French Election ‘17 (Mar 3-7): Initial Analysis

Introduction

Right Relevance (RR) provides curated information and intelligence on ~50 thousand topics. This includes:

  • Topic relationships including related topics & semantic information like synonyms.
  • Topical influencers (~2.5M) with score and rank.
  • Topical content and information in the form of articles, videos and conversations.

Additionally, Right Relevance provides an Insights offering that combines the above Topics and Influencers information with real time conversations to provide actionable intelligence with visualizations to enable decision making. The Insights service is applicable to emerging events like elections, conferences, product launches, breaking news developments, outbreaks like Ebola etc.

This report is a summary of graph analysis of engagements and conversations including retweets, mentions and replies of tweets from March 3rd to 7th on terms related to the French Presidential Election 2017.

Analysis Risks:

  1. This is our first effort at non-English so there are several learnings involved.
  2. Twitter uptake in France isn’t to the scale in USA and UK. Analyzed correctly, scale helps smoothen biases and errors. Scale still seems sufficient for analysis.

Data & Duration

The report leverages tweets sampled from Mar 3rd to Mar 7th 2017 and along with Right Relevance topics, topical communities’ and articles data form the basis for the analysis.

The phrases used for gathering tweets are: “frenchelection”, “frenchelections”, “french election”, “france election”, “#france2017”, “#Presidentielle2017”, “#Frexit”, “Fillon”, “FrancoisFillon”, “Le Pen”, “LePen”, “MarineLePen”, “#Marine2017”, “#hamon2017”, “Hamon”, “BenoitHamon”, “macron”, “#macron2017”, “Melenchon”, “Yannick Jadot”

Most of the summary report is extracted from the analysis collateral in the form of:

  1. Tableau Online Dashboard
  2. Gephi Communities Graph Visual: Extracts are shown below.

For access to the complete graphs please send email to biz@rightrelevance.com.

The analysis methodology is outlined at http://54.244.44.22/insights

Communities Graph

Community detection graph algorithms like Walktrap and InfoMap are used to identify communities (as sub-graphs) in our engagements graph built using Neo4j & R. Graph visualizations are done via Gephi.

All Engagements Graph

The communities’ graph (Fig 1) for the French Election, unlike our US election analysis and Brexit analysis graphs, shows multi-polarity instead of the stark bipolar split witnessed in the other two analyses.

 

Figure 1: All Engagements ‘French Election’ Communities Gephi Graph

In the all engagements graph (includes mentions), the purple/blue/pink colors mainly represent the right of center to the far right spectrum. Francois Fillon (purple) and Marine Le Pen (blue) communities are on this side as one would expect. The green/yellow/orange mix represents the center, left of center and far left. Emmanuel Macron (green), Benoit Hamon (yellow) and, to a lower extent, Jean Luc Melenchon (orange) form communities to the left of center. It is visibly smaller in size showing lower engagements.

Some basic stats from the all engagements graph:

Community/Subgraph

Percentage Nodes

Percentage Edges

Fillon (Purple)

34%

42%

Marine Le Pen (Blue)

8.5%

4%

Macron (Green)

19%

7%

Hamon (Yellow)

6.5%

2%

Mélenchon (Orange)

4.5%

2%

 

Points of note:

  1. The visibly lower size of the graph towards the left suggests lower excitement on the left of center.
  2. This noticeable difference could also be attributed to the large and seemingly successfully rally on March 6th by Francois Fillon, the French Republican party candidate for the presidency. Since data is from March 3rd to 7th, the Fillon rally injects a significant skew. Twitter shows good traction for the rally. Does this prove to be the moment signaling the end of the backward spiral of Fillon’s candidacy and reignite his campaign providing stability and forward progress vis-à-vis Emmanuel Macron and Marine Le Pen?
  3. Another interesting thing of note is the much larger than usual overlap in the center between the left and the right. We’ll discuss this further in context of the RTs-only graph.

ReTweets-only Graph

The retweets-only graph (Fig 2) is fairly interesting too.

Francois Fillon again dominates the graph due to the data having a big skew from his March 6th rally. It reinforces the impression about the traction and overall positive benefit of the rally for Fillion’s candidacy. It remains to be seen if Fillon can continue to build on the regained momentum.

Fig 2 shows the ReTweets-only Gephi graph.

Figure 2: RTs-only ‘French Election’ Communities Gephi Graph

The retweets-only graph (Fig 2) is usually much better at bringing out the differences. Stark and clear communities can be seen for the 5 leading candidates in this election.

Some basic stats from the RTs-only graph:

Community/Subgraph

Percentage Nodes

Percentage Edges

Fillon (Green)

22%

33.5%

Marine Le Pen (Blue)

9%

7%

Macron (Yellow)

6.5%

4.5%

Hamon (Pink)

4%

2%

Mélenchon (Orange)

5.5%

5%

Mixed (Violet)

27%

13%

 

One extremely interesting thing to note, is the fairly large sub-graph in the center of the RTs-only graph. This matches the big overlap we noticed in the mentions graph. One reasonable way to interpret this data is that there are a large number of undecided voters in the center, slight right of center and much more towards the left of center. This represents a large number of voters that can change the outcome in the first round by either not showing up or making up their minds by events closer to the voting day. This isn’t very dissimilar from the point of both Brexit and the US election, where voter no-shows were critical in deciding the eventual outcome.

Two extremely interesting accounts; @PrisonPlanet (Paul Joseph Watson, Infowars Editor-at-Large) and @vivelafra (ViveLaFrance); caught our attention. These accounts were extremely active and, we believe, effective in both the US election (in support of Trump) and UK’s EU Referendum (in support of Brexit). They’ve thrown their support and media skills behind Marine Le Pen in this election.

In Progress

The following analysis would be added in the next few reports:

  1. Influencers-only graphs
  2. Most trusted and verifiable sources of information for the French election based on Right Relevance influencers information.
  3. Sub-communities and the leading users driving each sub-community as we dissect Flocks from our Tableau tables and charts.

 

Write a Reply or Comment

Your email address will not be published. Required fields are marked *