Introduction

If you run a search on https://www.kaggle.com on Namibia, you will only find four data sets. One of them, is for the tweets posted from major party candidates for the 2019 Namibia Presidential Election. Sounds interesting. Lets have a look.

You can download the data set here. For more detail you can also check out my notebook.

Disclaimer

I am not a member of any political party. The below information is for educational purposes and I take no responsibility on how you interpret the below information.

Main goal

My main goal was to see which candidate had the most responses per tweet. Not necessarily who tweeted the most or who received the most interactions and responses overall.  I had to do a few arithmetic but nothing major.

We will explore the following:

1. Total tweets and responses;
2. Interactions per tweet;
3. Interactions when specific word is mentioned;

Total tweets and responses

‍

From the above graphs you could infer the following:

1. President Hage Geingob got the most interactions from his tweets;
2. President of the Popular Democratic Movement, McHenry Venaani tweeted the most and
3. The that time independent candidate, Panduleni Itula, tweeted the least and got the least responses.

You would be correct, but let's dive a little deeper

Interactions per tweet

So even though presidential candidate,Itula has tweeted the least, he has more likes per tweets than the other candidates. President Hage Geingob has the most replies per tweet while him and Italu both has 30 retweets per tweet.

Interactions when specific word is mentioned in tweet

Below will display five graphs showing the interactions when each presidential candidates uses a specific word in their tweet. The words I chose were:

1. "youth";
2. "land";
3. "health";
4. "corruption"; and
5. "money"

You can copy the code on my kaggle notebook and enter any word you like to see how the numbers look.

Tweet Results

Itula got the most likes per tweet for all of the five words. He also got the most retweets when he included the words health, corruption and money. President Hage Geingob got the most retweets for the other two words: youth and land.

For fun let's look at the frequency of words used for all three candidates.

Now lets look at the word frequency per candidate.

Conclusion

The above information shows that you don't have to tweet the most to have the best quality interactions from the public. Also, best interactions on twitter will not necessarily result in presidency as President Hage Geingob won with a 56.3% vote.

I did not go too deep in the replies because that would involve reading them. Not all replies are positive so getting lots of negative replies cannot be seen as positive interactions. I hope you liked the analysis. Thats it for now.

‍