MASENO UNIVERSITY
SCHOOL OF COMPUTING AND INFORMATICS
DEPARTMENT OF COMPUTER SCIENCE
UNIT CODE: CCS323
UNIT NAME: GROUP PROJECT
PROJECT TITLE: E-KULIMA SYSTEM
REGISTRATION NO.
NAME
CI/00099/016
ABRAHAM OMONDI
CI/00038/018
VICTOR KIOGORA
CI/00102/018
ONYANGO BENSONS
THIS PROJECT PROPOSAL IS HEREBY PRESENTED FOR EXAMINATION WITH THE
APPROVAL OF THE PROJECT SUPERVISOR.
DECLARATION
We hereby declare that this proposal is our original work and has not been presented for
development on any platform.
NAME
SIGNATURE
DATE
Abraham Omondi
…………………………
………...
………………………..
Victor Kiogora
…………………………
……….
………………………….
………..
…………………………
Onyango Bensons
………………………..
ii
ABSTRACT
The internet was opened to general users in 1994 and this new era of information and
communication technology has played an important role in the field of expert systems. The Web
technologies allowed the knowledge engineers and domain experts to build the expert systems that
were having dynamic knowledgebase capabilities.
The domain experts could update the
knowledge at the central server and the users had an access to the recent knowledgebase through
a Web interface.
Expert systems have the capability to transfer location specific technology and advice to the
farmers efficiently and effectively. This in turn will reduce losses due diseases and pests’
infestation, improve productivity with proper variety selection and increase income of the farmer.
The taxonomy and identification of crop infection have the foremost technical and economic
importance in the Agricultural Industry. However, disease detection needs incessant observing of
specialists which might be prohibitively costly in big farms region. Automatic recognition of plant
diseases is necessary to research themes as it may benefit in monitoring huge fields of crops and
thus automatically detect the symptoms of diseases as soon as they appear on plant leaves.
The goal of this application is to develop a system which will recognize crop diseases. In this user
will have to take an image through a mobile phone, Image processing starts with the digitized color
image of the diseased leaf. Finally, by applying the SVM (Support Vector Machines) plant disease
would be predicted:
 It will simulate human reasoning about a problem domain, rather than simulating the
domain itself.
 It will perform reasoning over representations of human knowledge.
 It will solve problems by heuristic or approximate methods. The expert system will apply
to the problems of diagnosing Soybean diseases, one of the earliest expert systems
developed in agriculture.
A unique feature of the system is that it will use two types of decision rules:
 The rules representing expert’s diagnostic knowledge
 The rules obtained through inductive learning from several hundred cases of disease.
iii
ACKNOWLEDGEMENT
I would like to express my deep gratitude to my project supervisor (Mr. Adongo) for his valuable
and constructive suggestion during the project.
I would like to appreciate the school of computing for offering us an opportunity to take the course
and develop this system.
Our next acknowledgement is to the group members for extensive research and dedication towards
developing the development of the project
iv
TABLE OF CONTENTS
Contents
DECLARATION ............................................................................................................................................... ii
ABSTRACT..................................................................................................................................................... iii
ACKNOWLEDGEMENT .................................................................................................................................. iv
CHAPTER ONE ............................................................................................................................................... 1
1.0 INTRODUCTION. ................................................................................................................................. 1
1.1 PROBLEM STATEMENT. ...................................................................................................................... 1
1.2 PURPOSE OF THE STUDY .................................................................................................................... 2
1.3 OBJECTIVES OF THE STUDY ................................................................................................................ 4
1.4 SIGNIFICANCE OF THE STUDY ............................................................................................................ 4
1.5 SCOPE OF THE PROJECT...................................................................................................................... 5
CHAPTER TWO .............................................................................................................................................. 6
2.0 LITERATURE OVERVIEW ..................................................................................................................... 6
2.1 Agriculture in Kenya ........................................................................................................................... 6
2.2 Agricultural extension ........................................................................................................................ 7
2.3 Existing information systems in Kenya .............................................................................................. 8
CHAPTER THREE.......................................................................................................................................... 12
3.0 METHODOLOGY................................................................................................................................ 12
3.1 METHODOLOGY................................................................................................................................ 12
3.1.1 Object-oriented analysis and design (OODA)........................................................................... 12
3.1.2 Object oriented analysis (OOA) ................................................................................................ 12
3.1.3 Object oriented design (OOD) .................................................................................................. 12
3.2 REQUIREMENTS GATHERING. .......................................................................................................... 16
3.2.1 Case Based Reasoning ............................................................................................................... 16
3.2.1 Non-functional Requirements .................................................................................................. 18
3.3 SYSTEM ANALYSIS. ........................................................................................................................... 19
3.3.1 Feasibility study ......................................................................................................................... 20
3.4 System Design. ................................................................................................................................. 22
3.4.1 Knowledge Base ........................................................................................................................ 24
3.4.2 Working Memory ...................................................................................................................... 24
v
3.4.3 Inference Engine ........................................................................................................................ 24
3.4.4 Diagnostic System ..................................................................................................................... 25
3.4.5 Backend (CBR) ........................................................................................................................... 25
3.4.6 Frontend .................................................................................................................................... 26
3.5 Coding and Debugging. .................................................................................................................... 29
3.6 Testing For The Prototype................................................................................................................ 30
3.6.1 Test on Functional Requirements ............................................................................................. 31
3.6.2 Module Testing .......................................................................................................................... 32
3.6.3 Regression Testing..................................................................................................................... 32
3.6.4 System Testing........................................................................................................................... 32
CHAPTER FOUR ........................................................................................................................................... 33
4.0 RESOURCES, BUDGET ESTIMATE AND PROJECT PLAN. ................................................................... 33
4.1 RESOURCES. ...................................................................................................................................... 33
4.1.1 Hardware Platform.................................................................................................................... 33
4.1.2 Software Development Platform: ............................................................................................ 33
4.1.3 Database: ................................................................................................................................... 33
4.1.4 Programming languages: .......................................................................................................... 33
4.1.5 Communication: ........................................................................................................................ 34
4.2 BUDGET ESTIMATE. .......................................................................................................................... 34
4.3 PROJECT PLAN. ................................................................................................................................. 34
4.3.1 ACTIVITIES. ................................................................................................................................ 34
4.3.2 GANT CHART.............................................................................................................................. 35
CHAPTER FIVE ............................................................................................................................................. 36
5.0 CONCLUSION .................................................................................................................................... 36
5.1 REFERENCES ...................................................................................................................................... 37
vi
CHAPTER ONE
1.0 INTRODUCTION.
1.1 PROBLEM STATEMENT.
In spite of the various approaches to improve production by clearing and diagnosing plant diseases
in Kenya, diseases are still a problem towards food production. One of the reasons for limited
impacts in farmers’ income is the modern farmer often relies on agricultural specialists and
advisors to provide information for decision making. Despite the deployment of extension officers
to offer agricultural diagnostic services to farmers, the ratio of extension officers to farmers still
stands at 1:1500 farmers. In addition, there are challenges encountered including: inadequate
budgetary allocation to agriculture, which eventually trickles down to the low budgetary allocation
to the operations and movement of extension officer to offer services to the farmers. Moreover,
Kenya is characterized by poor road infrastructure which raises the cost of ware and tare and
making some areas inaccessible to extension officers.
On the other hand, service providers in the telecommunication industry such as Safaricom have
expanded their networks to over 80percent of the country. The Government has facilitated internet
connectivity through fiber optic cables to major towns. The policy of provision of computers to
every child joining class one, is projected to transform the destiny of future generations in terms
of ICT education and application in all sectors of the economy. Food and income insecurity have
been attributed to limited access to production inputs such as diseases, pesticides and insects. This
has resulted in wastage of produce and low prices to smallholder farmers.
This study therefore seeks to establish technological efficacy of extension officers to guide the
development of an artificial intelligent system for agricultural diagnostics.
1
1.2 PURPOSE OF THE STUDY
Agriculture dominates the Kenyan economy, accounting for 70 percent of the workforce and about
25 percent of the annual gross domestic product (GDP). As the financial system of Kenya is
relayed on farming production, the extreme concern of food production is essential. The issue of
improving agriculture in order to increase its productivity has been given due weight and attention
in Kenya. A shift towards self-sufficiency in food production in Kenya depends to a greater extent
on the improvement of agriculture. Agricultural production has evolved into a complex business
requiring the accumulation and integration of knowledge and information from many diverse
sources. In order to remain competitive, the modern farmer often relies on agricultural specialists
and advisors to provide information for decision making. Unfortunately, agricultural specialist
assistance is not always available when the farmer needs it. In order to alleviate this problem,
expert systems were identified as a powerful tool with extensive potential in agriculture.
Expert Systems (ES) development is considered as a division of the Artificial Intelligence
fraternity. Expert system (ES) is an intelligent software which works in one particular domain. It
uses knowledge and inference capability to solve problems. The core idea of ES development is
to convert the available human knowledge into the computer. So, this knowledge can be used when
required. The expert system provides the knowledge in the usable form. The expert systems proved
powerful tool to solve many real-world problems of technological, social, agricultural and life
science spheres. This has been resulted in the ES development as a prominent area of research not
only for the Artificial Intelligence branch but also for many interdisciplinary research works.
As expert system is based upon knowledge so this is also known as knowledge-based system.
Knowledge residing in knowledge base of expert system consists of domain facts and associated
heuristics. This knowledge is collected from human experts who are not available every time.
Moreover, one or two experts are not enough for some working areas just like agriculture where
so many fields are part of one major field. That is why knowledge is gathered from more than one
expert and accumulated to a software, which is called expert system. As expert systems have been
using in different fields of life like medicine, process controlling etc. so agriculture field has also
not been left affected by these knowledge-based systems. These systems are being used by
agricultural decision makers at different levels: “operation level and planning level. On the
operation level, the extension workers in the village, district, and/or governorate can use the system
to support him in making his decision in giving the appropriate advice to the growers. On the
2
planning level, the decision makers can use the expert system to predict the needs of water,
fertilizers, and pesticides.
Baloyi (2010) showed that considerable changes would be required in small holders farming
operations if the economic benefits of increased incomes would be fully realized. These changes
entail producing good-quality, high-value crops on a large scale and accessing high-value markets.
This will only happen if smallholder farmers have access to comprehensive and holistic
agricultural support services. Pests like the germ, fungus, and microorganisms are the origin of the
disease to plants through a failure in excellence and extent of production. There is a great quantity
of loss of farmer in making the crops. Hence proper care of plants is necessary for same. Research
on the discovery of plant disease is gaining importance nowadays, which may prove useful in
1
observing the large area and consequently mechanically finding the symptoms as they come out
on plant. This project will give a general idea of with image processing technique to identify a
variety of plant diseases. It will offer more capable ways to discover infection created by fungus,
microorganisms or germ on plants. Simple interpretation by eyes to identify ailment is not precise.
An overindulge of pesticides create destructive chronic diseases in human beings as not cleaned
correctly. It also damages plants nutrient superiority. Which come out to be an enormous defeat of
production to the farmer. Hence the image processing techniques can bring into play to discover
and categorize diseases in agricultural applications is supportive.
3
1.3 OBJECTIVES OF THE STUDY
 The general objective of this study is to increase the food production by bringing a 24/7
access to agricultural advice to the farmer.
 The project also intends to increase the income of the modern farmer which inclines to
increase of the annual GDP to improve the living standards and the lifestyle of the farmers
in Kenya.
1.4 SIGNIFICANCE OF THE STUDY
The adoption of knowledge driven expert system will increase food production and income of the
modern farmer since plant pathology solutions are brought to the farmer without limitations of
distance or interaction with an agricultural officer. The program recognizes that increases in
agricultural productivity with a defined market will result in improved incomes of the farmers. The
assessment will be done on the production, pre-handling process of the goods, easy access to high
quality medications, credit facility and the acceptability of their produce to highly competitive
trading systems on the income of the smallholder farmers. The research work will ascertain the
effectiveness of the approach in easy access to production inputs, finance and markets in
augmenting incomes for smallholder farmers, the level of participation of both the large-scale
farmer and small-scale farmers, document the success factors for the sustainability of the approach
for up scaling in other projects and investigate, strengths weaknesses, opportunities and threats of
the approach via conventional approaches.
4
1.5 SCOPE OF THE PROJECT
The scope of the study was to develop and implement diseases/pest diagnostic system based on
knowledge-based reasoning where by previously solved cases are used to solve the current
problems. The system supports several languages including English and Swahili.
The system can be accessed through two user interfaces namely:
 Dynamic Website.
 PWA (Progressive Web Application
5
CHAPTER TWO
2.0 LITERATURE OVERVIEW
This study was a multi-disciplinary study touching on various academic areas ranging from
agriculture and agricultural extension, crop protection, communication and computerized expert
systems development. This section provides a comprehensive literature review and highlights the
gaps which exist on the mentioned sectors.
2.1 Agriculture in Kenya
The agricultural sector in Kenya has recently recoded improved performance especially in
important commodities and enterprises such as horticulture, tea, dairy and maize. The agricultural
sector is reviving and is on a trajectory of further development. However, challenges remain in
some commodities such as coffee, sugar, pyrethrum and exploiting livestock and fisheries
potential. Emerging constraints to agricultural growth need to be addressed. Challenges and
constraints facing the sector vary with commodity and region. The effects of some of these
challenges and constraints were accelerated by the worldwide food price crisis and its underlying
drivers in 2008. Some of the general challenges include inadequate budgetary allocation; reduced
effectiveness of extension services; low absorption of modern technology; high cost and increased
adulteration of key inputs; pre- and post-harvest crop losses; livestock losses to diseases, pests and
insecurity; limited capital and access to affordable credit; low and declining soil fertility;
Inappropriate legal and regulatory framework; Inadequate disaster preparedness and response and
inadequate infrastructure (GoK, 2010).
Some of the challenges outlined above are a pointer to the fact that Kenya has inadequate human
resource capacity. This could be attributed to first, insufficient budgetary allocation which leads
to reduced operation budgets and hence inability of extension officers to access all farmers.
Secondly was reduced effectiveness of extension services due to inappropriate methods. The
information flow from researchers to farmers through extension officers provides a room for
distortion. Thirdly, pre- and post-harvest crop losses due to mainly lack of correct and timely
diagnostic information services which is brought about by high extension officer to farmer ratio.
6
Lastly, heavy livestock losses to diseases and pests due to low surveillance and understaffing
(GoK, 2010).
Several of the above challenges are information and knowledge related. The agricultural sector
extension service plays a key role in disseminating knowledge, technologies, agricultural
information and linking farmers with other actors in the economy. Extension service is one of the
critical change mediums required in transforming subsistence farming to modern, commercial and
smart agriculture. The transformation is required to promote household food security, improve
income and reduce poverty.
2.2 Agricultural extension
Extension approaches in Kenya include Focal Area Approach (FAA)– (Use of common interest
groups (CIGs); Farmer Field Schools – Farmer to Farmer Extension; Commodity-based approach
- commercial enterprises; Multidisciplinary Mobile Extension Teams especially in arid and semiarid areas (FAO, 2011). Despite the various extension approaches employed by the private and
public sector, several challenges exist. These include reduced staffing and funding for operations
and maintenance (GoK, 2010), extension and dissemination of conflicting messages (World Bank
2011), unnecessary competition and duplication of efforts and lack of synergy (World Bank 2011).
Extension officers play an important role in provision of diagnostic services to farmers. Despite
the various extension approaches employed by the private and public sector, several challenges
are encountered in public sector. These challenges include: reduced staffing and funding for
operations and maintenance of extension service delivery. In private sector the challenges include
dissemination of conflicting messages, s; unnecessary competition, duplication of efforts, and
general lack of synergy among extension service.
These challenges have led to limited access to credible extension services in most parts of the
country. The national extension staff to farmer ratio is 1:1,500 (Africa Science News Saturday, 16
November 2013). An agricultural system is considered sustainable when it satisfies producers
‘needs and preserves natural resources for current and future generation. Development of the
system should be based on three pillars: economic feasibility, social fairness and environmental
sustainability (Carlos, 2006). This study was designed with the objective of enhancing provision
of extension services, by developing a computerized artificial intelligent information system that
7
meets economic, social, and environmental feasibility. The system was designed to improve
delivery of diagnostic and management services for maize diseases in Kenya.
An agricultural system can be considered sustainable when it satisfies producers ‘needs and
preserves the natural resources for the present and subsequent generations. Development of the
system should be based on three pillars: economic feasibility, social fairness and environmental
sustainability (Carlos, 2006). Given the challenges facing the delivery of extension services in
Kenya, the country is still far from achieving sustainability.
2.3 Existing information systems in Kenya
ICT in Kenya is growing at a rapid rate. The penetration of mobile service in Kenya reached 64.2
percent by 2012 (Okyere, 2012). Kenya ‘s high mobile penetration rate and subscription number
indicates that mobile technology is a promising business opportunity, and an indispensable tool
for empowering the country ‘s citizens, especially its rural poor. The services widely used include
use of MPesa which is a mobile money transfer service owned by Safaricom and launched in 2007.
It allows users to use their mobile phones to send, receive and transfer money; MFarm, an
agribusiness software solution which was started in 2010 and offers information to farmers on
farming and market information to improve their productivity through sending an SMS (short
messaging service) to 20225; iCOW. This is an SMS and voice-based mobile application launched
in 2013 and used by dairy farmers to access information on the cows ‘gestation period, veterinary
information and record keeping; Airtel Kilimo is a unique and innovative service aimed at
providing phone-based agricultural information, advice and support to smallholder farmers over
Airtel ‘s mobile network, launched in 2011. This service is utilizing Africa ‘s mobile network and
technologies to bridge the knowledge gap in rural areas. The service can be subscribed on *760#
for free SMS subscription where a customer is charged KSh3 per SMS and on Interactive Voice
Response (IVR) for KSh3 per minute (Maritz, 2011).
Another ICT program is the Kenya Agricultural Commodity Exchange (KACE), established in
1997, which has offers and bids. These services are prominently displayed on blackboard and are
disseminated via SMS and Internet. KACE collects, updates, analyses and provides reliable and
8
timely market information and intelligence on a wide range of crop and livestock commodities,
targeting actors in commodity value chains, with particular attention to smallholder farmers and
small-scale agribusinesses (KACE, 2011). The components of the KACE links are: market
resource centers, mobile phone short messaging service (SMS), interactive voice response service,
internet database system, radio and the KACE headquarters central hub in Nairobi. All these
applications help in accessing information on daily wholesale buying prices for over 20
commodities as well as offers to sell and bids to buy (KACE, 2011).
Mobile phones are also used to distribute agricultural insurance products to farmers, most of whom
cannot afford conventional insurance. A product called Kilimo Salama, Swahili for ‗safe
agriculture ‘, enables smallholder farmers in Kenya to insure their agricultural inputs against
adverse weather conditions, such as drought or too much rain. Developed by UAP Insurance, the
Syngenta Foundation for Sustainable Agriculture and Safaricom, Kilimo Salama allows
smallholder farmers to insure their inputs and produce. To be covered under the scheme, farmers
only need to pay an extra 5percent for a bag of seed, fertilizer or other inputs.
Mobile technology plays a central role in the scheme as it is used both for registration of new
policies as well as for payouts (Okeene, 2012). Kilimo Salama is distributed mostly through agro
dealers that have been equipped with a camera phone that scans a special bar code at the time of
purchase, which immediately registers the policy with UAP Insurance over Safaricom mobile data
network. This innovative application then sends a SMS message confirming the insurance policy
to the farmer ‘s handset. Payouts are determined by automated weather stations that monitor the
rainfall. Based on the stations ‘measurements and a predefined formula of crop rainfall needs,
payouts are automatically made to farmers using Safaricom mobile money transfer service, MPesa. Farmers do not have to fill out any claim forms. Since its official launch in 2010, the scheme
has already made payouts to numerous farmers (Okyere, 2012). It is expected that products like
Kilimo Salama will increase productivity since only about half of Kenyan farmers invest in
improved seeds and soil inputs (Okyere, 2012). A key reason for the low demand is the fear among
farmers that poor conditions, such as drought, will render their investment worthless, robbing them
of both their crops and their savings.
The literature indicates systems in various agricultural value chains; however, none of the existing
systems provide agricultural diagnostics. In addition, the SMS based systems have limited words
9
since an SMS cannot take more than 150 letters. Because of this reason, huge information on
diagnostics and prescription might not fit through the existing systems. On the other hand, the
existing systems are not user centered but only technologically center. It is important to develop
systems which are user driven for ease of adoption.
Mittal (2009) studied the role of mobile technology in improving small farm productivity in India
by looking at the potential solution mobile phones could have in information asymmetry in
agricultural sector. The study used focus group discussions and in-depth interviews with farmers
to find answers to the use and impact of mobile phones and mobile-enabled services on agricultural
productivity. The results showed that although mobile phones can act as catalyst to improving farm
productivity and rural incomes, the quality of information, timeliness of information and
trustworthiness of information are the three important aspects that have to be delivered to the
farmers to meet their needs and expectations. This implies that the major factors for adoption of
the technology were timeliness, quality of information and trust (Mittal, 2009)
The literature identified several gaps in Kenya agriculture which includes low budgetary
allocation, ineffective extension services, low adoption of modern technologies; pre and postharvest crop losses to diseases and pests, inadequate disaster preparedness, high cost and increased
adulteration of farm inputs, low and declining soil fertility and inadequate infrastructure. The
literature review identified several gaps in the extension system in Kenya which includes reduced
staffing and funding for operation and maintenance, dissemination of conflicting messages,
reduced mobility due to poor road infrastructure, and low usage of modern communication
technologies in extension among others.
The various types of research and indicated that evaluative and developmental research are two
research approaches directed toward solving problems (Ackoff, 2005). The developmental type of
research "involves the search for (and perhaps construction or synthesis of) instructions" that yield
a better course of action (Ackoff, 2005). Developmental research has largely been ignored by some
researchers. However, without research efforts directed toward developing new solutions and
systems, there would be little opportunity for evaluative research. Information system development
is a credited research methodology. This study will adopt the information system development
research methodology for the development of the artificial intelligent system. Use of modern ICTs
10
in the medical field; where ICTs have been used successfully to support diagnosis of disease.
However, these ICTs have only been used to provide information to physicians (and not the
patient) that eventually makes decisions on diagnosis and management. This study will be
designed to develop an artificial intelligent system that will use the symptoms observed by farmers
to diagnose disease and provide necessary prescription to be administered.
Literature also highlights the importance of maize crop to the Kenyan economy and society at
large, the challenges affecting maize production in relation to disease diagnostics and
management. The findings/outcomes of this study will eliminate most of the challenges through
provision of information to farmers. The study will aim at providing a fully functional prototype
of an agricultural computer aided diagnostic system for crop diseases. The study will be guided by
several theories/ models. The theory guiding the study will be the system theory. The umbrella
methodology for the study would be the system development research methodology. Within the
system development research methodology, traditional waterfall system development model will
be used to develop the system.
Literature identifies the gap in the existing agricultural extension communication model; long
communication chain which hinders the achievement of effective communication.
11
CHAPTER THREE
3.0 METHODOLOGY
3.1 METHODOLOGY.
Object oriented analysis and design will be adapted for the system design.
3.1.1 Object-oriented analysis and design (OODA)
This is a software engineering approach that models a system as a group of interacting objects.
Each object represents some entity of interest in the system being modeled, and is characterized
by its class, state (data elements) and behavior.
3.1.2 Object oriented analysis (OOA)
This applies object modeling techniques to analyze the functional requirements for a system.
Object-oriented design (OOD) elaborates the analysis models to produce implementation
specifications.
The outcome of object-oriented analysis is a description of what the system is functionally required
to do, in a form of a conceptual model.
3.1.3 Object oriented design (OOD)
This transforms the conceptual model produced in object-oriented analysis to take account of the
constraints imposed by the architecture and any non-functional requirements. The concepts in the
analysis model are mapped onto implementation classes and interfaces. The result is model of the
solution domain, a detailed description of how the system is to be built.
12
Database Schema
13
Entity relationship diagram
14
Sequence Diagram
15
Activity Diagram
3.2 REQUIREMENTS GATHERING.
3.2.1 Case Based Reasoning
1.Capture problems and solutions
This module provides an interface which allows the system manager to key in problems, their
corresponding solution and explanation as given by the expert. It also enables the manager to
perform administrative tasks for the system.
Inputs
The parameters include: Case description (symptoms, disease names, images of the symptoms),
Solutions to the described cases and explanation and locations on where to purchase the prescribed
solution.
Operations
a) This module provides the user with an interface to capture each of the inputs specified above.
16
b) The module validates the captured input.
c) The module provides an interface for deleting the unwanted cases and updating the existing
cases and their corresponding solutions.
Outputs
The module displays a message indicating that the status of the operation has been completed
2.Processing user problems
This module allows users to post problems through two interfaces namely: WAP application and
Website application
Inputs
User problem description in a case format.
Operations
a) Retrieve – Upon receiving the new case from the user, the case base reasoning searches the
database to find the most approximate case to the current problem. The retrieval starts with
description and ends when the best matching from previous case handled by the system.
b) Reuse – This is where the case-based reasoning uses the retrieved case and adapting it to the
new situation. The case-based reasoning proposes a solution for each retrieved case, the
corresponding strategy used and the results obtained are looked up, at the end of the process.
c) Revise – In case a case generated by the reuse phase is not correct, an opportunity for learning
from failure a rise. The case solutions are evaluated via expert feedback and if successful, learning
from the success, otherwise repair the case solution using domain-specific knowledge.
d) Retain – This process enables CBR to learn and create a new solution and a new case that
should be added to the case base. It incorporates what is useful to retain from the new problemsolving episode into the existing knowledge. The learning from success or failure of the proposed
solution is triggered by the outcome of the evaluation and possible repair. It involves selecting
which information from the case to retain, in what form to retain it, how to index the case for later
retrieval from similar problems, and how to integrate the new case in the memory structure.
17
3.2.1 Non-functional Requirements
Performance Requirements
The artificial intelligent system for maize disease diagnostic and management supports real time
processing and communication.
Security Requirements
Access to the artificial intelligent system for maize diagnostic and management is limited to
authenticated users whose roles have been defined at the database level.
Software Quality Attributes
Below are key expected software quality attributes of the artificial intelligent system for maize
disease diagnostic and management.
Availability
The artificial intelligent system for maize disease diagnostic and management is expected to
operate continuously from time of installation.
The artificial intelligent system for maize disease diagnostic and management is expected to run
24 hours a day, 7 days a week, 365.25 days a year and throughout its lifetime.
Reliability
The artificial intelligent system for maize disease diagnostic and management is expected to handle
system failures elegantly.
Flexibility
The addition of new modules or upgrading of existing sub system should not cause interference
with existing modules.
Assumptions and Dependencies

The users have the necessary technology which enables them to access and use the system
18

They farmers have access to the internet that is reliable, that is available all the time they
want it.
3.3 SYSTEM ANALYSIS.
This study will be conducted in the sample counties Uasin Gishu (the largest maize producer in
Kenya), Kericho (the largest tea producer in Kenya) and Bomet (the largest potato producing
county).
Uasin Gishu County one of the 47 counties and the second largest maize producer in Kenya. The
2009 population census in was estimated at 894,179 people and 202,291 households.
(http://www.scribd.com/doc/36672705/Kenya-Census-2009). There are two rainy seasons with an
annual rainfall ranging between 900 to 1200 mm. Seated on a plateau; the county has a cool and
temperate climate, with annual temperatures ranging between 8.4 °C and 27 °C. The wet season is
experienced between April and May, while the dry season is encountered between January and
February. The county is purposively selected due to its favorable climate to maize production, the
large number of experienced maize farmers and extension officers and the high
prevalence/incidence of diseases and pests affecting maize occurrence.
Kericho is endowed with ideal climate; tropical, volcanic red soils; well distributed rainfall ranging
between 1200mm to 1400mm per annum favorable for tea growing. The task of managing small
scale holder lies with the Kenya Tea Development Agency (KTDA). KTDA reports 66 tea
factories serving over 500,000 small scale farmers cultivating over 100,000 ha. Of the tea
produced over 60% is managed by KTDA.
Potato is the second most important food crop after maize in Kenya. It is cultivated by over
800,000 growers who are mostly small holder. Potato production in Kenya is currently about Sh
50 billion. This figure can be increased if production is optimized. Potato has capability of
providing staple food requirements since maize decline upon Maize Lethal Necrosis Disease
(MLND). Bomet has favorable climatic conditions for potato with annual rainfall of between 8501200mm and altitudes of between 1500m and 2800m above sea level.
Data to be collected using interview guide. The study sample size would be estimated using the
formula proposed by Mugenda and Mugenda (2003):
19
Where;
n = the desired sample size for N>10,000,
Z = constant associated with the required confidence level
p = proportion of the population that possessed the target characteristics. The target characteristics
will include farmer with and without access to extension services.
d = the level of precision estimated
Systematic random sampling technique would be applied to select the households.
The sampling interval (k) would be determined using the formula = /n, where N is the population
size and n is the sample size.
The operation of the proposed system is to collect data on the spot using automated system that is
being designed. All that the user needs are to login to the interface using the appropriate uniform
resource locator and upload the data at a time using the registration form and the data will be saved
and can be accessible immediately at the head office that is it is real time.
3.3.1 Feasibility study
The types of feasibility study include:
1.Operational feasibility
Operational feasibility is performed to determine whether the operations of the e-kulima system
will be acceptable to the users and if the intended users will be well trained and experienced to
work with the system efficiently or there is need to recruit the existing users and employ new
instructors. It is determined that there are computer literate personnel who needed only to be
introduced and guided on how to use the system so as to be able to navigate through it.
A good number of instructors and farmers we interviewed were in full support of the project, as
there will be improvement in the farming sector in general with time as more people begin to use
the system. The implementation of functional requirements and non-functional requirements of
the system ensured better easy learning was achieved.
20
2. Economic feasibility
The economic feasibility was carried to determine the cost of developing, operating and
maintaining the e-kulima System and its economic viability in relation to return on investment and
service delivery. It is proposed that the Complementary e-kulima System was feasible because the
developers of this project who are also the sole sponsors and managers have enough financial
resources to support the implementation of the new system and farming standards would greatly
improve with introduction of a new system that is well equipped hence the system is economically
feasible.
3.Technical feasibility
Technical feasibility is concerned with the availability of hardware and software required for the
development of the system, to see compatibility and maturity of the technology proposed to be
used and to see the availability of the required manpower to develop the system. After the study
we came up with the tools and development environment chosen by us. This was important in our
case as we were working together on two various phases of the department that will need to be
integrated future to make and extended system
21
3.4 System Design.
The waterfall model highlights processes which developers have to follow. These processes are
categorized in phase namely Requirements specification (Requirements analysis); Software
design; Implementation and Integration; Testing (or Validation); Deployment (or Installation) and
Maintenance. In a strict Waterfall model, after each phase is finished, it proceeds to the next one.
Reviews may occur before moving to the next phase which allows for the possibility of changes
(which may involve a formal change control process). Reviews may also be employed to ensure
that the phase is indeed complete; the phase-completion criteria are often referred to as a "gate"
that the project must pass through to move to the next phase. Waterfall discourages revisiting and
revising any prior phase once it is complete. This "inflexibility" in a pure Waterfall model has been
a source of criticism by supporters of other more "flexible" models (Kent, 2000).
This study therefore will use the traditional waterfall system development model because of the
nature of this project and the advantages the model has over all other models.
Diagnostics of crops pests and diseases is one of the major roles of extension officer. It always
involves extension officers going to visit the farms, viewing the symptoms on the crops and from
their knowledge of pests and diseases, identifying the specific disease. The extension officer then
proceeds to prescribe curative measures to the farmers for the plan based on their knowledge of
existing management option. This service has however been hindered by under staffing of
extension officer, low budgetary allocation to the agricultural sector, poor road infrastructure
among other. Because of these challenges, there is always more often delay or some time complete
failure in the delivery of the diagnostics services to farmers which eventually result into loose of
crop productivity. With the advances made in ICT in Kenya and the development of several
artificial intelligent system which can perform like humans on decision making, it was necessary
to rethink the delivery of these services to the farmers in a more speedy, efficient and cost-effective
manner.
A study will be done to establish a module which can drive an Artificial Intelligent System for
Agricultural Diagnostics. The study will involve a review of the modules of already existing
system for diagnostics in both agriculture and human medicine. Some of the systems reviews will
include Rice-Crop Doctor, Cattle Disease Diagnostic System (CaDDiS), Agricultural Pest
22
Diagnosis Using Imaging Technologies, Mobile Interfaced Crop Diagnosis Expert System
(MICDES), Computer assisted Medical diagnostic systems (CAMD) and Computer Aided
Diagnostics (CAD) in Medical Imaging. MOCDES module was adopted wholly without any
alterations
A comprehensive farmer and extension officer requirements and the electronic database for crop
diseases and their management provides satisfactory information for the artificial intelligent
system modeling.
An artificial intelligent is an application which contains the knowledge of a human expert about
solving problems in a bounded environment. Such applications have been developed such as
manufacturing, medicine, business procedures or engineering.
The most important goals of an expert system are:
 To conserve the knowledge of human experts.
 To provide help to laymen and experts in solving problems.
 To make information readily available that is difficult to recall.
Major Components of Artificial intelligent systems
An artificial intelligent system consists of three major modules:
 The knowledge base
 The inference engine
 Working memory
This portioning is deduced from how human expert problem solving works.
23
3.4.1 Knowledge Base
The highly specialized knowledge of the problem area is located in the knowledge-base. This
module will contain problem facts, rules, concepts and relationships. The first step to build the
knowledge base is to gather the knowledge from human experts. After this step, the knowledge
must be coded in a form which is useful for automatic processing. Several techniques of knowledge
representation are available including a rule-based representation and case-based representation.
3.4.2 Working Memory
The working memory contains the facts about a problem that are collected during one consultation
of the artificial intelligent system. When a new problem has to be solved, the user enters
information about the problem in the working memory. The artificial intelligent system then uses
this information together with the facts in the knowledge base to infer new facts. The content of
the working memory can also consist of facts that have been collected from external storage like
databases, spreadsheets, or sensors, beside the information that is taken from knowledge base.
3.4.3 Inference Engine
The reasoning of the expert is molded in the knowledge processor, usually named the inference
engine. The engine uses knowledge base as the source of information for reasoning. Together with
the other available information to draw conclusions and or give recommendations as to how
specific problem can be solved.
Another task for the inference engine is to reason with uncertain data. The certainty associated
with a value is usually described by a numeric value, and various mathematical rules are employed
to derive the joint confidence across sets of values.
24
In this section the focus is given on the previously developed artificial intelligent systems which
are used to disseminate information and SMS based systems.
3.4.4 Diagnostic System
The Diagnostic System will be divided into two components:
 The backend – Case-based reasoning (CBR) system
 The frontend – Interfaces for accessing the backend
3.4.5 Backend (CBR)
Case-based Reasoning means to use previous experience in form of cases to understand and solve
new problems. A case-based reasoning remembers former cases similar to the current problem and
attempts to modify their solutions to fit for the current case. The underlying idea is the assumption
that similar problems have similar solutions. Though this assumption is not always true, it holds
for many practical domains.
CBR consist of four modules:

RETRIEVE

REUSE

REVISE

RETAIN
When a new problem is matched against cases in the case base and one or more similar cases are
retrieved. A solution suggested by the matching cases is then reused and tested for success. Unless
the retrieved case is a close match the solution will probably have to be revised producing a new
case that can be retained and this will be done by the case Manager.
25
CBR directly addresses the following problems found in rule-based technology.
 CBR does not require an explicit domain model and so elicitation becomes a task of
gathering case histories
 Implementation is reduced to identifying significant features that describe a case, an easier
task than creating an explicit model
 By applying database techniques largely volumes of information can be managed
 CBR systems can learn by acquiring new knowledge as cases thus making maintenance
easier.
3.4.6 Frontend
The proposed system will provide two modules through which users will interact with it and these
include:
1. Website (response will be both in text format and picture format)
2. PWA (work as the usual offline mobile application)
Through these interfaces user diversities will be taken care of, making the system more accessible
and acceptable to the users.
Website: The system provides a web interface through which users will key in a new problem
either in text mode or a picture template. If a match to the posted problem is found the proposed
solution will be presented in either text format, picture or video format.
The process is shown in the diagram below:
26
The knowledge processing module will have Rule based Identification and Ontology based
Problem Identification sub modules. In Rule based Identification, the users will enter the question
answering mode where he/she will feed in the symptoms of the diseased plant. Here, the platform
is in text format. This session will yield the solution of the problem such as of disease diagnosis
or identification of pests.
In Ontology based Problem Identification the user will also enter the question answering session,
but here the platform is allowed in picture form. The user will be able to input by either capturing
a picture of the affected plant or will be an option of uploading a picture file from gallery or any
file from the phone storage. An Image recognition system is incorporated: the photo taken by the
farmer should be recognized by the system and critical accurate decisions made upon it. The image
data will be fed into the system once uploaded, the closest measurements will be used to classify
the image data and diagnosis given upon it.
The presented expert system will be based on the n-tier model of the web applications. This model
will allow different components of the system to be built by different experts, specialized in their
27
domain. Knowledge base and inference engine are the two most important components of an
expert system. The basic principal of the separation of the knowledge from its treatment is of
prime importance in the building of every expert system. The building and maintenance of an
expert system is greatly facilitated by trying to adhere to this principal as closely as possible.
1. The Knowledge Base Layer (KBL): the knowledgebase will be built using ontology. It
will contain knowledge about plant varieties, diseases and insect-pests.
2. The Database Layer (DBL): this layer will be implemented using MS SQL Server 8.0
database. This will contain the authorization information about users and crop specific
information.
3. The Reasoning Engine: the reasoning engine will accept user input queries and responses
to questions through the I/O interface and uses this dynamic information together with the
static knowledge stored in the knowledge base. The knowledge in the knowledge base will
then be used to derive conclusions about the current case or situation as presented by the
user’s input. JENA is used here for this purpose.
4. Server-Side Application Layer (SSAL): application layer will be built using Java Server
Pages (JSP). The JSP provides the web developers with a framework to create dynamic
content on the server using HTML, XML, Java classes, which is secure, fast and
independent of server platform.
5. Client-Side Interface Layer (CSIL): it will be implemented using HTML, CSS and
JavaScript. The CSIL consists of forms for accepting information from the user and
validation those forms using JavaScript. It also provides the explanatory interface to the
users of expert system.
28
Web container
SPARQL
KNOWLEDGE BASE
SQL
JEN
User interface (Web browser)
internet
DATABASE SERVER
3.5 Coding and Debugging.
Artificial intelligent system for agricultural diagnosis, a case study of farmers in Kenya is carry
home expert system which is based on reasoning for diagnosing crop diseases using the previous
resolved cases. The system supports two input platforms which include: text and picture to make
it more acceptable, usable and accessible. The system performs diagnosis of known crop diseases
and pests.
The system is accessible through the mobile devices for example WAP enabled mobile phones
and dynamic website. The system has the ability to learn from the unsolved cases. The system
alerts the expert when it fails to diagnose a case via email who then tries to solve the case so that
the next time users present the same the system will be able to diagnose.
29
Artificial intelligent system for agricultural diagnosis, a case study for farmers has been
implemented using PHP, SQL, Java and XML and uses an open source database.
The realization of the system was through interaction with farmers and extension officers in Kenya
through interviews and issuing of questionnaires. A validation of the system was conducted to
establish the soundness and completeness of the indicated solutions.
Expert systems have the capability to transfer location specific technology and advice to the
farmers efficiently and effectively. This in turn reduces losses due diseases and pests’ infestation,
improve productivity with proper variety selection and increase income of the farmer.
The taxonomy and identification of crop infection have the foremost technical and economic
importance in the Agricultural Industry. However, disease detection needs incessant observing of
specialists which might be prohibitively costly in big farms region. Automatic recognition of plant
diseases is necessary to research
3.6 Testing For The Prototype.
To ensure a reasoning veracity and completeness of the artificial intelligent system, a number of
tests were conducted. Testing was iterative and incremental in nature. Of concern in this testing
were both the functional and non-functional requirements which had been identified during the
analysis stage. The tests were performed on an emulator for the mobile client before deploying the
application on the mobile phone.
For the web client, it was done on a local server before hosting in a public domain.
One approach will be applied to test the system:
Explore the capabilities of the system adding and removing symptoms. At each stage, check the
results given along with the calculated probabilities of the diseases to see whether they match the
analysis of which diseases are most likely to be implicated in causing these symptoms. If at any
point inaccurate information is presented, corrections were made.
A set of possible observed symptoms were defined. The signs were used as input
specification and the results compared with the mental picture of sensible diagnosis. A set
of scenarios were evaluated.
30
3.6.1 Test on Functional Requirements
A sample test data shown below in table 7.1 was prepared to test the functional
requirements of the system. It shows the expected behavior before the test and the actual
behavior after the implementation of the system
Sample Tests for Functional Requirement
Test No.
Description
Expected Behavior
1.
Veracious diagnostic The
results
Actual Behavior
application The
should
diagnostic
results
generate match the intuitive opinion
diagnosis results that of the individual doing the
are
complete
and tests.
genuine
2.
Exhibit some learning The
application With
with a variance in should
vary
weights.
results
addition
symptoms
3.
of
of
the symptoms, the confidence
confidence rating of rating
the
addition
varied
and
the
with number
of
possible
new problems
fell
until
a
solution was found
Add or subtract input It should be possible The users could provide
specification
for a user to add or additional symptoms or
remove symptoms or remove and submit for a
input specifications as diagnosis.
they wish.
31
3.6.2 Module Testing
Each module will be tested on its own. This is done to ensure that every unit/module works
properly on its own. Some of the tests carried out include entering wrong data or submitting blank
fields for required fields in order to test the outcome e.g. submitting an empty form when the
administrator is entering crop name the following error message is displayed.
3.6.3 Regression Testing
In the process of cording, new changes are sometimes made and incorporated into a module. The
new changes and the entire module are tested again to ensure that the changes did not affect the
functioning of previously verified codes.
3.6.4 System Testing
The module will be integrated after the sub modules were completed, tested and found to function
as per the expectation. The entire system is then tested to ensure that the system performed
according to specification.
32
CHAPTER FOUR
4.0 RESOURCES, BUDGET ESTIMATE AND PROJECT PLAN.
4.1 RESOURCES.
4.1.1 Hardware Platform
The hardware specifications below are chosen for the successful completion of the project since
they are sufficient to meet the functional and non-functional requirements of the system, especially
with regard to processing and response times. A WAP enabled phone was used as a minimum
requirement since GPRS enabled phone could be used alternatively.
 Pentium V Dual Core 3.0 GHz Processor
 80Gb hard disk
 512 MB RAM
 100 Kbps LAN
4.1.2 Software Development Platform:
 Wamp 5_1.6.1 (Windows)
 Windows Vista environment
 Macromedia Dreamweaver 8
 Internet Explorer version 6.0
 NetBeans Open source
 Open wave SDK version 6.2.2
 Visual Studio.
4.1.3 Database:
 MySQL database
4.1.4 Programming languages:
 XML
33
 XHTML
 Java
 PHP
 SQL
4.1.5 Communication:
 TCP/IP, GPRS wireless connectivity
4.2 BUDGET ESTIMATE.
Required
Unit
Quantity
Unit Price
Total Price
Item
1.
Laptop
Each
1
40,000.00
40,000.00
2.
Backup Disk
TB
1
7,500.00
7,500.00
3.
Internet Usage
GB
10
2,000.00
20,000.00
4.
TOTAL
67500.00
4.3 PROJECT PLAN.
4.3.1 ACTIVITIES.
 Requirement gathering
 Analysis
 Design
 Coding and Debugging
 Testing
34
4.3.2 GANT CHART.
WEEK/DATE
1st August 2021
5th September 2021 3rd
to
to
October
2021
2nd September 2021 30th October 2021
to
21st Nov 2021
S. No. Project Activities
1
Requirement gathering
2
Analysis
3
Design
4
Coding and Debugging
5
Testing
1
2
3
4
5
35
6
7
8
9
10
11
12
13
CHAPTER FIVE
5.0 CONCLUSION
As an economical aspect, expert systems can clearly affect over all agricultural production.
Similarly, we can preserve our natural resources instead of being polluted because of excessive
and wrong usage of chemicals and fertilizers. Moreover, better training of extension workers
through these systems can also be beneficial. These systems are completely computer based.
The theory of Expert System is well developed and matured and can be applied to a wide spectrum
of agricultural problem. Perhaps one of the greatest hindrances in increasing crop production today
is that of transferring new agriculture technologies developed at laboratories to the farmer’s field.
Expert System technology is an ideal approach for transferring the crop production technologies
to the farmer’s level, the ultimate consumer of agriculture research. Expert systems are not static
but dynamic devices as there is always scope for improvement and up gradation. The approach for
the development of Expert System is not difficult to understand and tools for developing expert
system are readily available, even some are freely available on internet. The careful development
and logical use of Expert System in agriculture can help bridge the gap between research worker
and extension worker. This view of the future is the result of studies and experience gained through
the Expert System currently being developed and implemented. Therefore, „Expert System in
Agriculture‟ aims to train extension workers and distribute the Expert System to all extension sites
nation-wide. Taking the reducing prices of computers/mobiles into consideration, internet
connectivity at village level for the said purpose can be achieved.
36
5.1 REFERENCES
 Africa
Science
News
Saturday,
(16
November
2013)
http://www.africasciencenews.org/en/index.php?option=com_content&id=596:howtorevolutionise-kenyan-agriculture-the-greenhouse-way-&Itemid=113
 Ammenwerth, E., F. Ehlers, U. Kutscha, A. Kutscha and Eichstadter R., 2002: Supporting
patient care by using innovative information technology. Disease Manage. Health
Outcomes, 10: 479-487. DOI:10.2165/00115677-200210080-00004
 Anderson J., Feder G.C 2005: Agricultural Extension, Handbook of Agricultural
Economics; Agricultural Development: farmers, Farm Production and Farm Markets,
Volume 3
 Arden, B. W. (ed) 2006: "What can be automated?" The Computer Science and
Engineering Research Study (COSERS). Cambridge, MA: MTT Press.
 Arokoyo Tunji 2002: ICTs In the Transformation of Agricultural Extension, The Case of
Nigeria. International journal of Agricultural Extension Volume 3 Issue 5
 Basili, V. R.; Selby, R. W.; and Hutchens, D. H. 2004: Experimentation in software
engineering. IEEE Transactions on Software Engineering SE-12, 7, 733-743.
 Benbasat, I. 2008: An analysis of research methodologies. In the Information Systems
Research Challenge, Harvard Business School Press.
 Blake, S. P. 2009, Managing for Responsive Research and Development. San Francisco:
W. H. Freeman and Company.
 Blalock, A. B., and Blalock, H. M., Jr.2003 Introduction to Social Research, 11th edition.
Englewood Cliffs, NJ: Prentice-Hall.
 Boehm, B. W. 2005, Improving software productivity. IEEE Computer, 20, 9.
 Boehm, B. W. 2007, Software Engineering Economics. Englewood Cliffs, NJ: PrenticeHall.
 Booch, 2004. Object-oriented development. Interactions on software Engineering, SE-12,
 Chen, M. 2008, The Integration of Organization and Information Systems Modeling: A
Metasystem Approach to the Generation of Group Decision Support Systems and
Computer- Aided Software Engineering. Unpublished doctoral dissertation. University of
Arizona, Tuc- son, AZ,.
37
 Chen, M.; Nunamaker, Jr. J. F. and Weber, E. S, 2009. Computer-aided software
engineering: present status and future directions. Newsletter, ACM SIGMIS Database
Volume 20 issue 1, page 7-13
 Computer Science and Technology Board (CSTB). 2010, Scaling up: a research agenda
for software engineering. Communications of the ACM, 33, 3, 281-293.
 Creswell, J.W., 2011. Educational Research: Planning, Conducting and Evaluating
Quantitative and Qualitative Research. 4th Edn., Pearson Education, Boston, ISBN-10:
0132613948, pp: 650.
 Curtis, B. 2004. Tutorial: Human Factors in Software Development. Los Alamitos, CA:
IEEE Computer Society Press,.
 Curtis, B. 2007: Introduction to empirical research on the design process in MCC's software
technology program: MCC Technical Report Number STP-260- 87, Austin, TX,.
 Curtis, B.; Krasner, H.; and Iscoe, N. November 2008: A field study of the software design
process for large systems. Communications of the ACM, 31, 11 , 1268-1287.
 Davis, F. D. 2007, "Perceived usefulness, perceived ease of use, and user acceptance of
information technology", MIS Quarterly, 13(3): 319–340
38