Conference Paper

Causal Impact Model to Evaluate the Diffusion Effect of Social Media Campaigns

Loading...
Thumbnail Image

Fulltext URI

Document type

Text/Conference Paper

Additional Information

Date

Journal Title

Journal ISSN

Volume Title

Publisher

European Society for Socially Embedded Technologies (EUSSET)

Abstract

Organized information campaigns on social media platforms have influence on collective opinions through processes such as social influence and majority opinion formation. Evaluating the effect of such campaigns has become a critical question. We proposed a method by first characterize user engagement and the semantics in public discussions with social media data, then apply a causal impact analysis to measure the effect. We conducted a case study to examine the effect of the 16 Days Campaign (a campaign organized by UN Women) through changes in public discussions of the MeToo, which is a related topic the campaign was aimed to impact. Results showed there were significantly more discussions in MeToo after the launch of the campaign. Hashtags on 16Days topics were used more and by more people. The proposed methods evaluate the direct and indirect diffusion effect of a campaign by quantifying the difference had the campaign not taken place based on social media data. The method enables to evaluate the overall outcome of collaborative work in a social media campaign.

Description

Yu, Xinchen; Mashhadi, Afra; Boy, Jeremy; Nielsen, Rene Clausen; Hong, Lingzi (2022): Causal Impact Model to Evaluate the Diffusion Effect of Social Media Campaigns. Proceedings of 20th European Conference on Computer-Supported Cooperative Work. DOI: 10.48340/ecscw2022_ep10. European Society for Socially Embedded Technologies (EUSSET). PISSN: 2510-2591. Exploratory Papers. Coimbra, Portugal. 27 June - 1 Juli 2022

Keywords

Citation

URI

Endorsement

Review

Supplemented By

Referenced By


Load citations
Please note: Providing information about citations is only possible thanks to to the open metadata APIs provided by crossref.org and opencitations.net. These lists may be incomplete due to unavailable citation data.