Prayers : The graph of Sympathy

Maybe people pray for something when they really need it.
People ask others to pray when they want to share the issue.

Prayers is the final project of Social Data Analysis class at ITP NYU. I saw some people asked others to pray for something in Twitter. I wanted to figure out what twitter users ask other twitter users pray for and when/why they want to share the issues.

For the analysis, I pulled Twitter data from 11/20 to 11/29 and searched tweets including the words “pray for”.

Co-tag network graph : what people asked others pray for?

Center of the network graph

The network graph of various incidents happening all over the world

This co-tag graph shows what people asked others to pray for in Twitter. If users mentioned #prayfor with other hashtags, I connected them. I also defined communities based on modularity and marked different colors per each community.

Some of issues in the graph
- 12-year-old Tamir shot by police officers in Cleveland(11/22/14)
- 18-year-old boy Joey died by car accident before Thanksgiving
- Trayvon Martin was a 17-year-old who was shot by a nightborhood watch voluteer in 2012
- Soccer player Pele is in hospital
- Lost Malaysia Air plain

The result is one big network graph only showing serious events happening all around world. We can see different social issues in the world such as gun accidents, illness, protests happening or mentioned during research period. Since Ferguson was a big issue, many similar accidents are shown in this network graph. Also, we can see Malaysia with Ulama hashtags and Mexico protest in the graph as well. Below image shows some tweets about these issues.

Ferguson and Neymar

Below is the left part of the network graph. There are many nodes related to political protests near Ferguson node, which makes sense. We can find a group of hashtags about Brazilian soccer player Neymar because there was a soccer game (Brazil vs Austria) that Neymar came out on Nov. 18.

Collocation Analysis

Collocations are expressions of multiple words which commonly co-occur.
I found below words those are frequently co-occur with “prayfor” in tweet texts.

makki malaysia; malaysia weloveulama; naruddin marbu; pencintarasulallah; shiekh naruddin; weloveulama pencintarasulallah; daughter shell; frnds kindly; sinusitis todayjazak; biyaali frnds; indigenous gospelfilms; orphan peoples; remaining groups; good health; kindly pray;

Most words are also shown in co-tag graph which is composed of hashtags. Since people usually use hashtags to emphasize the main topics of their tweets, these words were duplicated to hashtags.

Timeseries

This image is Timeseries graph to show the trend of top rank hashtags. It doesn’t count how many times these hashtags are mentioned with “prayfor”. You can see the number of Ferguson (purple) hashtag keeps growing.
People used “Malaysia”, “WeLoveUlama” hashtags a lot from 11/20 to 11/22. Ulama means a teacher and a religious leader of the society. People talked about Shiekh Naruddin who is a popular Ulama in Malaysia. There was a social issue in Malaysia and people tried to promote him. In fact, people did a campaign in Twitter during this time period.

The Power of marketing campaign

MoeMoneyEnt is the one of the music companies and they released a new music called “PrayFor me” at 11/24. They sent plenty of tweets at 11/24 as a part of promotion. MyFarSinger is a Indonesian TV program about singers’ competition. People sent many tweets to encourage others to vote to the favorite, Kembalikan Semula in EwalAF 2014. We can see from second image of tweets that his fans were really asking others to pray for him in Twitter.

Even if people used these hashtags a lot, they are not related to pray. That’s why they don’t show up on previous network graph based on co-occurrence and relationship between pray for and the words. They are mostly one part of marketing campaign and are not serious social incidents, which is different from hashtags we can see in network graph.

Conclusion

I believe that asking others to pray in Twitter means that people want to share the issues with as many people as possible. Also, most of the issues were not personal because of accessibility and they were serious incidents such as gun incidents and big protests happening all over the world.
Since twitter is one of the good tool of marketing campaign, some hashtags show mass cultural events.

What’s next?

I will create a website to visualize a map of the prayers so that people can navigate a graph and see the tweet texts as well.


If you have any questions and comments, please email me (jiwon.yoon@nyu.edu)

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.