the Web browser.

As we did in Chapter 1, Network Analysis – The Six Degrees of Kevin Bacon, we'll use D3 (http://d3js.org/) and ClojureScript (https://github.com/clojure/clojurescript/).

The graph of the word counts earlier in this chapter as well as the ones that will come later are examples of this system. They're part of a static website. That is, the resources that load in the browser are read from the filesystem, not generated dynamically by a server-side web application. The data is read from CSV (comma-separated values) files that we'll create from the topic model data. Finally, the ClojureScript is compiled into a JavaScript file that's loaded by the browser.

We'll see later how to set up this site with ClojureScript as well as how to create the graphs. As usual, for the full code, refer to the source code download from the Packt Publishing website.