The document discusses the development of a blind/deaf communications API (bdc-api) aimed at creating an inclusive educational model for deaf and blind individuals participating in MOOCs. It outlines the architecture and technologies used to facilitate communication and access to digital educational content, highlighting the involvement of multidisciplinary teams in its development. The bdc-api offers solutions such as voice recognition, text-to-speech, and sign language translation to enable effective learning experiences for both communities.

1. EDUCATIONAL CONTENT USING
BLIND/DEAF COMMUNICATIONS API
Paula Escudeiro, Bertil P. Marques
Piedade Carvalho, Ana Barata, Patrícia Queirós
Ana de Sousa, Carlos Dias, Emanuel Rocha, João Ulisses
TEEM2018, 25 October 2018

2. Summary
• Introduction
• MOOC Pedagogical Model
– Implemented Pedagogical Model
– Implemented Solution
– QEF in digital content evaluation
• Educative contents architecture for the
deaf and blind
• Conclusion
2TEEM20182018-10-25

3. Introduction
•The constant need for teachers to be professionally
up-to-date in their knowledge areas and in
pedagogical approaches.
•Massive Open Online Course (MOOC) are inclusive
and innovating pedagogical model that started being
devised, having as premise the at distant
educational model.
32018-10-25 TEEM2018

4. Introduction
• In these specific MOOCs it is intended to create an
inclusive model which will allow the participation of
deaf and/of blind people. The authors are innovating
in the Educational field.
• This paper present the Blind/Deaf Communications
API (BDC-API) architecture which translates digital
educational content for the deaf and for the blind.
42018-10-25 TEEM2018

5. MOOC Pedagogical Model
5
• The development of the MOOC
pedagogical model introduced in this paper
comprehends two Postgraduation degrees:
– Informatics in Education (IE);
– Supporting Technologies for Education
(TAE).
2018-10-25 TEEM2018

6. Implemented Pedagogical Model
6
• The involvement of multidisciplinary teams, made up of
professionals with distinct training, allows the
construction of a balanced educational environment,
which fosters an accessible, inclusive and innovative
pedagogical environment.
• There was especial focus on the team to guarantee the
active participation of all individuals in accordance with
the objectives previously set and the defined plan.
• The cohesion and compromise of the team enable a
consistent development which will lead to a product
totally adapted to a specific target.
2018-10-25 TEEM2018

7. Implemented Pedagogical Model
7
Figure 1 – Active intervenients in the innovative MOOC pedagogical model
TEEM20182018-10-25

8. Implemented Solution
8
Figure 2 – Design solution
2018-10-25 TEEM2018

9. QEF in digital content evaluation
9
The QEF (Quantitative Evaluation Framework) approach
has been developed to highlight the strengths and
limitations of the pedagogical models. Clearly indicate
advantages, limitations and possible applications.
• This evaluation framework will enable controlling the
quality of the final product by controlling its
development process.
2018-10-25 TEEM2018

10. Educative contents architecture
for the deaf and blind
10
• For people with hearing impairment the problem arises through
the fact that the deaf community uses its proper grammar in any
written language, meaning the majority are unable to read
written text.
• It must also be noted that blind people rely on hearing cues to
communicate while deaf people rely on visual representations.
• These different channels prevent any type of communication
between these two communities, meaning that separate
solutions must be drafted for each community.
• The architecture presented in Ulisses et al. seeks to solve these
communication problems most particularly in the area of
education.
2018-10-25 TEEM2018

11. Educative contents architecture
for the deaf and blind
11
• The BDC-API extracts the best technology in order to create a
simple tool ready to be used by teachers, serious game
developers and students.
• The technologies used include IBM Watson for voice-recognition
and text-to-speech, Google’s Web Speech API for voice-
recognition and text-to-speech, VirtualSign for the 3D avatar
performing Sign Language, MeaningCloud for Sign Language
grammar recognition and translation, NLTK and Floresta for Sign
Language grammar recognition and translation, Unity and Godot
Game Engines to execute the application in all platforms and PHP
with PDO for data communication through all databases.
2018-10-25 TEEM2018

12. Educative contents architecture
for the deaf and blind
12
Figure 3 – BDC-API translation modules for educational content.
2018-10-25 TEEM2018

13. Conclusion
13
- This paper has presented the BDC-API architecture
applied to an integrative MOOC.
- BDC-API architecture translates digital educational
material for the deaf and for the blind, who otherwise
would be unable to have full access to its content.
- Educational content on PowerPoint can have official
Microsoft plugins that can call BDC-API translation
methods to translate automatically.
2018-10-25 TEEM2018

14. Conclusion
EL2018 14
• The blind BDC-API uses Web Speech API for Voice
Recognition to allow the blind to interact with the
machine, and text-to-speech to give information and
to generate sounds in BDC-API.
• Regarding the deaf, BDC-API makes use of Virtual Sign
avatar and sign language grammar to make the
translations so that the deaf can understand the
content.
• Both cases allow bidirectional communication in any
language, which could be used in classrooms.
2018-10-25 TEEM2018

15. Thank you!
 bpm@isep.ipp.pt
24-26 October 2018