The document summarizes discussions from a weekly meeting about analyzing social media data using different tools. It discusses using SBS (presumably a data analysis tool) to generate insights from a news article dataset about topics related to Ukraine. It also explores using Power BI to visualize trends from a Twitter dataset related to Xinjiang province in China. Key findings from the Twitter data include the popularity of hashtags like #Xinjiang but few mentions of forced labor controversies, and most engagement going to media outlets rather than individual accounts.
2. Weekly Meeting 6
XN2 (Informatics Team 2)
Lingyu Hu, Zhaode Ouyang, Ziyan Yan, Zixun Zhou
Mar 2, 2022
3. ● Use SBS as an analytical tool to generate insights
● Explore how Power BI can pool multiple data sources together, visualizing data based on the
themes/patterns, sentiments, etc.
This Week
4. SBS
Dataset:
Time: Early Feb 2022
From: Event Registry
Type: News articles
Location: Worldwide
Topics: Ukraine
Language: English
5. SBS
Topic 4: Nato, Ukraine, Russian, U.S.
Topic 11: Production, supply, energy, job, money
6. SBS
We can see NATO, Zelenskyy, Putin, and Biden are care about Topic 4
Xi is more care about his Winter Olympic
Topic 4: Nato, Ukraine, Russian, U.S.
Topic 11: Production, supply, energy, job, money
Biden seems care more about Topic 11 than other leaders
Relationship of Keywords and Topics
7. SBS
Topic 4 and Topic 11 have the strongest connection
Topic 4: Nato, Ukraine, Russian, U.S.
Topic 11: Production, supply, energy, job, money
9. SBS
● “Russian” and “Ukraine” appeared most in the dataset
● Specific words do not appear too much
10. SBS
● Ukraine, Russian, and Putin care Topic 1 most
● NATO, Russian, and USA care Topic 2 most
● Xi and Biden care about Topic 5 most
11. SBS
● T6 has a strong relationship with T2
● T5 has the second strong relationship with T2
Conclusion:
● Ukraine, Russian and Putin care Winter Olympics most
● Russian, NATO and USA care potential military activities
● Xi and Biden care relationship with other countries
● Covid has stronger relationship with potential military activities
● Relationship in countries could influence the potential military activities
13. POWER BI
The Twitter Dataset from the People’s Republic in China- Xinjiang Province was published in December 2021. Since
Twitter is officially banned in mainland China. Only a handful of accounts (2048) were registered in the region. There
were in total of 15635 Tweets send in the year 2021. Due to the nature of a small sample posted. Only a small amount of
Tweets were liked. It would be difficult to judge the authenticity of the posts but only a handful of users’ posts were liked
based on the data given.
14. POWER BI
When it comes to retweets of the information. #Xinjiang is retweeted the most indicating the increased attention to the
matter. 207 retweets among 1983 retweets indicating that Xinjiang is still one of the biggest attention onto the matter.
15. POWER BI
When it comes to comparison between the Like count and the hashtag counts. It’s clear that most of the hashtags were
created by the media outlets with Xinjiangonline as one of the media outlets drawing the most likes. Other relevant factors
such as Xinjiang being a nice place and Uyghurs being liked for the price. One of the obvious controversies of the Chinese
government forcing Xinjiang laborers to harvest cotton did not receive much attention with only 4 likes in the total posts.
16. POWER BI
The following count indicates similar pattern in media outlet with a small handful of people being followed showing
quality content while other outlets receiving a very small amount of attention.
17. POWER BI
Based on the above analysis, it’s clear that despite the Xinjiang Issue as one of major attention by the international society,
most information from Twitter didn’t draw a conclusion on the forced labor issue in Xinjiang. Rather, general perception
toward such trend is showing the nature of Xinjiang and a handful of fake news despite being spread by a few accounts did
not receive much attention in the whole picture.