لخّصلي

osition political parties closer together based on the program’’s algorithm. We can see the clear arrangement around blogs and state-run media sites.

In May 2011, Western social media and news outlets are still at the center of the online Egyptian political network. The majority of common links between Egypt’’s political parties are commercial, Western sites. Most central, we see: Facebook, Google, YouTube, CNN, Yahoo!, Blogger, BBC, Flickr, Twitter and Wordpress. Notably, none of the Websites crawled in November 2010
linked to al Jazeera, and there were only six outgoing links to al Jazeera when the crawl was repeated in May 2011.

The results of the May 2011 network generation show that the same Western media are still present, but are now oriented along the periphery of the Muslim Brotherhood’’s Websites. And while links to Western media are found on many Egyptian sites, the Muslim Brotherhood provides a surprising amount of new content in both its Arabic and English language sites. In terms of pages, the Arabic version grew
Figure 6: Structure and Content of Egypt’’s Online Political Sphere, May 2011

Note: See Appendix for technical details.

by almost 60 percent, and in terms of size it more than tripled. Considering how large the Muslim Brotherhood’’s Arabic language Website is, it is interesting to note how relatively few links it makes to outside news sources or content from other political actors.

The National Democratic Party's Website [ndp.org.eg] is no longer in service. The last publicly available versions of the site were cached in Google's search engine on February 26, 2011. There is no redirect, so it appears that the host servers have been taken offline. The April 6 Movement, which had a central role in the uprising, barely existed as a standalone URL in November 2010 because most of its content was not on its own Website but almost exclusively on social networking platforms like Facebook and Twitter. The National Association for Change and the National Democratic Party ceased to exist after the uprising.

The Muslim Brotherhood’s Changing Online Content Even today, Facebook remains one of the most central nodes in Egyptian networks of political information. In November 2010, before the political uprising, the Websites of major political actors had more links to Facebook and other Western media than they had to each other. By May 2011, Facebook was still central, but with Mubarak’’s departure, the volume of digital content produced by the Muslim Brotherhood has come to dominate these networks.

The most significant change in how political parties operate online involves
two of the Websites of the Muslim Brotherhood. In November 2010 these two Websites had 10,495 pages, with 962 outgoing links and 333 megabytes of content. By May 2011 this had evolved to 12,527 pages, 806 links and 845 megabytes of content. In other words, the number of Web pages had grown by 19 percent, the number of outgoing links had diminished by 16 percent, and the volume of content had grown by 154 percent.

The Muslim Brotherhood presence on English and Arabic Websites was dominant in Egypt’’s online political sphere before the November elections began and has grown significantly since then, particularly with respect to content. The Brotherhood's English site links to much more external content than its Arabic site, but it is a smaller site in terms of hosted content. This may indicate that when seeking to inform their English-speaking audience, the Muslim Brotherhood provides more links to external content to build legitimacy. This may also indicate that there is more English content than Arabic content available to link to. The Arabic version of the site has fewer external links (90 percent fewer) and more hosted content (159 percent more) than the English version.

The Muslim Brotherhood is actively developing its own social media sphere, with ikhwantube.com and ikhwanbook and ikhwanwikitube.com —— Websites that offer much of the functionality of Western namesakes like YouTube and Facebook. As regional experts might expect, the Muslim Brotherhood and Communist Party of Egypt share a number of links to the same kinds of

content. Both parties were the major opponents to Mubarak’’s ruling National Democratic Party. Since the November 2010 elections, both parties have increased the amount of content they have online.

Viral Videos Spread the Freedom Meme YouTube became a particularly important tool for spreading news and information of Egypt’’s uprising——in the form of user-generated videos —— around the world. Our research identifies the top viral videos as of June 2011 (see Appendix for basic statistics ). While it is difficult to measure the precise impact of these videos on audiences, some images of suffering certainly would have spurred protests and heightened moral outrage.

The first significant Egyptian video went viral on January 25, 2011. The video depicts thousands of protesters converging on Tahrir Square. The images are captured by an amateur cameraman looking out of a building near the main road. Based on the metadata reported on the uploader’’s YouTube account, the video was distributed by an account registered as located in the United States. Since then, this video has accumulated more than 600,000 views. Based on tracking of the embed code, it is most likely that the video received popular attention after being posted on AllMania.com, a sports commentary site that has experienced a 600 percent increase in traffic in January 2011.

RussiaToday’’s YouTube channel contributed 5 of the top-20 viral videos,
totaling 1,200,000 cumulative views. These videos were from citizen journalists and included live footage rebroadcasted through the news agency’’s outlets. Al Jazeera English’’s YouTube channel similarly contributed three videos, totaling more than 300,000 views. Reuters’’s YouTube channel contributed one video totaling more than 200,000 views.

There were four major types of content that typified Egyptian viral videos: raw protest and mobilization footage; citizen commentary; political punditry; and ““soundtracks for the revolution.”” Raw protest and mobilization footage was the most common, totaling nearly 5.5 million views from 23 videos. One video featured a detailed 20-minute dialogue between a religious scholar and political philosopher about the future of Egypt, totaling 100,000 views. Another featured a home-made video with a young girl’’s commentary about political events, totaling 275,000 views. But the most popular video, a music video, was heralded as a soundtrack to the revolution and served as a rallying cry of support for the Egyptian peoples’’ protests. This music video was the single most popular viral video for the Egyptian revolution, uploaded on January 27, and accounted for 25 percent of the top-20 video views. A list of the most prominent viral videos about the political uprising in Egypt appears in the Appendix

Conclusions
Social media played a crucial role in the political uprising in Tunisia and Egypt. Using original data from multiple social media sources, we can offer some concrete conclusions about what that role was. First, social media played a central role in shaping political debates in the Arab spring. Second, a spike in online revolutionary conversations often preceded major events on the ground. Third, social media helped spread democratic ideas across international borders.

Democratization movements existed in North Africa and the Middle East long before technologies such as mobile phones, the Internet, and social media came to the region. However, with these technologies people who share an interest in democracy learned to build extensive networks, create social capital, and organize political action. In both Tunisia and Egypt, these virtual networks materialized in the streets in early 2011 to help bring down two longstanding dictators.

Anecdotally, we know that social media played an important role at key moments in the events of this year. But what are the big-picture trends in social media use that explain why public demand for democratic reform rose now, and why events unfolded the way they did? Our unique datasets reveal much about the role of different kinds of social media. The Tunisian blogosphere provided space for open political dialogue about regime corruption and the potential for political change. Twitter relayed stories of successful
mobilization within and between countries. Facebook functioned as a central node in networks of political discontent in Egypt. During the protests, YouTube and other video archiving centers allowed citizen journalists, using mobile phone cameras and consumer electronics, to broadcast stories that the mainstream media could not or did not want to cover.

Social media alone did not cause political upheaval in North Africa. But information technologies —— including mobile phones and the Internet —— altered the capacity of citizens and civil society actors to affect domestic politics. Social media, such as Facebook, Twitter, and YouTube, have several kinds of impact on local systems of political communication. First, social media provides new opportunities and new tools for social movements to respond to conditions in their countries. It is clear that the ability to produce and consume political content, independent of social elites, is important because the public sense of shared grievances and potential for change can develop rapidly. Second, social media fosters transnational links between individuals and groups. This means that network ties form between international and local democratization movements, and that compelling stories, told in short text messages or long video documentaries, circulate around the region. The inspiration of success in Tunisia was not just a fast-spreading contagion, for civil society leaders in neighboring countries also learned effective strategies of successful movement organizing through social media.

Social movements are traditionally defined as collective challenges, based on shared purposes, social solidarity, and sustained interactions with elites, opponents and authorities. They support a public claim against target authorities and engage in political action by forming coalitions, organizing public meetings and demonstrations, and using the media to highlight their claims. Through such demonstrations and media use, social movements display their unity, numbers and commitment. Social media, social networking applications, and consumer electronics have not changed the purpose of social movement organizing —— economic opportunity and political voice are still the shared goals of social movements.

But in North Africa and the Middle East, relatively new youth movements have been surprised by the speed, size and success of protests they have organized over social networking Websites. Over several years they have found their political voice online and have held their meetings virtually. Each of the dictators in these countries has long had many political enemies, but they were a fragmented group of opponents. Now these opponents do more than use broadcast media to highlight their claims. They use social media to identify goals, build solidarity, and organize demonstrations. During the Arab Spring, individuals demonstrated their desire for freedom through social media, and social media became a critical part of the toolkit used to achieve that freedom.

Methods Appendix Analyzing Twitter Data The data for Figures 1, 3 and 5, comes from the analysis of Twitter feeds. This project is among the first to analyze the flow of text messages about the potential and strategy of democratization movements among multiple countries. In addition, we figured out how to distinguish between domestic, regional, and international contributors to the growing online consciousness about political crisis. Demonstrating Twitter’’s impact on regional conversations is an important contribution but was technically challenging. We processed more than 3 million tweets for their use of hashtags about events in North Africa and the Middle East. We purchased cloud computing time from Amazon to speed up the text analysis, and wrote automated scripts for identifying the relevant tweets. A significant number of the tweets provide longitude and latitude information, and that information was automatically converted into country location. Finally, we hired a translator to help with texts and location information that is in Arabic, French, Hebrew and Turkish.

This dataset was created using the Twitter archiving service TwapperKeeper (http://twapperkeeper.com/) to capture the flow of tweets from the Twitter Search API for Algeria, Egypt, Tunisia, Morocco and Yemen. The hashtags analyzed, in order, were ““#algeria””, ““#egypt””, ““#feb14””, ““#morocco””, ““#sidibouzid”” and ““#yemen.”” Since these archives were initiated by different users at different times, they
do not all cover the same time period. The earliest, #sidibouzid, begins on January 14, 2011, and the last tweets (in multiple hashtags) occur on March 24, 2011. TwapperKeeper experienced system overloading at several times during this duration, resulting in coverage gaps within some of the archives. But even for archives with no gaps, it is highly unlikely that TwapperKeeper’’s archive captured all relevant tweets due to limitations imposed by Twitter. All six archives combined contain a total of 3,142,621 tweets, some of which undoubtedly overlap because each tweet could contain multiple hashtags. Over 75 percent of these (2,363,139) are from #egypt. This method omits some unknown number of in-region tweeters due to blank location fields, deleted accounts, and uninterpretable information in the fields.

Twitter changed its terms of service on March 20, 2011, to disallow public sharing of tweets. The archives analyzed in this report were queued for downloading from TwapperKeeper on March 19, 2011, but due to the backlog of similar requests from other users, the downloads did not become available until several days later (which is why some of them include tweets added after March 20). The archive dates for specific hashtags vary, and the earliest data points come from #egypt on January 5, 2011. All tracking ends March 20, 2011 due to Twitter’’s terms of service change. TwapperKeeper, the service used to track hashtags, was crippled (See