www.culturemachine.net • 1 THE RESEARCH POLITICS OF SOCIAL MEDIA PLATFORMS (original) (raw)
Related papers
Since Manovich's The Language of New Media (1999), the computational processes that make digital media possible have been discussed as media patterns that shape sociocultural dynamics. Apart from Manovich's structuralist point of view (Galloway, 2012), software and algorithms have appeared in the humanities as infrastructural elements of everyday life (Fuller, 2008) that condition our very existence (Kitchin and Dodge, 2011) as a material complexification (Thrift, 2005). They encode a neighborhood of relations that “trace contemporary production, communication, and consumption.” (Mackenzie, 2006, p. 169). Specifically the notion of algorithm, as a powerful logic of command and control for solving computational problems, is at the core of the discussions about current smarter digital media environments. It is part of a rising sociotechnical phenomenon: the usage of personalization mechanisms in digital media. Since 2009 when Google changed search paths to return more contextual results, many other web services have adopted this strategy, which is generally associated with more intelligent systems that deliver a better user experience. These mechanisms represent a complex data-driven process based on algorithms that seek to predict user preferences, as well as to show “relevant” advertising. This highly mediated environment is impossible to disable. In digital media company discourse, personalization systems are generally evaluated by their capacity to keep people logged into a website, and the results of personalization processes are described in subjective terms including “better,” “right,” “genuine” and “important.” Thus this paper explores the specific case of the Facebook News Feed, investigating the ongoing, unstable and conflicted relationship between personalization systems and user’s disruptive interactions with these mechanisms. The Facebook News Feed creates a personalized flow of publications which promises to deliver “the right content to the right people at the right time so they don’t miss the stories that are important to them” (Backstrom, 2013, online). This standard functionality for all Facebook users is considered one of the most significant informational environments given Facebook’s wide user base (Pariser, 2011). The visibility constructed through Facebook News Feed is enacted within algorithmic processes in relation to a range of data such as their interrelationship with users (Bucher, 2012). However, this digital “right driven” standard pattern has encountered disruptive reactions which seek to act on its digital materiality. They may be seen in browser extensions which try to modify different Facebook News Feed features, such as the filtering process (Facebook Mood Manipulator), visual metrics of contents (Facebook Demetricator) or agreement buttons (I dislike button). They may even shut down the whole News Feed (Quiet Facebook, Kill News Feed and News Feed Eradicator for Facebook). These mechanisms show diverse but similar material-discursive practices (Barad, 2003) which subverting the Facebook News Feed. Along these lines, this study seeks to produce an in-depth description about dissonant forms of constructing a sociotechnical narrative. Understanding these can contribute important notes about the Facebook News Feed’s agency and, more generally, about a more complex relationship between technical objects’ agencies and users’ disruptive practices. This paper is part of my ongoing PhD research on the agency of personalization systems in social network sites. Ethnographic strategies form the basis for the construction of the fieldwork(Ardèvol and Gómez-Cruz, 2014; Hine, 2015). Combining the Multi-sited Ethnography (Marcus, 1995) with material semiotic strategies to conceive materially heterogeneous agency (Latour, 1992; Akrich, 1992; Law, 2009), I develop strategies to enable the production of deep and contextual understanding through the ethnographic description. Generically speaking, I feature it in four broad and interconnected movements of fieldwork construction: First, following the Facebook News Feed trajectory with a broad digital archeology; second, following controversies surrounding this mechanism; third, following everyday usage through observation participant seeking to understand how people make sense of their lives within the personalization processes (Hine, 2015); and fourth, following Facebook News Feed disruptive reactions, the methodological movement which is emphasized in this paper. Finally, I present some fieldwork notes, demonstrating that these browser extensions are constructed as an empowerment of subjects against the passive conception of the user enacted by these mechanisms. Simultaneously, the counter agency of these mechanisms provides insights about how Facebook enacts sociality and usage of its personalization features through the idea of “right content, people and time”.
Experimenting with How Facebook’s Algorithm Works
Teaching Journalism During a Disruptive Age, 5th World Journalism Education Congress Proceedings, 2019
This paper presents reflective feedback on a teaching experiment conducted over three consecutive years by the authors together with students pursuing a Master's degree in journalism. The experiment brought together students and teachers from three universities belonging to the same international network of journalism schools (ARPPEJ, Association internationale de recherche sur les pratiques et la pédagogie du journalisme). Together, we investigated the way the algorithm of the social network Facebook organizes the distribution of news reports on its "News Feed." This experiment tried to reflect some of the challenges that journalism schools are currently facing in their attempt to grasp the specific nature of digital journalism (
The Research Politics of Social Media Platforms
2013
This is the first time the world has seen this scale and quality of data about human communication', Marlow says with a characteristically serious gaze before breaking into a smile at the thought of what he can do with the data. For one thing, Marlow is confident that exploring this resource will revolutionize the scientific understanding of why people behave as they do. His team can also help Facebook influence our social behaviour for its own benefit and that of its advertisers. This work may even help Facebook invent entirely new ways to make money.
The World of Edgerank: Rhetorical Justifications of Facebook’s News Feed Algorithm
2016
Web algorithms like Facebook’s so-called Edgerank algorithm play an increasingly important role in everyday life. The recent surge of research in such algorithms often emphasizes algorithmic orderings as powerful but opaque. In this essay, we propose an alternative reading of the Edgerank algorithm as a self-justifying ordering of the world. Drawing on the pragmatist sociology of Boltanski and Thévenot, we examine Edgerank as not just a hidden logic, but a rhetor that actively constructs a rhetorical commonplace that can be drawn upon in order to justify the evaluations produced by the algorithm. We do so by examining three specific situations where the operations of Edgerank have been critiqued and defended: First, Facebook’s own response to the critique that social media produce echo chambers. Second, Facebook’s presentation of the main variables in the Edgerank algorithm. Third, social media marketing blogs about how to handle the algorithm in practice. Based on these events, we construct an ‘internalistic’ account of the rhetoric of Edgerank, opening for an exploration of its moral grammar and the world or dwelling place it assumes and enacts. We find that the world of Edgerank is ordered according to recent engagement, which means it has affinities with what Boltanski and Chiapello have termed the connectionist world. At the same time, the world of Edgerank is marked by a tension between authenticity and automation that is a result of the algorithmic standardization of relations. In the rhetoric that comes with Edgerank, this tension is not something to be overcome, but rather a self-justifying hybrid, which points to a potential displacement of moral grammars in an age of computational valuation.
Engaging information: How targeting creates more comments but less likes on Facebook
Annual meeting of the Association for Education in Journalism and Mass Communication
In the present study, I investigate the eff ects of targeted Facebook posts on audienceengagement. Conducting a content analysis of 1536 Facebook status updates, I find that targeting is related to an increase in the amount of comments, but a ffects the number of likes negatively. To better understand this result, I also explore two variables that potentially affect the likelihood of a newspaper using targeted Facebook posts: Circulation and social media use in the newspaper's print community. Based on the results, I discuss implications for the future measurement of engagement, stimulating contributions to online communities, and targeting information to specific user groups.
How Facebook Algorithms Show Posts on Users Pages: Survey Study
International Journal of Computer Science and Mobile Computing, 2017
We may have heard about so-called Facebook algorithms for the emergence of posts or news, and did not understand at the time what these Facebook algorithms mean? What are the most important variables that depend on it? How does it affect you as a business owner or as a person seeking profit through Facebook? And then how do you deal with it to get the highest visibility of your page's publications in front of members? In this article, we will answer these questions through which you will learn how to display posts in front of Facebook users. Then you will learn how to manage your Facebook page so that you get the best results through it.
Explaining the News Feed Algorithm: An Analysis of the " News Feed FYI " Blog
Facebook uses algorithmic curation—automated selection and ranking of content—to present a personalized News Feed to each user for consumption. However, the News Feed user interface provides little information to help users understand how the ranking algorithm works. We analyzed the company's " News Feed FYI " blog series to better understand the degree to which Facebook employs " how " and " why " explanations of its News Feed algorithm. These types of explanations have been used in other recommendation and intelligent systems as a means of promoting user understanding and acceptance. Our findings show that the " News Feed FYI " blog posts focus more on explanations that justify why the algorithm works the way it does, and less on explanations that describe how the system works. These findings suggest that the " News Feed FYI " series would be more helpful for increasing users' confidence in the system, but not improving their trust in the system.
Sharing News Online (Martin/Dwyer), 2019
Studying the uses of language in social media news sharing is central to understanding how its industrial ecology operates, and how its actors and processes might influence the production of news journalism. Google and Facebook have built empires on the acquisition of user-generated text, and the identification, patterning, and commodification of keywords in our searches, conversations and annotations. In those cases, the words we most often search for or discuss have intrinsic value to advertisers, based on their capacity to capture our browsing attention. News, too, has historically been shaped by particular rhetorical strategies, packaged with compelling headlines, summary leads and concise, active language that demands the reader's attention. Yet, little public research has been done on the language of most shared news. For our Share Wars partners, it was critical to explore what linguistic factors people value in the communication of news-and whether these conform to, or diverge from, the contemporary language of news and the traditional 'news values' that guide journalists in filtering and selecting information that will be newsworthy-that is, timely, relevant and interesting in other ways to their audience. From an industry perspective, if the words we share online most often emphasise particular criteria, like the unusual aspects of an event or positive emotional reactions, then these may be indicators as to how journalists could shape their stories for greater shareability. Here linguist Monika Bednarek, a co-investigator on the Sharing News Online team, uses computational techniques to examine how the language of a set of most highly shared stories constructs newsworthiness, providing insights into how news might be packaged for an era of social sharing.