2013 IEEE 37th Annual Computer Software and Applications Conference Workshops
Building Adaptive Accessible Context-Aware for User Interface Tailored to Disable Users Lamia Zouhaier, Yousra Bendaly Hlaoui and Leila Jemni Ben Ayed LaTICE Laboratory National High School of Engineers of Tunis University of Tunis- Tunisia
[email protected] [email protected] [email protected] Abstract— This paper presents an approach of user interface adaptation based on context awareness system and context modeling in order to ensure accessibility for disabled people in pervasive environment. The challenge is how to design and generate accessible software systems that accommodate dynamically to changing context.
customized or adapted to all types of devices hosting service they provide and to the physical environment (location, time, sound, light, etc.) where the user interacts with the system while considering the individual characteristics of the user (profile, preferences, abilities, etc..).
Keywords— Accessibility; Disable user; awareness; context modelling; ontologies
The challenge of accessibility needs to be addressed through more proactive and generic approaches, which account for all dimensions and sources of variation. These dimensions range from the characteristics and abilities of users, to the characteristics of technological platforms, to the relevant aspects of the context of use [8].
context-
I. INTRODUCTION The advance of new technology reflects an increment of users’ needs to use information in every place and at any time. Generally, the technological change delivers products and services requiring particular skills and abilities on the part of the human user [8]. Accessibility of user interfaces of computing systems in ubiquitous environment is considered as a recent and an explored issue of research [1, 19]. Sensory, cognitive or physical disable people are usually excluded from computing services since they are unable to use interactive systems that have been designed ignoring human diversity [21]. People with disabilities confront access barriers on their daily activities. They are almost unable to interact with user interfaces IUs of the emerged devices. There is a lack of enforcement for these people on how to proceed with such system as well as how to use easily and satisfactory. It is indispensable that computational system gives accessible interface for disable people in order to provide equal access and equal opportunity between all people. Handicap user profile is considered as a basic and centric challenge when designing accessible interactive systems but it is not the only one. In fact, the need of mobility of user during the interaction with the system and the ambient user environment that accommodate the interaction are indispensable to take into account when designing interactive systems.
Our contribution consists of the introduction of new techniques to represent context and the context model to then be able to apply the process of adaptation in order to generate the final interface from model-based interface following the approach Model Driven Architecture MDA [15]. Hence, we propose a novel approach that aims to support interactive application based on infrastructure of context awareness, monitoring by disable users, which must be aware of context changes and be adapted accordingly. We consider adaptation of various possible contexts such as handicap user profile, user preferences, interactive devices and environmental parameters. The remainder of this paper is as follows: in section 2, we present some relevant work in relation with our problematic. The third section details our approach and the fourth one presents how to model context. We finish with an illustration and a conclusion with future work. II. RELATED WORK A limited number of work have focused on identifying more general solutions able to adapt any user interface of applications for various combinations of context of use including disable people using context awareness mechanism.
These systems must be used in different contexts according to user profile, its environment and the terminal being used. Increasingly, to support the daily activities of any individual, engineering of interactive systems is facing the challenge of not only generate context-aware user interface but also that can be 978-0-7695-4987-3/13 $26.00 © 2013 IEEE DOI 10.1109/COMPSACW.2013.39
157
time. Occasionally, it is addressed in one of several ways: manual redesign of the interface, limited customization support, or by supplying an external assistive technology.
In the early period, accessibility problems were primarily considered to concern only the field of Assistive Technology (AT), and consequently, accessibility entailed meeting prescribed requirements for the use of a product by people with disabilities [8]. For some developments, it is necessary to apply also more specific guidelines such as sets of guidelines for specific application type (e-learning, tele-working), access device (mobile devices) or user type (elderly, children, blind, deaf) [4]. Web developers and designers can leverage standards such as WCAG to ensure the overall accessibility of a given Web application. This type of guidelines devises a set of conformance levels based on how loose or strict is a Web page’s support on accessibility issues, independently from any particular disability i.e., perceived as an overall status [13].
AVANTI [7] is the first project to employ adaptive techniques in order to ensure accessibility and high quality of interaction for all potential users [8]. It provides users with special needs with hypermedia information by adapting the content, navigation and presentation. It put forward a conceptual framework for the construction of systems that support adaptability and adaptivity at both the content and the user interface levels [8]. The distinctive characteristic of the AVANTI browser is its ability to dynamically tailor itself to the abilities, skills, requirements, and preferences of the end-users, to the different contexts of use, and to the changing characteristics of users as they interact with the system.
There are a wide variety of applications [8] to which personalization can be applied and a wide variety of different devices available on which to deliver the personalized information. Most personalization systems are based on some type of user profile, a data instance of a user model that is applied to adaptive interactive systems. Leonidis et al. propose in [3] a toolkit for rapid prototyping, in order to ease the design of adaptive widget-based interfaces.
EGOKI [2] is a system that generates accessible mobile user interfaces adapted for people with disabilities in order to grant them access to ubiquitous services. These interfaces are intended to provide access to ubiquitous services in intelligent environments. EGOKI dynamically creates an instance of the interface running on the user device. To adapt the interface to the user characteristics, it is necessary to take into account what the most suitable communication modalities are for each user, mapping them to the appropriate media.
Adaptive and adaptable interactions techniques are increasingly emerged in recent research [1, 8, 9, 10, 12, 22]. There are no generic solutions oriented towards accessibility of user interface, but different terminologies are employed as Universal Access [7], User Interfaces for All [9], Design For All [14] Unified User Interfaces [9] because of the range of the population which may gradually be confronted with accessibility problems extends beyond the population of disabled and elderly users to include all people [8]. Universal Access refers to the global requirement of coping with diversity in: (i) the characteristics of the target user population (including people with disabilities); (ii) the scope and nature of tasks; and (iii) the different contexts of use and the effects of their proliferation into business and social endeavors [8].
In literature, few works deal with adaptation of the content, presentation or the navigation scheme of the user interface to users with special needs [21]. Ubiquitous services are usually provided by means of generic interfaces that may contain barriers for people with disabilities [16]. To overcome this problem, the use of adaptable or adaptive user interfaces is recommended [1]. The authors of [14] summarize different type of user interface adaptation configuration based on four levels: • Initiation (I) decision of an actor to perform an adaptation. • Proposal (P) represents the suggestions or recommendations. • Decision (D) is the process of choosing among proposals reaction. • Execution (E) consists of the implementation of the chosen reaction. Dieterich’s taxonomy [14] is considered as a seminal reference for classifying adaptation techniques of user interface. A process of adaptation can take different settings by combining the characteristics of adaptability and adaptivity. An adaptive user interface
Universal accessibility system should be accessible for all users, although the design is focused on people with special needs. But none of these projects resulted in any concrete solutions for users with special needs. The scope of User Interfaces for All, as a perspective on HCI, is necessarily broad and complex, involving challenges, which pertain to issues such as context oriented design, diverse user requirements and adaptable and adaptive interactive behaviors. This diversity of needs is generally ignored at the present
158
changes dynamically in response to its experience with users [7] [9]. The system can recognize the situation (including multiple triggers set by the designer) and adopts reaction (recommendations) also set by the designer for the new situation. An adaptable user interface, however, provides tools, which allow the user to tailor certain aspects of the interaction, while using an interactive system [7] [9]. The system is customizable on explicit intervention of the user who can act on parameters set by the designer [9]. With the combination of the adaptivity and adaptability using the four levels indicated by Dey et al [13], different configurations of adaptation process are possible [14][7].
III.
A NOVEL APPROACH ADAPTATION BASED CONTEXT-AWARE SYSTEMS
OF ON
Our approach consists of proposing architecture and adaptation strategy of user interfaces of applications for new contexts of use based on the following objectives: -
Multiple targets can be determined in adaptation process. In [11] [9], authors have defined the target as an entity that is intended for adaptation, which can be: the user, the environment or the characteristic of the system.
Provide a generic and scalable architecture for adapting applications to new contexts of use. Provide an overall adaptation strategy (features, data and presentation) of an application to new contexts of use. Ensure scalability of available adaptive mechanisms that we can apply to an application.
To ensure these objectives, we propose to include recent techniques employed by context-aware systems for context acquisition and management for, afterwards, enabling the application of adaptation rules in order to generate the final interface using model-based technique following the approach Model Driven Architecture MDA [15].
Plasticity [11, 12, 13] is a recent and emerged technique of adaptation which is the capacity of an interactive system to withstand to context variations while preserving usability. In order to support the end-user preferences, adaptations rules can be changed according to user’s order [10] [12]. It results from a SituationÆReaction process where the situation denotes a context change that needs a reaction, and reaction denotes the procedures that the system and/or the user executes to preserve usability.
Three principal steps that characterize our approach as scheduled and represented by Figure 1: 1. 2. 3.
In pervasive computing environments, context consists of any information that can be captured from the surrounding environment in which the user is interacting with the application, gathered from a variety of sources. Context-aware systems offer entirely new opportunities for application developers the possibility to gather context data and adapting systems behavior accordingly [18]. Adaptation of user interface in mobile contexts is a topic that has recently stimulated various research contributions [19].
Context acquisition and Management (1) (2) (3) Application of adaptation rules (4) Generation of final interface based on model based development using the paradigm Model Driven Architecture (MDA) (5)
1
Context Sensitive System (Sensors)
Knowled ge Base (KB)
2
Context Management System (Ontologies)
Existing work in the area of context awareness focused on all aspects of capture, interpretation, modeling, storage and dissemination of context but there are no generic and global solutions that include all steps of adaptation from context acquisition to generated final interface.
User Interface Model (MDA)
Building context-aware applications is one of the solutions adopted but also still a complicated task due to the lack of an adequate and generic infrastructure that support in pervasive computing environments [14].
5
Adaptation Layer (Adaptation Rules)
Figure 1: Overall Aproach Steps
159
3
4
a. Architecture of Context - Aware system The majority of frameworks that support context sensitive use a layered architecture supporting the important aspects of sensor capture, context extraction and reasoning [22]. These layers are used in order to prepare an infrastructure for system adaptation to the current context. They are based specially on: - Layer of capture based on physical sensors. This layer acquires information from surrounding environment. - Layer for interpretation in order to interpret and make useful the context captured. - Layer for reasoning based on ontology responsible of reasoning about contextual information. - Layer of service which embodies a set of services to manage contextual information.
-
description. In some cases, context value might contain an absolute numerical value or feature describing context. Attributes specify the context expression and might contain any additional details not included in the other properties e.g Timestamp describing the date/time when context was sensed, Source containing how information was gathered, etc
In table 1, we show an example of context modeling vocabulary based on context type and it value. Table 1: Example of Context vocabulary (Source: [20]) Context type Environment:Sound:Intensity Environment:Light:Intensity Environment:Light:Type Environment:Location:Building
b. Context Management Models Context management system must be able to gather, organize, interpret, evaluate and distribute context information. It is processed at different level of abstraction with different types (single scalar, vector, abstract situation, etc.) and by different layers. It must respect some requirements of these environments such as distribution, mobility, device constraints, and heterogeneity of context sources. Although a number of context models have been presented in the literature, none of them supports all of these requirements to a sufficient extent at the same time.
Context value {Silent,Moderate, Loud} {Dark, Normal, Bright} {Artificial, Natural} {Indoors, Outdoors}
A. Adaptation Process The adaptation of interface is the capacity of personalization that takes into account all contextual information from different entities as environment attributes and user characteristics [8]. The multi facets is a multi-faceted process, which can be analyzed along three main axes, namely the source of adaptation knowledge, the level of interaction at which it is applied, and the type of information on which it is based [8]. The adaptation process (see Figure 1) is based on a set of adaptation rule which are specified according to the arrival of context events (context change). These rules correspond to the application of a set of principles according to the reception of any contextual information, about (a) eser characteristics (b) environment characteristics and (c) devices characteristics. The adaptation rule is triggered by context events. These context events represent the different changes in the context detected by more than one sensor (hardware or software). The transformation is applied on UI after application of adaption rules. We can apply at this level the approach of Model Driven Architecture MDA which is based on a set of transformation between Abstract UI, Concrete UI and Final UI.
The most relevant approaches that address context modeling are Key-value Model, Object oriented Model, Mark-up Scheme Model, Graphical Model, and Ontology based Model. These approaches differ on the scheme of data structure used to exchange contextual information in the respective system. A good context modeling approach must include modeling of context information quality to support reasoning about context and providing a good formalism to reduce the complexity of context-aware applications and to improve their maintainability and evolvability [6]. The different entities must have a common structure for representing information. Each context expression must contain at least Context type and Context value: - Context type: Each context must belong to a category (Sound, time, temperature, etc.). These types will be used in a context subscription or query. Context type concepts form a tree structure. - Context value refers to the semantic or absolute “value” of context type and is usually used together with context type, forming a verbal
160
Adaptation Rules
1..*
1..* Triggers
user ContextEvent +1..*
1..* applies
Produces 1..*
1 Transformation
Visual Impairments Auditory Impairments
Preferences: if a user wants to watch highlights
Sensor
Prioritising audio context, textto-speech transmoding Prioritising video context, audioto-text transmoding, subtitle presentation. Prioritising a selected area of visual content in a scene (e.g. focus on lecturer lips) Summarising the session
Figure 2: Metamodel for Adaptation Rules (extract from [17])
Device
Environment
B. Generation of the new interface based on MDA paradigm MDA separates the specification of system functionality from the specification of implementation. One specification of system functionality can be transformed into many specifications of implementation. The heart of model driven architecture (MDA) is its transformation. There are two types of transformations in MDA, i.e. Model (Plateform Independent Model) to Model (Plateform Specific Model) and the Model (Plateform Specific Model) to Code transformation.
UsingDevice InEnvironment
User HasPreferences
HasActivitie
Activities IdentifiedBy
Preferences
User Profile HasImpairement
Impairement
HasHistory
Usage History
Figure 3: User Context
IV.
ILLUSTRATION V. CONCLUSION AND FUTURE WORK In this paper, we have presented a novel approach of development of adaptive tailored user interface based supporting context-aware architectures. Context sensitive process is very crucial feature for ubiquitous computing applications. The specification of modelbased adaptation based on adaptation rules is a great interest in our contribution since there are more and more approaches supporting the model-based approach and also, there are many approach of context-awareness system. The context modeling, that we will proposed, must to support the ability to model different types of context, such us physical, social or computational context and unify the context representation. In addition, it provides a uniform way of representing and sharing context. Thus, a good context modeling approach must include modeling of context information quality to support reasoning about context and providing a good formalism to reduce the complexity of contextaware applications and to improve their maintainability and evolvability. We propose the use of ontology in order to support context modelling and to provide a formal semantics to context knowledge. In the developed ontology, we must to distinct between user preferences, user profile information (name, genre, age, etc.) and also user impairments information. User Profile is static
There is the need to consider the user as a central and fundamental part on context-aware applications. Through this consideration of user, it will be possible to adapt content according to user attributes. Users differ in a wide range of variables; (a) User profile including identity, name, gender, age, education, role, demographic characteristics, (b) impairment including deaf, visual impaired person, cognitive skills etc., (c) preferences including audiovisual, presentation, etc. (c) activities (d) usage history. Figure 3 depicts the user context as a central core of our context-aware system which is associated to relevant user concept elements such as Impairment, profile, preference, activities, history, etc. and attached to other context such as environment and devices used for interaction. A set of rules must to be defined at the arrival of context event which represent elementary contextual information. An illustration in table 2 is giving demonstrate how content can be adapted according to the arrival of a new user with impairment. Table 2: Illustration of possible content adaptation operations according to user context (extract from [5]) Context type
Context attributes
Content adaptation rules
161
information but a preference is dynamic information which dependent on the current situation or event.
[16].
Our future work will be dedicated to extending the architectural solution proposed in order to support various types of contextual information, and to give a context model based on ontologies in order to identify a set of adaptation rules for various types of disabilities.
[17].
[18].
[19].
REFERENCES [1]. A. Julio, A. Amaia, C. Idoia, G. Borja, G.V. Nestor, M. Raúl, “Automatically Generating Tailored Accessible User Interfaces for Ubiquitous Services”, ASSETS’11, October 24–26, 2011. [2]. A. Julio, A. Amaia, C. Idoia, G. Borja,G. Néstor, M. Raúl, “A modular approach to user interface adaptation for people with disabilities in ubiquitous environments”, [3]. A. Leonidis, M. Antona , and C. Stephanidis, “Rapid Prototyping of Adaptable User Interfaces”, International Journal of Human-Computer Interaction, 2011. [4]. A.Myriam, V. Markel, A. Amaia, A. Julio, “Accessibility Guidelines Management Framework”, HCI (7), pp: 3-10, 2007. [5]. B. Vitor Hugo, “Ontologies for Context-Aware Applications”, Porto, March 2008. [6]. C. Bettini, J. Indulska, D. Nicklas, A. Ranganathan, D. Riboni, et al, “A Survey of Context Modelling and Reasoning Techniques”, In Journal Pervasive and Mobile Computing (2): pp 161-180, 2010. [7]. C. Stephanidis, A. Paramythis, M. Sfyrakis, A. Stergiou, N. Maou, A. Leventis, G. Paparoulis, C. Karagiannidis, “Adaptable And Adaptive User Interfaces For Disabled Users In The Avanti Project”, 5th International Conference On Intelligence In Services And Networks: Technology For Ubiquitous Telecom Services, pp 153-166, 1998. [8]. Constantine Stephanidis, Anthony Savidis, “Universal Access in the Information Society: Methods, Tools, and Interaction Technologies”, UAIS, pp: 40–55, 2001. [9]. D. Akoumianakis, A. Savidis and C. Stephanidis, “Encapsulating Intelligent Interactive Behaviour in Unified User Interface Artefacts”, Interacting With Computers, Vol. 12, No. 4, pp. 383-408, 2000 [10]. D. Thevenin, “Adaptation En Interaction Homme-Machine : Le Cas De La Plasticité”, Phd Thesis, University of Joseph Fourier, Grenoble1, 2001. [11]. D. Thevenin, J. Coutaz, “Plasticity of User Interfaces: Framework and Research Agenda”, Proceedings of Interact99, 1999. [12]. G. Calvary, J. Coutaz, D. Thevenin, Q. Limbourg, N. Souchon, L. Bouillon, M. Florins and J. Vanderdonckt, “Plasticity Of User Interfaces: A Revised Reference Framework”, Proceedings of the first international workshop on Task Models And Diagrams For User Interface Design, 2002, pp 127–134. [13]. G. Calvary, J.Coutaz and D. Thevenin, “Supporting Context Changes For Plastic User Interfaces: A Process And Mechanism”, Proceedings of AFIHM-BCS Conference on Human-Computer Interaction IHM-HCI, 2001. [14]. J.V. Lopez, F. Montero, C. A. Fermandez, and M. Lozano, “Towards Adaptive User Interface Generation: One Step Closer to People”, In Proceedings of ICEIS (3), 2003, pp.97103.
[20].
[21].
[22].
[15]. J.Vanderdonckt, Model Driven Engineering of user interfaces: Promises, Success and failures, Proc of 5th Annual
162
Romanian Conference on Human –Computer Interaction ROCHI2008 , Bucarest, pp 1-10, 2008. K.Z. Gajos, D.S Weld, J.O. Wobbrock, “Automatically generating personalized user interfaces with Supple”, Artificial Intelligence Journal 174(12-13), pp: 910–950, 2010. L.J. Victor, M.S. Francisco, R. Fernando, “Designing user interface adaptation rules with T: XML”, IUI, page 383388, ACM, 2009. M Baldauf, “A survey on context-aware systems”, Int. J. Ad Hoc and Ubiquitous Computing, Vol. 2,. No. 4, pp.263–277, 2007. M. Marco Manca,P. Fabio Paternò, S. Carmen, “Adaptive Mobile Multimodal User Interfaces for Dynamic Accessibility”, Workshop on Mobile Accessibility, CHI 2013, 2013. P. Korpipää, J. Mantyjarvi, J. Kela, H. Keranen, E-J. Malm, “Managing context information in mobile devices”, IEEE Pervasive Computing, Vol. 2, No. 3, July–September, pp.42– 51, 2003. R. Miñón, J. Abascal, A. Aizpurua, I. Cearreta, B.Gamecho, and N. Garay, “Model-Based Accessible User Interface Generation in Ubiquitous Environments”, INTERACT (4) 2011: 572-575. Tarek Chaari, “Adaptation d’applications pervasives dans des environnements multi-contextes”, Phd Thesis, The National Institute of Applied Sciences in Lyon, September 2007.
