A tool for visualizing and exploring research publications. Given the vast collection of research literature that exists and is being produced, searching for the right documents is a challenging problem. This is especially true for someone exploring a new area with little background. ResearchVisualizer aims to fix that by using visualization techniques to make the search process easy.

The tool stores publication metadata as a json file with the following schema:

{ "id" : Int, "title" : String, "authors" : List, "abstract" : String, "keywords" : Option[List], "type" : Int, "venue" : String, "year" : Int }

Type encodes the type of publication (0-conference, 1-journal, 2-book)

Citations are stored as links between papers

{ "source" : Int, "target" : Int }

Running locally: Open a terminal and cd to the project directory. Then start a local server. A simply python server should suffice: python -m SimpleHTTPServer 8888

Then open your browser and go to localhost:8888/publications.html

This will open the primary publications visualization.

Dependencies: This tool uses the following Javascript libraries:

1. d3.js (http://d3js.org)
2. underscore.js (http://underscorejs.org/)