University of BonnFraunhofer IAIS
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Mirette Elias, Steffen Lohmann, Sören Auer

The 8th International Conference on Knowledge Engineering and Semantic Web (KESW 2017)
November 08, 2017SZCZECIN, POLAND
https://slidewiki.eu/wp-content/uploads/2016/04/cropped-slidewiki-1.png


  • Collaborative OpenCourseWare (OCW) authoring, multiplatform delivery and learning analytics
  • 17 European partners (platform development, contents, trials, etc.)
  • Collaborative creation (and hosting) of OCW material
  • Wikipedia for Slides (OCW)
https://slidewiki.eu/wp-content/uploads/2016/04/cropped-slidewiki-1.png
https://slidewiki.eu/wp-content/uploads/2016/02/csm_SlideWikiHome_4906ccf19e.jpg
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

SlideWiki overview 1 and 2 slides are  simple introduction to SlideWiki platform. After listing the main concept and functionality, we should focus on Accessibility as an important concern and that's why this paper was written.


  • Collaborative editing
  • Multi-device delivery
  • Semantic annotation
  • Semi-automatic translation
  • Commenting, discussion
  • Question generation
  • Accessibility
https://slidewiki.eu/wp-content/uploads/2016/04/cropped-slidewiki-1.png
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

SlideWiki overview 1 and 2 slides are  simple introduction to SlideWiki platform. After listing the main concept and functionality, we should focus on Accessibility as an important concern and that's why this paper was written.


Problem Definition

Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

Open Educational Resources (OER) are openly licensed and freely accessible learning materials that can be used in e-learning contexts and beyond. Often, OER are published on the Web in the form of OpenCourseWare (OCW) organized in courses and complemented by tools for collaboration and evaluation. OCW systems thus provide means for distributing free educational content to awide range of learners over the Web. These learners include people with disabilities who have diverse needs, in terms of the type and severity of their disabilities, which must be addressed by OCW systems that aim to be inclusive. 

Designing one system that meets the needs of all learners is usually not possible, as learners (disabled) have different needs and preferences. Even users with same disability have different preferences. For example, one blind user might want to use a screen reader, while another blind user might prefer a braille display or both might want to use the same device but with different configurations (e.g., different text reading speeds).


Problem Definition

How to define and represent these needs?

How can OCW reflect these needs?

Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

We have two research questions:

1. how to define and represent the needs and preferences of these type of learners, and put this knowledge in a structure that can be processed later? 

2. how to used this knowledge to customize OCW with respect to the learner preferences?

This paper answers the first question while the second question will be considered as a future work


Main Objective

Using ontologies to represent accessibility needs and preferences
of learners to structure the knowledge and to access the information
of recommendations and adaptations in OCW context
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

We are using ontologies to represent the accessibility concepts in a structured way in order to enable access to this knowledge for recommendations and adaptations in OpenCourseWare systems.

This is done by extending the ACCESSIBLE ontology  with concepts of IMS AfA. We decided to work on this ontology and guidelines based on a survey we established previously in the paper " Fostering accessibility for OpenCourseWare with semantic technologies - a literature review. In: 7th International Conference on Knowledge Engineering and Semantic Web (KESW ’16). pp. 241–256. Springer (2016) "

ACCESSIBLE ontology is the output of an EU project that comprises characteristics of disabled users according to the “International Classification of Functioning, Disability and Health (ICF)” of the WHO, descriptions of assistive devices and software applications, web accessibility standards and guidelines (WAI-ARIA and WCAG 2.0), as well as assessment rules for mapping user requirements and constraints. 

IMS AfA is a guideline and metadata specification, based on the ISO/IEC24751-1:2008 standard, for developing accessible e-learning applications and resources with respect to the learner needs and preferences. It links the accessibility preferences of a learner through the AfA Personal Needs & Preferences (PNP) model to the learning objects defined by the AfA Digital Resource Description (DRD). 

 


System Architecture

  • AccessibleOCW ontology
  • Learner profile
  • OCW components
  • Personalization module
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

The overall system architecture of the envisioned adaptive accessibility component of the SlideWiki system is thus composed of an ontology (which we call the AccessibleOCW ontology), learner profiles, representations of the OCW components, and a personalization module.

1. AccessibleOCW ontology: our proposed ontology that contains the relevant accessibility knowledge required in OCW contexts (e.g., disability types, assistive technologies, accessibility guidelines, e-learning standards, etc.) 

2.  Learner profile: are presentation of learners, including information about their disabilities. The disabilities can be either automatically detected or manually entered by the learners (or their caregivers) using a form. Other needs and preferences of the learners can then be inferred from their disabilities using the ontology.

3. OCW components: these are the website, educational resources, assessments, and collaboration tools. The OCW components are adapted according to the learner profiles.

4. Personalization module: processes the profile and ontology as an input to retrieve the preferences, assistive technology requirements, accessibility guidelines and standards related to the learners. Based on this information, it adapts the OCW content to the individual needs and preferences of the learners, and suggests the most appropriate educational resources to them.
 

 


Methodology

  1. Defining and representing learner needs and preferences
  2. Representing web accessibility needs
  3. Creating and managing educational resources
  4. Adapting the website
  5. Assessing design
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

This slide is not important, may be removed


AccessibleOCW Ontology

Extend ACCESSIBLE ontology with the IMS Access for All (AFA) concepts
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

The ontology is available on VoCol and GitHub.  I will use VOWL link to visualize and explain the ontology main concepts.

We have two main Classes:

1. Learner (which extends the User class of the ACCESSIBLE ontology in addition to the IMS AfA PNP properties) . The Learner class is defined by a number of properties: hasRequiredAccessMode is the access mode required by a learner; here, the learners define their preference for the resource representation. For instance, a learner may define that if the resource is a visual one, she would prefer a textual representation. 

2. DigitalResource  defines the properties of an individual educational resource. The properties are designed based on the IMS DRD specification. They describe the resource access mode (e.g., visual, auditory) and define if a resource has adapted versions (e.g., if there is an alternative transcript file to an auditory file). 

The Learner class is linked with the DigitalResource class through an hasAccess relation, which is used to filter the accessible educational resources according to the user-defined disabilities and preferences


Learner Example

:Learner_1 a owl:NamedIndividual , :Learner;
  GenericOntology:hasName "Ms. Martinez"ˆˆxsd:string;
  GenericOntology:hasAge "62"ˆˆxsd:int;
  GenericOntology:User_has_Disability :Deafness;
  :hasLanguageOfAdaptation "English"ˆˆxsd:string;
  :hasLanguageOfInterface "English"ˆˆxsd:string;
  :hasReqAccessMode :auditory_textual;
  :hasReqAdaptationDetail :auditory_verbatim;
  :hasReqAdaptationType :auditory_caption.
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

The Learner instances are created from the examples in 

Learner 1:  Ms. Martinez. The properties and classes of the ACCESSIBLE ontology start with the term GenericOntology. For example, we use the property ”User has Disability” from the ACCESSIBLE ontology to define the user’s disability. This property has a well-defined list of disabilities with respect to the ICF standard classification, as mentioned before. When this property is defined, other properties can be concluded from the ACCESSIBLE ontology, such as the devices that can be used by this type of disability, and the limitations resulting from this disability. The remaining properties are defined for IMS purposes; some of these properties require input by the user, while others can be concluded and recommended from the context or ACCESSIBLE ontology. For example, the hasReqAccessMode of this persona requires textual representations for resources that are of auditory type. 


Digital Resource Example

:digitalResource1 rdf:type owl:NamedIndividual ,
    :DigitalResource ;
  :hasAccessMode :visual ;
  :hasControlFlexibility :fullKeyboardControl ;
  :hasDisplayTransformability :backgroundColour;
  :hasEducationalComplexityOfAdaptation_dr
    :enriched ;
  :hashazard :flashing ;
  :isAtInteroperable_dr "true"ˆˆxsd:boolean ;
  :hasAdaptation "digitalResource2_URI"ˆˆxsd:anyURI ,
    "digitalResource3_URI"ˆˆxsd:anyURI .
 
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

 The source file digitalResource1 is a video with a visual access mode property. This digital resource is available in two alternative forms (in the paper) for better accessibility: digitalResource2 is a textual resource that can be tactually accessed, and digitalResource3 is an auditory resource with full keyboard control. 


Applying the Ontology

Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

We designed an input dialogue to collect the disability information of learners. The learners are supposed to enter their profile information, which could also be automatically inferred from the context in parts. The form of the input dialogue is generated from our AccessibleOCW ontology using SPARQL queries. It asks the learners to input their disabilities, available devices, and personal preferences.

We implemented a prototype of the dialogue using the JavaScript library React.

We used SPARQL to query the ontology, using Fuseki as the SPARQL server. We created a SPARQL client in React that accesses our ontology on the Fuseki server. 


Learner Profile

PREFIX acc: <http://www.Acces[...]ogy.com/GenericOntology.owl#>.
PREFIX acc_ocw: <http://purl.org/accessible_ocw#>.
SELECT ?impairment ?disability ?device ?ims_AT
WHERE {
?impairment a acc:Impairment. ?disability acc:Disability_belongsTo_Impairment ?impairment.
OPTIONAL {?device acc:Device_belongsTo_Disability ?disability}.
OPTIONAL {?disability acc:Disability_has_Device ?device}. OPTIONAL {?ims_AT a acc_ocw:isAtInteroperable}.
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

This is the interface of the learner profile . This is a sample of a SPARQL query which feeds the Impairment/Disability/Assistive Technology dropdowns


Conclusions & Future work

  • AccessibleOCW ontology was proposed to address the accessibility need of an OCW.
  • A Learner profile was created a a proof of concept
  • Future work
    1. Integrating further accessibility guidelines and techniques into the ontology (i.e., users with cognitive disabilities)
    2. Working on the personalization module to adapt the content presentation according to the learners’ needs.
Ontology-based Representation of Learner Profiles for Accessible OpenCourseWare Systems

Speaker Notes

we presented an ontology that addresses accessibility in OCW systems. We reused and extended the ACCESSIBLE ontology to represent learner needs and preferences with respect to the accessibility requirements of the IMS AfA specifications. IMS AfA is concerned with annotating digital resources and learner preferences to achieve a better accessibility. Combining it with the ACCESSIBLE ontology makes it more extendable and does not limit it to special types of disabilities. The combination of IMS AfA and the ACCESSIBLE ontology provides more detailed descriptions of disabilities, assistive technologies, and user preferences. Furthermore, it allows to add concepts of other disabilities, such as cognitive impairments, which are relevant in learning contexts, and suggests mappings to educational resources. 




Creator: mirette

Contributors:
clange


Licensed under the Creative Commons
Attribution ShareAlike CC-BY-SA license


This deck was created using SlideWiki.