SMARTPHONE BASED E-LEARNING

∗

Anandha Gopalan

Department of Computing, Imperial College London, 180 Queens Gate, London SW7 2RH, U.K.

Spyridon Karavanis

Hellenic Army, Athens, Greece

Thomas Payne

Omnifone Ltd., Island Studios, 47 British Grove, London, W4 2NL, U.K.

Morris Sloman

Department of Computing, Imperial College London, 180 Queens Gate, London SW7 2RH, U.K.

Keywords:

Virtual learning environment, Mobile learning, e-Learning, Voice-based framework.

Abstract:

Children often attend schools intermittently in rural areas in Africa and India due to socio-economic conditions

which make pupils augment their family income by working. An e-Learning solution could aid in raising the

level of education by making it easier for children to ﬁt schoolwork into the day, acting as a complement

to when they are able to attend school. Traditional distance learning solutions based on computers are not

suitable due to lack of infrastructure support. In this paper, we evaluate both text and voice based smartphone

prototype environments which could provide the tools and services for pupils to download educational content,

interact with teachers as well as other pupils to discuss topics. These have been implemented as a proof-of-

concept and the initial evaluation feedback, although not from target users, was very promising. We intend to

re-implement the prototype and do a proper evaluation with rural-area school children.

1 INTRODUCTION

Access to education remains an impediment to-

wards economic development of rural communities in

Africa and India. The reasons include lack of infras-

tructure, lack of teachers, poor attendance and ﬁnally

cost; the main problem among these being the low

attendance rate – (Azim Premji Foundation, 2004)

states that the gap between enrolment and attendance

is signiﬁcant. Due to the prevailing socio-economic

conditions, most pupils have to work to augment the

family income, thereby affecting their ability to at-

tend classes (Jean Dr

eze and Geeta Gandhi Kingdon,

2001). This paper presents an interactive e-Learning

environment that would allow pupils to learn in a col-

∗

We gratefully acknowledge support from EPSRC Grant

EP/EO251881 AEDUS2: Adaptable Environments for Dis-

tributed Ubiquitous Systems.

laborative manner without constant supervision. The

proposed system will deliver content in a manner that

takes into account restrictions such as poor IT infras-

tructure, high cost of bandwidth and poor knowledge

of electronics.

To achieve this goal, the high availability of more

affordable cellular mobile devices in these areas will

be leveraged to provide interactive content via mobile

phones (Marsden, 2003; Kam et al., 2009a; Parikh

and Lazowska, 2006). Mobile phones need very lit-

tle infrastructure, are low-power devices that can be

used in places where the availability of electricity is

not very reliable and are the fastest growing technol-

ogy platform in the developing world (CNN. Weapon

against epidemics: Cell phones, 2009). In addition,

mobile phones are inherently interactive in nature, as

compared to TV and radio, so are the most suitable

technology to use to access and share content in rural

161

Gopalan A., Karavanis S., Payne T. and Sloman M..

SMARTPHONE BASED E-LEARNING.

DOI: 10.5220/0003334901610170

In Proceedings of the 3rd International Conference on Computer Supported Education (CSEDU-2011), pages 161-170

ISBN: 978-989-8425-50-8

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

areas.

Our proposed e-Learning environment will pro-

vide educational content in the form of lessons, sim-

ple tests or quizzes, as well as allowing pupils to in-

teract both with teachers and their peers as they would

in a normal school environment. Connectivity cannot

always be guaranteed so the system should support

downloading educational content for the pupil to read

at a later time, and recording input from pupils to up-

load when a connection is available.

A couple of Masters student projects were used to

implement proof-of-concept systems to evaluate suit-

able implementation technology. The ﬁrst project - a

Virtual Learning Environment for Mobiles (VLEM)

used an iPhone and a fairly sophisticated open-source

web-based educational system called Moodle (Moo-

dle, 2010) as the back-end for management of text

documents. The second project explored the use of

an Android smartphone for a Voice-Based Framework

(VBF) for delivery of content and interaction with

teachers and peers. This was primarily because of in-

creased familiarity with voice systems in rural areas.

The proposed methods of educational delivery and in-

teraction were intended to act as a complement (and

not a substitute) to attending school. The emphasis of

the two projects was on evaluating technology for im-

plementing such a system rather than evaluating the

effectiveness of delivering educational content using

smartphones. No real educational content was used in

the project. Both implementations were evaluated by

a few volunteers to obtain some feedback on feasibil-

ity and usability. In the future, we plan to merge the

different implementations into one to provide a com-

plete solution. The evaluation results are very encour-

aging and prove that this is a viable means of provid-

ing distance education in rural areas. We plan to in-

stall the proposed system in a village in India and run

ﬁeld tests to determine the effectiveness of the sys-

tem and gather valuable data that would help improve

the overall solution. Although smartphones can be

criticised as currently being too expensive for the tar-

get environment, their development environments are

much easier to use for a proof-of-concept implemen-

tation. Given the rise of the Android open platform,

it is very likely that these smartphones will be more

affordable in developing countries within a few years.

There has been research on using mobile phones

as a means of providing information and services in

rural areas (see related work), but as far as we know,

this is the ﬁrst project of its kind to try and provide

a comprehensive solution that would allow pupils in

rural areas to be able to keep up with their educa-

tion while not forsaking their livelihood. According

to the World bank, “Despite growing hype, there are

still precious few widespread examples of the use of

[mobile] phones for education purposes inside or out-

side of classrooms in developing countries that have

been well documented, and fewer still that have been

evaluated with any sort of rigor” (World Bank, 2009).

The rest of the paper is organised as follows. Sec-

tion 2 gives an overview of the proposed environ-

ment and the requirements we have identiﬁed. Sec-

tion 3 describes the text-based Virtual Learning Envi-

ronment, while Section 4 describes the Voice-Based

Framework. Section 5 provides a detailed evaluation

of the two frameworks, while Section 6 outlines areas

of future work. Section 7 outlines the related work,

while Section 8 concludes the paper.

2 AN E-LEARNING

ENVIRONMENT USING

SMARTPHONES

2.1 Overview

Figure 1 shows the overall architecture of the pro-

posed e-Learning framework. It consists of (a) a sim-

ple user interface on the phone for displaying and

recording information and (b) technology and ser-

vices at the server for content management and dis-

semination. The required content is developed using

normal workstations connected to the internet, and

delivered by the content providers (schools, educa-

tors) to the content manager. The technology for de-

veloping the content will use off-the-shelf tools that

are readily available to educators and schools. Each

school will be connected to the central server. Pupils

will use phones to access content such as available

lessons and to collaborate with peers via a forum sup-

ported by an educational portal which can easily be

extended by adding additional capabilities, such as

announcements, tests etc. To show how the proposed

framework would work, an example scenario is de-

scribed in Section 2.2.

Figure 1: System Design.

CSEDU 2011 - 3rd International Conference on Computer Supported Education

162

2.2 Example Scenario

A teacher updates the content system with the new in-

formation based on the lessons that were used in class

during the day and also deletes any obsolete informa-

tion. A pupil living in a remote village goes about his

errands in the morning and is unable to attend school

for the day. On returning home in the evening, he uses

a mobile phone to access his classs information. The

pupil checks for any update since last access and if

any, downloads the required content on to the mobile

phone. After completing required tasks, the pupil will

have caught up with some of the material covered in

school during the day. Some material may not be ap-

propriate for delivery to a mobile phone. Queries can

be sent to his teachers, and he can ask questions or

discuss issues with his fellow pupils via the phone.

2.3 Requirements

The following requirements were identiﬁed for an e-

Learning environment:

• Content management (addition, deletion, updates)

must be easy for non-technical educators. A

widely used content management system (CMS)

is preferable, as educators may already be famil-

iar with it and the system will be maintained and

developed to keep it upto date.

• The User Interface on the mobile phone must be

intuitive, simple and easy to use. For example,

having a very elaborate user menu on the phone

to access the system may increase its complexity

and make it hard for pupils to use the system.

• The system must enable downloading content for

ofﬂine perusal and ofﬂine recording of questions,

discussions etc. by pupils for uploading when a

connection is available.

• The system must facilitate peer-to-peer discussion

between pupils as well as with teachers to simu-

late the classroom environment.

• The system must take into account bandwidth us-

age. This requires recording “changes” to the

data; only the data changed since the last request

is transferred to synchronise the phone and CMS.

• The system should facilitate peer-to-peer sharing

of downloaded content between pupils when they

come into contact with each other, using bluetooth

or wiﬁ to reduce usage of comparatively expen-

sive cellular networks. Note that neither of the

projects had time to implement this.

• The code-base developed must be modular and

easily extensible to incorporate future additions.

3 VIRTUAL LEARNING

ENVIRONMENT FOR

MOBILES (VLEM)

The VLEM system used an iPhone and evaluated

Moodle (Moodle, 2010) as the back-end software.

Moodle is a well known open-source Course Man-

agement System, with a large user community. The

functionality is thus continuously being enhanced and

the code is well maintained. Another advantage is the

ability to send bulk data which will help in bandwidth

reduction. The iPhone was chosen for its user inter-

face. The overall architecture along with the different

components is depicted in Figure 2.

3.1 Moodle Component

The Moodle component consists of a back-end for

content storage and a workstation interface for teach-

ers and administrators, which contains all the internal

functionality to facilitate registering users, creating

and updating lessons etc. Each teacher has an account

to access Moodle and is associated with a number of

courses. A teacher uses Moodle to create the lesson,

which is stored in a MySQL database. To access this

information from outside the Moodle domain, the sys-

tem was extended in the following ways to support re-

quirements identiﬁed in Section 2: (i) implement the

external functions used by Moodle using its naming

and development conventions, (ii) add functionality

required in the workstation to support web-based ac-

cess to Moodle back-end, (iii) create a table, called

ﬁrechanges in the Moodle database to manage the

changes. This table keeps a record for each user of

what changes he has received related to his lessons,

forums etc. by labelling them with 0 (clean) for re-

ceived and 1 (dirty) for not received, (iv) create the

triggers (Insert, Delete, Update). For every table of

interest in the Moodle database we have created trig-

gers to inform the ﬁrechanges table of any change that

has been made, and (v) create the webservice external

functions to retrieve the ﬁrechanges data. After the

retrieval of the data the tables are reset to 0 and a new

request is made to retrieve the changed data.

3.2 iPhone Application Component

The iPhone Application was implemented in Objec-

tive C and Figure 3 shows its architecture. The view

module contains a main view and many sub views.

The observer module monitors the network on the

mobile phone and informs the component of the net-

work’s “connectivity”. The action module is a virtual

SMARTPHONE BASED E-LEARNING

163

Figure 2: VLEM Architecture.

Figure 3: iPhone Application Architecture.

module that is included to ensure that proper empha-

sis is attributed to the actions performed during the

execution phase. The service client supports func-

tions to access the web service. The storage manager

sets up every database request and it uses the query

builder to create the appropriate query which needs to

be executed. All requests to the database are via the

ﬁle manager. In order to use VLEM, pupils have to

be authenticated using the iPhone application (due to

space constraints we do not go into detail).

Figures 4 and 5 show example application screens

on the iPhone. Figure 4(a) shows the starting screen

of the Virtual Learning Environment. Figure 4(b)

shows the help menu that explains the functionality of

each of the buttons. Figure 5(a) shows the list of avail-

able courses in the VLEM and choosing one of them

will bring the user to the interface shown in Figure

5(b), which gives the user the options that are avail-

able for that particular course.

(a) Main Screen. (b) Help Screen.

Figure 4: VLEM iPhone Screencaptures.

(a) List of Courses. (b) One Course.

Figure 5: VLEM iPhone Screencaptures.

4 VOICE-BASED FRAMEWORK

(VBF)

Unlike VLEM, the voice-based framework was not

meant to provide an entire e-Learning framework, but

to provide a means for pupils to gain some of the inter-

action that they are missing if they are studying from

CSEDU 2011 - 3rd International Conference on Computer Supported Education

164

home. The idea is for a forum where pupils will be

able to interact with their teachers and peers, asking

questions and ﬁnding out answers. Several studies

have discussed the success of voice-based mobile ap-

plications in rural areas of the developing world due

to the fact that they are so used to accessing informa-

tion and communicating orally (Parikh, 2010; Agar-

wal et al., 2009). For this reason, the forum was de-

signed and implemented as an asynchronous audio

space where pupils can post questions or comments

to teachers and listen to those of others.

While using the criteria stated in Section 2, the

available software and APIs that best suited our pur-

pose was the Android open-source OS for the mobile

phone and Django for the back-end. Android runs

on a wide variety of handsets, from top-of-the-range

devices to some at the lower end of the smartphone

market (Open Handset Alliance, 2010). Another key

reason for choosing the Android platform was the

availability of an API for sound recording. Django

was chosen since it’s a fast, extremely reliable open-

source framework that comes with a built-in modiﬁ-

able administration interface.

4.1 Client Component

The aim of the client is to provide a simple interface to

the user that will allow them to access the information

involved in the system via the following methods: (i)

record the user’s voice as they ask their question and

save it as an audio ﬁle, (ii) browse other users’ ques-

tions and responses, (iii) ﬁlter these questions and re-

sponses based upon subject area or user, (iv) browse

responses to a particular question, and (v) post an au-

dio response to a particular question.

The diagram shown in Figure 6 represents a sim-

pliﬁed ﬂow of control between the different compo-

nents of the client application. There are two main

ﬂows from the welcome screen, either the user wants

to record a new question or they want to browse ex-

isting posts. The former takes the right-hand path

through the process of recording the audio for their

question and onto the Entry Edit activity which will

take in additional metadata from the user before sav-

ing everything to the Local Database and starting up

the Synchronisation Service. This is a background

service which synchronises the local database with

the master database on the server.

The left-hand path leads into a browser that lists

posts on the system; selecting one will take the user

into the Audio Detail activity showing more informa-

tion about the post and allowing playback. There is

an option to leave a response to the post, which takes

the user to the Sound Recorder activity.

Figure 6: Flow between components of the client.

The audio ﬁle is encoded using the Adaptive Multi-

Rate Narrowband codec, a data compression scheme

optimised for speech. It is recorded with a bitrate of

approximately 12kbit/s which is higher than the aver-

age telephone quality (8kbit/s), but still low enough to

ensure that the ﬁles will not be too large. An average

question is expected to last approximately 30 seconds

generating an audio ﬁle of roughly 45kB which even

on an early GSM network would upload in roughly

ﬁve seconds.

4.2 Server Component

The server stores and maintain all the necessary data

for the application via the following methods: (i) re-

ceive an audio ﬁle and metadata and store these in the

master database, (ii) provide metadata in a machine-

readable format about questions that are stored in the

master database, ﬁltered by criteria such as subject

area and user, (iii) provide metadata about responses

to a particular question, and (iv) serve a speciﬁc audio

ﬁle after providing its URL.

The server has been designed to provide its func-

tionality through a RESTful (Representational State

Transfer) API above HTTP to enable it to be used by

a wide variety of potential clients. Interactions are

stateless, so any client submitting the same request

to a particular URL will receive the same response;

this means that these URLs can be bookmarked and

shared among users. This simpliﬁes creating a new

audio post that can then be accessed and responded to

SMARTPHONE BASED E-LEARNING

165

by other clients, thus creating a thread of audio con-

versation.

The server-side components have been built based

on the model-view-controller principles, leading to a

completely modular environment with in-built sup-

port for being run in a variety of environments. The

built-in administration interface was also extended to

provide a robust web-based interface for the teacher.

(a) Main Screen. (b) Recording Screen.

Figure 7: VBF Android Screencaptures.

(a) Options after Record-

ing.

(b) Available Tags.

Figure 8: VBF Android Screencaptures.

Figures 7 and 8 show some application screens on

the Android phone. Figure 7(a) shows the welcome

screen of the Voice-Based framework. If the user

wants to record a new question, they will be taken

to the recording interface that is provided by Android

(Figure 7(b)). After recording their question, the user

can either tag the question, play the question back or

submit it (as shown in Figure 8(a)). If the user chooses

to tag the question, a list containing the different sub-

jects is presented to them (shown in Figure 8(b)).

5 EVALUATION

5.1 VLEM

We evaluated VLEM using two approaches. In the

ﬁrst approach, we conducted a survey, asking 16 users

to use the system and to ﬁll in a questionnaire. 12%

of the users were in the age group 8-18, 63% were

between 19-35, while 25% were older than 35. In

order for the users to use the system, we created a

few lessons apriori. The users browsed the available

lessons and participated in the forums. Most of the

users were competent in the usage of a mobile phone.

The questionnaire contained personal as well as ques-

tions relating to the usability of the system. In the

second approach, we asked an educator (teaching sec-

ond grade in a school) to use and evaluate the system.

Due to the space constraints, we only present the re-

sults from the user survey, but the educator did say

“The convenience and support of the pupils is the pri-

mary goal of this application. I believe that ideas such

as this must be carried out because it will help pupils

and in general the whole education sector”.

To begin with, we wanted to evaluate the idea of

a VLEM before users actually saw the implementa-

tion. We asked users to indicate whether they thought

the concept of a VLEM was interesting or not. As

we can observe from the results in Figure 9, an over-

whelming majority thought that the idea would make

a difference in education.

Figure 9: Opinion on VLEM principle.

For the next part of the survey, we asked users to

use the application and ﬁll in the questionnaire. Our

ﬁrst concern was the impression the VLEM applica-

tion created on the users, especially after their opin-

ion about VLEM as a concept. Figure 10 shows that

the results were very promising with 25% saying that

they were very excited by the application and only 6%

saying that they were disappointed. This augurs well

for the functionality as well as the implementation.

Our next step was to ﬁnd out how easy it was to

use the VLEM. Figure 11 shows that 13% found the

CSEDU 2011 - 3rd International Conference on Computer Supported Education

166

Figure 10: VLEM Experience.

application hard to use. Though this is not a very large

number, it is still not small enough to ignore. One of

the reasons for this higher number is revealed later.

13%$

56%$

31%$

How$easy$was$VLEM$to$use?$

it$was$very$hard$to$use$$

it$was$average$to$use$

it$was$excellent$to$use$$$

Figure 11: Ease of Usage.

To ascertain why some users had a more positive

experience overall as when compared to others, we

asked them which feature of the VLEM they liked

or disliked. The results for the “best features” and

“worst features” of the VLEM are shown in Figures

12 and 13 respectively. As can be seen from the ﬁg-

ures, the best feature of the VLEM was considered to

be the forums. Users stated that forums were essential

because they enabled communication and collabora-

tion. As for the worst features, 25% of the people se-

lected the user proﬁle. One of the reasons for this very

high number is the fact that the current system lacks

the functionality that allows a user to update his/her

proﬁle on the iPhone. The button size was consid-

ered as the next worst feature. This suggestion was

taken into consideration and has since been incorpo-

rated into the application.

Finally, we asked the users their opinion as to

what was missing in the current implementation of

the VLEM. From Figure 14, we can see that most of

the users wished for additional functionality, such as

a library. This is a very interesting suggestion, which

will be investigated further in the future. Many others

wished for quizzes and pictures. The functionality to

support quizzes already exists (in terms of communi-

cation). As for the ability to have pictures, it may not

be feasible because of an increase in trafﬁc. An inter-

Figure 12: Best features of VLEM.

!"#$

%&#$

'#$

%(#$

(%#$

What%do%you%believe%is%the%worst%

feature%of%the%applica5on?%

)*+,-.$

/01+2$345.$

6+*073$

-.33+23$

8)).8*829.$

2+:;42<$

Figure 13: Worst features of VLEM.

esting addition that was also suggested was having a

help menu for the buttons. As we see in Figure 4(b),

this suggestion was incorporated into the application.

Figure 14: Missing Features of VLEM.

We believe this survey was helpful in providing

a direction towards improving speciﬁc aspects of the

application. It provided us with valuable experience

concerning usability aspects and it also gave us fresh

ideas concerning implementation issues.

5.2 VBF

Nine fellow-students evaluated our Voice-Based

Framework and completed a a survey on the appli-

cation. The application was explained to the volun-

teers and each one chose to use the system as they

saw ﬁt (a few of them adopted a passive approach

by just downloading and listening to existing infor-

mation, while others were active in posting and an-

SMARTPHONE BASED E-LEARNING

167

swering questions). This survey was structured so that

opinions could be harvested on the topics deemed im-

portant for the target audience in an attempt to coun-

teract some of the differences between the test au-

dience and the target audience. Much like the Hu-

man Access Points (HAPs) discussed in (Gary Mars-

den, Andrew Maunder and Munier Parker, 2008), the

test users have knowledge of what is possible with

technology, they just lack the cultural and contextual

knowledge that is desired from a HAP. For this rea-

son, they were given a survey to answer, after they

had used the application for a period of time, rather

than just asking for open-ended feedback. The aim of

this was to encourage them to think about the applica-

tion in terms of what has been deemed important for

the target audience based upon existing literature.

First of all the simplicity of the User Interface (UI)

was looked into and users were asked questions about

how intuitive the interface and structure of the appli-

cation was. Although most users were able to get to

the desired task without much trouble, several users

did point out that it was not always obvious what a

button would do before they pressed it. This feedback

was taken on board, and although all the buttons re-

main as they are, the layout ﬁles have been simpliﬁed

for ease of editing at a later date and plenty of screen

space has been left on most pages to allow some in-

structional text to be placed.

The next questions concerned the quality of the

audio. As might be expected, all responses to the lat-

ter question were positive, since the bitrate and qual-

ity are higher than that of a telephone call.

The users were next asked about the speed and re-

sponsiveness of the application. Although they were

using different handsets, with varying components,

they all reported that the application was very respon-

sive upon button presses. This is due to the sim-

ple interface that has been implemented and the fact

that all expensive operations are handed off to a sepa-

rate thread which will not impede the UI. Users were

speciﬁcally asked about the time taken for new posts

downloaded from the server to appear on screen after

the browser was started and what kind of network they

were connected to at the time. The results showed that

even on the slower 2G network, updates did not take

a long time (usually under 3 seconds).

Overall the test was very useful; the interface

has been made more ﬂexible to accommodate users’

suggestions and some potential extensions have been

identiﬁed. As stated at the beginning of this evalua-

tion, the test was limited in its scope, but provides a

basis for moving forward to test the framework with

the intended audience in the future.

6 FUTURE WORK

We aim to integrate the text and voiced based systems

on one platform containing both features of VLEM

and VBF. This would be based on the use of Moodle

as we think its focus on educational content manage-

ment would prove very useful for teachers. We will

also focus on the use of Android phones as, although

the user interface development support is not as so-

phisticated as the iPhone, it is cheaper and so is more

likely to be available as a smartphone platform in the

future in developing countries

One problem with both applications occurs when

a user loses connectivity during an upload or a down-

load. In order to combat this problem, the aim would

be to break up the data being transferred into smaller

chunks that are monitored so that if the network con-

nection did fail then the transfer would only need to

be restarted from the most recent chunk of data.

One of the possible extensions of VBF is a voice

recognition system. There are a couple of ways that

voice recognition could be used to improve the func-

tionality of the framework; the ﬁrst involves imple-

menting a voice-based menu system, where a user is

prompted to enter the number on the keypad. Another

area where voice recognition could really add func-

tionality to the application is in using it to create text

transcriptions of the audio posts that have been made.

With these text-based questions a decent search func-

tionality becomes a real possibility. It would also give

users the option to read the posts themselves rather

than listening or use speech synthesis to listen if they

are on a particularly slow connection and do not want

to download the audio ﬁle.

7 RELATED WORK

Mobile phones have been used for implementing ru-

ral computing applications e.g. (Derenzi et al., 2008)

presents a system for providing the IMCI (Integrated

Management of Childhood Illness) protocol using a

PDA, and (Maunder et al., 2008) provide a cost-

effective way of accessing relevant public information

by sharing it using a Bluetooth enabled camera phone.

A mechanism to interact with paper documents and

automate paper-intensive information processing for

micro-ﬁnance groups using an interface toolkit is pro-

vided in (Parikh, 2005; Parikh et al., 2006). De-

sign principles that address the challenges in design-

ing rural computing applications are outlined in (Gary

Marsden, Andrew Maunder and Munier Parker, 2008;

Parikh and Lazowska, 2006). All of the above work,

while focusing on support for people in rural areas,

CSEDU 2011 - 3rd International Conference on Computer Supported Education

168

does not address educational applications. Most of

the information ﬂow is one-way and the interface used

is passive where the local community do not interact

with the system on an everyday basis.

There has been some work on using technology

for education in rural areas in Africa through a World

Bank survey (Farrell, Glen and Shaﬁka Isaacs, 2007)

as well as the SchoolNet project (SchoolNet Africa,

2008). These differ from our approach in that they

aim to establish high quality education through the

use of Information Technology in existing schools. In

(Kam et al., 2008; Kam et al., 2009b) games (based on

traditional village games) on cell phones are used to

improve the English language skills of the children.

In (Jones and Marsden, 2004), the authors use mo-

bile phones and PDAs to allow pupils to communicate

with the lecturer during classes. This project though is

only helpful in the scenario when pupil attend classes.

The BBC Janala project (BBC Janala, 2009) uses mo-

bile phones, TV and radio to improve the English lan-

guage skills of people in Bangladesh, however there

is no feedback from users and the emphasis is on lis-

tening to spoken English over the phone. The Mo-

bilED (Ford and Leinonen, 2006) project provides a

learning and teaching environment that is enhanced

with the use of mobile technology and services. The

framework consists of an audio wikipedia that can be

accessed by pupils by sending an SMS with a key-

word and in return they get to hear the article relevant

to the keyword. Though this idea has its merits, it is

very restrictive since it assumes an advanced level of

knowledge. Although the above works uses technol-

ogy for education in rural areas, none of the projects

provide for a complete environment wherein pupils in

rural areas can learn at their own time without forsak-

ing their livelihood. Also, these projects do not allow

for collaborative learning, which is a very important

learning tool.

There has been quite a lot of work on developing

voice-based interfaces for rural areas. This is primar-

ily due to the fact that rural communities have sig-

niﬁcantly different communication needs and patterns

as compared to urban communities (Seshagiri et al.,

2007; Kolko et al., 2007). Given the fact that people

are comfortable using telephones, a voice-based in-

terface would work very well with such communities.

(Cervantes and Sambasivan, 2008) is an audio-based

classiﬁed advert service, where users call in to post,

listen or delete an advertisement. An Audio Wiki

that acts as a repository of spoken content that can

be accessed and modiﬁed through the use of any tele-

phone is presented in (Kotkar et al., 2008). The World

Wide Telecom Web (Kumar et al., 2007) is a project

that allows people to create their own “spoken” web-

pages. Users navigate using a simple speech-based

interface and the “pages” are organised by user (akin

to a web-page). VoiKiosk(Agarwal et al., 2009) is a

voice based kiosk that provides access to information

in rural areas. This was extended in (Patel et al., 2010)

to provide a message board (coupled with some radio

broadcast) that serves small farmers and is used pri-

marily as a forum for exchanging agricultural advice.

Most of the related work mentioned above use

voice-based solutions to cater to and solve certain

problems. None of them, however cater for educa-

tion. Given the level of technology penetration in ru-

ral areas, voice-based systems offer a promising solu-

tion due to the fact that people are used to using tele-

phones. However, most of the related work tackles the

problem with respect to a particular topic and hence

both the complexity of the system and the amount of

data managed, is less. Also, having a pupil phoning

in and having to scroll to multiple questions belong-

ing to multiple topics will only add to their cost.

8 CONCLUSIONS

In this paper, we have prototyped smartphone sys-

tems to enable e-Learning in rural areas in developing

countries such as in Africa and in India. One system

focused on text-based access to a sophisticated educa-

tional content management system, while the voice-

based system focused on supporting interactions with

teachers and other pupils. These frameworks serve a

dual purpose, with the former acting as an educational

portal that allows a pupil to be able to keep up with

their lessons at their own time, without necessarily

forsaking their livelihood, while the latter provides a

forum for interactive communication that is achieved

through voice messages (since users are more used to

using their mobile phones as a tool for calling). The

proposed systems were implemented as a proof-of-

concept and evaluated. Although the systems have

not been tested with the target audience, the frame-

work has been set up as much as possible to allow

ﬂexibility in the interfaces. The results were very en-

couraging and some of the suggestions have been im-

plemented. We also identiﬁed areas for future work.

We are hoping to be able to deploy these systems and

test them out in the ﬁeld.

The implementations were based on smartphones,

which could be criticised as being currently too ex-

pensive for developing countries. One reason for our

use of smartphones was that their development envi-

ronments are much easier to use for proof-of-concept

systems. Simpler phones do not have the ability to

display suitable learning material although they might

SMARTPHONE BASED E-LEARNING

169

be adequate for voice based-interactions. It is also

likely that the Android open platform smartphones

will drop sufﬁciently in cost to be affordable in de-

veloping countries within a couple of years. Though

the motivation for this work was for e-Learning in ru-

ral areas in developing countries, the ideas suggested

can also be used to enhance learning experiences in

developed countries.

REFERENCES

Agarwal, S., Kumar, A., Nanavati, A., and Rajput, N.

(2009). Content Creation and Dissemination by-and-

for Users in Rural Areas. In Proc. 3rd IEEE/ACM Int.

Conference on Information and Communication Tech-

nologies and Development (ICTD2009), pages 17–19.

Azim Premji Foundation (2004). The Social Context of El-

ementary Education in Rural India.

BBC Janala (2009). http://www.bbcjanala.com/.

Cervantes, R. and Sambasivan, N. (2008). VoiceList: user-

driven telephone-based audio content. In Proceedings

of the 10th international conference on Human com-

puter interaction with mobile devices and services,

pages 499–500. ACM.

CNN. Weapon against epidemics: Cell phones (2009).

http://www.cnn.com/2009/TECH/science/06/16/

cellphones.health.disease/index.html.

Derenzi, B., Lesh, N., Parikh, T., Sims, C., Maokla, W.,

Chemba, M., Hamisi, Y., Hellenberg, D. S., Mitchell,

M., and Borriello, G. (2008). e-IMCI: Improving Pe-

diatric Health Care in Low-Income Countries. In CHI

2008, ACM Press, pages 753–762.

Farrell, Glen and Shaﬁka Isaacs (2007). Survey of ICT and

Education in Africa: A Summary Report, Based on 53

Country Surveys. Washington, DC: infoDev / World

Bank.

Ford, M. and Leinonen, T. (2006). Mobiled: A mobile tools

and services platform for formal and informal learn-

ing. In mLearn 2006, the 5th International Conference

on Mobile Learning.

Gary Marsden, Andrew Maunder and Munier Parker

(2008). People are People, but Technology is not

Technology. Philosophical Transactions of the Royal

Society, October 2008, pp 1-10.

Jean Dr

eze and Geeta Gandhi Kingdon (2001). School Par-

ticipation in Rural India. Review of Development Eco-

nomics, 8(2):198–222.

Jones, M. and Marsden, G. (2004). “Please Turn ON Your

Mobile Phone” - First Impressions of Text-Messaging

in Lectures. In Mobile HCI, pages 436–440.

Kam, M., Agarwal, A., Kumar, A., Lal, S., Mathur, A.,

Tewari, A., and Canny, J. (2008). Designing e-

learning games for rural children in india: a format for

balancing learning with fun. In DIS ’08: Proceedings

of the 7th ACM conference on Designing interactive

systems, pages 58–67. ACM.

Kam, M., Kumar, A., Jain, S., Mathur, A., and Canny, J.

(2009a). Improving Literacy in Rural India: Cell-

phone Games in an After-School Program. In pro-

ceedings of 3rd IEEE/ACM International Conference

on Information and Communication Technologies and

Development (ICTD2009).

Kam, M., Mathur, A., Kumar, A., and Canny, J. (2009b).

Designing digital games for rural children: a study of

traditional village games in india. In CHI ’09: Proc.

27th Int. Conference on Human factors in computing

systems, pages 31–40. ACM.

Kolko, B. E., Rose, E. J., and Johnson, E. J. (2007). Com-

munication as information-seeking: the case for mo-

bile social software for developing regions. In WWW

’07: Proceedings of the 16th international conference

on World Wide Web, pages 863–872. ACM.

Kotkar, P., Thies, W., and Amarasinghe, S. (2008). An Au-

dio Wiki for Publishing User-Generated Content in the

Developing World. In HCI for Community and Inter-

national Development (Workshop at CHI 2008).

Kumar, A., Rajput, N., Chakraborty, D., Agarwal, S., and

Nanavati, A. (2007). WWTW: the world wide telecom

web. In Proceedings of the 2007 workshop on Net-

worked systems for developing regions, page 7. ACM.

Marsden, G. (2003). Using hci to leverage communication

technology. Interactions, 10(2):48–55.

Maunder, A. J., Marsden, G., and Harper, R. (2008). Sna-

pandgrab: accessing and sharing contextual multi-

media content using bluetooth enabled camera phones

and large situated displays. In CHI ’08: extended ab-

stracts on Human factors in computing systems, pages

2319–2324. ACM.

Moodle (2010). http://moodle.org/.

Open Handset Alliance (2010). http://www.openhandset

alliance.com.

Parikh, T. (2010). Voice as Data: Learning from What Peo-

ple Say. In Proceedings of AAAI Artiﬁcial Intelligence

for Development (AI-D’10).

Parikh, T. S. (2005). Using mobile phones for secure, dis-

tributed document processing in the developing world.

IEEE Pervasive Computing, 4(2):74–81.

Parikh, T. S., Javid, P., Sasikumar, K., Ghosh, K., and

Toyama, K. (2006). Mobile phones and paper doc-

uments: evaluating a new approach for capturing mi-

croﬁnance data in rural india. In CHI ’06: Proceed-

ings of the SIGCHI conference on Human Factors in

computing systems, pages 551–560. ACM Press.

Parikh, T. S. and Lazowska, E. D. (2006). Designing an ar-

chitecture for delivering mobile information services

to the rural developing world. In WWW ’06: Proceed-

ings of the 15th international conference on World

Wide Web, pages 791–800. ACM.

Patel, N., Chittamuru, D., Jain, A., Paresh, D., and Parikh,

T. S. (2010). Avaaj otalo - a ﬁeld study of an interac-

tive voice forum for small farmers in rural india. In

CHI ’10: Proceedings of the SIGCHI conference on

Human Factors in computing systems.

SchoolNet Africa (2008). http://www.schoolnetafrica.org/.

Seshagiri, S., Aman, S., and Joshi, D. (2007). Connect-

ing the ”bottom of the pyramid”: an exploratory case

study of india’s rural communication environment. In

WWW ’07: Proc. 16th int. conference on World Wide

Web, pages 855–862. ACM.

World Bank (2009). http://blogs.worldbank.org/edutech/

surveying-the-use-of-mobile-phones-in-education-

worldwide.

CSEDU 2011 - 3rd International Conference on Computer Supported Education

170