Hello, this is CYRAM.

With the upgrade of NetMiner's News Data Collector to News Data Collector Plus, tasks like network analysis, topic modeling, and creating analytical reports, which previously took at least a week, can now be completed in just 10–20 minutes with a few clicks.

In our previous session, we collected news articles related to "Samsung Electronics" and explored the AI-generated analysis report results. This time, we will use News Data Collector Plus to examine how media coverage trends have shifted during the U.S. presidential election. Additionally, we’ll uncover the key keywords emphasized in news articles about each candidate and compare the topics covered in articles for each candidate.

• Collection Keywords : Harris, Trump

• Collection Period : 2024-11-01 ~ 2024-11-11

• Sorting : By relevance

• Number of Articles per Channel : 2,000

• Full Text Collection : Yes, even if it takes longer, full text will be collected.

The total number of articles mentioning 'Harris' was 1,253, while 'Trump' was mentioned in 3,023 articles. This indicates that domestic media showed more interest in articles related to Trump.

Articles related to both 'Harris' and 'Trump' generally showed a gradual increase in volume starting from November 1, peaking sharply the day before the U.S. presidential election results were announced (November 6).

Following the announcement, articles about the winner, 'Trump,' reached their highest volume on November 7, the day after the results were declared. However, the number of articles dropped sharply on November 8 before starting to rise again from November 9 onwards.

On the other hand, articles about the runner-up, 'Harris,' showed a continuous decline in volume after the election results were announced.

[Changes in Mentions by Media Outlet : (Left) Harris, (Right) Trump]

When analyzing by media outlet, it was observed that for articles related to 'Harris,' only KBS reached its peak on November 4, while the remaining 8 outlets peaked on November 6 and then showed a consistent decline.

For 'Trump'-related articles, all media outlets recorded high article volumes from November 6 to November 7. Post-election, most outlets continued to report more than before the election, except for MBC, which showed a consistent decline in coverage following the election results announcement.

2. Analysis of Main Key words

In articles related to 'Harris,' the AI identified 'Joe Biden' as a central keyword, ranking high in both degree centrality and betweenness centrality, indicating its pivotal role.

'Economic Power' ranked high in weighted degree centrality, suggesting that economic factors play a significant role in discussions surrounding Harris's election campaign.

'North Carolina' also ranked high in weighted degree centrality, highlighting the importance of this region to Harris's candidacy in the election.

['Voter' Keyword Association Network Map (Harris)]

By reviewing the association network map of the keyword 'voter', which ranks high in degree centrality, the AI interpreted that the mentions of 'exit poll' and 'discontent' together indicate discussions on voter dissatisfaction with the election results and the reliability of exit poll data.

Additionally, 'racism' and 'labor market' appeared as associated terms with 'voter.' This suggests that racial issues and labor market changes emerged as key concerns among voters.

The simultaneous mentions of 'target' and 'resentment' imply discussions about voter rejection of specific policies or candidates and the strategic responses to address these sentiments.

In articles related to 'Trump,' the AI identified 'North Korea' as a significant keyword, ranking high in terms of the number of appearances, degree centrality, and weighted degree centrality. This highlights its importance in the context of Trump's discussions.

'Enterprise' ranked high in degree centrality, suggesting that economic factors were a critical influence on Trump's election issues.

'Russia' ranked high in weighted degree centrality, indicating that international relations played a key role in the discourse surrounding Trump.

['Enterprise' Keyword Association Network Map (Trump)]

From the association network map of the keyword 'enterprise', which ranks high in degree centrality, the AI identified 'semiconductor' as the most frequently associated term. This suggests that the re-election of the Trump administration is anticipated to bring significant changes to the semiconductor supply chain, leading to heightened tensions in the domestic semiconductor industry and increased market uncertainty.

'Investment' was the second most mentioned term, reflecting the active investment activities by enterprises to drive growth and expansion.

'Subsidy' ranked third, indicating efforts by enterprises to strengthen their competitiveness through government support.

'Tariff' was highlighted as an issue affecting enterprise activities, indicating the influence of international trade policies.

3. Issue Analysis

The main topics of articles related to 'Harris' and 'Trump,' as analyzed by AI, are as follows.

The common topics of interest in news articles about both 'Harris' and 'Trump' included themes like 'polling stations' and 'vote counting', reflecting predictions and updates on the question, "Who will be the next U.S. president?" Additionally, topics like 'interest rates' and 'tariffs' highlighted economic and industrial trends, including anticipated changes in U.S. tariff policies and responses from domestic enterprises based on the election results.

However, while articles about Harris included figures like 'Joe Biden', who could influence the election, articles about Trump focused on topics like 'North Korea', emphasizing potential impacts on U.S. foreign relations and diplomatic policies.

Let's examine the evolution of topics in articles related to 'Harris' and 'Trump' from November 1 to November 11.

The changes in topics over time for articles related to 'Harris' and 'Trump' are as follows:

In the results of the time-based topic changes for 'Harris' articles, the number of articles per topic showed an overall increasing trend as November 6 approached, with the 'polling station' topic recording the highest value during the entire period on the day of the election results announcement. However, after the announcement of the election results, the numbers plummeted. The 'Joe Biden' topic recorded the highest number of articles on November 7, the day after the announcement of the election results. The 'interest rates' topic had the lowest number of articles among the three topics and did not show significant changes by date.

For the results of the time-based topic changes for 'Trump' articles, as with 'Harris,' the number of articles per topic surged as November 6, the height of election result counting, approached, but declined after the election announcement. However, starting from November 9–10, the number of articles began to rise again.

The 'vote counting' topic recorded the highest value during the entire period on November 6, the day of the election results announcement. However, after the announcement, the numbers plummeted. The 'North Korea' and 'tariffs' topics both surged the day after the election announcement, November 7, then declined again but began to rise from November 10 onward.

Let’s examine the topic proportions of 'Harris' and 'Trump' articles across nine media outlets.

Here are the results of the topic proportions by media outlet for articles about 'Harris' and 'Trump.'

In the topic proportions by media outlet for 'Harris' articles, all nine media outlets had the 'polling station' topic as the most dominant. Among them, Kyunghyang Shinmun had the highest proportion at 70.23%, while Chosun Ilbo had the smallest proportion at 52.48%. For the second-ranked 'Joe Biden' topic, SBS showed the highest proportion at 31.68%, while Kyunghyang Shinmun had the lowest at 19.85%. Regarding the third-ranked 'interest rates' topic, Chosun Ilbo recorded the highest proportion at 21.99%, while KBS showed the lowest at 6%.

Similarly, in the topic proportions for 'Trump' articles, the 'vote counting' topic was the most dominant across all nine outlets. MBC recorded the highest proportion at 49.59%, while KBS had the lowest at 38.09%. For the 'North Korea' topic, KBS had the highest proportion at 36.82%, while Chosun Ilbo had the smallest at 19.35%. Regarding the 'tariffs' topic, Chosun Ilbo showed the highest proportion at 39.58%, while SBS had the lowest at 21.52%.

As identified in previous analyses, both 'Harris' and 'Trump' articles shared common economic topics such as 'interest rates' and 'tariffs'. Therefore, let's compare the economic issues covered in articles about Harris and Trump.

First, here is the visualization of the 'interest rates' topic keyword network map from the 'Harris' articles.

['Interest Rates' Topic Keyword Network Map]

The AI interpreted that the keyword 'enterprise' had the highest number of associated words, emphasizing its importance as an economic entity in articles about 'interest rates'. The keywords 'investment' and 'policy' also showed a high number of associations, indicating the impact of interest rate changes on investment and policy decisions. 'Semiconductor' and 'influence' were also prominently linked, suggesting that interest rate changes significantly affect the semiconductor industry.

​

Next, here is the visualization of the 'tariffs' topic keyword network map from the 'Trump' articles.

['Tariffs' Topic Keyword Network Map]

In the 'Trump' articles, the keywords 'enterprise', 'investment' and 'semiconductor' also appeared prominently, similar to the 'Harris' articles. Among these, the keyword 'enterprise' recorded the highest number of associations within the tariff topic, as in the 'Harris' keyword network results. The AI interpreted this as indicating that tariff policies have a significant economic impact on enterprises.

Keywords like 'investment' and 'concern' also showed high association counts, reflecting the influence of tariffs on the investment environment and economic uncertainty. The keyword 'semiconductor' was linked to discussions about the impact of tariffs on the technology industry.

In conclusion, domestic media displayed a notable interest in articles related to Trump, the winner of the election. For both candidates, articles covering real-time updates and predictions of the U.S. presidential election results were the most common. Additionally, reports explored how the U.S. election would affect the global economy and South Korean enterprises. However, a key difference in media focus was observed: articles about 'Harris' often discussed the Biden administration and its policies, while 'Trump' articles highlighted diplomatic issues, such as those involving North Korea, revealing distinct areas of interest in domestic media coverage for the two candidates.

The AI analysis report feature of News Data Collector generates a substantial amount of analysis and interpretation results in just about 10 minutes. Additionally, the analysis results and reports automatically generated by AI are provided in multiple formats, including image files (.png), Excel files (.xlsx), and NetMiner files (.nmf). This allows users to reprocess the results into their desired formats, conduct additional analyses as needed, and modify report content, making it a highly flexible tool for work applications.

Don't miss the opportunity to significantly reduce your work time and maximize efficiency in your tasks!