You can use InfraNodus to analyze comments to an article to the article content itself. This approach will help you reveal the main topics in the public discussion and discover what either the public comments or the article is missing.
In this example, we will use an article from The Financial Times newspaper: "Why are interest rates not taming inflation?". We will demonstrate how the article is missing a crucial point which is revealed in the comments.
The first step is to analyze the article. We can do this by simply copying and pasting its text into InfraNodus. As a result, we see a graph with all the main topics inside and connections between them:
We can quickly reveal the main factors that influence the global inflation outlined in the article: its relation to the housing market (fixed rates do not reflect the current borrowing rates, so it takes longer to react) and service economy growth (lower unemployment rates driving the salaries higher).
These two reasons are cited in the article as the key factors that keep inflation rates high despite the government increasing the interest rates.
The second step is to analyze the comments on the article. In order to do that, we load all the comments and either copy and paste them or use the scraping technique to extract them in text form.
As a result, we get the following topics with InfraNodus:
As we can see, the picture is very different here. People are discussing Covid-related monetary policy and the financial crisis that was triggered by irresponsible borrowing and lockdowns. They see this as the main root cause of the problem and actively discuss this in the comments. The original article has only a fleeting mention of this as a cause for high interest rates and inflation.
We can further highlight this discrepancy between the article and the comment by using the compare graph feature of InfraNodus. In order to do that, we open the graph of the article again and then click compare the graphs button and choose to show how the graph with the comments is different to the graph of the article.
The resulting visualization will show us the terms and the topics that are present in the comments but are not present in the original article:
As we can see, the two main concepts here are
These concepts are present in the comments to the article but not in the original article. If we click on those items we can see the comments that contain those terms. We can then use the built-in GPT-3 AI to summarize those comments and, as a result, we get the following insight:
During the Covid, the governments borrowed huge amounts of money and injected cash into the economy that was crippled by lockdowns. They now have to service these debts (borrowed at the lower rates in 2020/2021) so it's cheaper for them to keep the interest rates high as they have to pay less and, thus, shift the burden of these debts on taxpayers and people who have to deal with the rising prices and inflation. This could be one of the reasons why they are not proactively taming inflation with other means because they are simply not interested in reducing the interest rates just yet, before they pay out the bulk of the payments on debt.
Interestingly, but also something that can be expected, this is not what the newspapers are telling us. The comments to this Financial Times article are highlighting this aspect though, which confirms, once again, why having a public platform for discussing the news in the media is so important.