Microsoft Word fidis-wp2-de models doc

FIDIS 

Future of Identity in the Information Society (No. 507512)

 

D2.3 

  

[Final], Version: 2.0 

File: fidis-wp2-del2.3.models.doc 

Page 42 

 

5  Summary and Future Work 

 

5.1 Summary 

In the previous section, the different facets of representation of a person have been examined, 

and for each of them the application domains that are the most likely to utilise each of these 

facets has been identified. To date and to our knowledge, a similar work had not been 

conducted on a large scale, although some tentative attempts, specific to some domains (in the 

domain of e-learning, for instance, Diogene (2002) tries to identify other user models in 

different domains in order to provide guidance in the definition of the student model), are 

worthy of mentioning. 

 

This document has revealed some notable findings. Firstly, it has been observed that several 

application domains (human resource, e-learning, justice) have already started to create quite 

complete and holistic models of the person that fulfil the needs of this domain (see the 

example of HR-XML, IMS/LIP, or JXDM in the annex). In some cases, domains have 

independently developed different representations (in some cases very elaborated) for 

representing the same objective (such as for the representation of the address of a person), 

raising the question of the duplication of work. 

In other cases, attempts to reuse some of the existing specifications have been observed. For 

instance there is a clear similarity between the specifications developed for e-learning in 

IMS/LIP and in Human resources with HR-XML that indicates some clear interdependence 

between the two specifications. Also, the Liberty ID-SIS Service Specification explicitly 

mentions the use of another specification (vCard) as part of the specification. 

Currently, the modelling of identities appears to be composed of dispersed initiatives, is still 

very much under construction (and incomplete), and very complex for the most advanced 

initiatives. Such initiatives would probably benefit from additional cross-disciplinary studies, 

such as this, even if the benefit of a holistic approach still needs to be demonstrated (because 

of its complexity). It has been noted in particular that there is relatively limited interaction 

between different “Universes” (such as identity management and intelligent adaptive systems) 

for which representation of persons is very important. 

Finally, and not surprisingly, it can be noted that some aspects of the person modelling, such 

as the behavioural or social characteristics, are still largely underdeveloped, especially 

considering that they will be essential in the new technologies of tomorrow, such as 

intelligent mobile and ubiquitous applications. 

 

5.2 Future work 

 

In Chapter 4, we provided a first preliminary application perspective of identity 

representation. For instance, we have mentioned the work of different application domains 

such as human resources (HR-XML), Education (IMS/LIP - Learner information package), 

government, police and justice (JXDM), healthcare, commerce (CIQ - Customer Information 

---

FIDIS 

Future of Identity in the Information Society (No. 507512)

 

D2.3 

  

[Final], Version: 2.0 

File: fidis-wp2-del2.3.models.doc 

Page 43 

 

Quality), banking, business (business card), mobility or leisure. We have also found that some 

of the application domains such as human resources (with HR-XML) and justice (with 

JXDM) are already pretty advanced in the elaboration of relatively holistic models, although 

it remains to be seen what the impact is on the complexity of the systems that will have to 

implement them. 

 It is clear that a more in-dept and systematic work of identifying the attributes of identity in 

different application domains, and an analysis of the commonalities between the different 

domains would be useful. 

 

Finally, other routes that deserve studies in their own right are related to: 

 

1.

 

The control and the distribution of the personal information.  

People may not want in the future to see their personal information managed within a single 

centralised system. Federated identity management systems such as the Liberty Alliance may 

have to deal with several standards of representation of person’s information.  

 

2.

 

The management of multiple identities.  

The grouping of attributes may have to be considered more dynamic in order to address a 

larger diversity of contexts (an application may for instance have to deal with several identity 

representations varying over time) 

 

3.

 

The legal and ethical issues associated to the representation, the control and the use of 

person’s information.  

A distinction should indeed be formed between the characteristics of the person that can 

technically be represented and captured (this limit is continuously pushed forward with 

progress in the technology and in knowledge modelling), and the one’s that should be 

represented and captured - society has to determine limits on information that can be 

represented and processed in Information Systems. 

 

---