D2.3: Models

 

Modelling the Person in Information Systems

 

IMS, or systems that integrate an IMS component, use a variety of attributes to represent (model) a person and to later manage this person’s information. For instance attributes can be used to represent the identifiers of a person (such as name or pseudonym), her biological characteristics (gender, hair colour), her location (permanent address or geo-location at a given time), competences (diploma, skills), social characteristics (affiliation to groups, friends), and even behaviours (personality or mood).

In some cases, standards and specifications have even been elaborated to facilitate the design and the interoperability of such systems. For instance LDAP schemas have been defined to specify how to represent person’s information in directories. In the human resources domain, the HR-XML specification has been elaborated to standardise the way information about employees are represented in the management software (see the Annex for an overview of different standards and specification for people representation).

Actually, an important strand of research has been conducted for many years in user modelling, aiming at enhancing the interaction between users and systems via the design of adaptive systems (Fischer, 2001; Brusilovsky, 2001; Stephanidis, 2001; Kay, 2000; Andre et al., 2000; Fink and Kobsa, 2000, etc.). The goal of research on personalisation is to improve the efficiency and effectiveness of user interaction by taking into account the specificity of the end-user (such as his cognitive style, or his competence) as well as the context of activity of this user (for instance the current tasks in which he is engaged or the organisational context (Nabeth, Angehrn, and Balakrishnan, 2004)). Practically, adaptive systems are able to support the user better by filtering the irrelevant information (reducing cognitive load), by delivering this information at the right time, by choosing a form of delivery that maximises its impact on this user, or by proposing very contextualised help. Research on adaptive systems has been conducted for applications in a number of domains such as e-learning Diogene (2002), e-commerce (Kobsa et al., 2000) or knowledge management (Razmerita, 2004). 

 

In this document, we will not enter into the details of these theories or standards (which would be out of the scope of this document) but just make a tentative attempt to find some way to categorise these attributes.  

In particular, we are going to present categorisations according to a 

1. temporal perspective 

2. functional perspective 

3. domain perspective 

 

Temporal categorisation

The different attributes can be first categorised by the level of permanence of the information they represent: 

1. permanent – given 

2. permanent – acquired 

3. persistent situations 

4. temporal state 

 

Permanent – given: Some attributes are used to represent some permanent (given) characteristics that were given to a person and on which he usually has no influence. Examples include for instance the biological characteristics (gender, eye colour, fingerprint, etc.), some socio-cultural-economical characteristics (parents, country of birth, etc.), basic personality traits (for some psychologists such as Hans J. Eysenck, personality has an important genetic basis), etc. Some exceptions such as gender changing have to be made regarding the person’s non-influence.

Permanent – acquired: Some other attributes are used to represent permanent (acquired) characteristics that the person was able to acquire because of some circumstances or because of a deliberate action. Examples include qualification (either because of a deliberate action like graduating at a University or because of circumstances like learning a new foreign language during the stay in a country), behavioural characteristics.

Persistent situations (or states): Other attributes are used to represent a situation that is not permanent, but that has some persistence (for instance several years). Examples include the address of a person, a job position (title, employer, etc.), marital status, social status, or a network of friends.

Temporal states: Finally, other attributes are used to represent very temporary situations that are attached to a particular context. Examples in this case include for instance the geographical position of a person at a given time or the mood of the person.

 

Functional categorisation

The attributes can also be categorised according to some functional characteristics. 

Examples of such categories of attributes include: 

1. identification (such as a name, the social security number, password, …) 

2. location (geographical location, addresses, …) 

3. biological characteristics (biometrics, age, …) 

4. personal - psychological (personality, psychological state, preferences, …) 

5. group - sociological (affiliations, social group, social networks…) 

6. … 

 

This categorisation will be detailed in the following chapter. 

 

Categorisation by domain / spheres

The attributes can also be grouped according to their application domain / activities in which these attributes are used such as: 

1. work (employer, title, roles, expertise, acquaintances, work context / tasks, …) 

2. education (university, degrees, …) 

3. leisure (pseudo used in chat spaces, friends, sexual preferences, …) 

4. government (registration information, tax services, …) 

5. justice and police (criminal files, …) 

6. health (social security number (ssn), medical information, …) 

7. … 

 

A subsequent chapter of this document will present more in detail how the categories of attributes are managed in different domains of application. 

 

 

8 / 53