Abstract: Community Organizing in the Digital Age: An Exploratory Analysis of the 2017 #Womensmarch Using Deep Neural Networks (DNNs) (Society for Social Work and Research 24th Annual Conference - Reducing Racial and Economic Inequality)

Community Organizing in the Digital Age: An Exploratory Analysis of the 2017 #Womensmarch Using Deep Neural Networks (DNNs)

Schedule:
Friday, January 17, 2020
Mint, ML 4 (Marriott Marquis Washington DC)
* noted as presenting author
Maria Rodriguez, PhD, MSW, Assistant Professor, City University of New York, Hunter College, New York, NY
Heather Storer, Ph.D., Assistant Professor, University of Louisville
Gleneara Bates, MSW, Doctoral Student, Silberman School of Social Work, New York, NY
Sebastian Hoyos-Torres, Doctoral Student, City University of New York, NY
Jama Shelton, PhD, Assistant Professor, Hunter College, New York, NY
Background and Purpose: Social media has been indispensable in the emergence, development, and mobilization of global and local mass protest actions and sustained movements (see, for example Tufecki, 2017).  Digital Organizing is an emerging area of research and practice within macro social work which seeks to reconcile historical organizing theories and frameworks of community organizing with digital social movements of today. One key obstacle in the sub-field’s development has been the ability to compile and analyze the sheer volume and variety of data required to understand social media based social movements. While fields such as communications, political science, and sociology have harnessed the tools of computational social science towards this very end, social work is only now beginning to explore its possibilities. The current paper builds on this nascent work by performing an exploratory computational theoretical analysis one of the largest protests seen in history: the January 21st, 2017 Women’s March. The protest was a global response to the presidential inauguration of Donald Trump. 

Methods: Using the Twitter streaming API and twitteR package, the research team extracted 24 hours of the first #WomensMarch, resulting in over 4 million tweets. We then employ a deep neural network (DNN) algorithm, the most popular machine learning algorithm currently in use for nearly all types of prediction and classification of large datasets, to conduct sentiment and thematic analyses to understand the various ways in which tweeters employed the #WomensMarch hashtag on the day of global protest.

Results: Preliminary results indicate a variety of hashtag usage, from logistics and calls for legislative action, to trolling and plans for the next series of off and online actions. Primarily, we find that the #WomensMarch hashtag was used in three ways key to understanding it as the generation of a new social movement: 1) a way of connecting and mobilizing protesters (and would-be protesters) across the globe to local actions, 2)  drawing specific parallels between the feminist movement and the Women’s March as a continuation of that movement; and 3) discussions on ways to harness the momentum of the day towards sustained pressure on the highest elected office of the United States.  

Conclusions and Implications: Social movements in the 21st century are harnessing social media and concomitant technologies to build on the lessons of successful movements by, arguably, picking up where they left off. We find that social media based social movements may best be understood as transformational social movements, in that they seek to bring about the structural level changes sought by previous generations of movement organizers at simultaneously individual and collective levels.  At the individual level, social media based social movements  rely on individuals liking, sharing, commenting on tweets, toward the end of disseminating and improving messaging frames, recruiting members, and mobilizing individuals to action. At the collective level, these movement actions are broadcast on a global scale, inviting individuals within the collective (and the collective by virtue of the sheer numbers of people reached) to (re)examine their position on the issues at hand.