Reading people's hearts has long been a concern of politicians.

This is because when you read the public sentiment, you can respond to the public sentiment and gain public sentiment.

However, it is not easy to know public sentiment.

There are polls, but the cost is insignificant.

It takes quite a bit of time, and it is difficult to see an immediate response.

Recently, many people avoid the response itself or hide their thoughts, such as'Shy ○○'.



On the other hand, portal news comments are immediate.

It doesn't cost anything to check.

As anonymity is guaranteed, opinions can be expressed freely.

It is not an answer to passive or hypothetical questions like public opinion polls, but rather an opinion that is'expressed' about a phenomenon or issue.

This is the reason why portal news comments are often introduced in public opinion despite the fundamental limitations of not being able to confirm who wrote it or whether it was written by several people.

But what if the fundamental limitations overshadow the features (advantages)?



The SBS <Actually> team and the data journalism team <Missing Stumble> prepared in-depth planning related to portal news comments.

With support from the <Korean Journal of Journalism> and <SNU Fact Check Center>, 75.8% of Koreans read the portal news that they use to get current affairs information, and about 80% of portal news users read our society through comments that they see It's an attempt to see.



To this end, comments were collected on <Most Viewed Articles>, which are frequently exposed to users.

Articles from August 2017, when Daum's <Most Viewed News> list is released to September 2020, just before Naver's <Most Viewed News> reorganization, are subject to analysis.

The basic ingredients are 208,000 Naver News articles, 58,000 Daum News articles, and 190 million comments on these articles.

Big data analysis company <Dive> was in charge of the analysis.



● Are the following news users more active in writing comments?



People who see portal news comments are common around.

However, it is not common for people to comment.

Therefore, the statistical analysis of the collected comments was expected to be a simple result of confirming the rule of thumb as data.

However, the results of the actual analysis were serious at some point and broke some conventional wisdom.

And some were new questions rather than results.



Naver and Daum are not disclosing traffic such as the number of daily users of the news.

However, based on various data in the industry and academia, it is estimated that there are several times more users of Naver News than the following.

There are also studies that show that there is a difference of up to 8 times or more.



It is reasonable to expect that the more users there are, the more the number of comments or commenters will be.

However, when comparing Naver and Daum, this prediction is wrong.

From January to September 2020, there were 26.5 million comments posted on Naver articles, and about 23 million next.

If you look at the absolute number, Naver has more.

But this is an optical illusion.



As mentioned earlier, there are far more articles to be analyzed by Naver than in the next page.

As of 2020, Naver's analysis targets 499,320 articles, which is 3.6 times the next 13,700 articles.

However, the total number of comments does not differ significantly.

Considering that Naver News users are more than 8 times as many as Daum News, Daum News has more than 10 times more comments than Naver News.



In particular, this trend is getting stronger and stronger.

Daum's average number of comments per article was 1.85 times more than Naver in 2017, but the gap grew to 2.35 times in 2018, 2.73 times in 2019, and 3.12 times in 2020.

In particular, in the case of social news, the average number of comments per article on Naver in 2017 was higher than Daum, but in 2020, Daum was 2.16 times more than Naver.

Political news also showed a similar increase as in the social field, but it can be interpreted that the increase in comments in the political and social fields led to a change in the whole.



● Comment authors, Daum'increased tax' vs Naver's'decreased tax' The



above results imply that Daum News users are more active in writing comments than Naver.

However, there is not enough to confirm this.

This is because we cannot rule out the possibility that some users have commented on the news in an extreme number.

So you need to check the number of commenters.




The graph above shows the number of people who commented on Naver and Daum News at least once per quarter.

The overall trend of Naver News is decreasing, while the next is increasing.

The same goes for politics, economy and society.



Now we can speak more confidently.

Considering the number of users and the number of articles to be analyzed, we can reach the conclusion that'Daum News users are much more active in commenting than Naver News users'.



● More users, fewer authors, and more attention.



However, this clearer conclusion can conflict with conventional wisdom.

In the parliamentary audit in October last year, Rep. Park Seong-joong, a member of the current People's Power, raised the possibility of manipulating the portal public opinion.

Rep. Park's raising of the issue was especially focused on the next side, he said, "The reason Daum is a prey for manipulation of public opinion is that it is much more advantageous to manipulate it because it is smaller than Naver and the reason why Daum is the prey for manipulation of public opinion is that all news is arranged in the number of recommendations and it is smaller than Naver."



It is not necessary to associate Park's concerns with illegal activities, such as manipulating drucking comments or manipulating NIS comments.

It is more appropriate to think of it as raising the question of whether it is possible to change the direction of portal public opinion through active activities of a small number of organizations.

In this regard, since Naver has so many users, it was a general idea that public opinion manipulation through comments, i.e., the outcome of the activity, would not be large.



However, comment statistics provide a very different answer from this conventional wisdom.

'Naver News users are more passive in writing comments than Daum News.

This trend is getting stronger.

In other words, if you actively write comments on Naver News, you can raise your attention even more in Daum.'

This answer, on the one hand, also means that the voices of people who comment on Naver News are more overrepresented than the following.



● The gap



in participation in comment writing Daum, with more and more people participating in comment writing.

Then, can Daum's comment space be said to be healthier in terms of diversity of opinions and active participation?

To do this, it is necessary to check the gap between participants, the so-called participation gap.

This is because even if a large number of people are involved, a certain minority may be in control of the public sphere.




The graph above shows the share of comments made by the top 10% of the authors based on the number of comments.

Although it is said to be better than in 2019, the proportion in 2020 is over 70% for both Naver and Daum.

The top 10% are Naver and Daum, respectively, 13.36 million and 103,000.

Less than 0.3% of our population in our teens or older writes most of the comments.



The top 1%, that is, 13,600 (Naver) and 1,300 (next), posted comments up to September last year, 25% (Naver) and 24% (next), exceeding 1/5 of the total. .

In 2018, they accounted for 31% and 30%, respectively.

0.03% of the population in their teens and older accounted for 30%.

On the other hand, among the people who commented on Naver and Daum in 2020, 31% (Naver) and 33% (Daum) each wrote only one comment for 9 months, and 75% each wrote less than 10 comments.

There is a phenomenon of extreme concentration, and by field, the degree of political, social, and economic news was generally severe.



Lee Won-jae, a professor at KAIST Graduate School of Culture and Technology, whose main interest is diagnosing our society through data, compared this phenomenon to a situation where a small number of IT companies dominate the global economy.

Professor Lee said, "Big IT companies such as Google and Apple do not reflect the interests of various industries in the world such as oil refining and agriculture, but the boom of five IT companies such as Google, which accounts for more than 20% of the US economy, is a problem in other industries. "You can hide it," he pointed out, "Similarly, the current trend in which a small number of commenters write a significant number of comments needs to be taken seriously (in terms of diversity)."



● What happened to Naver Political News comments in the third quarter of 2019?




However, what is positive is that the share of comments made by the top 10%, so-called heavy commenters, is slightly decreasing.

The graph above shows the proportion of comments made by the top 10% heavy commenters on a quarterly basis, and the dotted line represents a trend line.

It can be seen that Naver's share of comments in the top 10% is rapidly decreasing compared to Daum.

In particular, it can be said that Naver's focus is improving faster than the next one, as the proportion of comments in the top 10% declined despite the decline in the number of commenters.



However, there is something to pay attention to in the graph above.

This is Naver's 3Q 2019 graph.

The proportion of the top 10% of news commenters in the political sector was the only one in the analysis period lower than in the social and economic sectors.

It also fell below 50%, the only one during the analysis period.

This phenomenon is not found in the following news.

This period is also a period when the number of Naver commenters increased by 90,000 from the previous quarter.

What happened?



In the third quarter of 2019, that is, from July to September 2019, the so-called motherland war took place.

Various suspicions related to the candidate for the Justice Minister Cho Kook were poured out, and the prosecution investigation began in full swing during the personnel hearing.

Candles were burning in Seocho-dong and Gwanghwamun, Seoul, respectively, shouting'protecting the motherland' and'resigning the motherland'.

This is a time when many people consider it a turning point for strengthening the logic of our social camp.



During this period, the number of commenters in the political field surged from the previous quarter in both Naver and Daum.

It can be seen that a person who hesitated to express opinions or was passive about expressions actively expressed their opinions on the situation in the motherland or as an opportunity to the situation in the country.

However, a sharp decline in the proportion of the top 10% of comments is not observed in the following.




This phenomenon can also be seen in the change in the proportion of comments of the top 1% authors, who can be called'super heavy commenters'. As shown in the graph above, Daum's overall share of the top 1% is stable, while Naver's fluctuations have been large since the third quarter of 2019.



In response, Prof. Wonjae Lee raises the possibility that a comment war broke out on Naver News due to the so-called motherland crisis. But why is it Naver, not Daum? Professor Lee analyzes, "When Naver's weight in the portal news market is several times larger than that of the following, it may be the result of people concentrating on Naver and debating when a sharp conflict arises such as the motherland crisis." Because of the market dominance or representation in the portal market, the purpose of the war is to take place in Naver, the so-called big game.



This analysis implicitly presupposes that the comment space of Daum News is specifically homogenized. Because people with general opinions are gathered, there is a limit to arguing over different opinions. It will be reported later on whether it is, and what stories came out in the Naver News comments in the third quarter of 2019.



Although the result of asking questions about the third quarter of 2019, the message of this analysis is clear. 'The portal news comment space is led by a certain minority.' But there are also limitations. Even if not a few people write comments, they express their intentions with'empathy' or'disapproval' about the comments, but that part was not considered in this analysis.



SBS's <Actually> team and the data journalism team <Surprise> will analyze and report whether the so-called'super heavy commenter' is leading the news comment space both quantitatively and qualitatively. In addition, we plan to continue an article looking into our society through news comments by analyzing whether news comments actually affect people's thoughts and what comments were posted by issue and period.