🎥 Case Study: Analyze the News Discourse with Graphs

Using InfraNodus, you can represent the current news discourse as a graph in order to reveal the main topics and trends, estimate the level of urgency in a discourse, and find some hidden connections and parts of the discourse you would not normally see before. 

Below we present a quick tutorial showing you how it works using the example of crypto market news. You can, of course, apply this same approach to any other topic as well. 

Part 1: Importing from Various News Sources via RSS feeds, Twitter, or manually via copy + paste 

Option 1: RSS feed. Use the InfraNodus RSS app (Apps > News > RSS) to import the news from the sources you trust or use some of the default ones we provide. Note, you can use a service like RSSGround or Feedspot to create your own RSS feed from multiple news sources.

Option 2: manual copy + paste. If your news resource does not provide an RSS feed or it's not free, you can simply copy and paste the data from the website directly. For instance, if you want to grab the trading suggesting from TradingView.Com, you can simply copy and paste the text and add it manually using the InfraNodus Add Text app.

Option 3: CSV / spreadsheet export. Some resources allow you to search for the news on a certain topic and to then export them as a CSV spreadsheet. You can then import this spreadsheet to InfraNodus to get the results you need.

Option 3: Twitter or social media. You can import the news from Twitter using the InfraNodus Twitter app or also an RSS feed from subreddit (use InfraNodus RSS import app and add ".rss" to your subreddit link. 

Part 2: Advanced Graph Analysis: Detecting the Main Topical Trends, the Latent Topical Brokers, Iterating through Nodes

Hint 1: click on the main topics to see which context they are used in. This is a great way to explore the news in a way that is not chronological but where you start with the stuff that are most relevant to you.

Hint 2: look at Insight > Latent Topical Brokers. These are the nodes that have the highest influence per degree and will often reveal some interesting and not very evident concepts.

Part 3: Revealing the Urgency of the Discourse using its Structure, Identifying the Trend

Step 1: Look at the structure of the discourse and its mind-viral immunity (correlated). If it's diverse (mind-viral immunity high), there are multiple topics present, the urgency and the bias are low. If it's focused or biased (mind-viral immunity low), the discourse is too focused on one topic. This measure helps you identify the urgency of a discourse. 

Step 2: Check the current trend using the dynamic graph feature in the Filter panel. Drag the knob to see how the discourse evolved over time. If you see that the main topics stayed the same over some time, this means there was not so much change happening. If you see that it's jumping all over the place, there's been a lot of topical volatility in the discourse.