1、Community Experiences of Tropical Cyclone Alfred Natural Hazards Research Australia 2025Dr Julian MarxLecturerThe University of MelbourneA Social Media Analytics ApproachCyclone Alfred&Social MediaDataStage#FacebookInstagramXStage 1(keyword+group-based tracking)10 groups/event pages267 posts458 comm
2、ents2,307 posts254 comments26 posts90 commentsStage 2(account-based tracking)18 agencies(37 accounts)450 posts1,500 comments300 posts1,200 comments200 posts800 commentsMethodology and High-Level FindingsTopic Modeling(automated)Thematic Analysis(manual)Bottom-upTop-downTemporal bracketing(automated)
3、#1:Requesting crisis-related information#2 Sharing crisis-related information#3 Support and Wellbeing#4 Humor as a Coping Strategy#5 Turning to unofficial information sources#6 Exploiting the crisis#1 Emergency warnings#2 Household and business preparedness#3 Trusted information sources#4 Insurance
4、and financial preparedness#5 Extent of damage#6 Initial recovery and clean upFindings“legacy”communicationsofficial sourceshigh information densitytext-heavy,static visuals medium-targeted,but low reachFindingsalgorithmic audiencingVarious sources(influencers,media companies,community members)low in
5、formation densityshort-form video contentlow-targeted,but extensive reachFindings”dark”socialClosed and semi-public community spaceshyper-localised and community-based targetingImplicationsThere is a gap between how EMA social media and how communities consume itAlgorithmic Algorithmic AudiencingAud
6、iencingDark SocialDark Social(Riemer&Peter,2021)(Marx et al.,2024)“Legacy”“Legacy”CommunicationsCommunicationsStrategic Recommendations Explore visual,short-form content(understand algorithms!)E.g.Instagram Reels,TikTok,Stories Partner with trusted community influencers Collaborate with group admins