D2.2: Set of use cases and scenarios

Title: CONCLUSION

Conclusion

The conceptualization of identity in DSEs (Digital Social Environments) that we have presented in this paper relies to a large extent on a more informal and abstracted perspective of identity than the concept of identity manipulated by the information system or security specialists. Whereas in the last case the identity is managed principally with people representations (as a set of attributes and characteristics that can be stored for instance in an identity card) and their authentication (usually happening once at the beginning of a session), the management of identity in social environments is more diffused, and in particular its control is not granted to a central authority but based on the idea of providing transparency about the behaviours and the actions of people, and is socially regulated (for instance with social pressure). 

Still, a closer look indicates that the two worlds (formal and informal perspective) are not totally alien but are on the contrary complementary, and are subject to cohabit more and more in the future, in particular as the identity issues in the Information Society are addressed more holistically. Indeed, as the frontier between the physical and the digital worlds is becoming blurred (this is best illustrated by the advent of ambient intelligent environments) and is converging, as this new world is becoming very complex and difficult to manage by traditional methods (via explicit identity and identifier, and one time authentication), and as the Information Society is becoming interested in supporting more widely human aspects (privacy, social dimension, etc.), new methods (in particular more flexible and more robust) will need to be activated to manage the identity in our societies. 

Social mechanisms (reputation, social control, etc.) represent an effective means of regulation for complex systems, and should be considered (at least for a partial use) in every identity management solution. In particular, social engineering approaches to systems that include an important human component (including management of risk associated with inevitable human errors and biases), a category to which identity management systems belong to, are often more effective and more robust approaches than the technical engineering approach alone. This paper has also indicated that the use of social mechanisms also has its limits, and that the management of identity in DSEs could benefit from the work of the more formal management of identity. 

In conclusion, we can consider as suggested in (Jordan, Hauser, and Foster, 2003), that the two approaches are complementary and should be combined in order to implement systems that are more flexible, more robust (in particular concerning human error) and more reliable, for designing the next generation of Internet systems that will be more socially aware, and more humanly friendly than they are today.