Disability Advocacy using a Smart Virtual Community

Bushra Kundi

, Dhayananth Dharmalingam

, Rediet Tadesse

, Alexandra Creighton

Rachel Gorman

, Pierre Maret

2,5 f

, Fabrice Muhlenbach

, Alexis Buettgen

Enakshi Dua

, Thumeka Mgwigwi

, Serban Dinca-Panaitescu

and Christo El Morr

School of Health Policy and Management, York University, Keele Street, Toronto, Canada

Université Jean Monnet Saint Etienne, France

School of Gender, Sexuality and Women's Studies, York University, Canada

Student Learning and Academic Success Department, York University Libraries, York University, Toronto, ON, Canada

The QA Company, Saint Etienne, France

dhayananth.dharmalingam@etu.univ-st-etienne.fr, Gorman@yorku.ca, pierre.maret@univ-st-etienne.fr,

fabrice.muhlenbach@univ-st-etienne.fr, a.buettgen@gmail.com, edua@yorku.ca, thumekam@yorku.ca, serband@yorku.ca,

elmorr@yorku.ca

Keywords: Health Informatics, Machine Learning, Semantic Web, Natural Language Processing, Disability, Critical

Disability, Wikibase, Disability Advocacy, Disability Rights.

Abstract: The lack of readily available disability data is a major barrier for disability advocacy globally. The collection

and access to disability data is crucial to address social inequities, discrimination, and human rights violations

within the disability community. The Disability Wiki project intends to use AI techniques such as Machine

Learning and Semantic Web to extract and store existing disability-related data into one platform (Wikibase)

and to provide a multilingual natural language enabled search engine and a screen-reader-accessible for its

users.

1 INTRODUCTION

The lack of availability of disability data is a major

barrier for disability advocacy and disability rights

monitoring globally. Disability advocacy information

and data is not readily available which can further

delay important projects and commissions by human-

rights and non-governmental organizations (NGO)

(Loeb, 2013).

It is vital to have readily available structured and

unstructured disability data from a critical disability

https://orcid.org/0000-0001-9581-8531

https://orcid.org/0000-0002-4682-8992

https://orcid.org/0000-0002-7179-5555

https://orcid.org/0000-0002-7573-221X

https://orcid.org/0000-0003-3144-1733

https://orcid.org/0000-0001-5437-7725

https://orcid.org/0000-0002-1825-4290

https://orcid.org/0000-0002-7370-723X

https://orcid.org/0000-0003-2554-2028

https://orcid.org/0000-0001-9757-8084

https://orcid.org/0000-0003-2620-4389

https://orcid.org/0000-0001-6287-3438

perspective to track systemic discrimination, social

exclusion, adverse socioeconomic outcomes, and

social inequity for persons with disabilities. Health

informatics (El Morr, 2014, 2018) and artificial

intelligence techniques (Akerkar et al., 2012;

Diefenbach et al., 2017; Diefenbach et al., 2019; El

Morr & Ali-Hassan, 2019b; Gorman et al., 2021;

Vercouter & Maret, 2012) provide helpful tools to

organize, store and search for disability data using

Natural Language Processing (NLP). Also, virtual

communities that allow remote members to

316

Kundi, B., Dharmalingam, D., Tadesse, R., Creighton, A., Gorman, R., Maret, P., Muhlenbach, F., Buettgen, A., Dua, E., Mgwigwi, T., Dinca-Panaitescu, S. and El Morr, C.

Disability Advocacy using a Smart Virtual Community.

DOI: 10.5220/0010748600003123

In Proceedings of the 15th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2022) - Volume 5: HEALTHINF, pages 316-319

ISBN: 978-989-758-552-4; ISSN: 2184-4305

collaborate over a task have been used successfully in

health (El Morr, 2010; El Morr et al., 2017) and search

techniques benefited from artificial intelligence to

uncover bibliographical data (Muhlenbach & Lallich,

2010; Muhlenbach & Sayn, 2019).

The Disability Wiki project aims to create a virtual

community consisting of a multilingual web site with

screen-reader-accessibility and a semantic data

storage (wikibase). The system is endowed with a

document upload function with hybrid (automated

and manual) paragraph tagging and the querying

function implements an intelligent natural language-

based search system.

The Disability Wiki project includes 3 phases:

ontology development and fine tuning (Phase I),

Wikibase and search engine design, development,

and testing (Phase II), use and use analysis (Phase

III). Phases I and II are presented below

2 METHODS

A computer ontology describing the Convention on

the Rights of Persons with Disabilities (CRPD) was

developed through constant deliberation and

coordination between experts in the field of critical

disability studies, health informatics and computer

science. The ontology development process was

iterative and allowed the critical disability studies

experts and health informatics and computer science

professionals to fine tune and clarify information

description and user requirements. The development

of the ontology served as a blueprint for the Disability

Wiki website.

Background technology used included use of

QAnswer, a platform that makes Resource

Description Framework (RDF) data accessible via

natural language. QAnswer enables intelligent search

using intelligent natural language questioning

(Diefenbach et al., 2020; Diefenbach et al., 2019;

QAnswer, 2021). QAnswer is the first artificial

intelligence (AI) driven platform to query RDF data

stores in natural language using semantic

technologies (QAnswer, 2021). Python and R were

both used to prepare and program machine learning

models(El Morr & Ali-Hassan, 2019a).

2.1 Glossary & File Tagging

A set of disability rights reports were chosen to train

the software for knowledge acquisition, i.e. paragraph

splitting and tagging. A glossary of tag terms was then

created by experts in the field of health informatics and

critical disability to aid in tagging the disability rights

reports. Synonyms were then added to represent a wide

variety of tag terms. The glossary was continuously

edited and fine-tuned to meet the standards of the

disability rights reports and critical disability studies

field until expert satisfaction was reached.

To ensure knowledge acquisition, paragraphs

were extracted from disability rights reports and a

machine learning model was trained to predict the

paragraphs’ semantic meaning and tag it with the

appropriate glossary terms.

Extracted paragraphs were initially automatically

tagged using the glossary, then manually re-checked,

and re-tagged to train the machine learning model

responsible for automatic paragraph tagging.

2.2 Disability Wiki Website

The Disability Wiki virtual community website was

developed on top of an instance of a Wikibase

dedicated to the data storage). It allowed for

authenticated users (i.e., information producers) to

upload disability documents and use the machine

learning to tag their paragraphs. Information producers

could delete or add tags if they deem necessary, and the

new tag set can serve for new training of the machine

learning model responsible of tagging.

Once tagging is done the administrator can push

the documents and tags to the wikidata storage.

Using the website search feature, public users can

access disability data and information with Natural

Language questions. No authentication is needed for

public users.

3 RESULTS

The CRPD ontology described relationships between

entities in the CRPD. For instance, the CRPD is a

convention developed by the United Nations and is

composed of several articles, each article has one or

several topics. Figure 1 shows a partial view of the

ontology.

Figure 1: Partial view of the CRPD ontology.

Disability Advocacy using a Smart Virtual Community

317

Figure 2 shows part of the glossary that was

created for tagging the disability rights reports.

Column A shows the primary tags that were used to

tag paragraphs while the synonyms representing the

terms are in the columns that follow.

Figure 2: Glossary.

Figure 3 shows a page of the Disability Wiki

project web site. The image shows the search bar with

examples of questions.

Figure 3: Disability Wiki Web Page.

The disability wiki website uses QAnswer API

endpoints to send the questions to and retrieve

answers from the disability wiki.

Figure 4: Search function flow of the website.

4 DISCUSSIONS

The multidisciplinary nature of critical disability

studies, health informatics and computer science

required the researchers to deliberate and coordinate

on every stage of the project. To reach a mutual

agreement, the project required multiple meeting

sessions and re-iteration between researchers for

unexpected delays, deadlines, clarification, or

challenges.

In the domain of critical disability studies, we are

not aware of any other tool or effort for collecting and

giving easy access to related data. Our proposal

therefore responds to this need.

Giving access to data through the Web is of course

not new. However very topic-focused data is too

specific to be adequately made accessible by

generalist search engines such as google. Also, the

online encyclopedia Wikipedia.org is not suited

because it requires editing documents, which is not

the case on our platform (document upload).

The web site www.enslaved.org is an example of

domain related data collection, however document

upload is not proposed, nor automated tagging or

semantic free text search.

Future evaluation for the use of the virtual

community using technology Technology

Acceptance Model or the RE-AIM Framework

would be important to understand users acceptance

(El Morr et al., 2017). A user interface evaluation is

underway to ensure accessibility of the platform.

Finally, the platform supports English and French,

and is already expanding in scope as we are currently

adding data and documents pertaining to Black

Disability Studies and Disability Justice.

5 CONCLUSIONS

Readily available and easy to access disability data is

an essential need today. The social model of critical

disability allows us to see and address the

discrepancies, discrimination, and inequities in the

system for persons with disabilities.

The Disability Wiki virtual community uses AI

techniques to easily integrate and index existing

disability data into one accessible platform as well as

allowing natural language searching for its’ users.

The platform is currently in a pilot stage.

ACKNOWLEDGEMENTS

The authors acknowledge and thank The QA

Company for providing access to their technology.

FUNDING

The work reported in this paper was funded by the

Social Sciences and Humanities Research Council

(SSHRC), Insight Development Grant; Grant number

430-2020-00819.

REFERENCES

Akerkar, R., Maret, P., & Vercouter, L. (2012). Web

intelligence and communities Proceedings of the 4th

HEALTHINF 2022 - 15th International Conference on Health Informatics

318

International Workshop on Web Intelligence &

Communities, Lyon, France. https://doi.org/10.1145/

2189736.2189738

Diefenbach, D., Amjad, S., Both, A., Singh, K., & Maret,

P. (2017). Trill: A Reusable Front-End for QA Systems.

In E. Blomqvist, K. Hose, H. Paulheim, A.

Ławrynowicz, F. Ciravegna, & O. Hartig, The Semantic

Web: ESWC 2017 Satellite Events Cham.

Diefenbach, D., Giménez-García, J., Both, A., Singh, K., &

Maret, P. (2020, 2020//). QAnswer KG: Designing a

Portable Question Answering System over RDF Data.

The Semantic Web, Cham.

Diefenbach, D., Migliatti, P. H., Qawasmeh, O., Lully, V.,

Singh, K., & Maret, P. (2019). QAnswer: A Question

Answering prototype bridging the gap between a

considerable part of the LOD cloud and end-users The

World Wide Web Conference, San Francisco, CA,

USA. https://doi.org/10.1145/3308558.3314124

El Morr, C. (2010). Health Care Virtual Communities:

challenges and opportunities. In M. M. Cruz-Cunha, A.

J. Tavares, & R. Simoes (Eds.), Handbook of Research

on Developments in E-Health and Telemedicine (pp.

278-298). IGI Global. https://doi.org/10.4018/978-1-

61520-670-4.ch013

El Morr, C. (Ed.). (2014). Research Perspectives on the

Role of Informatics in Health Policy and Management.

IGI Global. https://doi.org/10.4018/978-1-4666-4321-5.

El Morr, C. (2018). Introduction to Health Informatics: A

Canadian Perspective. Canadian Scholars’ Press.

https://www.canadianscholars.ca/books/introduction-

to-health-informatics

El Morr, C., & Ali-Hassan, H. (2019a). Analytics in

Healthcare: A Practical Introduction. Springer

International Publishing. https://doi.org/10.1007/978-

3-030-04506-7

El Morr, C., & Ali-Hassan, H. (2019b). Descriptive,

Predictive, and Prescriptive Analytics. In Analytics in

Healthcare: A Practical Introduction (pp. 31-55).

Springer International Publishing. https://doi.org/

10.1007/978-3-030-04506-7

El Morr, C., Maule, C., Ashfaq, I., Ritvo, P., & Ahmad, F.

(2017, February 16 -19, 2017). A Student-Centered

Mental Health Virtual Community Needs and Features:

A Focus Group Study. Information Technology and

Communications in Health (ITCH 2017), Victoria, BC,

Canada.

Gorman, R., Maret, P., Creighton, A., Kundi, B.,

Muhlenbach, F., Buettgen, A., Dua, E., Reaume, G.,

Mgwigwi, T., Dinca-Panaitescu, S., & El Morr, C.

(2021). The Potential of an Artificial Intelligence for

Disability Advocacy: The WikiDisability Project. Stud

Health Technol Inform, 281, 1025-1026.

https://doi.org/10.3233/shti210338

Loeb, M. (2013). Disability statistics: an integral but

missing (and misunderstood) component of

development work. Nordic journal of human rights,

31(3), 306-324. https://www.ncbi.nlm.nih.gov/pmc/

articles/PMC4766593/

Muhlenbach, F., & Lallich, S. (2010). Discovering

Research Communities by Clustering Bibliographical

Data Proceedings of the 2010 IEEE/WIC/ACM

International Conference on Web Intelligence and

Intelligent Agent Technology - Volume 01,

https://doi.org/10.1109/WI-IAT.2010.117

Muhlenbach, F., & Sayn, I. (2019). Artificial Intelligence

and Law: What Do People Really Want?: Example of a

French Multidisciplinary Working Group Proceedings

of the Seventeenth International Conference on

Artificial Intelligence and Law, Montreal, QC, Canada.

https://doi.org/10.1145/3322640.3326722

QAnswer. (2021). Accessing your Knowledge via Natural

Language. QAnswer.eu. Retrieved July 15, 2021 from

https://www.qanswer.eu/

Vercouter, L., & Maret, P. (2012). Introducing Web

Intelligence for communities. Web Intelligence and

Agent Systems: An International Journal, 10, 91-92.

https://doi.org/10.3233/WIA-2012-0233

Disability Advocacy using a Smart Virtual Community

319