Our OSINT analysts prepared a brief overview of the media mention trends surrounding the upcoming French presidential elections. This short analysis uses a small sample of data collected by our Media Mining System between January and March 2022 and covers selected topics, including Sentiment Analysis, Social Media Analysis, as well as visualizations of media mentions.
Background Information – Brief Overview
In April 2022, French voters will elect their next president. In addition to the incumbent President Emmanuel Macron, seven other candidates are running for office. It appears that, apart from Macron, the two far-right candidates, Marine Le Pen and Éric Zemmour, as well as far-left candidate Jean-Luc Mélenchon and center-right candidate Valérie Pécresse, would receive the most votes. According to the current poll figures, Le Pen has 18% of the vote, Mélenchon 13%, and Zemmour and Pécresse 11% each. Macron is currently leading with 29% of the vote.
If no candidate has an absolute majority after the first round of voting on April 10, there will be a run-off between the two candidates with the most votes on April 24. If there is a run-off between Macron and Le Pen, the situation could be similar to the last presidential election in 2017. Back in 2017, Macron won the second round of voting against Le Pen with around 66% of the vote. Thus, it remains to be seen whether history will repeat itself.
Trend Analysis – Mentions of Candidates in the Past Three Months
Through strategic mining of multi-media sources – including audio and text processing – it is possible to monitor and identify trends in real-time. The trend graph below shows mentions of the five most relevant candidates in French traditional and social media in the past three months.
It can be observed that all candidates have seen an increase in French media mentions: while Macron dominates them, the other candidates are also subject to a steady increase in mentions. The significant jump in mentions of Macron in recent weeks correlates with the Russian troop buildup and subsequent invasion of Ukraine. The increase can be explained by Macron’s role as head of state and European mediator in this crisis.
Trend Analysis – French Mentions of Macron in the Past Month
The more detailed analysis of French media in the past month reveals two significant increases in mentions of the keyword Macron:
- The spikes in mentions on February 21 and 22 can be explained by a phone call between Macron and Putin on the evening of the 21st. During the call, Putin agreed to a (later cancelled) summit meeting between Russia and the United States.
- On March 3, Macron formally announced that he will run for a second term in the April presidential election.
Trend Analysis – Mentions of Other Candidates in the Past Month
Looking at the other candidates, apart from Macron, another increase in mentions can be observed:
- The joint increase of mentions of Pécresse and Zemmour on March 10 correlates with a TV debate between the two candidates.
- Interestingly, besides Macron, Pécresse and Zemmour have the most mentions in French media. Le Pen, although in second place according to the polls, has about the same number of mentions as the left-wing populist candidate Mélenchon.
Bot Detection – Heatmap
The heatmap below shows social media posts by bot accounts (as detected by the HENSOLDT Media Mining System) mentioning French parties and candidates in the past month. The map shows that bot accounts which mention French parties and/or candidates (in French) originate mainly in France, Belgium and West Africa. The sentiment of the posts is predominantly neutral or negative.
Bot Detection – Candidate Analysis
The gauges below display the following information:
- Total French media mentions of Macron, Le Pen, and Mélenchon (past month)
- The mentions by bot accounts of Macron, Le Pen, and Mélenchon (past month)
The data shows that, in the past month, around 10% of all French media mentions of Macron originate from bot accounts. By contrast, only about 4% and 3% of all mentions of Le Pen and Mélenchon, respectively, originate from bots.
Word Clouds – Candidate Analysis
The word cloud tool allows for visualising names, entities, organisations and countries. The size of the font represents the number of times an entity has been mentioned. The word clouds indicate that Macron is mentioned significantly more often on his own, without the context of specific issues or people, as seen by the size difference between his name and other entities. Le Pen and Mélenchon on the other hand are often associated with other topics or candidates. Mélenchon is frequently mentioned in combination with the keyword Ukraine.
Sentiment – Candidate Analysis
The Sentiment Analysis (SA) aims to determine the attitude of the speaker/writer with respect to the overall contextual polarity of a document. In this analysis, only French media from the last month is being monitored.
The data indicates that Macron is perceived most negatively, with only 13% positive, but 39% negative mentions. Le Pen, on the other hand, is comparatively more likely to be mentioned neutrally, as her mentions are 9% positive and 14% negative. Lastly, Mélenchon appears to be the most polarising: he has the most positive mentions with 19%, and almost as many negative mentions as Macron with 37%.
Presidential Elections in France - Outlook
Although most of the polls predict a run-off between Macron and Le Pen, an alternative could still be possible: the far-left candidate Jean-Luc Mélenchon has experienced a growing favoritism in the polls recently, while some polls even see him above the 15% mark. If the radical right-wing candidate Éric Zemmour wins over more voters and thus inevitably takes votes away from Le Pen, Mélenchon is predicted to have a fighting chance of making it into the run-off. It remains to be seen who will be up against Macron in the second round. However, the chances of another candidate becoming the new French president appear to be rather slim.
HENSOLDT Analytics will continue to monitor the media
landscape around this issue.
This is a brief and unbiased overview of the data collected by the HENSOLDT Analytics system and does not represent the entirety of the information gathered on the topic thus cannot be treated as singular and final analysis of the event(s).
