Paper Title: What are People Talking about in #BackLivesMatter and #StopAsianHate? Exploring and Categorizing Twitter Topics Emerging in Online Social Movements through the Latent Dirichlet Allocation Model
Check out the working paper version on arXiv: https://arxiv.org/abs/2205.14725
Figure 1: Common topics between BlackLivesMatter and StopAsianHate, and unique topics in the two movements
Figure 2: Blacklivesmatter: Time Series for Google Trend, Tweet Volume, and Events. The red line illustrated the relative interest on Google for the keywords "black lives matter" in the U.S, whereas the blue line represents tweet volumes, the number of tweets per day, and the red line illustrates google trend scores (Timezone: Universal Time Coordinated)
Figure 3: Stopasianhate: Time Series for Google Trend, Tweet Volume, and Events (spearman correlation coeffient=0.683, p-value$<$0.001)
Figure 4: Sample Dataset
Figure 5: Time Series for Average Tweet Length (daily)
Figure 6: Overall Word Frequency Ranking
Figure 7: Time Series of Google Trend and Tweet Volume for #BlackLivesMatter and #StopAsianHate
Figure 8: Coherence Score of Different Number of Topics for #blacklivesmatter and #stopasianhate
Figure 9: Blacklivesmatter: Networks of top 50 co-occurring words in tweets
Figure 10: Stopasianhate: Networks of top 50 co-occurring words in tweets
