#JoinTheCountDown (10.10.2020)
Count Down is a global initiative that aims to engage communities globally to help cut greenhouse gas emissions by half by 2030. On October 10th, 2020 a series of 600 events ran Globally across the world on the same day in the different countries in Figure 1. Although the events are under the umbrella of TEDX, actually the team is composed of more than 50 organizations that partner in this initiative (sample of the partner organizations: Youtube, BCG, WWF, and The Weather Channel). Furthermore, public figures such as Prince William, Duke of Cambridge spoke in the opening video to share the importance of the cause.
The objective of this task is to evaluate and analyze the reaction of the crowd to that Global event. Twitter was used to promote this global event to encourage people to join. The Slogan of the event was #JoinTheCountDown. So, that is why Twitter was used to retrieve tweets from this hashtag and see what could be inferred about the reaction of the crowd.
There are 3 motivations behind working on this topic. First, this event has a strong cause as it is related to critical issues that impact the sustainability of this planet, and analyzing the reaction of the crowd may reflect the existence of gaps between organizations that are trying to find a solution and the public who may have a negative impact on the planet if they are not familiar with the severity of the problem. Second, this is a unique event in terms of having it in different places at the same time. So, analyzing that hashtag may provide a global perspective which is hard to attain in other cases. Finally, That event was launched Oct 2020 but will be running until OCT 2021, so analyzing the reaction of the crowd from Twitter which was used by them for promoting the event ad posting it, may be noticed by any of the event organizers and may add any value (even if minor), so that would be a minor contribution toward supporting the cause.
Network Structure
Tweets were collected using NodeXL Pro. The query was set to retrieve 5000 relationships (Tweets, Mentions, Replies), but only 1369 were retrieved, as the hashtag is new and doesn't have more than that. The Network is composed of 520 nodes and 1369 edges. Figure 2 shows the big picture view for the network before grouping or trying to find clusters. Figure 3 shows the top nodes with respect to in-degree, page rank, or betweenness centrality. Each of these top nodes was represented with their account image. Some of the notable public figures interacting about this event are Alexander Verbeek who is a Dutch environmentalist and former strategic policy advisor at the Netherlands Ministry of Foreign Affairs. António Guterres, a Portuguese politician and diplomat serving as the ninth secretary-general of the United Nations. Nigel Topping who is CEO of “We mean Business” and was appointed as the UK climate action Champion in January 2020. Figure 4 shows the edges of the 15 vertices in Table one in a different color (Red) to reflect the spread and level of interaction with respect to the nodes surrounding them.
Content & Sentiment Analysis
Figure 5 shows the top 4 clusters detected based on the Wakita-Tsurumi algorithm, which is based on the modularity metric as a measure of division in a network. Group 1= Blue, G2= Orange, G3 = Pink, and G4 = Green.
The most repeated words by clusters:
G1: Change(20 times), Join the countdown(15 times), Climate and Crisis (each 13 times).
G2: Most of the words were either non-English names or words. The only recognizable word is “Eco-Warriors” and it appeared 5 times.
G3: ”#tedx and #manipal”(each 41 times), “Join the countdown and climate change “( each 40 times), “thank” (39 times).
G4: “hong kong” (5 times), “#sdgs” (5 times), and “#esg”(5 times).
Figure 7, may give an indicator that there is a positive sentiment toward the event, which is true, but after checking the keywords and the tweets from which the “Negative” sentiment was implied, it turned out that the tweets are pointing out the severity and the crisis that the planet is in. So it is not a negative sentiment about the event but mostly referring to the challenges related to Climate or Environment. Some of the positive words were (“Thank”, 43 times), (“Resilient”, 22 times), (“Brilliant”, 12 times), and (“Inspiring”, 10 times). While some of the negative words are (“Crisis”, 25 times), (“Fear”, 6 times), (“Catastrophe”, 5 times), (“Alarming”, 3times). As mentioned earlier, negative words are related to the situation and climate change and not the event.
Figure 6 shows the sentiment of the four clusters. Group 2 and 4 have less positive or negative words detected, which could be due to the language in which the tweet was posted, as all languages were kept since this is a global event. Also, many times, the users have a certain language on their profile, and then they tweet in English. Overall the positive sentiment is more in all 4 groups and as mentioned above, most of the negative words are more about the crisis and not the event.
Table 2 shows the top tweets in the entire network. Table 3 shows the most repeated words across the whole network, definitely, the name of the event was repeated the most (577 times). The Top mentioned accounts in the Entire Graph were: Tedcountdown(130 times), ted talks(58), tedxMahe-India(53), António Guterres(A Portuguese politician)(22), Maryam Pasha (17), and Ben Hurst who is a former rugby player in New Zealand(17).
Limitations
The data was retrieved via NodeXL pro and Twitter API, and there is a limit set by Twitter according to the number of records that may be retrieved. Furthermore, Protected tweets are not retrieved unless the users are in my network. So that analysis above covers the opinion and interaction of accounts that have public sharing settings, and also the segment of the population who have accounts on Twitter and active as well on Twitter. So the sampling may not represent all the population (specifically the users who do not have accounts on Twitter). All the data retrieved despite the language were kept, as the data set is small and filtering by English will exclude and not provide a correct view for the network. Also, that is a global event, so excluding non-English tweets will lead to biased results.
References:
Yuliana, I., Sukirman, S., & Sujalwo, S. (2017). A Comparison of Community Clustering Techniques: Fruchterman-Reingold and Wakita-Tsurumi. Proceedings of ISETH 2017 (The 3rd International Conference on Science, Technology, and Humanity).
Hansen, D., Shneiderman, B., & Smith, M. A. (2010). Analyzing social media networks with NodeXL: Insights from a connected world. Morgan Kaufmann.
