DH Methods

Archival Practices

Canadians Read is simultaneously an archival and digital analysis project. We are archiving the discourse around Canada Reads while attempting to analyze that discourse and understand it from a range of perspectives.

Topic Modeling

Topic Modeling is a method of “distance reading” that employs statistical algorithms and software to identify latent themes, or topics, within a collection of data. While this method has been applied in humanities research in a range of fields, it is relatively unproven in literary analysis. We are using topic modeling algorithms on the entirety of the Twitter data related to Canada Reads in order to identify these themes and study how they overlap with, or differ from, the stated themes of the program.

Force-Directed Network Graphs

Force Directed Graphs are a method of demonstrating the links between nodes in a network. We employ this method to identify the networks of communication that occur between Twitter users around discussions of Canada Reads. By asking “who is talking to who” we can better understand the relationships between readers and the program and have an understanding of the users that occupy central and marginal positions within the field of discourse.

Digitally Assisted Textual Analysis

We employ a range of digitally assisted forms of textual analysis. This can range from as simple a method as calculating the average number of tweets per user or identifying the top users who write about Canada Reads to more complicated analyses of the word concordances (what words tend to show up together in this online discussion) and n-gram matching (what phrases repeat in this online discussion). In this digital/numerical form of textual analysis we are cautious to

Voyant Methods

We use Voyant Tools for quick data visualizations and word concordances.