Skip to main content
placeholder image

Social Media User Archetypes during Extreme Events: Fuzzifying Dynamic Membership of Crisis Communication Participants (SMUCC)

Grant


Scheme


  • Australia Germany Joint Cooperation Scheme

Abstract


  • The joint research project consolidates dynamic and ongoing discussions on social media during enterprise crises. We plan to conduct case studies on impactful crises such as Volkswagen's "Dieselgate" or the 2011 Qantas industrial disputes. To investigate online conversations, current methods of sentiment analysis are not sufficient, as they are limited to a merely positive/negative polarity. We introduce an improved approach to this by incorporating new sentiment categories theorised using Systemic Functional Linguistic perspective. We apply those categories (e.g. appraisal, appreciation, judgement) to large datasets of enterprise crises, which we already gathered or will collect and process during the project. On this basis, we develop a prototype based on existing software to systemise this kind of analysis. Furthermore, we visualise our data analysis to derive practical implications for decision makers. Needed assets and skills are provided by both project partners: University of Duisburg-Essen: Social media analytics infrastructure (data tracking, processing), datasets, expertise on enterprise crises (i.e. prior publications, analysis methods, network of partners and access to the research training group "User-centered Social Media") University of Wollongong: Collaboration Laboratory (Co-Lab), visualization tools and skills, expertise in semantic sentiment analysis, (i.e. prior publications, partner network including the SMART Infrastructure Research Centre)

Date/time Interval


  • 2020

Geographic Focus


Sponsor Award Id


  • 57511909

Local Award Id


  • 131662

Scheme


  • Australia Germany Joint Cooperation Scheme

Abstract


  • The joint research project consolidates dynamic and ongoing discussions on social media during enterprise crises. We plan to conduct case studies on impactful crises such as Volkswagen's "Dieselgate" or the 2011 Qantas industrial disputes. To investigate online conversations, current methods of sentiment analysis are not sufficient, as they are limited to a merely positive/negative polarity. We introduce an improved approach to this by incorporating new sentiment categories theorised using Systemic Functional Linguistic perspective. We apply those categories (e.g. appraisal, appreciation, judgement) to large datasets of enterprise crises, which we already gathered or will collect and process during the project. On this basis, we develop a prototype based on existing software to systemise this kind of analysis. Furthermore, we visualise our data analysis to derive practical implications for decision makers. Needed assets and skills are provided by both project partners: University of Duisburg-Essen: Social media analytics infrastructure (data tracking, processing), datasets, expertise on enterprise crises (i.e. prior publications, analysis methods, network of partners and access to the research training group "User-centered Social Media") University of Wollongong: Collaboration Laboratory (Co-Lab), visualization tools and skills, expertise in semantic sentiment analysis, (i.e. prior publications, partner network including the SMART Infrastructure Research Centre)

Date/time Interval


  • 2020

Geographic Focus


Sponsor Award Id


  • 57511909

Local Award Id


  • 131662