First Monday

Information affordances: Studying the information processing activities of the core Occupy actors on Twitter by Sikana Tanupabrungsun, Jeff Hemsley, and Bryan Semaan



Abstract
Emergent social movements benefit from technologies that support the information activities that core actors perform to achieve the goals of the movement. Starting with an information-centric perspective, we use the theoretical lens of information processing, to examine the roles of Twitter affordances during the Occupy movement in solving information problems. Our data is a corpus of tweets from the core actors of the movement, as identified by our extended 1/9/90 rule. Using inferential statistics and network visualization, we show how different Twitter affordances act as mechanisms to resolve information problems in many different ways. Our network visualization shows the ecology of the interactions amongst the core actors, and between the core actors and the public. Our work contributes to the discussion around recent social movements on Twitter and advances the theory of information processing.

Contents

1. Introduction
2. Literature review
3. Method
4. Results
5. Conclusion

 


 

1. Introduction

Emergent social movements such as the Arab Spring and Occupy Wall Street have appropriated social media, like Twitter, which support and extend the action repertoires of an amorphous crowd. Social media have increasingly become a site through which people come together in solidarity to practice online activism in aid of social change. Drawing on Vegh’s (2003) typology of online activism, we suggest that the key mechanisms underlying and facilitating the crowd’s work are community organization — getting together under a common purpose — and information processing — the tasks that members perform to sustain and grow the movement.

Research suggests that the Occupy Wall Street movement provides an excellent case from which to study the roles of Twitter in mediating online activisms (Agarwal, et al., 2014; Hemsley, 2016). Twitter was a key platform in facilitating the community organization of the movement by serving as a key stitching mechanism. Specifically, Twitter stitched together a network of both human and technology-based networks (Agarwal, et al., 2014; Bennett, et al., 2014). The result is a network of networks (Hemsley, 2016) that enables information processing activities necessary for the members to move the movement forward. Yet little is known about these information processing activities, nor about the roles of Twitter affordances in supporting these activities.

Drawing on the theoretical lens of information processing in organizations (Daft and Lengel, 1986), we suggest that the information activities of the Occupy movement on Twitter are high in both uncertainty (lack of information) and ambiguity (ambiguous information) forces. Of our particular interest are those activities of the core actors, who were largely instrumental in moving the movement forward. Here, we define the core actors as those who committed a great deal of time and effort, and successfully gained attention from the public. These actors needed to engage in various information processing activities to solve the problems of uncertainty and ambiguity as a means to achieve the goals of the movement.

Using statistical methods and network visualizations, our results show that the core actors engaged in activities related to solving the problems of ambiguity more often than dealing with uncertainty. They accomplished this by using rich content features of Twitter (e.g., @mentions, retweets and hashtags) to make messages richer, thus reducing ambiguity. We show that the core actors tended to use multiple hashtags as a way to reduce ambiguity in a large network of networks by organizing information into sub-streams. The core actors also reduced ambiguity by selectively promoting tweets from the crowd through retweeting. We visualize the @mention network and show that only half of the core actors actually talked to one another, while the rest only interacted with non-core actors.

As our work is based in information processing theory, it makes a contribution by bridging organizational and behavioral studies on social media. We make a methodological contribution by offering a new way to study crowd information work on social media through the lens of information processing, and contribute to the growing body of literature concerning the emergent role of social media (Gerbaudo, 2012; Hemsley and Eckert, 2014; Starbird and Palen, 2012; State and Adamic, 2015; Tarrow, 2011) and specifically Twitter affordances (Hemsley, 2016; Hemsley and Eckert, 2014; Starbird and Palen, 2012), in recent social movements.

 

++++++++++

2. Literature review

2.1. Online activism: Social movements meet social media

Social media platforms, unlike traditional one-way communication channels, afford horizontal communication networks between like-minded actors across geographic boundaries (Hemsley and Eckert, 2014; Park, et al., 2015; Thorson, et al., 2013). Twitter has increasingly become a site through which people come together, under a common purpose, to engage in social movements (Gerbaudo, 2012; Hemsley and Eckert, 2014; Starbird and Palen, 2012; State and Adamic, 2015; Tarrow, 2011). This typically involves a range of actors, comprised of different constellations of individuals and organizations irrespective of geographic location, and is facilitated by technological affordances enabled by Web 2.0 technologies, such as Twitter and Facebook (Bennett, et al., 2014; Hemsley and Eckert, 2014; Park, et al., 2015; Thorson, et al., 2013).

The underlying organization of online movements often comprises a core-periphery structure (Borgatti and Everett, 2000; Hojman and Szeidl, 2008). Simply put, in such a network structure, “a small number of central agents or ‘hubs’ gather a disproportionate amount of connections, while most other agents maintain few relationships.” [1] Across a range of self-organized, emergent crowds, there often exists a small committed core of participants who have much influence within the network, whereas a large peripheral group also exists, though they tend to be far less active and are often viewed as being individually less critical to the movement’s success (Alba and Moore, 1978; Gomez Rodriguez, et al., 2010; Guo, et al., 2009). Thus, underlying a social movement is a networked community of individuals.

Research examining the Occupy movement has found that a community on Twitter was formed with a hub-and-spoke structure (Park, et al., 2015). Park, et al. created a small communication network of participants as indicated by either 1) interacting with an account @OccupyWallSt by @mentions or @replies; 2) using a hashtags #OccupyWallSt; or 3) simply including the term ‘OccupyWallSt’ in their tweets (Park, et al., 2015). They suggested that the network was comprised of small sub-networks, each of which had a core actor occupying the central position. That is, the network was occupied by a relatively small number of core actors whose roles were to serve as key informants, brokers of information, and bridges (Burt, 2004) between key network clusters.

Indeed, such core actors are instrumental in moving a social movement forward (Bennett, et al., 2014; Massung, et al., 2013; Park, et al., 2015). In this work, we are interested in the emergent online information activities of such actors from an information centric perspective, and the role of Twitter affordances in supporting their information activities. We operationalize the core actors as the highly-engaged actors, or those who dedicated a great deal of time and effort to the movement, and successfully gained attention from the public.

Specifically, our research questions are:

RQ1: In what ways did the Occupy core actors use Twitter to engage in information activities tasks?

RQ2: How did the mechanisms afforded by Twitter facilitate the information activities of the Occupy core actors?

2.2. Information processing in social movements

Drawing on Daft and Lengel’s (1986) work on information processing in organizations, we untangle the complex ecology of the information activities of the Occupy core actors on Twitter. In their work, Daft and Lengel identified two forces underlying information processing: uncertainty and ambiguity.

Uncertainty refers to a lack of information needed in order to act. The problem of uncertainty is resolved by providing more information. In the context of a social movement, actors may not know how many “boots” are going to be on the ground. As such, they may engage in various information activities in order to ameliorate this uncertainty, such as broadcasting information to inform or raise public awareness.

Ambiguity refers to when too much, or conflicting information exists. The solution for ambiguity tends to be using media with suitable richness to clarify the message. Daft and Lengel suggested that some media channels were richer than other channels. Face-to-face was the richest channel, and memos were the leanest. While Twitter probably ranks low in the richness scale, the platform does offer a range of affordances that can make any individual tweet richer than others. For example, the inclusion of a hashtag may help clarify the context of the message.

Whereas Daft and Lengel’s perspective is typically associated with the information needs of organizations, here, we apply their conceptual lens to the study of social movements. We reveal how the core actors utilized Twitter’s affordances to solve the problems of ambiguity and uncertainty as a means to move the movement forward. Before diving into their information processing activities, we present a discussion about the information needs of the Occupy movement through the lens of information processing.

2.3. Occupy on Twitter: A high uncertainty and high ambiguity network

The Occupy movement emerged in response to social and economic injustice, and became a worldwide movement (Conover, et al., 2013; Tremayne, 2013). As a movement centered around the core principle of creating a “leaderless democratic consensus” [2], Occupy was a crowd-enabled organization where activism practices were achieved via dispersed assemblies and digitally assisted communication infrastructures. For example, while face-to-face gatherings, camps, and offline events were critical to the identity of the movement, such practices were only part of the story (Agarwal, et al., 2014). That is, several dense movement networks emerged vis-à-vis the appropriation and uses of various technologies, such as SMS, e-mail and Web sites, as well as different social media platforms, e.g., Facebook, YouTube and Twitter (Agarwal, et al., 2014).

Due to the lack of formal bureaucracy, recognized authorities (leaders), or formal structures (Bennett, et al., 2014; Thorson, et al., 2013), we suggest that the information processing tasks related to the activism practices of the emergent Occupy network are high in both uncertainty and ambiguity forces.

One source of uncertainty was from the inclusiveness nature of the movement. Any actions were welcomed as long as they were in line with the core principles of the movement (Agarwal, et al., 2014). This inclusiveness could result in different sets of actors being under-informed about the actions of other actors. To solve this, the core actors needed to engage in information negotiation (Thorson, et al., 2013), which involved a large amount of information processing work. Even then, in a crowd-enabled network, protest activity outcomes are not easily predicted nor controlled, thus creating a high degree of ambiguity. Similar to other bottom-up or crowd-driven events, such as viral information events, protest activities are not easily controlled; by their very nature there exists a high degree of fuzziness (Nahon and Hemsley, 2013).

Another source of uncertainty and ambiguity resulted from the large amount of information needed to develop collective actions. Information flowed within and among city-based networks as well as up to and down from the national level (Conover, Davis, et al., 2013; Hemsley and Eckert, 2014). Being a multi-centered movement (Bennett, et al., 2014), the information circulated within and between each group differed, and could be ambiguous, particularly to outsiders or those at the national level, especially when related to the events in a specific physical location, e.g., camp evictions, and police brutality. Thus, for the core actors to solve the information problems on Twitter, they would sometimes need to reduce uncertainty by providing informative tweets, while at other times they may need to resolve ambiguity by posting rich tweets.

2.4. Twitter: A mediating tool for social movement

Twitter is a microblogging service that allows users to broadcast 140-character messages (tweets) to groups of other users who subscribe to their accounts (followers). Twitter offers a range of affordances, all of which support different information needs of the site’s users. In addition to text, users can include additional information resources i.e., URLs, that make them more informative and so reduce uncertainty (Bennett, et al., 2014). Referencing URLs is a common practice employed by Twitter users (boyd, et al., 2010). As users include more text or embedded features, there will be less room in tweets for other ‘rich content’ features. Thus, the users make a trade-off that emphasize informativeness over richness.

The rich content features work as cues to make a tweet richer and easier to interpret, thus reduce ambiguity. First of these, the hashtag convention is used as a way of creating thematic, contextualized conversations that are easily searchable (Golder and Huberman, 2006; Huang, et al., 2010; Marwick and boyd, 2012). Marwick and boyd (2012) suggest that Twitter users experience ‘context collapse’, meaning they often merge multiple audiences into one. Without knowing the context or topic of a tweet, the message can be easily misinterpreted. Hashtags allow users to label the tweets with relevant topics or keywords (boyd, et al., 2010), to direct tweets to specific streams (Huang, et al., 2010), and to signal the context within which the tweets occur (Huang, et al., 2010; Marwick and boyd, 2012). As such, the uses of hashtags reduce ambiguity by directing tweets to the intended audience and providing context necessary for the interpretation.

For users whose accounts are not explicitly set as private, their tweets are also posted to a public, searchable timeline, and can be retweeted by other users (boyd, et al., 2010). The retweet emerged as a technique for forwarding and interacting whereby Twitter users pass on tweets with attribution to the original author. By retweeting, users become part of the conversation and signal that they are listening to, acknowledging, or trying to curry favor with, the person who tweeted (boyd, et al., 2010). People also retweet to associate themselves with different organizations, communities or peers (Seidel, et al., 2016) and before they spread content they consider their audience (boyd, 2008). People are selective about whose tweets they retweet (Kwak, et al., 2010). Retweets are also regarded as a curation process, which includes the selection, promotion, and validation of the content. It signals what the crowd considers more valuable or less counterproductive contributions. Another way to think about it is that the crowd tends to filter out noise, spam, and misinformation by not propagating it, thus selectively signaling credibility of content and accounts (Nahon and Hemsley, 2013).

Twitter users began to use the @-sign preceding another user’s account name as a way of addressing tweets to a specific user. This convention is known as @mention and allows users to engage in one-on-one, yet still public, conversation (Honeycutt and Herring, 2009). The use of @mentions functions as a signal to tagged users that they are being talked about (boyd, et al., 2010), as well as a form of ‘addressivity’ to directly address, or indicate intended recipients of the tweet (Honeycutt and Herring, 2009).

These rich content features (e.g., @mentions, retweets and hashtags) reduce ambiguity by enhancing the richness of a tweet. For example, tweets with an @mention are interpreted in the context of addressing someone to specify the recipients. Likewise, hashtags, which often signal topical information, indicate that the message ought to be interpreted within the context of the given topic. Indeed, the 140-character limit of Twitter restricts ones from constructing a tweet that is both rich enough to reduce ambiguity and, at the same time, highly informational. Consciously or not, the core actors must make tradeoffs that emphasize either richness of informativeness. Part of this work explores whether core users lean towards reducing ambiguity or being more informative, which may help us understand the information needs of social movements like Occupy.

Twitter does not employ a mutual relationship-based dynamic. For example, user A can follow user B, but it is not required that user B follows user A. Twitter users receive tweets from the set of users they elect to follow. Thus, users can follow activists or organizations to receive timely updates. Twitter users with many followers are often considered more influential and more credible (Xu, et al., 2013). The number of followers, along with other measures such as retweets, essentially becomes a highly visible sign of Twitter status. This study uses these signals for identifying core actors.

Daft and Lengel (1986) argue that the goal of communication, or of an information processing task, is to reduce uncertainty and ambiguity. One way to reduce uncertainty would be by providing information, whereas in order to reduce ambiguity, a communication medium should provide suitable richness. As a communication medium, a measure of a tweet’s informativeness/richness would refer to its ability to alter understandings or communicating meanings that reduces uncertainty/ambiguity.

In this work, we think of the uses of informational features (tweet text and URL) as a way to enhance tweet’s informativeness, and rich content features (@mentions, retweets and hashtags) to enhance tweet’s richness. The 140-character limit of Tweets means that when users include numerous hashtags, e.g., #OccupyWallSt and #OccupySeattle, they are then more limited in terms of including informative features such as URLs. Alternately, the inclusion of a few URLs would limit how many hashtags users could include. While it is possible to include a URL and a hashtag in the same tweet, core actors cannot construct a tweet that is both highly informational and rich enough to reduce ambiguity, at the same time. They must make tradeoffs, consciously or not, that emphasize either richness or informativeness.

 

++++++++++

3. Method

3.1. Twitter data collection process

This study uses a corpus of tweets collected from Twitter’s streaming API which allows developers to collect real-time tweets from a specific set of keywords. Three researchers curated a list of 355 keywords. The terms included popular hashtags, keywords and Occupy city accounts. The collection includes tweets from mid-October 2011 to June 2013, and contains approximately 104 million English tweets from 18 million unique users.

3.2. Identifying the core actors of the Occupy movement

With our research questions around the uses of Twitter by the core actors of the Occupy movement, we first aggregated our tweet corpus by Twitter user IDs. To identify core and peripheral members, we relied on the 1/9/90 rule, an unofficial standard for analyses of user communities where tweet frequency follows a long tail distribution (Bruns, et al., 2013). Specifically, the rule sorts users by their tweet count and categorizes the most active one percent as lead users, the next nine percent as highly engaged users, and the remaining 90 percent as least active users.

In this study, we extended the 1/9/90 rule by including two other measures — engagement period and change in followers — to conform with our definition of the core actors. Specifically, we define core actors as those who dedicated a great deal of time (engagement period) and effort (tweet frequency) to the movement, and successfully gained attention from the public (change in followers). Specifically, the engagement period refers to the duration in which a user appeared in our collection, tweet frequency is the total number of tweets created by the user in our collection, and change in followers is the increase or decrease in the numbers of followers over the period that the user appeared in our collection. The change in followers can be seen as a curation process in which Twitter users help filter spammers, and selectively add credibility to users — that is, users who gain more followers are seen by the participants as being credible and valuable.

We identified the 99th percentile of each measure separately. Figure 1 presents the log-log plots of tweet frequency, engagement periods (in days), and changes in followers per user, of all users in our collection. This figure shows that, except for the engagement period, the other two measures follow the same long tail distribution. That is, most of the users did not tweet very frequently and gained only small number of followers. Interestingly, the engagement period distribution suggests that the majority of the users participated in the Occupy network for extended periods of time.

 

Distributions of tweet counts, engagement periods and changes in followers per user, of all users
 
Figure 1: Distributions of tweet counts, engagement periods and changes in followers per user, of all users.
Note: Larger version of figure available here.

 

Using the three criteria of tweet frequency, engagement period and change in followers being in the top one percent, we identified 364 core actors of the Occupy movement. In other words, for a user to be considered part of the core, they had to be in the top one percent of all three measures. Using multidimensional measures, we overcome the limitations of each individual metric. Specifically, the metric of tweet frequency could potentially include troll or bot accounts who flooded the Twitter streams with promotional or spam messages. The metric of engagement period could be in favor of those who only participated at the very beginning and the end, but not in between. The change in follower counts metric, by itself, would have the disadvantage of giving prominence to accounts that are already popular in other contexts, such as news channel and celebrities. Being the top users in all three metrics indicate that they were highly engaged actors who dedicated a great deal of time and effort to the movement, and successfully gained attention from the public.

Examples of our core actors are @doctorow, who described himself in Twitter’s profile as “Writer, blogger, activist. If you want a reply, use email. Blog suggestions here: http://boingboing.net/s” and @MiaFarrow, whose description is “Every time they try to vote in a bad law- RESIST. Call your House of Reps switchboard: 202-225-3121 — they will connect you. Persist”. Table 1 presents the descriptive statistics of our identified core actors.

 

Table 1: Descriptive statistics of the core actors.
 Min.MeanMedianMaxStandard deviation
Tweet count3771.29521,127272.5
Engagement period (days)34142340455854.25
Change in followers4434,218.495724,770680.47

 

 

Table 2: Descriptive statistics of the classified tweets per user, of our core actors.
 Min.MeanMedianMaxStandard deviation
Informational tweet010.39517316.83
Rich tweets067.08481,12476.84

 

Next, we examine the large corpus of tweets sent by our identified core actors. We break the tweets into two groups. Rich tweets are any tweets with @mentions or hashtags, or are either @replies or retweets. All remaining tweets are classified as informational, which would be either plain text or tweets with URL. In total, we have 28,199 tweets from 364 core actors. Table 2 presents the descriptive statistics of the counts of informational and rich tweets per user, of our core actors.

To address the question about the types of information activities of the core actors (RQ1), we use statistical methods to test if they lean towards reducing ambiguity or reducing uncertainty. We report the statistics and use visualizations to examine how the core actors utilize Twitter’s rich content features in their information activities (RQ2). Our deeper analysis focuses primarily on the word frequencies of informational tweets and the @mention behavior of rich tweets.

 

++++++++++

4. Results

In general, the core actors tend to have higher numbers of rich tweets, compared to their numbers of informative tweets. We perform a paired t-test to test whether the mean difference in tweet types of core actors differs from zero. The test statistic of 13.7 with 363 degrees of freedom associated with p-value of 0.01 indicates that the differences are significant. Thus, consciously or not, the core actors make tradeoffs that emphasize richness, thus reducing ambiguity, rather than providing more information through the information features (tweet text and URLs).

We look further into the corpus of informational tweets to get a sense of topics core users may have been trying to reduce uncertainty around. From 3,782 tweets, approximately 35 percent contains at least one URL and the average length of tweets is 77 characters. We generate a word cloud to visualize the kinds of content created by the core actors.

Figure 2 illustrates the top 150 words in the informational tweets after removing hashtags and keywords used for constructing the collection. The most frequent words by far are police and protester. We also see several city/country names such as Oakland, London and India. Another group seems to relate to critical issues such as arrest, dead, hunger and ill. Perhaps core users were informing activists about these critical issues within certain places. Additional work may shed light on this.

 

The word cloud visualization of informational tweets
 
Figure 2: The word cloud visualization of informational tweets.

 

Next, we examine the larger corpus of rich tweets with an eye towards understanding the ways core users were attempting to reduce ambiguity. From 24,417 tweets, approximately 32 percent are retweets, 53 percent contain @mention(s), and 46 percent contain hashtag(s). Note that we did not count @mention and hashtag in retweets.

Retweets: Our core actors created 7,785 retweets. Of those, only seven percent was retweets of tweets from the other core actors, none of them was self-retweet. This suggests that the core actors primarily retweeted the greater public. When we consider retweeting as a mechanism to filter noise, spam, and misinformation, retweeting the greater crowd would help resolve ambiguity in the information ecology of the Occupy Twitter network. Being retweeted by the core actors would add credibility to the content and help spread the message.

Hashtags: Of the 24,417 tweets from the core actors, 11,171 tweets contain at least one hashtags. On average, a tweet has two hashtags (mean = 1.9 and median = 2). Among these, there are 4,297 unique hashtags. The most popular hashtag is #ows, which was used 1,766 times. This hashtag co-occurred with #occupy 201 times, with #P2 151 times, and #TCOT 97 times. Interestingly, the last two hashtags represent two different political alignments. The #P2 hashtag is a liberal hashtag which stands for ‘Progressive 2.0’, and #TCOT is a conservative hashtag and stands for ‘Top Conservatives on Twitter’. This suggests that the core actors used multiple hashtags to organize information into sub-streams. Specifically, using the generic hashtag like #ows indicates that a tweet is relevant to the Occupy movement. Adding #TCOT would direct it to the sub-stream of conservative audience. Using multiple hashtags thus reduces ambiguity by de-collapsing the contexts and signaling how the tweets should be interpreted.

@Mentions: The uses of @mentions are particularly interesting because nearly half of the tweets contained @mentions. Of the 364 core actors, only 17 users did not use @mentions at all. In their attempts to reduce ambiguity through addressivity, were the remaining 347 core actors interacting with each other, the greater public, or both. Figure 3 shows that the @mention network of the core actors comprises of a large connected component. While all of the core actors (yellow nodes) indeed interacted with the greater public (green nodes), only half of them (48 percent) actually interacted with one another. We argue that this group of actors served as a key stitching mechanism (Agarwal, et al., 2014). That is, they stitched together sub-networks of the Occupy movement, and collectively formed a larger network. This kind of focused interaction — one actor directly addressing another — could have played a role in reducing ambiguity by making sure the right information went to the right place. Otherwise, there could be a cacophony of information with no clear way of indicating who ought to be paying attention to it.

 

The @mention network of our core actors
 
Figure 3: The @mention network of our core actors.
Note: Larger version of figure available here.

 

 

++++++++++

5. Conclusion

Social media platforms provide mechanisms that support and facilitate the core actors in engaging in the information activities necessary for the movements. In this study, we offer an analysis of how the mechanisms afforded by Twitter facilitate the information activities of the Occupy core actors from an information-centric perspective. Through the lens of information processing, we argue the Occupy network is high in uncertainty and ambiguity forces, and that the core actors need to engage in various kinds of information activities in order to solve the problems of uncertainty and ambiguity.

Using the extended 1/9/90 rule, we identified a small set of 364 users at the top one percent in three dimensions: tweet frequency, engagement period, and change in followers. We suggest that using the three dimensions helps us overcome the limitations of each individual metric. Being the top users in all three metrics indicate that they were highly engaged actors who dedicated a great deal of time and effort to the movement, and successfully gained attention from the public.

We then classified their tweets as either ‘rich’ or ‘informational’ and found that they created more rich tweets than informational tweets. We answer RQ1 that the core actors used Twitter to resolve ambiguity more often than reducing uncertainty. This implies that reducing ambiguity either required more work, or was a more important information processing task for the core actors.

We answer RQ2 by delving into each of the rich content features. We found that core actors primarily retweeted the greater public. We suggest that retweeting resolves ambiguity by filtering noise, spam, and misinformation by adding credibility to a piece of information. When a core actor retweeted a message, it signaled the greater public that the message was trustworthy and worth reading.

We found that the core actors tended to use multiple hashtags. We suggest that Occupy was a large-scale network and comprised several sub-networks. Using multiple hashtags would de-collapse the contexts by sorting information into sub-streams where the tweets belong. The uses of liberal and conservative hashtags, i.e., #P2 and #TCOT, along with the more generic #ows hashtag, are an excellent example.

The uses of @mentions accounted for half of the tweet corpus. We constructed the @mention network and found the formation of information network where all of the core actors interacted with the greater public but only half of them actually interacted with one another. We see this as an activity aimed at clarifying information or making sure that the correct information reaches the right people, thus reducing ambiguity.

We note a few important limitations. First, we only looked at the English tweets, as indicated in the meta-data. Although the Occupy movement had essentially emerged worldwide, we do not claim that our findings could be applied to the Occupy networks elsewhere. Also, because our paper analyzes Twitter data, we do not claim that our results are generalizable to platforms such as Facebook, or to other social movements. Although our work essentially is primarily theory driven, it does introduce a new mechanism through which scholars can explore online behavior in social media through the lens of information processing, especially with respect to the information activities of the crowd. An example might include uncovering information exchanges or identifying community leaders on social media during mass emergencies or large-scale crises. End of article

 

About the authors

Sikana Tanupabrungsun is a Ph.D. candidate in the School of Information Studies at Syracuse University.
E-mail: stanupab [at] syr [dot] edu

Jeff Hemsley is Assistant Professor in the School of Information Studies at Syracuse University.
E-mail: jjhemsle [at] syr [dot] edu

Bryan Semaan is Assistant Professor in the School of Information Studies at Syracuse University.
E-mail: bsemaan [at] syr [dot] edu

 

Notes

1. Hojman and Szeidl, 2008, p. 295.

2. Agarwal, et al., 2014, p. 648.

 

References

S. D. Agarwal, W. L. Bennett, C. N. Johnson, and S. Walker, 2014. “A model of crowd enabled organization: Theory and methods for understanding the role of Twitter in the Occupy protests,” International Journal of Communication, volume 8, at http://ijoc.org/index.php/ijoc/article/view/2068, accessed 22 January 2018.

R. D. Alba and G. Moore, 1978. “Elite social circles,” Sociological Methods & Research, volume 7, number 2, pp. 167–188.
doi: https://doi.org/10.1177/004912417800700204, accessed 22 January 2018.

W. L. Bennett, A. Segerberg, and S. Walker, 2014. “Organization in the crowd: Peer production in large-scale networked protests,” Information, Communication & Society, volume 17, number 2, pp. 232–260.
doi: https://doi.org/10.1080/1369118X.2013.870379, accessed 22 January 2018.

S. Borgatti and M. Everett, 2000. “Models of core/periphery structures,” Social Networks, volume 21, number 4, pp. 375–395.
doi: https://doi.org/10.1016/S0378-8733(99)00019-2, accessed 22 January 2018.

d. boyd, 2008. “Taken out of context: American teen sociality in networked publics,” Ph.D. dissertation, School of Information, University of California, Berkeley, at https://www.danah.org/papers/TakenOutOfContext.pdf, accessed 22 January 2018.

d. boyd, S. Golder, and G. Lotan, 2010. “Tweet, tweet, retweet: Conversational aspects of retweeting on Twitter,” HICSS ’10: Proceedings of the 2010 43rd Hawaii International Conference on System Sciences, pp. 1–10.
doi: https://doi.org/10.1109/HICSS.2010.412, accessed 22 January 2018.

A. Bruns, T. Highfield, and J. Burgess, 2013. “The Arab Spring and social media audiences: English and Arabic Twitter users and their networks,” American Behavioral Scientist, volume 57, number 7, pp. 871–898.
doi: https://doi.org/10.1177/0002764213479374, accessed 22 January 2018.

R. S. Burt, 2004. “Structural holes and good ideas,” American Journal of Sociology, volume 110, number 2, pp. 349–399.
doi: https://doi.org/10.1177/0002764213479374, accessed 22 January 2018.

M. D. Conover, E. Ferrara, F. Menczer, and A. Flammini, 2013. “The digital evolution of Occupy Wall Street,” PloS ONE, volume 8, number 5, e64679.
doi: https://doi.org/10.1371/journal.pone.0064679, accessed 22 January 2018.

M. D. Conover, C. Davis, E. Ferrara, K. McKelvey, F. Menczer, and A. Flammini, 2013. “The geospatial characteristics of a social movement communication network,” PloS ONE, volume 8, number 3, e55957.
doi: https://doi.org/10.1371/journal.pone.0055957, accessed 22 January 2018.

R. L. Daft and R. H. Lengel, 1986. “Organizational information requirements, media richness and structural design,” Management Science, volume 32, number 5, pp. 554–571.
doi: https://doi.org/10.1287/mnsc.32.5.554, accessed 22 January 2018.

P. Gerbaudo, 2012. Tweets and the streets: Social media and contemporary activism. London: Pluto Press.

S. A. Golder and B. A. Huberman, 2006. “Usage patterns of collaborative tagging systems,” Journal of Information Science, volume 32, number 2, pp. 198–208.
doi: https://doi.org/10.1177/0165551506062337, accessed 22 January 2018.

M. Gomez Rodriguez, J. Leskovec, and A. Krause, 2010. “Inferring networks of diffusion and influence,” KDD ’10: Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1,019–1,028.
doi: https://doi.org/10.1145/1835804.1835933, accessed 22 January 2018.

L. Guo, E. Tan, S. Chen, X. Zhang, and Y. E. Zhao, 2009. “Analyzing patterns of user content generation in online social networks,” KDD ’09: Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 369–378.
doi: https://doi.org/10.1145/1557019.1557064, accessed 22 January 2018.

J. Hemsley, 2016. “Studying the viral growth of a connective action network using information event signatures,” First Monday, volume 21, number 8, at firstmonday.org/article/view/6650/5598, accessed 22 January 2018.
doi: http://dx.doi.org/10.5210/fm.v21i8.6650, accessed 22 January 2018.

J. Hemsley and J. Eckert, 2014. “Examining the role of ‘place’ in Twitter networks through the lens of contentious politics,” 2014 47th Hawaii International Conference on System Sciences (HICSS).
doi: http://dx.doi.org/10.1109/HICSS.2014.233, accessed 22 January 2018.

D. A. Hojman and A. Szeidl, 2008. “Core and periphery in networks,” Journal of Economic Theory, volume 139, number 1, pp. 295–309.
doi: http://dx.doi.org/10.1016/j.jet.2007.07.007, accessed 22 January 2018.

C. Honeycutt and S. C. Herring, 2009. “Beyond microblogging: Conversation and collaboration via Twitter,” HICSS ’09: 42nd Hawaii International Conference on System Sciences, 2009, pp. 1–10.
doi: http://dx.doi.org/10.1109/HICSS.2009.89, accessed 22 January 2018.

J. Huang, K. M. Thornton, and E. N. Efthimiadis, 2010. “Conversational tagging in Twitter,” HT ’10: Proceedings of the 21st ACM Conference on Hypertext and Hypermedia, pp. 173–178.
doi: http://dx.doi.org/10.1145/1810617.1810647, accessed 22 January 2018.

H. Kwak, C. Lee, H. Park, and S. Moon, 2010. “What is Twitter, a social network or a news media?” WWW ’10: Proceedings of the 19th International Conference on World Wide Web, pp. 591–600.
doi: http://dx.doi.org/10.1145/1772690.1772751, accessed 22 January 2018.

A. Marwick and d. boyd, 2012. “I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience,” New Media & Society, volume 13, number 1, pp. 114–133.
doi: https://doi.org/10.1177/1461444810365313, accessed 22 January 2018.

E. Massung, D. Coyle, K. F. Cater, M. Jay, and C. Preist, 2013. “Using crowdsourcing to support pro-environmental community activism,” CHI ’13: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 371–380.
doi: https://doi.org/10.1145/2470654.2470708, accessed 22 January 2018.

K. Nahon and J. Hemsley, 2013. Going viral. Malden, Mass.: Polity Press.

S. J. Park, Y. S. Lim, and H. W. Park, 2015. “Comparing Twitter and YouTube networks in information diffusion: The case of the ‘Occupy Wall Street’ movement,” Technological Forecasting and Social Change, volume 95, pp. 208–217.
doi: https://doi.org/10.1016/j.techfore.2015.02.003, accessed 22 January 2018.

S. Seidel, N. Berente, S. Debortoli, and N. Srinivasan, 2016. “Tweeting Like Taylor Swift? Affordances, status production, and online platforms,” 2016 49th Hawaii International Conference on System Sciences (HICSS), pp. 1,861–1,869.
doi: https://doi.org/10.1109/HICSS.2016.236, accessed 22 January 2018.

K. Starbird and L. Palen, 2012. “(How) will the revolution be retweeted? Information diffusion and the 2011 Egyptian uprising,” CSCW ’12: Proceedings of the ACM 2012 Conference on Computer Supported Cooperative Work, pp. 7–16.
doi: https://doi.org/10.1145/2145204.2145212, accessed 22 January 2018.

B. State and L. Adamic, 2015. “The diffusion of support in an online social movement: Evidence from the adoption of equal-sign profile pictures,&edquo; CSCW ’15: Proceedings of the 18th ACM Conference on Computer Supported Cooperative Work & Social Computing, pp. 1,741–1,750.
doi: https://doi.org/10.1145/2675133.2675290, accessed 22 January 2018.

S. G. Tarrow, 2011. Power in movement: Social movements and contentious politics. Revised and updated third edition. New York: Cambridge University Press.

K. Thorson, K. Driscoll, B. Ekdale, S. Edgerly, L. G. Thompson, A. Schrock, L. Swartz, E. K. Vraga, and C. Wells, 2013. “YouTube, Twitter and the Occupy movement,” Information, Communication & Society, volume 16, number 3, pp. 421–451.
doi: https://doi.org/10.1080/1369118X.2012.756051, accessed 22 January 2018.

M. Tremayne, 2013. “Anatomy of protest in the digital era: A network analysis of Twitter and Occupy Wall Street,” Social Movement Studies, volume 13, number 1, pp. 110–126.
doi: https://doi.org/10.1080/14742837.2013.830969, accessed 22 January 2018.

S. Vegh, 2003. “Classifying forms of online activism: The case of cyberprotests against the World Bank,” In: M. McCaughey and M. D. Ayers (editors). Cyberactivism: Online activism in theory and practice. New York: Routledge, pp. 71–95.

B. Xu, Y. Huang, H. Kwak, and N.Contractor, 2013. “Structures of broken ties: Exploring unfollow behavior on Twitter,” CSCW ’13: Proceedings of the 2013 Conference on Computer Supported Cooperative Work, pp. 871–876.
doi: https://doi.org/10.1145/2441776.2441875, accessed 22 January 2018.

 


Editorial history

Received 15 May 2017; accepted23 January 2018.


Copyright © 2018, Sikana Tanupabrungsun, Jeff Hemsley, and Bryan Semaan.

Information affordances: Studying the information processing activities of the core Occupy actors on Twitter
by Sikana Tanupabrungsun, Jeff Hemsley, and Bryan Semaan.
First Monday, Volume 23, Number 2 - 5 February 2018
https://firstmonday.org/ojs/index.php/fm/article/download/7888/6641
doi: http://dx.doi.org/10.5210/fm.v23i2.7888