How to use

Step One: Data Collection

• Specify a valid subreddit name stub (i.e. spacex)


• Reddit has a few types of submission listings we can scrape data from: New, Top and Hot. Top can be filtered by day, week, month and year.


• Including comments will greatly increase the available data but at the tradeoff of making a large number of API calls to Reddit.
• The quality of the data associated with comments is generally lesser than that of the posts themselves.

Step Two: Data Visualization

• The tool will process the retrieved data and attempt to create a network between frequently used sets of meaningful words.


• If no frequent itemsets are found, try lowering the minimum support threshold.
• Too low of a threshold will result in longer computation times.


• Remove undesired or subreddit-specific stop words by selecting them in the meaningful word listing.
• A list of common English stop words have already been loaded.


• If the tool becomes unresponsive, refresh the page and try again.

Meaningfulness

Our proposed meaningfulness score will focus not only on the frequency of a given word over the subreddit, but it will also be scored based on the karma of the post that the word occurs in. This helps distinguish between a common word and a meaningful word.

This data visualization tool was made as a course project for CSCI4210U Information Visualization taught by Dr. Christopher Collins @ UOIT.

Created by

• Lachlan Johnston
• Michael Lombardo
• Spencer Bryson

View the source @ GitHub