Data Flows and Crises in Online Reputation Economies

From Cyborg Anthropology
Jump to: navigation, search

Prior to network culture, traditional news outlets were the first reliable source for news concerning major events. This was because traditional news media outlets have established reputations for providing a certain level of credibility and reliability.

In a global, ever-connected economy, it is finally possible to rely on citizen media outlets to receive news almost as soon as it happens, however, people often have a limited basis on which to determine validity. Online, time and space for information gathering is compressed. This also means that time and space for decision making is also reduced. This is why online social networks try to use online metrics to establish validity in as short amount of time as possible. Take, for example, critical situations like wars, attacks, accidents or natural disasters:

  • In emergency situations, traditional media sources are often too slow in providing clear, relevant information.
  • In delicate political environments, standard news outlets are often blocked from transmitting relevant information.

These situations call for non-traditional data points. These data points exist in the form of social nodes in networks. The wired, network of the online world allows anyone close to the news source to have the same power as ones with bigger budgets, bigger political power or better transmission equipments like a traditional news source. Reputation in critical moments like these (such as earthquake reporting, or terrorist attack information and safety instructions) must be negotiated almost instantaneously. Unlike traditional offline news identities, there are no presuppositions of identity.

In a space where news sources are both distributed (in both sense of the word: “distribution of power” and “fragmentation”) and largely anonymous, reputation becomes the sole metric for validity. This is the problem that this paper tries to address.

Reputation is extremely complex. There is no single way to calculate it:

  • It can take the form of a hyperlink between two places, abilities, or powers. In other words, reputation is a way of describing the link between two entities.
  • It can be transitory, especially online, where reputation serves as a social construction only as long as it’s needed, depending on data flows, proximity to events, or distance between individuals. In other words, reputation is a dynamic system of situated knowledge that sorts social interactions.
  • It can be a handshake, in a sense that both parties must agree to open up and exchange something valuable for a trust-relationship to happen. In the business realm, for instance, this action has been formalized in the act of exchanging business cards.
  • It can be measured or tracked as an overlay on a series of data points showing relations and trust.
  • It can be measured or tracked as factors that individuals share in common. More shared things will lead to more shared beliefs, value systems and judgment, and generally could better reputation.

Measuring Reputation

A new metric is thus needed in order to quickly determine credibility and reputation in the event of a crisis. Note that this paper does not aim to search for and establish the most accurate metric, but rather, one that provides the user with an idea about the situation, then leaves the ultimate value judgment in her hand. In other words, to be both economically and timely achievable, the metric has to have enough ‘fuzziness.’ “What you want is a durable perception of person”, says programmer Anselm Hook, “one that allows one to quickly understand whether a piece of information from a source is reputable or not in the fastest way possible”. One way is to wait for a backup vote. Robert’s Rules of Order say that a statement must be seconded before it can be voted on by many. But in some cases, waiting for a second is difficult, because there may be only one person next to a data source or event that is capable of reporting it. In order to determine a valid metric, one must define a few key elements of the online experience: Interest and Power

Power is created by interest. This is the most easily observed in online environments, where the creation of value and interest is most fluid. The fluidity of value creation and exchange.

Interest Groups

One could call an interest group a demographic. Demographics are those with specific lifestyles that influence interest, and also support those who create products or services that fulfill these interests.

Crises and Social Networks

During a crisis, interest groups tend to converge upon a single topic or news source. The creation of validity in a news source in an online social network is often very fast, and generally not a traditional news source. Network users who were formerly low-level nodes can suddenly become major nodes of traffic if they begin to provide data that has proxemic, relational, or newsworthy value.

Those nodes that can provide the fastest information have tremendous power over those who have recently turned to follow them.


Point A marks the status of normal social network conditions and interest groups.

B marks the first appearance of crisis in the social network.

C signals the ramp-up of information awareness among social groups not in the social interest group of the initial reformers.

At point D, the crisis becomes a topic of collective interest. Networks of trust re-broadcast the news to un-informed groups until the network is saturated with information from all groups capable of absorbing the information.

At E, the discussion of crisis decreases due to crisis resolution of exhaustion of topic. The crisis falls out of common interest and formerly melded interest groups diverge once again.

F marks the final resolution or disappearance of the crisis. The crisis falls almost completely out of social network conversation.

One of the problems with social networks during crises is quickly finding the nodes with the most valuable information a voice in an efficient way, and promoting them to the top of a social network so that all that need that information can find it.

Disaster Reporting

On May 11th, 2008, a earthquake that measured 7.8 on the Richter scale hit China. Several of those who experienced the earthquake Twitter user @dtan Tech Reporter Robert Scoble was able to rebroadcast the message to (at the time) approximately 40,000 followers.

Robert Scoble's Earthquake Tweet

But how did Robert Scoble know that @dtan’s Tweets were valid?

Was it the architecture of Twitter? A trust economy, established by the rapid exchange of everyday data on Twitter helped to. But Scoble’s reputation process takes a while. He has to first follow @dtan and through direct or indirect exchange determine the user’s reputation to report on an emergency.Of course, later on, additional reports from other people in China who also experienced the quake arrived. But it took CNN hours later to report on the event. This demonstrates the agility, relevancy and accuracy of non-traditional nodes as news sources.

As an aside, tools such as Google’s translation engine allowed @dtan’s Tweets, which were written almost entirely in Chinese, to be translated into English, and passed on to a more global audience.

Improving Data Flows in Crisis

All individuals have social bases. There are an increasing number of individuals who use social networks as social bases. However, these bases are not necessarily the same. Social networks record relationships in different ways.One who uses the photo-sharing website Flickr as a social base interacts with data differently than a Facebook or Twitter user. Robert Scoble was able to transfer authority and power to @dtan very quickly, but rapid, local news of the earthquake was constrained to Twitter.

There was no system that looked at Twitter as a database and pulled out information. Neither was there a system that added Twitter’s earthquake updates to other relevant information coming out of mainstream news sources and other social networks.

To improve data flows in crisis, there is a clear opportunity to transcend data silos and aggregating data streams into a more accessible and unified databases, so that users of different social networks, or limited social networks, can quickly access relevant information.

This calls for either:

  • The establishment of an open standard for disaster reporting across networks.
  • The use and appropriation of existing open standards for reporting.

For instance: the DiSo project is an initiative to facilitate the creation of open, non-proprietary and interoperable building blocks for the decentralized social web.

Another other alternative (besides traditional media) is to rely on many ‘Scoble’s’ on each social network who talk to and inform each other on current happening at all times. This is highly impractical and very costly.

Additional Sources:

For more information and a full analysis of the Twitter Earthquake reporting, please visit: