Twitter: A Social Information Network

Αυτό το άρθρο γράφτηκε στα γαλλικά για τον ερευνητικό και ενημερωτικό ιστότοπο
Μετάφραση στα αγγλικά François Couture

A highly original service, Twitter wears several functionality hats at once: social networking, micro-blogging, link recommendations/sharing, chat, and professional networking. Its success can be explained in large part by its multifaceted nature allowing a wide range of uses and its simple interface. Twitter is currently one of the most dynamic services on the global Internet: the number of registered accounts crossed the one hundred million threshold in April 2010, with 300,000 new accounts created each day, and there are now fifty million messages exchanged daily on the network – including 500,000 originating from Francefor an average of six hundred messages per second.

Twitter was created within the Odeo Company in 2006 by Jack Dorsey, soon joined by Biz Stone and Evan Williams. It raised funds for the first time in 2007 to separate itself from Odeo. Since then, Twitter has raised over 60 million euros from major venture capital firms in the United States. Twitter allows members to post short messages (140 characters or less) on their personal page (timeline); these messages are available to all subscribers (followers). Members can also follow messages streams produced by other users. But Twitter provides more than these basic functionalities: you can message specific members (direct message), retransmit (retweet) other users’ messages, create, track and share account lists, make advanced searches, etc.

From social network to “information network”

The initial concept behind Twitter is based on a simple idea: allow users to say what they are doing in real time – hence its first slogan: “What are you doing?” Gradually, the service evolved from interpersonal or small-group communication to mass networking. In time, Twitter has somehow become a watchdog community tool to monitor a variety of topics. In order to reflect the new uses that had developed, a new slogan appears on the homepage since 2009: “Discover what’s happening right now, anywhere in the world.” According to its founders, Twitter is more than just a social network, it is an actual “information network.”

Twitter’s reputation in the public and the media (Uscali, 2009) has built from the coverage of current events and “scoops” produced by its users, and not, as in the case of Facebook, through a very high penetration rate, especially among young people. For example, Twitter users have reported on the spot the terrorist attacks in Mumbai(November 2008) and the urban riots in Greece (December 2008). More recently, a picture of the emergency landing of U.S. Airways airplane on the Hudson River in New York (January 2009), broadcasted by a single user, has been seen around the world.

But the real efficiency of the service as a news media was established with the Iranian presidential election of June 2009: Twitter, one of the few media left uncensored by the Iranian regime, was used by opponents to communicate with the outside world and show the extent of the mobilization and repression – such activist usage had already been observed in Africa (Mäkinen & Kuira, 2008). These events highlighted the benefits of the service: speedy information dissemination; near real-time coverage from where “it’s happening”, via mobile phones; multimedia content distribution through associated services; and the inability of authorities to censor the flow of information relayed by Twitter, unless they completely block access to the Internet.

Open API and usage-driven innovation

The effectiveness of Twitter as a means of news broadcasting can be explained by both its technical features and level of enactment among users. From the beginning, Twitter was designed as a mobile device service. From the 140-character message length limit (same as an SMS[+] NoteAccording to Wikipedia, SMS length is 160 characters in 7-bit encoding or 140 characters in 8-bit encoding. When Twitter started in 2007, SMS accounted for a large part of all messages exchanged on the system. Things have obviously changed since, with the widespread use of mobile Internet, but you can still send an SMS on Twitter.X ) to the stripped-down interface, everything had been thought out to benefit from the exponential growth of mobile phone services. Moreover, the choice of implementing open source tools and making the service’s programming interface (API) available to third parties has been crucial to Twitter’s success, while fostering the emergence of an ecosystem comprising dozens of interoperable services allowing users to access Twitter without a web browser: geolocation (Foursquare), link shorteners (, picture/video hosting (Twitpic), search engines (Topsy), and clients allowing you to use Twitter without a web browser. According to Jack Dorsey, Twitter’s API receives 20 times more requests from external applications than from – in other words, a vast majority of members access the information circulating on Twitter through means other than the website itself. Therefore, Twitter is not primarily a closed platform like Facebook or a portal, but an information stream updated in real time and viewable on an extremely diverse assortment of supports, which makes it almost impossible for an external body to impose censorship on it.

Twitter officials have also been able to use “innovation by usage,” i.e. “innovative technology or services arising from how users are using the services and spread through exchange networks between users” (Cardon, 2005). In fact, many basic features on Twitter – designating a user’s name by putting the @ sign before it, using hashtags (keywords preceded by #) to define the subject of a message, retweeting – were invented by its users. Twitter’s engineers simply had to follow the trend and adapt the service interface to features already adopted by many users.

Information propagation and categories of users

Many quantitative studies conducted by automated data extraction from the Twitter API are designed to map the flow of information through the network and classify user groups and message groups. For instance, the French agency Spintank tried to follow the propagation of informationaround the strike-down of the Creation and Internet (Hadopi) Law announced on Wednesday, June 10, 2009. The study shows that word of mouth around a current event, commonly known as a buzz, is often intense but short lived. If we classify the messages followed by the #hadopi hashtag into three categories (information, analysis, satire/irony), we see that as we move away from the announcement of the strike-down, the ratio of the two latter categories grows at the expense of purely informative tweets. However, overall, the messages announcing the event dominate the other two categories (half of the tweets labeled #hadopi), which suggests that Twitter is primarily a medium for the dissemination of raw information.

The same study also compared acts of public expression (tweets) and rebroadcasts (retweets), and it shows that rebroadcasting is very high when proprietary information is provided by a few well-informed decision-makers (in this case IT-specialized journalists), and it gradually decreases as the information spreads in the media. The study also notes that the impact of a message broadcasted on Twitter depends less on the information it contains than on the status of the individual broadcasting it, since the number of retweets varies dramatically for identical information. Yet, retweeting is a very powerful dissemination tool: when a message is retweeted, it is likely to reach an average of 1,000 users, regardless of the original broadcaster’s number of followers (Kwak et al., 2010). Users who benefit most from the rebroadcast of their messages are those with a high “digital authority,” in this case renowned web journalists and techies, but not necessarily those with a large number of followers (Cha et al., 2010). This feature of Twitter is not surprising, as it is in line with previous research on the media that have highlighted the role of opinion leaders in the approval and propagation processes of messages (Katz & Lazarsfeld, 1955).
On this last point, Spintank’s results mirror the results of a study conducted in May 2009 (Heil & Piskorski, 2009) on a sample of 300,000 Twitter accounts, which showed that 10% of users generate 90% of all messages. This suggests a rather centralized dissemination of information. These results also coincide with those of the Sysomos firm’s research: among over 11.5 million Twitter accounts analyzed, 5% of users generate 75% of the network activity. As shown by Albert-László Barabási in his reference book (Barabási, 2002), such a high concentration of activity around a small number of nodes is a common feature of many networks of various nature.
Analyzing the information streams on Twitter by categorizing users according to their influence and activity (Krishnamurthy et al., 2008, Java et al., 2007) confirms this uneven structure. The group labeled information sources or broadcasters, i.e. more or less Twitter’s elite, both benefiting from a large audience and a high retweet factor, is indeed mostly comprised of users who were renowned prior to their arrival on Twitter: official accounts (media, companies), journalists, and celebrities in their respective fields… This imbalance is also emphasized by low levels of reciprocal monitoring and direct communication between users (Kwak et al., 2010; Huberman et al., 2010).
The most influential category of Twitter users is an important marketing target, because these users are a key relay in terms of brand and product awareness, but also in terms of traffic being redirected to other websites. This last element is particularly relevant to news websites who are trying to tap into social networks to increase their audience.
Η συνέχεια εδω…

The role of Twitter in the news

Journalists and Twitter

A (future) giant with feet of clay?


Barabási Albert-László, (2002), Linked: The New Science of Networks, Basic Books, New York.
Boczkowski Pablo Javier, « Ethnographie d’une rédaction en ligne Argentine. Les logiques contraires de la production de l’information chaude et froide », Réseaux n° 160, 2010/2-3, p. 43-78.
Cardon Dominique, (2005), « Innovation par l’usage », in Alain Ambrosi, Valérie Peugeot et Daniel Pimienta (dir.), Enjeux de mots : regards multiculturels sur les sociétés de l’information, C & F Éditions, Caen.
Cha Meeyoung, Haddadi Hamed, Benevenuto Fabricio, Gummadi Krishna P., (2010), « Measuring User Influence in Twitter : The Million Follower Fallacy », Association for the Advancement of Artificial Intelligence, 4th International Conference on Weblogs and Social Media, May 23-26, George Washington University, Washington, DC.
Granjon Fabien et Le Foulgoc Aurélien, « Les usages sociaux de l’actualité. L’expérience médiatique des publics internautes », Réseaux n° 160, 2010/2-3, p. 225-253.
Heil Bill, Piskorski Mikolaj, 2009, « New Twitter Research: Men Follow Men and Nobody Tweets », Working Paper, Harvard Business School.
Huberman Bernardo A., Romero Daniel M., Wu Fang, (2009), « Social networks that matter: Twitter Under the microscope », First Monday,Volume 14, Number 1 – 5 January.
Java Akshay, Finin Tim, Song Xiaodan, Tseng Belle, (2007) « Why We Twitter: Understanding Microblogging Usage and Communities », 9th WebKDD and 1st SNA-KDD workshop on Web mining and social network analysis, August 12, San Jose, California.
Katz Elihu, Lazarsfeld Paul, Personal Influence, The Free Press, New York, 1955.
Krishnamurthy Balachander, Gill Phillipa, Arlitt Martin, (2008), « A few chirps about twitter », 1st ACM SIGCOMM Workshop on Social Networks, Seattle, WA, August 2008.
Kwak Haewoon, Lee Changhyun, Park Hosung, Moon Sue, (2010), « What is Twitter, a Social Network or a News Media? », 19th International World Wide Web Conference, April 26-30, Raleigh NC (USA).
Mäkinen Maarit, Kuira Mary Wangu, (2008), « Social Media and postelection Crisis in Kenya », The International Journal of Press/Politics, 13 (3), p. 328-335.
Uskali Turo, (2009), « Weak Signals in Innovation Journalism – Cases Google, Facebook and Twitter », Innovation Journalism, vol.6, n°6.

Leave a Reply

Your email address will not be published.