Conference PaperPDF Available

A Voting System for AAC Symbol Acceptance



This paper aims to illustrate how an innovative voting system has been developed to allow AAC users, their therapists, carers and families show their degree of acceptance for newly developed symbols and their referents. The approach, taking a participatory model of research, occurs via an online symbol management system using a set of criteria that provide instant feedback to the developers and the project team. Scores and comments regarding the symbols are collated and where a majority vote has occurred, symbols are added to the Arabic Symbol Dictionary with lexical entries in both Arabic and English.
A Voting System for AAC Symbol Acceptance
E.A. Draffan, Mike Wald,
Nawar Halabi
University of Southampton
Southampton, SO17 1BJ, UK
+44 (0)23 8059 7246
Amatullah Kadous
Amal Idris
Speech Therapy Dept., Hamad
Medical Corporation (Qatar)
Nadine Zeinoun,
David Banes, Dana Lawand
This paper aims to illustrate how an innovative voting system has
been developed to allow AAC users, their therapists, carers and
families show their degree of acceptance for newly developed
symbols and their referents. The approach, taking a participatory
model of research, occurs via an online symbol management
system using a set of criteria that provide instant feedback to the
developers and the project team. Scores and comments regarding
the symbols are collated and where a majority vote has occurred,
symbols are added to the Arabic Symbol Dictionary with lexical
entries in both Arabic and English.
Categories and Subject Descriptors
H.5.2 [User Interfaces]: Graphical User Interfaces; K.4.2 [Social
Issues]: Assistive Technologies for Persons with Disabilities
General Terms
Design, Human Factors
Augmentative and Alternative Communication (AAC), language,
voting, symbol acceptance, Arabic, dictionary
The voting system for symbol acceptance presented in this paper
has become part of a participatory approach adopted for the
development of an Arabic Symbol Dictionary for Augmentative
and Alternative Communication (AAC) users. The contents of the
dictionary are intended to support not only communication skills
but also literacy skills, general signage and language learning.
Crowdsourcing methods have been used in the past to ‘create
fictional but plausible AAC communications’ [1] but never to
engage AAC users, their families, carers and therapists in
evaluating the design of the visual representation of a symbol in
relation to its referent in a bilingual and diglossia situation.
The development of an Arabic Symbol Dictionary was felt to be
essential for the support of those working in the field of AAC with
Arabic speaking families, who at present are mainly supported by
English speaking therapists and specialist teachers using
westernized symbols. There is also a limited public availability of
Modern Standard Arabic lexicons or colloquial Arabic phrases for
communication for AAC use. AAC centers tend to develop their
own libraries of symbols which provide perfect personalization
for individual users, but do not offer options for symbols to be
shared across networks or for their use by others in the area.
Huer [2] points out that communication across cultures reveal
that nonsymbolic as well as symbolic forms of communication are
culturally dependent” and that “participants’ cultural/linguistic
experiences may be significant elements to consider when
selecting graphic symbols and when teaching consumers to
represent meaning through them.” [3]
An online forum of AAC users, their therapists, carers and
families was set up and discussions held about the best way to
encourage participation in choices for newly developed symbols
and an Arabic core vocabulary to complement the often used PCS
Symbols from Mayer-Johnson. An overarching Symbol
Management system was developed to host lexical entries and
culturally adapted symbols along with definitions, various
categories and a filtering system. A symbol rating system was
added with four voting criteria each with a 1-5 Likert scale;
participants logged in to cast their votes and make comments.
Individual symbol scores and comments were made available via
the Symbol Management System and Microsoft Excel was used to
aggregate the data.
3.1 Development of the Symbol Management
The online Symbol Management system was developed using the
Nodejs framework, written in JavaScript using an open source,
cross-platform framework for building network applications. On
top of Node, the Expressjs plugin was used which turns Node into
a Model-View-Controller (MVC) framework for web
development. The database used was Mongo dB, which is also
open source and can easily be interfaced from Node.
3.2 Symbol voting
At the outset participants were made aware of the differences
between the newly developed Arabic culturally sensitive symbols
with which they were presented, in comparison to the symbols in
daily use. Examples of differences came in the Arabic dress, food
types, places, religion, use of a right hand for actions and arrows
for past and future in opposite directions to English. There was an
initial trial voting session that used PCS, ARASAAC and the new
symbols on one page as a comparison. This proved to be too
complex and further discussions with the offer of alternative
activities and some rapid prototyping [4] resulted in a final much
Permission to make digital or hard cop ies of part or all of this work for personal
or classroom use is granted without fee provided that copies are not made or
distributed for profit or commercial advantage and that copies bear this notice
and the full citation on the first page. Copyrights for third-party components of
this work must be honored. For all other uses, contact the Owner/Author.
Copyright is held by the owner/author(s).
ASSETS '15, October 26-28, 2015, Lisbon, Portugal
ACM 978-1-4503-3400-6/15/10.
simplified voting system. This had a simple ‘flashcard’ interface
with a large version of a single symbol plus increased sized fonts
for lexical entries in MSA, Qatari and English above the voting
criteria. The four criteria on which participants voted using the
check boxes with a scale of 1-5 (where 5 was completely
acceptable) were:
1. Feelings about the symbol as a whole
2. Represents the word or phrase
3. Color contrast
4. Cultural sensitivity
Plus the addition of any comments to help improve the symbol
followed by large submit, previous or next buttons (Figure 2) so
that voters could choose to retrace their steps to change votes (the
final vote being the only one recorded). The whole system was
also translated into Arabic via localization files. Each symbol was
offered as one web page and could be activated via touch screen,
mouse, keyboard or switch access.
Figure 2. Voting system showing a black and white symbol for
a prayer time with the criteria on a scale of 1-5 where five is
completely acceptable.
Two young AAC users (who had cerebral palsy) aged 8 and 14
were also introduced to 21 adapted symbols using the same
interface described above but via Grid 2 as their preferred AAC
software. Rather than checkboxes they were presented with
thumbs up, halfway and down to represent a simplified rating of
1-3 or with two intermediary thumb positions for a rating of 1-5.
More AAC users and their families have only recently joined the
3.3 Results of individual symbol voting.
The first batch of symbols to be adapted had 63 participants
logging into the Symbol Manager resulting in 2341 votes for
65 symbols. The votes collected for the four different criteria, as
illustrated in Table 1, showed that all mean ratings were
significantly greater than a rating of 3.5 denoting general
acceptance of the initial batch of symbols. Comments also showed
most preferred colored symbols and wanted gender illustrated for
verbs rather than stick characters and liked different clothing
styles for example it was deemed “less distracting”, “I like both,
but prefer option 1 for Qatar” (voted for just Qatari dress) “one
uncovered”, “make one of them dressed in Abaya”, “Make one of
the girls wear abaya and one of the males wear a thowb”
Table 1. One Sample T test for Difference of Mean Ratings
from 3.5
Criteria Number
of voters Mean rating
2 tail P
Value for
from 3.5
1 63 3.94 <0.0001
2 63 3.90 <0.0001
3 63 4.07 <0.0001
4 63 4.10 <0.0001
Where the average scoring for an individual symbol was lower
than 3.5 out of 5, it was redrawn, taking into account any
comments and resubmitted to a following batch of symbols for
another vote.
The voting system produced speedy results and has encouraged
participants to complete batches of up to 60 symbols within 40
minute sessions. The AAC users have had the support of a parent
or teacher during their sessions and used eye gaze equipment to
access their computers. It took longer for them to complete their
21 symbols and the results were collected on paper and kept
separate from the general voting system. Stripping out interface
complexities and providing good feedback with speedy results has
been essential to ensure voter uptake remains high. The team
learnt much from the process with comments such as: “It is
amazing to see how one symbol can mean so many different
things to so many people and such small details can be offensive
to one ethnicity and small changes can be made to please
another.” Finally it is felt that this participatory methodology
could be used for a wide range of projects where consensus needs
to be achieved quickly with participants from different cultures,
countries, abilities and skills.
This research was made possible by the NPRP award [NPRP 6 -
1046 - 2 - 427] from the Qatar National Research Fund (a member
of The Qatar Foundation) and thanks must go to all those
participants in Doha and ARASAAC for allowing their symbols
to be used in this project. The statements made herein are solely
the responsibility of the authors.
[1] Vertanen, K. & Kristensson, P.O. (2011) The Imagination of
Crowds: Conversational AAC Language Modeling using
crowdsourcing and large data sets. Computational
Linguistics, pp.700711.
[2] Huer, M. B. (1997). Culturally inclusive assessments for
children using augmentative and alternative communication
(AAC). Journal of Children’s Communication Development,
19 (1), 23–34.
[3] Huer, M. B. (2000). Examining perceptions of graphic
symbols across cultures: Preliminary study of the impact of
culture/ethnicity. Augmentative and Alternative
Communication 16 (3): 180185.
[4] Kane, S. et al. (2012) What We Talk About: Designing a
Context-Aware Communication Tool for People with
Aphasia. ASSETS '12, ACM Press (2012), 49-56
... Project 1 engaged participants in discussions about the design of the symbol acceptance systems and presentation models. Initial trials showed that both users and experts required a very simplistic flash card type system to make choices with an easy 5-point Likert scale for the experts and an even easier 3-point scale for AAC users (Draffan et al, 2015). Both options allowed for additional comments and could be used online and on paper. ...
Purpose – A number of participatory research methodologies can be used to assist with developing assistive technologies. These methods vary in the amount that users lead and contribute to the work. Selecting the correct method can be important to ensure the overall success of the project and the engagement of users. The purpose of this paper is to explore factors that can impact on the degree of user participation. Design/methodology/approach – The paper considers whether criteria, that might influence assistive technology (AT) selection made during an assessment of need, review or purchasing process, could also be used to clarify the appropriate strategies for user involvement when developing assistive technologies. It outlines how this approach has been applied to two research and development projects which aimed to improve AT provision within niche markets. Findings – The paper demonstrates that it is possible to apply a decision making process to selecting the best participatory research method, based on factors affecting AT need. It reports on the outcomes of the user participation in the two research and development projects and discusses how this design approach has been applied to a third project. Originality/value – By examining a possible framework for identifying appropriate user-participation approaches, this paper will aid those designing research and development AT projects, whilst encouraging user participation within similar projects.
Pictographic symbols may be used as an alternative form of communication by individuals with complex communication needs. These symbols are a collocation of drawings that depict concepts often associated with glosses that express the meaning in spoken language. This research investigates the problem by focusing on one particular language, Modern Standard Arabic, and one set of symbols, the ARASAAC set. The outcomes are generalisable to other symbols sets and spoken languages. Symbols can be used as part of an electronic speech generating devices. Users may use symbols to express messages or to understand received messages. Thus, translating to and from text is an important task that increases communication with a wider community. Translating text to symbols requires awareness of the exact sense of the textual units that are part of the input message, to determine relevant symbols. Translating from symbol to text, on the other hand, requires in addition to associated glosses, an awareness of the grammar, and word sequence likelihoods, to be able generate fully-formed sentences. Machine translation has often been tackled by using methods that require large amounts of data. This data needs to match the domain and cover the same source and target registers that are expected by a translation system. However, a parallel corpus of pictographic symbols and MSA is currently unavailable. This research addresses this issue by proposing an approach that creates the training data needed by making use of existing multilingual textual resources to resolve ambiguity. The outcome was a corpus that had been automatically tagged with morphological annotations and pictographic symbols, the approach followed, and an investigation of the data involved and produced. The availability of this symbol tagged corpus is a step towards Arabic symbol/text translation. This has the potential to enhance communication for those requiring Arabic speech output from symbol messaging, and provide a better understanding of the complexities of automated symbol/text translation processes in Arabic.
Perceptions of three aided graphic symbol sets used in augmentative and alternative communication (AAC) (Blissymbols, DynaSyms, and Picture Communication Symbols) were examined across four groups of adult participants with differing cultural histories and life experiences. One hundred and forty-seven individuals who identified themselves as African American, Chinese, European American, or Mexican and who ranged in age from 30 to 64 years participated in the investigation. Graphic symbols with translated referents (in English, Mexican Spanish, and Mandarin Chinese) from a 41-item lexicon were rated on a 7-point scale of iconicity. Results suggest that individuals from different cultural/ethnic groups perceive graphic symbols differently. Methodological issues related to the study of the impact of culture/ethnicity on graphic symbol recognition are described, and implications for the practice of AAC across cultures are discussed.
Conference Paper
Many people with aphasia experience difficulty recalling words extemporaneously, but can recognize those words when given an image, text, or audio prompt. Augmented and alternative communication (AAC) systems can help address this problem by enabling people with aphasia to browse and select from a list of vocabulary words. However, these systems can be difficult to navigate, especially when they contain large amounts of content. In this paper, we describe the design of TalkAbout, a context-aware, adaptive AAC system that provides users with a word list that is adapted to their current location and conversation partner. We describe the design and development of TalkAbout, which we conducted in collaboration with 5 adults with aphasia. We then present guidelines for developing and evaluating context-aware technology for people with aphasia.
Towards the goal of appropriate evaluation of all children having disabilities, a four part AAC assessment protocol is presented. Two aspects of the protocol are unique: the protocol is "inclusive", emphasizing the need to address cultural features directly; and the protocol includes a self assessment for all practitioners before they engage in the delivery of services to children within cultural communities which may differ from their own. In addition to the self assessment for practitioners, the protocol examines a child's communication partners, communication needs, and capabilities. A historical review of the "roots" of AAC reveals that many current practices are predominately White, European, and Western in origin. The proposed protocol is intended to expand our practices, encourage the development of broader clinical competencies and redirect the focus of assessment so that the processes become more culturally inclusive.
Conference Paper
Augmented and alternative communication (AAC) devices enable users with certain communication disabilities to participate in everyday conversations. Such devices often rely on statistical language models to improve text entry by offering word predictions. These predictions can be improved if the language model is trained on data that closely reflects the style of the users' intended communications. Unfortunately, there is no large dataset consisting of genuine AAC messages. In this paper we demonstrate how we can crowd-source the creation of a large set of fictional AAC messages. We show that these messages model conversational AAC better than the currently used datasets based on telephone conversations or newswire text. We leverage our crowdsourced messages to intelligently select sentences from much larger sets of Twitter, blog and Usenet data. Compared to a model trained only on telephone transcripts, our best performing model reduced perplexity on three test sets of AAC-like communications by 60--82% relative. This translated to a potential keystroke savings in a predictive keyboard interface of 5--11%.