Can Mediaworks Predict the I'm a Celebrity 2020... Champion?
I’m a Celebrity… returns to our screens and is grabbing the UK’s attention once more. Last year, more than 10 million viewers tuned in each night on average, making it one of the most-watched TV programmes in 2019.
The 2019 series was one of the most-tweeted shows of the year too, and that trend looks set to continue.
Mediaworks will analyse a sample of 500 tweets about each celebrity every day to gauge public opinion on this year’s contestants. With this data, we can track which of I’m a Celebrity… 2020’s stars are the most popular.
Last year, Mediaworks’ social sentiment analysis showed the winner, Jacqueline Jossa was the most popular heading into the final. The data also showed eight out of the 9 contestants voted out were in the bottom three of the social sentiment on eviction night.
Will the sentiment scores in this year’s competition give us an indication of who will be crowned king or queen of the Gwrych Castle?
What Is Social Sentiment Analysis?
Social sentiment analysis identifies whether social media posts are positive or negative. Mediaworks will use AI technology to analyse 500 tweets for each I’m a Celebrity… contestant to understand their individual levels of online viewer support. From this, we’ll predict who will be eliminated from the show and who will come out on top.
What Is the Sentiment Score Based On?
The sentiment score evaluates the emotion shown in each tweet and attributes a score to the post, which is then automatically calculated as an average for that day. Individuals can score between -10 (extremely negative) and 10 (extremely positive). We will assign a score for each contestant every day based on new tweets to track the progress of their popularity.
Check out the latest Twittersphere reaction for each contestant below—this is updated daily based on the most recent episode. Click on each celebrity to compare their daily sentimental score with the rest of the camp:
The data presented by Digital Marketing Agency, Mediaworks is a sample and does not reflect the views and opinions of Mediaworks, nor is it representative of the entire population. The data presented is subject to third party machine learning algorithms, as well as human interpretation.