WP5. User modelling and interface design

Description of work

T5.1 – Persuasive design principles specification (TU/e) [M1-M12]

This task will describe design principles based on a novel persuasive technology framework for personalized interface design. It will define a taxonomy of motivators and a set of principles for deriving new motivators that can be used to change the behaviour or attitude of a person in terms of social inclusion, increasing trust, and diminishing reactance effects. This taxonomy will categorize motivators (indicating their suitability and effectiveness) for different types of users (differentiating for relevant characteristics which are specified in T5.2, like type of (dis-)abilities, sensitivity to persuasive strategies, personality, intentions, etc.), different types of situations (hospital, revalidation, home), and availability of other motivators (social: personnel, relatives, friends; financial: income, insurance; illness related, etc). Fundamentally, this framework for personalized interface design will be based on existing, well-accepted models of persuasive technology (Fogg, 2002; Oinas-Kukkonen & Harjumaa, 2009) taking into account the psychological implications delivered by the user (dis-)abilities. So, for example, (based on the motivator taxonomy) this framework might indicate that for a person who is sensitive to authority-based persuasive strategies, the interface should contain prominent references to relevant authorities (e.g., display very clearly a physician and his or her authority, and the fact that instructions to use the technology came from this person). Also, the framework might indicate that for users who feel lonely, persuasive strategies that include social connectedness (e.g., including forums and connections to other patients) will be effective in motivating the user to use the technology.

T5.2 – User modelling and profile generation (TU/e, SHEBA, AUTH, MDA Hellas) [M8-M12, M24-M27]

This task specifies user models and profiles, describing users with regard to relevant characteristics and limitations, such as (dis-)abilities, interaction behaviour, sensitivity to specific persuasive strategies, emotions, intentions, social abilities, extent to which training is needed, social network, sensitivity to persuasive strategies, based on knowledge of user modelling and the specific target groups. Also, it will adapt these models to the pilot trials and evaluation of WP6, as these will provide knowledge necessitating re-balancing these user characteristics, adding relevant characteristics, and new knowledge on these users not previously available in the literature. A set of profiles (personas) will be used to concretize the outcomes of this task that will be used to drive the design of the prototype interface applications.

T5.3 – Design and development of prototype interface applications (UNI KO-LD, CERTH, TU/e) [M13-M20, M28-M32]

Based on the persuasive design principles (T5.1) user models (T5.2), this task will develop prototype applications that will be used in lab evaluations to examine the actual effectiveness of the persuasive principles (personalized, stimulating trust in the interface) for influencing the behaviour and/ or attitudes of users with characteristics comparable or equal to those of the target population. So, using the framework developed in T5.1, this task will validate the usefulness of the psychology-inspired design principles to the non-conventional target group of people with disabilities. Nourishing on the output of T5.2, this will be done first of all (M13-M20) based on the first version of the user models (literature and clinical specialist based), and secondly on an adaptation of those models based on trial cycle 1 (M28-M32). All the above will be reflected in the design of a set of prototype interface applications that will be used during the pilot trials.

In order to develop these prototype applications, we will rely on the middleware and the Interaction SDK developed in WP4 by allocating our efforts along two tracks. In the first track, we will pursuit the scenario of incorporating our widgets for novel interaction paradigms as part of an existing software tool that is amenable to extensions. We will investigate both open source tools like Gallery[1] and vvvP[2] for managing and GIMP[3], Audicity[4] and Blender[5] for creating/editing multimedia content, but also closed-source tools that can be extended through third-party plugins like Photoshop[6] and Adobe Premier[7]. Our goal in this case would be to enable the highest possible of the control functionality incorporated in these tools through the user’s eyes and mind. In the second track, we will develop prototype applications from scratch by relying on existing powerful standards like HTML5. Our goal in this case is to have full flexibility over the control functionalities that we would like to offer, allowing the design of radically novel interfaces. In this way we will also have the opportunity to aim for the user’s behavioural change as a result of the employed persuasive design principles. Particular attention will be placed on integrating with our prototype applications the functionality to share content in social networks like Facebook, Instagram and YouTube as well as to facilitate blogging functionalities. Finally, detailed instructions on how to setup, configure and launch these prototype interface applications will be provided in order to increase the potential of adopting these applications in a home environment.

T5.4 — Training cycle development (TU/e, SHEBA, AUTH, MDA Hellas) [M12-M20, M27-M32]

This task describes training cycle specifications, based on performances of real users, including descriptions of the extent to which training is needed, differentiations for level of (dis-)abilities, and cognitive functioning. That is, the prototype applications developed in T5.3 will include training cycles for the user to get familiarized with the capabilities of the interface, to enhance the user’s performance in using MAMEM’s interfaces, and to increase trust and liking of the interface.  Also, T5.4 provides “gold-standard” data of the performances of these real users on the defined training cycles.

[1] http://galleryproject.org/

[2] http://vvvp.sourceforge.net/

[3] http://www.gimp.org/

[4] http://audacity.sourceforge.net/

[5] http://www.blender.org/

[6] http://www.photoshop.com

[7] http://www.adobe.com/products/premiere.html