/geotweets

Primary LanguageJupyter Notebook

Lab 2: Geo-tagged Tweet Collection And Visualization

Name: Joshua Zhang

Course: Geog 458

Section: Section AC

Date: Aprill 22 2022

This github repository is a work based on guidelines from Geog 458 Advanced Digital Geographies created by Professor Bo Zhao from University of Washington.

The specific instructions for making this work is on this page.


The Purpose of Comparison

The topic of this work is comparing the tweets with geographical location tags from two areas with the same roots of culture which developed into their own paths. Considering the requirement of doing analysis on English speaking areas, I choose the whole USA territories and part of the Western European countries(because English isn't the only dominant language in the area). The differences between the spatial patterns on the tweets from two areas can show the shift of culture on space after two centuries. After that, beside covering the Hawaii, Alaska, and other states in USA, the tweets also include Western European countries include countries within (35.3761155416293, -25.835924951807126) south west corner and (70.78168810314003, 29.535168417107514) north east corner.


The Maps

USA tweets at 3pm(Pacific Time Zone GMT-7)

image map of USA tweets

West European tweets at 00:00(Central European Summer Time GMT+2)

image map of European tweets

Comparision of two maps

In the first USA tweet map, the area with the most tweets is around east coast area and southwest corner. It can be considered as result of the popular density when it's interesting to see the pattern of middle states also concentrates on the eastern part of the country. Considering the east coast is 3 hours ahead of the west coast, there is a possibility the eastern population spends more time on twitter because they have finished the work of the day. Also, California and Washington state twitter users might spend their time on twitter for connecting with international business because their locations have more techinical companies. There are patterns on different areas in USA, but the overall usage of twitter is covering the whole country even in the Alaska and Hawaii.

In the West European map which is based on tweet data at the same time of USA tweets(3pm at Seattle is the midnight of Berlin and Paris), the data shows English tweets are not as influencial as the ones from USA. Still, the most tweets are made by people near to the coasts. There are hardly people using twitter with English speaking tweets in these countries. At the same time, the most tweets are concentrated at the British territory. Since large numbers of the Europe don't speak in English, the tweets are most likely made by people on vocation on business trips. This can also explains why the tweets are made in coasts ------ cities around the coast area have more people deadling with international affairs. Also, there are fewer and fewer tweets made if you go the east area. This is a pattern of ripples. Thus, large part of English tweets are in Britain when small amount of the tweets are from coastal areas of West Europe.

The differences between the two maps come from the popularity of English in these two areas. In USA, the frequency of English tweets depends on their own use of free time which implies the preference of social media and spread of population. In West European countries, the English tweets are concentrated at Britain and spread out to the rest of Europe like a ripple which means fewer influences in the area far away from Britain. The purpose of using twitter in the latter is possibily less indvidual than American twitter.


The word clouds

The word cloud of USA tweets

word cloud of USA tweets

The word cloud of West European tweets

word cloud of West European tweets

Comparision of two word clouds

In the word clouds, there are obvious usage of specifc words in each cloud. The USA tweets have a high frequency of using "Job", "See", "Out", or "Time". It's easy to combine them with other keywords such as "Work", "Look", "Go", or "Take". For West European tweets, the popular keywords are "Thank", "Great", "Love", or "Day". There are similaries on key words in active movements such as go out and see something. This implies the purpose of using twitter is to update one's own condition. But the difference in the most popular words is also indicating the major tweets's usages. In USA tweets, "Out" is a word opposing to the quarantine can demonstrate the influence from coronavirus is still negatively affecting American users. The American users seek the fresh air of outside when there are still necessary policies preventing them from the removing all the protection to virus. At the same time, work and job indicates the people are still in the time of working when it's 3pm to 6pm in America and midnight in West Europe. For European tweets, the policies must be weaker than the American tweets since the size of “Out" is much smaller. On the other hand, "Thank" and "Love" is overwhemingly huge in the middle of images. After looking at the specific content of tweets, they are functioning of communication between people. Most twitter users from this area use the word to say thank you to others in individual and business evenets. Also, it was Queen Elizabeth 96th birthday on April 21st 2022 and the postive words could be relevant to it.

There are similarities between tweet contents from USA and West European countries when it's easy to see the differences on the most popular keywords. The differences can imply the purposes of using twitters, time, and even specific events going on within the area. Word cloud makes the step of visualization these patterns easy to see and is helpful for us to find the direction of research when necessary.


The copyright of tweet data content are from twitter

(back to top)