P. 1
Using Mnemonics as part of Pictorial Interface for Self Identification of Illiterate Villagers

Using Mnemonics as part of Pictorial Interface for Self Identification of Illiterate Villagers

|Views: 2|Likes:
Published by Dinesh Katre

More info:

Published by: Dinesh Katre on Aug 11, 2010
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





Using Mnemonics

as part of PictoriaI Interface
for SeIf-identification of IIIiterate ViIIagers
Dinesh S. Katre
Centre Ior Development oI Advanced Computing (C-DAC)
Ministry oI Communications and InIormation Technology. Government oI India.
Agriculture College Campus. Near District Industries Centre.
Shivaii Nagar. Pune 411005. India.
Tel/Fax: 91¹020¹5533250
Personal Homepage: www.hceye.org

Various government departments in India are thinking oI
oIIering personalized services through touch screen kiosks
to the illiterate villagers. Establishing the identity oI a
person and its veriIication becomes very critical while
oIIering such services. This paper proposes use oI
mnemonic techniques as part oI a pictorial interIace Ior
selI-identiIication oI illiterate villagers. It banks upon the
existing visual literacy oI villagers. The experiments
documented in this paper help in identiIying the
eIIectiveness oI visual imagery mnemonics and other
complementing mnemonic techniques Ior recalling the
sequence oI pictures by the villagers. Terms like pictorial
identitv Ior login name and pictorial signature Ior password
are coined Ior the proposed selI-identiIication process. The
metaphorical aspects oI using a series oI pictures as one`s
identity and signature are also explained. The paper also
compares the proposed ideas with PassIaces.

Illiteracy. Visual Literacy. Pictorial Identitv. Pictorial
Signature. Touch Screen Kiosk. InterIace Metaphor.
Recall. Visual Imagery Mnemonic. Mnemonic Techniques.
Mental Model. SelI-identiIication. PassIaces

Several government departments in India are considering
the possibility oI oIIering variety oI services to the villagers
in rural parts oI India through touch screen kiosks. These
services include inIormation and guidance pertaining to
health. agriculture. market rates Ior agricultural products.
opportunities Ior artisans. education. voice messaging.
Iinance and governance. Maior impediment in the
eIIectiveness oI these initiatives is the illiteracy oI people.
Secured and personal space Ior every villager becomes
important iI the services are to be personalized.

Indian government planned to issue three million Kisan
Credit Cards Ior loan disbursement to the Iarmers in rural
parts oI India in 1998. But Iinally. the scheme has been
absolutely ineIIective |Business Line. 2003|. |The Tribune.
2000|. Articles published in news papers ironically mention
about a bare-chested man dressed merely in Dhoti using a
debit card. One oI the reasons behind the ineIIectiveness oI
Kisan Credit Card scheme was ensuring the identity oI
credit card holder.

The question arises that how will they establish their
identity iI they were to seek any services Irom the kiosk?

Lack of Process Orientation
Taking care oI a delicate smart card does not match their
hectic and rough liIestyle. Also the chances oI illiterate
villagers misplacing the smart cards are very high. It should
be noted that most villagers are not used to managing
documents. They are not so much process oriented.

Lack of Fine Motor SkiIIs
Another alternative Ior establishing the identity is to type
the login name and password. But the touch screen kiosks
usually do not provide a keyboard. Even iI we assume that
the keyboard is made available. still the typing by illiterates
is a problem.
Katre Dinesh (2004). Using Mnemonics as part of PictoriaI Interface for SeIf Identification of IIIiterate ViIIagers,
In Proceedings of InternationaI Conference on Human Computer Interaction (I-HCI 2004) organized by IESUP
and CSI, BangaIore, India.

In our experiments at Darevaadi (a village in Ahmednagar
District. Maharashtra State) we observed that the villagers
are unable to use the computer keyboard. We had asked
them to use only Iorward. backward. leIt and right keys Ior
navigating through a terrain walkthrough. Their Iine motor
skills are very poor. In many cases. we saw that while
pressing a key. they ended up pressing two/three more keys
around it. Their hands shiver while pointing a Iinger at the
key as they usually perIorm extremely laborious tasks.

This shows that a secured system Ior oIIering services to
illiterate villagers through touch screen kiosks is a
paramount need oI the country.

IIIiteracy, Diversity and Large PopuIation
Another idea is to provide assistance in the beginning while
creating the user account. This sounds all right. But how
will they recall the keys Ior typing their name and password
the next time. And how will they decide the password and
then type it? II they take assistance Irom someone then the
security oI password is questionable. This is a huge
problem iI considered the approx. 400 million illiterate
population oI India |Statistics by Education Department.
Government oI India. 2001|. On top oI it. India speaks over
300 diIIerent languages and dialects (oIIicially recognized).
Some oI them do not have scripts at all. Not all are
supported on computer Irom among those that have scripts.

ThereIore a simpler mechanism needs to be devised so that
an illiterate villager is able to establish his/her identity and
log into the system. OI course. veriIication oI identity could
be perIormed using the latest Iingerprint recognition
systems. iI they are aIIordable enough. They are
prohibitively expensive Ior a developing country like India.
This technology is still to be introduced in revenue
generating organizations like banks and airports oI India.
Proposing it Ior oIIering services to illiterate villagers is
quite unrealistic |PWW. 2004|.

ThereIore it is proposed to design a pictorial interIace.
which will enable the illiterate villagers in deIining their
pictorial identitv (login name) and pictorial signature
(password). The pictorial signature is a metaphoric
concept. The idea is to avoid usage oI text and rely on the
visual literacy oI villagers. Such interIace is proposed as
one oI the many alternatives Ior establishing the identity oI
a person.
The discussion so Iar can be summarized as under-
Illiteracy. Very large population
Using the visual literacy oI people to counter the
impediment oI script illiteracy in the given scope oI

1. Our idea is to use a collection oI pictures that will
be selectable Irom a touch screen kiosk. The
pictures should be Iamiliar to the villagers.
2. A villager should be able to select a series oI
pictures (around 4 to 8 oI them) oI his/her choice
to serve as a login name.
3. (S)he should be able to select another series oI
pictures (around 4 to 8 oI them) oI his/her choice
to serve as a password.
4. (S)he should be able to easily recall the pictorial
identitv (login name) and pictorial signature
(password) Ior inputting it in the system through
tough screen kiosk Ior availing the services.

At the outset oI this experimentation some issues were
already identiIied to be addressed. These are enlisted in the
Iorm oI questions.
Which type oI pictures will be easily recognizable
to villagers?
How many (minimum and maximum limit)
number oI pictures are required Ior deIining the
login name and password?
Will the villagers be able to recall the sequence oI
What is the solution iI they are unable to recall the
sequence oI pictures?
How large the collection oI pictures should be?
How will one browse through a large collection oI
How many unique login names and passwords can
be generated with such pictorial interIace?
It should be possible to create minimum 2000
unique login names and passwords. As the average
population oI an Indian village is approx. 2000
people |e-Census India. 2003|.

In order to Iind answers to these questions. a couple oI user
interIace prototypes are designed. They have been tested
over several subiects. Based on the results oI testing the
guidelines Ior such pictorial interIace are evolved.

Representing login name` and password` as a series oI
pictures is a metaphorical idea. As maiority oI subiects
select only those pictures. that are meaningIul to them. that
are associated with their activities and that reIlect their
aspirations. Basically they choose the pictures with which
they can relate themselves. The series oI pictures in the
Iorm oI pictorial identitv and pictorial signature is indeed a
metaphorical representation oI the person. It reIlects his/her
mental model. Using pictorial interIace Ior selI-
identiIication is quite an unconventional |Katre. 2002|
application oI interIace metaphor.

Such pictorial interIace can prove most suitable in India. as
most oral cultures tend to use visual metaphors unlike the
Modern Western European ways oI thinking inIluenced by
print culture that tends to use verbal metaphors |Clair.

HereaIter. login name` and password` will be reIerred as
pictorial identitv and pictorial signature respectively.

It was evident that we needed to take help oI existing
visual imagery mnemonic techniques` |Baddeley. 2000| or
devise new mnemonic techniques.

Mnemonic Experiment 1.
Task: Select two series of seven pictures each from a set
of 50 pictures and recall them in same sequence as they
were selected after a gap of 15 minutes. Subjects were
expected to remember both series of pictures.
We showed around 50 pictures in one screen to both
illiterate (Irom a village) and literate (Irom a metro city)
subiects. They were asked to select two series oI seven
pictures each (totally 14 pictures) out oI 50 pictures as seen
in Figure 1. Let us say the Iirst series oI pictures
represented the pictorial identitv and the second series oI
pictures is the pictorial signature oI the subiect. We chose
both illiterate and literate subiects in the initial test to Iind
the diIIerence in mnemonic techniques used by them. The
observations oI the test are categorized below.

Both type oI subiects preIerred the pictures that
represented their worlds.
A very prominent number oI literate subiects
selected pictures oI washing machine. television.
car. aircraIt. space ship. mango iuice. bike. etc.
Whereas the illiterate subiects showed distinct
inclination Ior pictures oI cattle. bullock cart.
tractor. well. gods. etc. Several oI them also chose
the picture oI television.

RecaII of pictures
In this experiment. many illiterate and literate
subiects did recall all the Iirst series oI 5/6 pictures
selected by them aIter a gap oI 15 minutes.
Those who chose pictures that lacked concrete
theme or the theme was not evident enough or the
picture was ambiguous to them got conIused e.g.
herd oI buIIalos. barren land. etc.
All pictures with concrete themes were recalled
perIectly e.g. television. bike. bullock cart. well.
hut. gods. cow. etc.
Many subiects recalled the pictures based on their
size. shape and location.
More precisely. the pictures they could identiIy and
verbalize were recalled better. ThereIore. the spoken
language did come into play even though we wanted to
avoid it. The pictures that mapped with their existing
vocabulary oI words were easy to recall. Illiterate subiects
did not choose pictures like spaceship and Einstein`s
portrait. as these are unIamiliar to them.

Figure 1. All pictures exposed in one screen

RecaII of sequence
Many oI them recalled the sequence oI Iirst series
oI seven pictures to a certain extent. But maiority
oI them iumbled up the sequence oI second series
oI pictures.
This resulted as most oI them started the selection
oI pictures Irom top row and then aIter initial 5/6
steps they chose the remaining pictures quite
The percentage oI illiterate subiects Iailing to
recall the sequence was much higher.
Literate subiects could recall the sequence to a
certain extent as they are trained to remember and
write the spellings oI words Irom leIt to right.
Furthermore. they are also used to pronouncing
the password aIter coining it. Illiterate subiects
have no clue oI it.
Average human limit to recall ¹-7 items |Miller.
1956| also seemed to be a reason Ior Iailing to

Visual Quality
Exposing all 50 pictures in one screen made the
task tougher due to visual chaos.
The pictures appeared too noisy in terms oI color
and cognitive load on the subiects.
Pictures with Ilat backgrounds received greater
preIerence Irom all subiects.

What we had to achieve now is to modiIy the
picture selection process in such a way that the
sequence oI selection gets registered in the minds
oI illiterate subiects. Or the system itselI takes care
oI the sequential selection.
We also needed to articulate the complementing
mnemonic techniques used by the subiects.

Figure 2. Pictorial identitv as my own photograph

Mnemonic Experiment 2.
As subiects could not remember 14 pictures at a time. we
decided to consider other alternatives. We thought that a
person`s photograph itselI could be used as his/her pictorial
identitv. In this option. only the signature will be
represented in the Iorm oI pictures. The visualization oI
interIace can be seen in Figure 2. But in this case we
realized that a villager would Iind it extremely diIIicult to
identiIy his/her photograph Irom hundreds oI other
villagers` photographs. It would be almost impossible to
search it without use oI text input. which is ruled out in
case oI illiterate villagers.

As a result oI this realization we decided to represent the
identity oI a person also as a series oI pictures.

Mnemonic Experiment 3.

Sequential Peg Screens
In this experiment. we distributed the images over 4
screens. Each screen included approx. 3x3 grid oI 9
pictures. The Iorth screen included a 3x3 grid oI 9 distinct
colors. The subiects were expected to select any one picture
Irom each screen. AIter selecting a picture. the application
automatically switched to the next screen. We also ensured
that the selected image appeared in the vacant slots at the
top portion oI the screen. Figure 3. shows the arrangement
oI peg-screens.
Length of Pictorial Identity
This time the length oI every visual identity is only 3
pictures long.
Length of Pictorial Signature
The length oI visual signature is 4 pictures long. The 4

Irame was to be Iilled by selecting a distinct color.

Figure 3. Screens as visual pegging device

Both pictorial identitv and pictorial signature put together
it becomes totally 7 pictures. This reduced the memory load
to remembering oI only 6 pictures and 1 distinct color. The
color is automatically applied to both series oI pictures. As
per Miller`s human memory limit. we Iitted everything
within the limit oI 7 items. Figure 4. shows the arrangement
oI slots created Ior inputting the pictorial identitv and

Pictorial Identity
Pictorial Signature

Figure 4. Picture Selection Mechanism
But in this case the question arises how many unique login
names and passwords can we generate using 4 screens oI 9
pictures each? The answer is simple.
9x9x9x9 ÷ 6561 unique combinations oI 4 pictures each.
We have not considered the repeated use oI pictures as
villagers may or may not be able to count.
As mentioned in the introduction oI this paper. the average
population oI an Indian village is approx. 2000 people. The
proposed system can deIine 3 times greater number oI
unique pictorial identities and signatures.
This experiment was conducted with illiterate subiects
only. The sequential presentation oI peg screens Iorced a
sequence while selecting the pictures. Most subiects did
very well. Most oI them recalled the entire sequence oI
pictures properly. This Iinalized the Iramework Ior picture
selection. The visual peg screens really worked very well.

Pictorial Identity
Pictorial Signature
System View
ViIIager's Perception

Figure 5. System View and Villager`s Perception
However. we observed another peculiar aspect oI villagers
as shown in Figure 5. They did not perceive the duality oI
the series oI pictures they were selecting. From their
perspective. they were selecting only one sequence oI 6
pictures and 1 color as their identity. It is only in system`s
view the series 7 pictures is divided in terms oI pictorial
identitv and signature.

We have witnessed that the subiects chose variety oI
mnemonic techniques in combination. ThereIore the
pictures should also IulIill requirements oI Iollowing
commonly used mnemonic techniques.

Location Mnemonic
The villagers should be able to recall the position oI a
selected picture (both in space and time).
Shape, Size and CoIor Mnemonics
All pictures need not be oI same size and shape. The
villagers should be able to recall the picture by its
uniqueness oI size. shape and color.

CIassification Mnemonic
The villagers should be able to recall the pictures belonging
to a Iamily e.g. pet animals. Ilowers. Iruits. etc.

Personification Mnemonic
The villagers should be able to select pictures that can
represent their belongings. liIestyle. likes and dislikes.

Succession Mnemonic
The villagers should be able to select pictures that indicate
the logical progression oI stages oI an activity or a

The idea oI PassIaces |BrostoII. 2000| does rely on recall
oI human Iaces as shown in Figure 6. But restricting the
choice oI pictorial identity and signature to a series oI
human Iaces has many weaknesses. Not many mnemonic
techniques as mentioned in the earlier discussion can be
applied with PassIaces. Our recall oI images is anchored to
various visual and conceptual aspects like shape. size.
color. location. content. classiIication. associated meanings.
personal choice. etc.

Also. the human Iaces tend to have racial characteristics.
For example. one might Iind the Iaces oI all Chinese.
Japanese. Mongolian. Indian people very alike respectively.
You start noticing the minute diIIerences between the
Chinese or Japanese Iaces only aIter becoming Iamiliar
with the individuals.

Another weakness oI PassIaces is that our recall oI human
Iaces is good only iI we personally know the individuals
whose Iaces are selected as part oI a pictorial password. We
do not like to remember the anonymous Iaces oI people that
we come across in public places.

Figure 6. Passfaces Grid of Brostoff
ThereIore as proposed in this paper. the pictures that reIlect
entire cultural context oI the targeted users and support
various mnemonic techniques are most usable Ior pictorial
authentication interIace.

It is possible to distribute the pictures theme wise e.g. we
could have agriculture. pet animals. daily activities. local
Iestivals. gods and goddesses. and even the regional and
national Iilm stars. The present set oI pictures does not
cater to the choice oI both male and Iemale subiects. But
this should be addressed.
The pictures should reIlect the mental model oI villagers.
The pictures should be localized. as the geographical
conditions and liIestyles are diIIerent in many states oI
AIter a villager has selected his /her pictorial identitv and
pictorial signature. a printout oI the same could be given to
them. They may preIer to use it initially.

It is possible to counter the illiteracy barrier by taking
help oI existing visual literacy oI villagers.
Sequential peg-screens are helpIul in registering the
sequence oI pictures in the minds oI villagers.
Combinations oI variety oI mnemonic techniques
enlisted in this paper can help in recalling the sequence
oI pictures.
Pictorial identitv and pictorial signature can be truly
helpIul in establishing the identity oI an illiterate
Proposed pictorial interIace Ior selI-identiIication can
be a very cost eIIective solution Ior providing
personalized and secured services through touch screen
kiosks to illiterate villagers.

We propose to enhance and expand the applications oI this
interIace in near Iuture. The tasks include-
Applications Ior children (who are iust beginning to learn
the scripts) can be explored. Various themes oI pictures
need to be tested on villagers. The gender based choice oI
pictures need to be addressed.

The idea oI using pictures as login name` and password`
is originally conceived by Dr. Mukul K. Sinha. Managing
Director. Expert SoItware Consultants. New Delhi. He is
also the Ph.D. supervisor oI the author oI this paper. Under
his guidance the author has experimented with the idea Ior
Iurther scientiIic clarity. The author also expresses
gratitude to C-DAC. Pune Ior oIIering the inIrastructure
and the Ireedom Ior research explorations.

1. Baddeley. A. Your Memory` A Users Guide.
Published by PRION. UK. (1996). 340-341.

2. BrostoII. S.. & Sasse. M.A. (2000). Are PassIaces
more usable than passwords? A Iield trial
investigation. In Proceedings oI HCI 2000. Sept. 5-8.
Sunderland. U.K.. 405-424 Springer.

3. Clair R. N.. Visual metaphor. Cultural Knowledge and
the Rhetoric. Learn In Beauty: Indigenous Education
Ior a New Century. Published by Northern Arizona
University. (2000). 85-101

4. Department oI Education`s Report. 2001
Retrieved on Sept. 18.


5. Katre. D. S.. Unconventional Inspirations Ior Creating
SoItware InterIace Metaphors`. Proceedings oI
International ConIerence on Media and Design (ICMD
2002). Volume I. Mumbai. India. pp. 1-15.

6. Miller. G.A.. The magical number seven. plus or minus
two: Some limits on our capacity Ior processing
inIormation. The Psychological Review. (1956). 81-97.

7. RBI pulls up banks over kisan credit cards. Business
Line. Dec. 13. 2003.
Retrieved on Sept. 17.


8. Smart ATMs - Serving the Poor ProIitably in Bolivia
Retrieved on Sept. 16.


9. Watershed in India`s history. Peoples Weekly World
July 26. 2004 Issue
Retrieved on Sept. 16.


10. Number oI villages and towns in India. e-Census oI
India. Issue 14. 2003. pp. 3. Retrieved on Sept. 17.


11. http://www.censusindia.net/results/eci14page3.html
Retrieved on Sept. 17.


You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->