Tutorial: Visualizing a 2-mode Semantic Network in Communalytic

New (July 2021): You can now generate a Two-mode Semantic Networks with any dataset, even if they came from sources other than CrowdTangle.

Step 1: Overview

This tutorial will demonstrate how to use Communalytic’s built-in network visualizer module to generate an interactive two-mode semantic network with data with Facebook or Instagram public posts collected from CrowdTangle. (Note: This feature can also work with data from other platforms such as Reddit, Telegram and Twitter.)

Step 2: Collecting Data 

To collect data from CrowdTangle, open the CrowdTangle Data Import module in Communalytic and enter a URL of interest, e.g., a link to a news article or a popular blog post. 

In this tutorial, as an example, we will collect public Facebook posts that shared a link to one of the pages on the Ryerson University’s website between January 1 and May 18, 2021. Ryerson University (now Toronto Metropolitan University) is a Canadian public research university located in downtown Toronto. 

Step 3: Access Network Analysis 

Once the data has been collected (and the toxicity analysis has been completed – optional), you’re ready to generate the network. Click on the star-shaped icon under the “Network Analysis” column on the My Datasets page. 

Step 4: Generate Network

To generate a network, click the “Generate Network” button on your screen. After this is completed, you will use the network visualization tool to analyze and understand the different properties of the network to answer any questions you might have. 

Step 5: Two-mode Semantic Network

Unlike other data sources available in Communalytic such as Twitter or Reddit, information about who replies to whom is not provided by the CrowdTangle API. As a result, you will not be able to create a communication network with CrowdTangle data; however, Communalytic can create another type of network commonly referred to as a Two-mode Semantic Network

As the name suggests, a Two-mode Semantic Network is a graph that describes connections between two types of nodes, where one of the node types represents Social Actors and the other node types represents Semantic Concepts. A connection from Social Actor to Semantic Concept in this network usually implies some form of endorsement, association or affiliation between the two nodes.

Step 6: Progress Bar

You will receive an email once the network has been generated. You can also watch the progress bar, which automatically refreshes every 30 seconds.  

Step 7: Visualize the Network

Once the network has been generated, navigate to the network analysis by clicking on “Visualize Network.”

Step 8: Different Node Types 

In this visualization, Facebook groups or pages that posted at least one message in the dataset (also known as Actor Nodes) are displayed as circles with different colours. In our case, examples of actor nodes are “Ryerson University Career & Co-op Centre,” “Ryerson School of Journalism,” and “Ricch Canada Inc.” 

Named entities mentioned in the posts (also known as Semantic Nodes) are represented here as star-shaped black colour nodes. In our case, there are a number of semantic nodes referring to Ryerson University faculties, student clubs and other affiliated organizations. 

Because of the nature of this network, connections are going from Actor Nodes to Semantic Nodes only. 

Step 9: Control Panel

On the right-hand side of the browser, you will find the Control Panel to filter and adjust different components of the network visualization which will allow you to better understand the context of link sharing behaviour by Facebook accounts in the dataset. 

Step 10: Network Properties 

The first section in the panel is called “Network Properties”. It reports the number of Actor nodes, Semantic Nodes and Edges which are present in the network. For example, in our case, there are 150 actor nodes connected to 816 semantic nodes by 1245 connections. 

Step 11: Selecting and Viewing Nodes 

To view all named entities mentioned in posts shared by a specific Facebook page/group, click on the desired actor-type node (colour circle). For example, we can click on the “Science Literacy Week” node which represents a Facebook page that describes itself as “a Canada-wide week-long event that celebrates science-based activities and highlights our outstanding scientists and science communicators from coast to coast”.

By examining the semantic nodes (black-shaped figures) that “Science Literacy Week”  is connected to, we can see that this page promoted a number of relevant events hosted by Ryerson University, including “Science Rendezvous 2021” which offered virtual science activities on May 8 and the “2021 Soapbox Science Toronto” event on May 31 and June 1.

Step 12: Reading related Posts

When selecting a node in the visualization, we can also see all posts where the named entities linked to the selected actor node appear. For example, in the case of the “Science Literacy Week” Facebook page, we can see that it is linked to 13 different named entities, listed under the “Posts” tab in the Control Panel. 

Step 13: Accessing the Original Post on Facebook or Instagram

To view how the selected post is displayed on the platform like Facebook, where it was originally shared (if the post is still available), click on the blue arrow located beside the post, which will open a new window with the post as the platform presents it.

Step 14: Unselecting Nodes  

To unselect a previously chosen node, click anywhere on the white space. To go back to the unselected node, a box in the Control Panel under the Posts tab will appear and clicking on this will bring you back to displaying only nodes directly linked to the previously selected node. 

Step 15: Locating Corresponding Connections

While scrolling through the posts from/to the selected node, you might want to identify where an edge corresponding to a particular post is located in the network visualization. To do that, click on any post inside the left-hand side of the Control Panel (under the Posts tab) and the corresponding edge will be bolded and highlighted using the orange colour in the main network visualization.