Evolution of a Digital Ecosystem for Knowledge Services to Indian Agriculture Jayanta Chatterjee, TV Prabhakar and Runa Sarkar Indian Institute of Technology, Kanpur

pogonotomygobbleAI and Robotics

Nov 15, 2013 (3 years and 10 months ago)

43 views




Evolution of a Digital Ecosystem for Knowledge Services to Indian
Agriculture

Jayanta Chatterjee, TV Prabhakar and Runa Sarkar

Indian Institute of Technology, Kanpur


Abstract

Cooperative efforts of seemingly unrelated domain experts such as farmers
and ag
ricultural scientists
with
computer scientists or economists can lead to e
ffective knowledge creation and
knowledge growth.
Relevant information at the right time could equip farmers
with appropriate
tools to make more
economically sound decisions enhanci
ng their competitiveness and, as a result, improve their well being.
This is the objective of the digital ecosystem for agricultural livelihood

project
.

At one end, d
eveloping such an ecosystem requires the development of
peer to peer networks,
classifi
cation
schemes, controlled vocabularies, thesauri, authority files, and glossaries as well as creation of semantic
standards for exchange of high quality metadata
. The semantic framework would comprise shared data
exchange standards and instruments that w
ould allow services exchange

(interoperability) between
collection

of information and knowledge
.

At the other end, there is a need to consult with, inform, orient, and involve
stakeholders such as farmers,

extension workers and NGOs

in developing, sharing
and refining the content in the open knowledge space.
Without their involvement,
a large number of knowledge networks, aimed at facilitating interaction
between peers on
all relevant

issues and sharing of resources and experiences

cannot be operationalize
d.
Towards this end, this paper explores the crucial elements which lead to
the creation of relevant content for
effective deployment and use of
socio technical networks
in the context of Indian agriculture.

Identifying
and applying alternative roadmaps f
or self
-
sustainability and growth of socio technical networks for
enhancing knowledge sharing would lead to our ultimate goal of regional development.



1. Introduction

Whether large or small,
information and communications technologies (ICTs)

are presen
t and a
d
vancing in
every area of economic, social, and political activity. Due to the networking possibilities they enable,
they

reduce transactions costs changing the structure of markets and institutions, resulting in an i
m
mediate
increase in the potent
ial value of human capital. Further they embody enormous knowledge and can serve
to empower people at community and national levels.


In the field of Indian agriculture, the adoption and development of ICTs takes place via thousands of
specific initi
a
tive
s led by communities, development, donor, and business organizations. As it r
e
quires
local know
l
edge, literacy, skills development, technical capability and effort
, effective ICT deployment

can
be a challenge to i
m
plement for a diffuse netwo
rk
1

of local i
nnovation systems
.


There is, however, a
government established top
-
do
w
n network of
agricultural extension counters called

Krishi Vigyan
Kendras


(KVKs)

which could be used to link India’s
geographically and culturally dispersed

rura
l
community
.



Agr
icultural and food security policy makers clearly see the need for knowledge co
n
nectivity from the
academic/research institutes to villages and then on to the world to close the loop so that the ‘best’ practices
can enhance India’s agricultural efficiency,

create the “next” practices and create new avenues for rural
livelihood. There is a national agenda for creating
k
now
l
edge centres

in ev
ery village. But the soft side

of
this challenge needs more attention. There is no concerted effort to create a n
a
tiona
l agricultural knowledge
repository in digital form which is alive and is nurtured daily through fee
d
ing, weeding, & pruning
-

or
enriched by interactive usage. Lot of good knowledge nuggets remain at local level and as unstru
c
tured
information or tacit kno
wledge. Moreover, agriculture is among the most complex commercial sy
s
tems
requiring inputs from myri
ad sources including soil, water, env
i
ronment, goods, asset and labour markets.
A detailed study conducted by the
Asia
-
Pacific Research Centre

of
the
Sta
nford
University tried to assess
the socio
-
economic impact of 9 major ICT iniatives in India to conclude that the usage of ICT was sparse
compared to its potential. Results of a questionnaire survey administered to the potential users of ICT and



1

It is estimated that there are over 104 million farm families spread over more t
han 590 rural di
s
tricts and
six lakh villages (Rai, 2006).




ICT prov
iders (termed infomediaries) to explore the gap between actual and potential usage is summarized
in Figures 1 and 2.

0
5
10
15
20
25
30
35
40
45
Illiteracy/ computer
illiteracy
Irrelevant Content
Too Few
Programmes
Infrastructure
(power,
connectivity)
Hardware/software
Centre site
Location
Unware of Usage
Government
regulations
# of respondents
Figure 1: ICT Service Challenges as Identified by Users
, adapted from Dossani
, Misra and Jhaveri

(2005)



0
5
10
15
20
25
30
35
40
45
50
Connectivity &
Power Supply
Relevant content
creation
Content
maintenance
Hardware
inadequacy/maintain
ance
Community
awareness
Users lack
skill/education
User acceptamce
Govt/pvt sector
cooperation
Location/Building
space
# of respondents

Figure 2
:

ICT Service Challe
nges as Ide
ntified by Project Authorities and
Operators
, adapted from
Dossani
, Misra and Jhaveri

(2005)


As can be seen, while more than 50 of approximately 140 user responses indicated the availability of useful
content and programmes as

significant imped
ing
factor
s

in use of ICT
, fewer

infomediaries or providers of
information
were of a similar opinion.

The creation, dissemination and enhancement of appropriate, timely
and relevant content for the farmer is the focus of the ‘Digital Ecosystem for Agricul
ture & rural
Livelihood’ (
www.dealindia.org
) project, experiences from which are shared here.


The digital ecosystem (DE) is one approach through which
one can ensure relevant and timely content
availability to the

rural community through dynamic and amorphous interaction among a multiplicity of



small entities to support knowledge sharing, co
-
creation of knowledge and developing new business
models. Moreover, the
diffusion and use of ICT can be made self sustaining

and self enabling despite
technological and literacy barriers.


In this paper, we document our experience from being involved in developing and implementing a DE for
knowledge diff
u
sion in rural India where sustainability of the initiative is wrought wi
th challenges due to
language and literacy barriers, r
e
source scarcity, dominance of top
-
down solutions and limited existence of
successful pa
r
ticipative business models. A DE for agriculture offers farmers from less developed and
remote areas opportuniti
es to participate in the global economy, resulting in d
y
namic knowledge sharing
and global cooperation among the farmers and the world community fostering l
o
cal economic growth. Co
creation and self
-
management of digital contents to support agriculture &
rural livelihood d
e
velopment
activities would result in access to the right kind of information at the right time, resulting in inclusive
growth as well as competitive agriculture. It also facilitates two way inte
r
actions among the farmers and
agricultur
al scientists which is critical for further technolog
i
cal progress in agriculture, whether with
respect to innovation or technology ado
p
tion.


2. A Pathway to Information Design for Knowledge Diffusion in Rural India

Q
uick dissemin
a
tion of technical inf
ormation from the agricultural research system to the farmers
, and
adapting it to different soil and climatic conditions will result in improved agricultural productivity
. T
hus,
th
e one
-
way route of
India’s
conventional
a
g
ricultural extension sy
s
tem needs

rapid transformation to a
‘real time and adaptive’ knowledge exchange network. This network needs to build real time fee
d
back
routes from the ‘fields to the laboratory’ and can derive necessary traction from other i
n
dustrial and
business knowledge managem
ent technologies and pro
c
esses like user to user exchange, expert to expert
exchange and KM oriented standards for inform
a
tion storage, r
e
trieval and aggregation with analytics.

Unless there is a mechanism by which the small and marginal farmer can co
n
vey

his needs to the extension
counter and there is an adequate perception of their circumstances, there appears no other way to improve
his competitiveness. Limit
a
tions of the physical face to face Transfer of Technology (TOT) model,
however, r
e
mains a chal
lenge for the public and private extension systems as there are at least 400,000
m
e
dium and large villages spread over a subcontinent that need to be reached. With the availability of
telephone and Internet, it is now possible to bridge this gap to quite
a large extent but only if an appropriate
mix of technologies can d
e
liver ‘dynamic content’ in response to ‘user pull’. Unless the content is problem
-
solving oriented to help farmers to take risks in venturing out to crop diversification or adopt novel
pro
cesses, the TOT can not make adequate impact on alleviating rural po
v
erty t
hrough improved
competitiveness. A

digital ecosystem
, which
entails a s
e
ries of interconnected and intra
-
dependant digital
platforms, created at key institutional levels (internati
onal, national and l
o
cal/Community) augmented by
technical (ICT) and social networking processes
can

help break down barriers to both horizontal and
vertical knowledge.


2.1 The Agricultural Ecosystem

An agricultural ecosystem is a unique and reasonably
stable dynamic arrangement of farm enterprises that a
household manages in response to the physical, biological and socioec
o
nomic environments. There could
be several interacting subsystems within this large ecosystem such as on the regional level, equall
y relevant
non agr
i
cultural systems such as the market system, the rural credit sy
s
tem etc. Agricultural subsystems
include the crop ecosystem, animal ecosystem, soil, weed and i
n
sect ecosystem, which of which are co
-
dependent. Thus, both farm related ci
rcu
m
stances such as weather conditions, type of soil, stage of
incidence or intensity of weeds etc. and socio
-
economic ci
r
cumstances such as availability and nature of
credit, costs of agricultural inputs, price of end product, farmers personal goals and
r
e
sources etc. feed into
the agricultural ecosystem. An ideal knowledge ecosystem for agriculture would be able to capture all
these intricacies and build a large know
l
edge sharing data base to ensure that the implicit knowledge or
experience of one farme
r is shared with many others without necessitating the re
-
invention of the wheel
over geographically or temporally separated regions.


2.2 Implementation

Figure
3

shows the information flow diagram for rural development activities. It is o
b
vious that an o
ntology
driven semantic interoperability through this maze can effe
c
tively network the di
f
ferent actors, while they
pursue their micro objectives. Given this network, successful implementation of a knowledge system
required develo
p
ment of digital content
from the tacit knowledge
bases of Krishi Vigyan Kendras

and

other



frontline entities through multiple media like landline phone, mobile phone, audio
-
video recording and
digitiz
a
tion of paper documents. There was a need to develop a common ontology, a sema
ntic
interoperability that facilitates knowledge storage, r
e
trieval and exchange within the network among the
various stakeholders so that a knowledge ecosystem can develop. This required open content and open
source optimization so that the technology to
ols are affordable and remain available while evolving. To
bridge the language and education divide, “citizen interfaces” to the e
x
tensive knowledge base were
r
e
quired. These could be iconic, graphical, symbolic user interfaces (that relate to the ontolog
y) for rural
citizens’ ease of access. Tec
h
nology application included touch screen, text to speech, screen reader,
visualization & animation, interactive voice response system co
m
puter
-
telephony integration and
application of wireless data services like M
MS. Digital co
n
tent architecture and tools for easy telephone,
mobile data and FM radio based interactivity and backend int
e
gration of such tran
s
actions into the
knowledge base was also developed.



Figure 3: Typical flow of in
formation among rural development age
n
cies

(Source: CRISP group, National Informatics Centre)


Partnerships were created with existing telecenters in rural institutes, village schools and Krishi Vigyan
Kendra. These had an inherent advantage that an exis
ting phys
i
cal infrastru
c
ture only has to be extended
and some of the ICT
-
relevant training can be cost
-
effectively i
n
tegrated into the mainstream curriculum of
these institutions. This partnership has succes
s
fully worked in our Digital Mandi project
(
www.digitalmandi.net
). Several brain storming se
s
sions of the stakeholders in the Digital Mandi pr
o
ject
generated a conceptual architecture of the desired knowledge
-
net. This is shown in Figure
4
.


Ministry of Rural Develo
p
ment

State Rural Development Department

Village Panchayats/ NGOs

Rural Beneficiaries

Block Panchayat/ Dist
rict Rural
Develo
p
ment Agency

Zila Parishad/ District Rural Deve
l
opment
Authority

(1) Consolidated Progress


Reports

(2) Consolidated Annual Plans


(1) Consolidated Progress


Reports

(2) Consolidated Annual Plans


(1) Consolidated Progress


Repo
rts

(2) Consolidated Annual Plans


(1) Details of loan/subsidy utilized, details


of achievement of intended benefits


(infrastructure, housing, income generated


etc.)

(2) Needs/Requirements etc.


(1) Details of fund allocations,
san
c
tions, sche
me guidelines by
Go
v
ernment of India


(1) Details of Fund Allocations, san
c
tions,
scheme guidelines by State Go
v
ernment

(2) Details of distribution of funds
all
o
cated/sanctioned by Govt. of I
n
dia,
scheme guidelines


Details of distribution of funds
all
o
cated/sanctioned by Go
v
ernment of India
/ State

Govt. , scheme guid
e
lines


Details of distribution of funds
all
o
cated/sanctioned by Government of I
n
dia /
State Govt. , scheme guidelines


Details of beneficiaries (village, group, family,
individual) selected, benefits (e
m
ployment, house
s,
roads etc.) funds (loan, subsidy etc.) avai
l
able etc.


(1) Progress Reports (Physical &


financial)

(2) Annual Plans





It was clear

that to acquire the characteristics of a self
-
managed ecosystem, in this know
l
edge
-
net, the
digital contents created in various forms by the stakeholders needed ‘interoperability’. Interoperability
provides potential for automation and sy
s
temic self
-
mana
gement. Initial experiments across the digital
repositories of the stakeholders in the project showed that syntactic interoperability can be achieved for
transfer, exchange, mediation and integr
a
tion of content by adopting compatible forms of encoding and

access protocols and design guidelines. Identification and naming schemas are impo
r
tant at this stage for
pulling together related information


2.3 Lessons

While implementing the
D
igital
M
andi project, we validated the existence of several barriers to in
formation
access, which have often been reported in literature (see Kra
l
isch and Mandl, 2006 for exa
m
ple). These
included physical, economic, intellectual or technological barriers that impede rural user participation in the
activities that add to the dig
ital knowledge repository. The architects and system designers did not a
c
tively
impose the barriers but they crept in through their lack of action or lack of u
n
derstanding of the critical user
conditions. Such critical user conditions may arise due to par
ticular demographic, geographic, cultural,
social, psychological, economic or other factors. Although issues related to Information system u
s
ability
such as ease of use, usefulness (Davis, 1989), decision effectiveness, user response, user satisfa
c
tion (Do
ll
et al, 1988) and many other aspect of usability have been studied in great detail, interactions with focus
groups at various agricultural market places around Lucknow
-
Kanpur showed the need of a more detailed
study on with a loca
l
ized set of prior
i
ties.




Figure 4
: Conceptual architecture of knowledge
-
net

Source: Chatterjee and Prabhakar (2005)


A general framework for web design keeping in mind the human
-
computer intera
c
tion the
o
ries (Pirolli,
2001), web site usability principle
s (Huang, 2003), information inte
n
sity paradigm (Palmer and Griffith,
1998), e
-
customization models is already in place and is assumed to sufficiently address the question of
defining broad guid
e
lines for designing any successful website. It was therefore,

assumed that a website
with rel
a
tively high
-
level of accurate, up
-
to
-
date and pertinent content, deployed in a user
-
friendly way,
customized to particular user groups, and tailored to specific ge
o
graph
i
cal needs should be universally
successful and hence,

accepted in India too. Ho
w
ever, we found that the challenges to agricultural and rural
livelihood website usabi
l
ity for rural India arose mainly due to the highly specific local needs and the great
d
i
versity in local conditions. The major challenges ident
ified were:







Poor literacy rate


low use of textual information in daily life and high rel
i
ance on verbal
communic
a
tion for knowledge transfer.



Remote village locations
-

physical distances compounding problems of d
e
pendence on middlemen
and a nexus of ex
ploitation through inform
a
tion asymmetry.



Absence of content in vernacular languages (both a cause and an e
f
fect)



Unavailability of economic, low
-
cost solutions
-

any technology sol
u
tion aimed at benefiting the
masses in rural India must be a
f
fordable and

low
-
cost so that the perceived economic benefits of
such an e
n
deavour are much more than the cost of switching over to a different technological
sol
u
tion.


Another lesson related to the sustainability of the DE in the agricultural and rural liv
e
lihood spa
ce. The
project soon revealed that without a self managed, evolving, ec
o
system like knowledge repository, where
users can co
-
create content and the co
n
tent can be so “tagged” that it can be recalled and reused in multiple
context, the ed
i
torial over head
remains high & expensive.


The finding from the initial research at Digital Mandi showed that the presence of a number of desired
features in any ICT system design for rural India that leads to higher user satisfa
c
tion. Such features are
broadly aimed at
satisfying one or the other of the following immed
i
ate user objectives:




Ease of access.



Up
-
to
-
date content.



Layout, design, consistent themes.



Easy navigation.



Higher interactivity.



Access through multiple media (particularly voice).



Higher use of non
-
tex
tual information.



Language options



Lower cost of transaction.


Since most of the farmers are quasi
-
literate, content in textual form becomes a cha
l
lenge, especially at
content creation stage. Content in audio form is often the only way we can ope
r
ate. Apar
t from its ease in
creation it has other advantages as well


it is more natural, there is a personal touch making it more
acceptable to both the creator and the listener and community 'viewing' (or in this case liste
n
ing) is easier.
But indexing and searc
h audio content poses problems and requires manual inte
r
vention.


Figure 5

gives a sample page depicting the user interface addressing some of these issues. The user ids are
iconic, and so are the passwords. In other words, the alph
a
bet consists of images
of fruits and vegetables
and the user can 'spell' her user name and password with this alphabet.

That is, a user can choose a tomato
with two onions and a potato together as the 'name' of the user and another such comb
i
nation as a
password.







Figure
5
: Iconic Logic in the Web interface for the Digital Mandi Project


A computer based platform appears difficult to maintain for various reasons. Apart from the cost of the
computer, due to the erratic power situation, one needs to thin
k of backup power sources like batte
r
ies, un
-
interrupted power supplies, generating sets and so on, making the whole solution quite unte
n
able. A mobile
device, like a phone or a PDA appears to be the most workable delivery platform.




The Digital Mandi pr
oject thus revealed that ICT tools and technologies could make knowledge and field
experiences (in the form of digital content) widely available. Et
h
nographic observation guided design
principles, which improved access and acce
p
tance by rural citizens. But

the maintenance, d
y
namic update
and enhancement of the digital content needed regular editorial intervention and the process of finding and
assembling information remained largely a manual task. It was clear that to a
c
quire the characteristics of a
self
managed ecosystem in the knowledge
-
net, the dig
i
tal contents created in various forms by the
stakeholders needed interoperability which would lead to automation and systemic self management.
While initial e
x
periments showed that such syntactic interoperab
ility can be achieved and e
n
forced in a
corporate extranet, prevalent socio
-
technical diversities and existence of a mult
i
plicity of
hardware/software in the network pose problems in the domain of agricu
l
ture and r
u
ral livelihood.


2.4 Benefits

Although

the benefits accruing as a result of the
D
igital
M
andi project have not been formally studied or
documented, some observations in this context are in order. First, it was quite ev
i
dent that the ‘ecosystem’
approach speeds up the process of identif
i
cation
, development and uptake of innovation. Second, rural
entrepreneurs benefited because the DE helped to i
m
prove access to markets or supply chains and pr
o
vide a
broader base for decision
-
making, thus making risk more calculable.





Moreover, it has been re
ported by several researchers that many local communities have e
x
perienced that
ICT have increased bottom
-
up participation in the governance processes and may expand the reach and
accessibility of government services and public infrastructure (Dossani, Mis
ra and Jhaveri, 2005). We have
not been able to test this in the digital mandi project yet, primarily because the ma
n
date of the project was
more focused on creating a self sustaining ICT platform rather than conduc
t
ing a social experiment.


3. Conclus
ion

A digital business ecosystem as a platform to foster business networks based on a dynamic and amorphous
interaction among a multiplicity of firms to support know
l
edge sharing and skill development is a self
sustaining mechanism of ICT adoption and deve
lopment. This p
a
per reported on the learning from using
semantic web technologies to construct agricultural portals to address the need for customization and
localization at the rural level. The digital ecosystem for agriculture and rural liv
e
lihood (DEA
L) project is
an ambitious web based initiative at coordinating back end infrastructure, media technology and knowledge
bases to make agricultural content easily accessible through multiple channels in rural India. It attempts to
overcome language and lit
eracy barriers to knowledge networking and dissipation by the deve
l
opment of
iconic, symbolic and visual overlays on knowledge maps. Existing Krishi Vigyan Kendras serve as nodes
and catalysts for knowledge driven self
-
generative socio economic developmen
t to nurture grassroots
innovation in rural livelihood models. By a
c
tivating and/or strengthening knowledge, skill, technology and
market links, thereby increasing the returns on investment for farmers, such a DBE would be instrumental
in preserving and n
urturing the wisdom of the farmers while improving agricultural co
m
petitiveness at the
same time.



4. References

Chatterjee, J and Prabhakar, T.V. (2005) “On to Action
-

Building A Digital Ecosy
s
tem for Knowledge
Diffusion in Rural India, Proceedings of

the 2005 International Conference on Knowledge Manag
e
ment,
North Carolina, USA , available at
http://emandi.mla.iitk.ac.in/deal/other/deal_paper.doc

Davis F. (1989) “Perceived usefuln
ess, perceived ease of use and user acceptance of info
r
mation
technology”, MIS Quarterly 13(3), 319

340.

Doll W.and Torkzadeh J.(1988) “The measurement of end
-
user computing satisfa
c
tion”, MIS Qua
r
terly 6,
259

273,

Dossani, R, Misra, D.C. and Jhaveri, R

(2005) Enabling ICT for Rural India, Asia P
a
cific R
e
search Center,
Stanford University and National Informatics Centre, downloaded from
http://iis
-
db.stanford.edu/pubs/20972/ICT_full_Oct05.pdf#search=%22ict%20governance%20india%20rural%22
,
last accessed 30th September 2006

Huang W.(2001) “Using Information Technology to Enhance Communications among Agribusiness
Organizations”, IAMA World

Food and Agribusiness Symp
o
sium, Sy
d
ney, NSW, Australia

Kralisch, A and Mandl, T (2006) “ Barriers to Information Access across Languages on the Internet:
Network and Language Effects”, in Proceedings of the 39th Hawaii International Conference on Sy
s
tem
s
Science (HICSS
-
39, 2006)

Palmer J. W. and Griffith D. A., (1998) “ Information Intensity: A paradigm for unde
r
standing web site
design”, Journal of Marketing Theory & Practice, 6 (1), 38
-
42

Pirolli, P., Card S. K. and Van der Wege, M. (2001) “Visual Inf
ormation Foraging in a F
o
cus+Context
Visualization”, CHI 2001, Seattle

Rai, M (2006), Foundation of National Strategy, The Hindu Survey of Indian Agricu
l
ture 2006, Chennai