expert system for effective extension service
flow of presentation
genesis of extension service
limitation of existing extension services
earlier expert systems and web and mobile based expert systems
2. Flow of Presentation
•Introduction
•Expert system and its concept
•Components of Expert System
•Application of Expert System in
Agriculture
•Working of Expert System
•Research Studies
•Conclusion
5. Genesis of Extension Service
Individual effortsBefore 1947s
Extension Workers
Multilevel workers
Not specialist
Individual contacts
Demo
Small group meetings
Community approach
Area approach
Clientele approach
Demonstrators
Field Supervisors
Specialists
(VLWs, BDOs)
Dealers
Friends & Relatives
Individual contacts
Group
Mass
1947 - 1979
T&V
ATMA
ICT initiatives
AEOs
Scientists, SAUs
Input Agency
Dealers
Friends & relatives
Print media
Individual contacts
Group
Mass / ICT
1980s to
2011…..
6. Source of Information
Before 1947s
Traditional Farming
Fore farmers knowledge
Small knowledge from unskilled workers
Community workers
Demonstrators
Friends & Relatives
VLWs, BDOs, Print Media
1947 - 1979
AEOs, AAOs,
Scientists, SAUs
Private players, Dealers
Friends & Relatives
Mass Media
1980 to present
Green
Revolution
Subsistence
Farming
Commercial
Agriculture
8. Poor
ratio of SMS
to agents
Lower level of
education of
extension agents
Short supply
of extension
agents
More number of
farmer per extension
worker
More area to
be covered by
agents
Less number
of female
extension
agents
Human Resource
of Extension
9. Poor
transportation
facility to agents
Lower pay to
extension agents
Less availability
of
programme cost
Poor housing
to extension
workers
Poor
communication
facility to agents
Very little
expenditure
per farmer
Financial Resource
of Extension
13. Which choice
is best?
Which choice
is income
generating?
Which choice
requires less
labour?
Which choice
requires less
land area ?
Which choice
is not much
affected by
season?
Which choice
requires less
input?
15. Concept OF Expert system
Expert systems were introduced by researchers under Stanford
Heuristic Programming Project.
Principal contributors to the technology were Bruce Buchanan,
Edward Shortliffe, Randall Davis, William vanMelle, Carli Scott, and
others at Stanford.
An expert system is software that attempts to provide an answer to a
problem, or clarify uncertainties where normally one or more human
experts would need to be consulted. Expert systems are most common
in a specific problem domain, and is a traditional application and/or
subfield of artificial intelligence.
16. Is an intelligent computer program that uses knowledge,
procedures and inferences to solve problems.
Is a system that employees human knowledge captured
in a computer to solve problems that ordinarily require
human expertise.
( Daniel Hunt,1986 )
An expert system is simply a computer software programme that
mimics the behavior of human experts.
(Ahmed Rafea, 2002)
EXPERT SYSTEM – MEANING
17. Components of expert system
Knowledge
acquisition
Knowledge
representation
User interface for
query, explanation,etc.
Inference/control
mechanism (e.g.
forward chaining.
Knowledge base
(Devraj et. al.,2001)
19. CONCEPTUAL DESIGN
Expert System
of Extension
Knowledge
Base
Domain
Expert
Knowledge
Engineer
Knowledge, Concepts, Solutions
Data, Problems, Question
Structured
Knowledge
Knowledge Acquisition Module
Technical &
Extension bulletins
Textbooks
Facts
Research Findings
Bahal et.al.,2004
20. Knowledge acquisition for knowledge engineer
Structured interviews
Unstructured interviews (tape recording,
video taping)
Note-taking and memory
Gestures
Knowledge Representation
•A method to represent the knowledge about the domain
• Knowledge about an area of expertise is encoded
21. INFERENCE ENGINE
A computer program to process symbols that represents
objects.
It can interpret knowledge in the knowledge base and
perform logical deduction and manipulation
22. USER INTERFACE
Allows the end-users to run the expert system and interact
with it.
Allows query, advice, explanation and interaction
23. APPLICATION OF EXPERT SYSTEM IN
AGRICULTURE
• Crop production estimates
• Crop selection
• Soil management
• Plant diseases and pests mgt
• Weed management
24. EARLIER MODULES OF EXPERT SYSTEM IN
AGRICULTURE
Specification Field of application
COMAX Integrated crop management in cotton
SOYEX Soybean oil extraction expert system
PLANT/ds Diagnosis of soybean diseases
SEMAGI Weed control decision making in sunflowers
RICE-CROP Diagnose pest and disease for rice crop and suggest
preventive/ curative measures
COTFLEX Cotton crop management coupled with SOYGRO model
CVSES Wheat crop variety selection
25. EARLIER MODULES OF EXPERT SYSTEM IN
HORTICULTURE
Specification Field of application
POMME Pest and insect management in
apple
CUPTEX Cucumber expert system
CITEX Citrus expert system
LIMEX A multimedia Expert system for
Lime production
TOMATEX Tomato expert system
26. Web based expert system
Maize Agri Daksh
Wheat expert system
RICE Doctor by IRRI
TNAU AgriTECH PORTAL
Digital mandi for the Indian kisan
mKisan Agri portal
Barley expert system
Rice knowledge management portal
Expert system for Agriculture and Animal husbandry
by DWCRA, Bhubaneshwar
Developed by
IASRI
27. Mobile based expert system
:
Crop insurance mobile app by Ministry of
Agriculture,GOI Agrimarket mobile app
mKissan app
RainbowAgri app
Manditrades app
Mpower social app
IFFICO kissan app
eSAP app….etc
28. Web based and mobile based Expert
systems
Agriculture and horticulture
Paddy expert system
Banana expert system
Sugarcane expert system
Ragi expert system
Coconut expert system
Animal husbandry
Cattle and Buffalo
Sheep and Goat
Poultry
Source: http://www.agritech.tnau.ac.in
•Developed by
TNAU
collaboratively
with ICAR
•Available in 4
languages
English
Kannada
Tamil
Malayalam
40. POTENTIAL ADVANTAGES OF
EXPERT SYSTEM
Solves critical problems by making logical deductions without
taking much time
It combines experimental and conventional knowledge with the
reasoning skills of specialists
To enhance the performance of average worker to the level of an
expert
42. An information technology enabled
Poultry Expert system: Perceptions of
veterinarians and veterinary students
Karuppasamy and Sriram
(2013)
Methodology:
Study Area : Nizamabad district, Hyderabad
Method of Sampling : Random Sampling
Sample Size: 30+30
43. Table 1: Response of veterinarians and students on
Utility of PES
Item Veterinarians( n=30) Students( n=30)
Agree undecide
d
Disagree Agree undecid
ed
Disagree
F % F % F % F % F % F %
Very much useful 30 100.
0
0 0.0
0
0 0.00 27 90.0
0
0 0.0
0
3 10.0
0
Handy to use 26 86.6
6
2 6.6
6
2 6.66 26 86.6
7
2 6.6
6
2 6.66
Saves time and
money
30 100.
0
0 0.0
0
0 0.00 23 76.6
7
4 13.
33
3 10.0
0
Advantageous
over the traditional
methods
30 100.
0
0 0.0
0
0 0.00 28 93.3
3
2 6.6
6
0 0.00
44. Table 2:Response of veterinarians and students on
Complexity of PES
Item Veterinarians( n=30) Students( n=30)
Agree undecide
d
Disagree Agree undecid
ed
Disagree
F % F % F % F % F % F %
Simple to operate 26 86.6
6
2 6.6
6
2 6.66 19 63.3
3
6 20.
00
5 16.6
7
Simple language 25 83.3
3
5 16.
67
0 0.00 24 80.0
0
5 16.
67
1 3.33
Easy navigation 27 90.0
0
3 10.
00
0 0.00 18 60.0
0
10 33.
33
2 6.66
Simple to
understand
29 96.6
7
1 3.3
3
0 0.00 30 100.
0
0 0.0
0
0 0.00
45. Table 3:Response of veterinarians and students
on Compatibility of PES
Item Veterinarians( n=30) Students( n=30)
Agree undecide
d
Disagree Agree undecid
ed
Disagree
F % F % F % F % F % F %
Replacement of
expert
9 30.0
0
15 50.
00
6 20.0
0
11 36.6
7
9 30.
00
10 33.3
3
Supplement to the
existing practice
23 76.6
7
7 23.
33
0 0.00 22 73.3
3
7 23.
33
1 3.33
Substitution of an
expert
12 40.0
0
11 36.
77
7 23.3
3
13 43.3
3
9 30.
00
8 26.6
7
46. Table 4:Response of veterinarians and students on
Technicality of PES
Item Veterinarians( n=30) Students( n=30)
Agree undecide
d
Disagree Agree undecid
ed
Disagree
F % F % F % F % F % F %
Credibility 25 83.3
3
5 16.
67
0 0.00 18 60.0
0
11 36.
67
1 3.33
Accuracy 23 76.6
7
7 23.
33
0 0.00 18 60.0
0
6 20.
00
6 20.0
0
In line with the
agreement of
experts
21 70.0
0
8 26.
67
1 3.33 22 73.3
3
6 20.
00
2 6.67
No discrepancy in
the message
20 66.6
7
7 23.
33
3 10.0
0
17 56.6
7
8 26.
67
5 16.6
7
47. Table 5:Response of veterinarians and students on
Feasibility of PES
Item Veterinarians( n=30) Students( n=30)
Agree undecide
d
Disagree Agree undecid
ed
Disagree
F % F % F % F % F % F %
Suitable to existing
information needs
of farmers
26 86.6
7
3 10.
00
1 3.33 25 83.3
3
5 16.
67
0 0.00
Affordability / cost
effective
20 66.6
7
6 20.
00
4 13.3
3
13 43.3
3
8 26.
67
9 30.0
0
New aid for transfer
of technology
26 86.6
7
3 10.
00
1 3.33 28 93.3
3
1 3.3
3
1 3.33
Can be used at
farmers level
18 60.0
0
8 26.
67
4 13.3
3
16 53.3
3
2 6.6
7
12 40.0
0
48. Table 6:Response of veterinarians and students on
Design of PES
Item Veterinarians( n=30) Students( n=30)
Agree undecide
d
Disagree Agree undecid
ed
Disagree
F % F % F % F % F % F %
User friendly 24 80.0
0
5 16.
67
1 3.33 27 90.0
0
2 6.6
7
1 3.33
Aesthetic 24 80.0
0
6 20.
00
0 0.00 25 83.3
3
3 10.
00
2 6.67
User centered
design
22 73.3
3
8 26.
67
0 0.00 24 80.0
0
5 16.
67
1 3.33
User centered
interactiveness
25 83.3
3
5 16.
67
0 0.00 25 83.3
3
2 6.6
7
3 10.0
0
49. INFORMATION EFFICIENCY OF AGRICULTURAL
EXPERT SYSTEM
Helen and
Kaleel (2009)
Methodology:
Study Area : Pallakkad district , Kerala
Method of Sampling : Purposive Sampling
Sample Size: 60
50. Table 1: Treatment wise Information Efficiency
Index of AES as assessed by extension personnel
SI NO DIMENSIONS AES alone
T1 (n=30)
AES+HES
T2 (n=30)
1 Retrievability 61.76 68.16
2 Relevancy 79.33 80.00
3 Practicability 84.00 86.00
4 Information content 68.74 78.21
5 Knowledge gain 49.35 60.44
IEI=obtained total score X 100/Maximum possible score
51. Perception of Prospective Users about the Performance
of Agricultural Expert System
Helen and Kaleel (2009)
Methodology:
Study Area : Pallakkad district , Kerala
Method of Sampling : Purposive Sampling
Sample Size: 60
52. Table 1: Perception of TOT researchers about the
Performance of Agricultural Expert System
SI NO Performance related attributes Researchers in
TOT
n=40
MEAN RANK
1 Settings in the AES 9.45 I
2 Retrievability of information 6.30 IV
3 Serviceability of the system 6.18 V
4 Relevancy of information 2.86 VII
5 Practicability of information 8.23 III
6 Information content 2.81 VIII
7 Mode of presentation 2.54 IX
8 Information treatment 2.37 X
9 Provision for updating information 5.05 VI
10 Future Prospects 9.27 II
53. Effectiveness of Paddy Expert System in terms of
Knowledge Gain, Skill Acquisition and Symbolic
Adoption Behaviour among the Paddy Growers of
Thoothukudi District in South Tamil Nadu
Karuppasamy and Sriram
(2013)
Methodology:
Study Area : Thoothukudi district, (Tamil Nadu)
Method of Sampling : Purposive Sampling
Sample Size: 105
54. Table 1 :Effectiveness of treatment towards
knowledge gain due to exposure to PES
Sl
no
Treatments Mean knowledge gain Mean
knowledge
gain
Percentag
e (%)
Before
exposure
Immediately
after
exposure
1 Marginal & Small
Farmers (T1)
5.11 9.94 4.83 13.80
2 Medium Farmers (T2) 4.71 10.23 5.52 15.77
3 Large Farmers (T3) 4.97 10.28 5.31 15.17
Total 14.79 30.45 15.66 44.74
55. Table 2 :Effectiveness of treatment towards knowledge
related due to exposure to PES
Sl
no
Treatments Mean Skill Acquisition Mean Skill
Acquistio
n
Percentag
e (%)
Before
exposure
Immediatel
y after
exposure
1 Marginal & Small
Farmers (T1)
3.40 10.17 6.77 19.34
2 Medium Farmers
(T2)
3.28 9.80 6.52 18.63
3 Large Farmers
(T3)
3.86 9.46 5.60 16.00
Total 10.54 29.43 18.89 53.97
56. Table 3 :Effectiveness of the treatment PES in terms of
Symbolic Adoption behaviour
Sl no Treatments No . of
Respondents
Percentage (%)
Marginal & Small Farmers
1 Low 4 11.43
2 Medium 23 65.71
3 High 8 22.86
Total 35 100.00
Medium Farmers
1 Low 4 11.43
2 Medium 29 82.86
3 High 2 5.71
Total 35 100.00
Large Farmers
1 Low 5 14.28
2 Medium 19 54.28
3 High 11 31.44