Crop Suitability Predictor Chapter 1 Introduction Maharashtra underwent several fluctuations last year with respect to the retail price of onions. The price increased from Rs. 26 per kilo in the first half of the year to a whopping Rs. 50 per kilo in August [1]. Observing the shoot in the price, many of the farmers in the state decided to grow onions on their farm, in the hope of making exorbitant profits. While this resulted in abundant supply in certain regions of Maharashtra, many other regions suffered failed crop output due to unfavorable conditions for growing onions. A subsequent shortage again in the following months had harsh ramifications on the lives of common man, as middleclass households could no longer afford onion- a frequently used commodity in their
kitchen. This example just goes on to show that a farmer’s decision about which crop to grow is generally clouded by his intuition and other irrelevant factors like making instant profits, lack of awareness about market demand, overestimating a soil’s potential to support a particular crop, and so on. A very misguided decision on the part of the farmer could place a significant strain on his family’s financial condition. Perhaps this could be one of the many reasons contributing to the countless suicide cases of farmers that we hear from media on a daily basis. In a country like India, where agriculture and related sector contributes to approximately 20.4 per cent of its Gross Value Added (GVA) [2], such an erroneous judgment would have negative implications on not just the farmer’s family, but the entire economy of a region. For this reason, we have identified a farmer’s dilemma about which crop to grow during a particular season, as a very grave one. Department of Information Technology, DJSCE Page 1 Crop Suitability Predictor The need of the hour is to design a system that could provide predictive insights to the Indian farmers, thereby helping them make an informed decision about which crop to grow. With this in mind, we propose Crop Suitability Predictor- an intelligent system that would consider environmental parameters (temperature, rainfall, farm’s latitude, longitude, altitude
and distance from the sea) and soil characteristics (pH value, soil type and thickness of aquifer and topsoil) before recommending the most suitable crop to the user. This model would take input from another recommendation system, called Rainfall Predictor, which would predict the month-wise rainfall of the next twelve months for the particular user’s district.
Department of Information Technology, DJSCE Page 2 Crop Suitability Predictor Chapter 2 Literature Review In this chapter, we will look into the related existing systems proposed by other authors.
We will also discuss the drawbacks of these systems, and subsequently define the scope of the proposed system. 2.1 Existing Systems More and more researches have begun to identify the importance of machine learning concepts to enhance agriculture. They are increasingly dedicating their time and efforts to further explore this field. We have identified four existing systems that have been presented and implemented by scholars in their technical papers. These systems closely resonate with the model that we intend to make, each one focusing on a separate parameter of crop prediction. Table 1 summarizes the works presented in these papers. Paper Title Department of Information Technology, DJSCE Page 3 Author(s) Publication / Conference Features Algorithm Used Advantage Limitation Develop ment of Yield Predictio Meteorological data was used Haedong IEEE as the inputs to Lee, Xplore/ model and Aekyung 16th generate the Moon Internation secondary
Non- parametric regression (Kernel Model is trained to identify that that high temperatures and excessive P rediction is g iven only on the basis of the weather Crop Suitability Predictor n System al meteorological smoothing) rainfalls are both conditions. Based on Conference data for method deterrent for the Several other Real- on reflecting the production of determining time Advanced characteristics apples. The effect factors, like Agricult Communic of apple. The of temperature on soil conditions ural ation
model then the sugar content and market Meteorol Technolog predicts the of apples is also value that ogical y final yield of taken into account affect a Informat apples on the to prevent farmer’s ion [3] basis of damaged yield. decision monthly making are weather not taken into patterns. consideration. Crop Selection The model can be Crop Method (CSM) RGF extremely Selection retrieves all (Regularized beneficial for Method
possible crops Greedy farmers to help to t hat are to be Forest), solve their Maximiz sown at a given GBDT conundrums in a e Crop time stamp. (Gradient situation when Yield Yield rate of Boosted more than one Rate these crops are Decision option to plant a using evaluated, if Tree), crop at a single Machine yield rate per regularization point in time using Learning day of these and regression limited land Techniq crops are fair problem resource is
ue [ 4] (within available. tolerance) then those crops are selected for crop sequences. Further, time after harvesting Department of Information Technology, DJSCE Page 4 Since the IEEE/ farming 2015 location is not Rakesh Internation taken as an Kumar , al input, the M.P. Conference prediction is Singh, on Smart not accurate, Prabhat Technologi especially for Kumar es and a place like and J.P. Manageme India where, Singh nt for at different
Computing place, the , weather Communic conditions ation, may Controls, drastically Energy and differ at a Materials given point of (ICSTM) time. Furthermore, a sequence of Crop Suitability Predictor time of crops that considered gives highest crop is taken as yield might a given time not always the stamp for optimum further choice for selection of farmers. crop. The model Efficient proposed is Crop intended to
Yield select a crop and based on Pesticide prediction yield Predictio rate influenced n for by multiple Improvi parameters and ng a lso focuses Agricult pesticide ural prediction and Economy online trading using based on Data agriculture Mining commodities. Techniq ues [5] Department of Information Technology, DJSCE Page 5 It takes into account not only Artificial the seasonal and Internation Neural meteorological al Journal Network, K- data for predicting
of Modern Nearest the most optimal Trends in Neighbors, crop for sowing, Engineerin Decision Tree but also the pH of g and Learning, soil and its Science Regularized nitrogen and (IJMTES) Greedy sulphur content. Forest, Since it accounts Gradient for multiple Boosted parameters, Decision Tree accuracy of prediction for real- time use by farmers is inevitably increased. The model is merely proposed, and not implemented. Soil Data Analysis Using
Classific T.R. Lekhaa The model was able to Jay calculate the Gholap, nutrient value Anurag of only Ingole, phosphor as it The model allows for analysis of soil dataset. It N aive Bayes, J 48 (C4.5), Linear conteCnrto, pwhSiuchitacbailnity Predictor It allows for the content, which can Internation content, which can prediction of soil al Journal depends on attributes such as depends on of depends on phosphorous depends on depends on ation depends on cuses on cuses on cuses on Regression, Techniq Regression, Regression, n Regression, n content, which can content, which can
assification Soil Least Median Least Median Simple Least Median Simple Least Median Simple be very beneficial be very beneficial decide which crop be very beneficial decide which crop be very beneficial decide which crop be very beneficial decide which crop least number least number However, least number However, least number However, least number However, least number However, ues and Attribute soil and can Regression soil and can Regression soil and can Regression Square, to grow. Square, to grow. Square, to grow. Square, to grow. for farmers to other soil for farmers to other soil for farmers to other soil for farmers to other soil for farmers to other soil of parameters. Predictio attributes like of parameters. of parameters. of parameters. of parameters. of parameters.
n [ 6] attributes like soil (such as tributes of the attributes like content and nitrogen phosphorous). nitrogen carbon Table 1. Literature Survey of Existing accounted for. Systems 2.2 Drawbacks of Existing Systems One shortcoming that we identified in all these notable published works was that the authors of each paper concentrated on a single parameter (either weather or soil) for predicting the suitability of crop growth. However, in our opinion, both these factors should be taken together into consideration for the best and most accurate prediction. To eliminate the aforementioned drawbacks, we propose Crop Suitability Predictor- which takes into consideration all the appropriate parameters, including temperature, rainfall, location and soil condition, to predict crop suitability. This is because, a particular soil type may be fit for supporting one type of crop, but if the weather conditions of the region are not suitable for that crop type, then the yield will suffer. Similarly, there may be a case where the weather conditions are favorable but soil characteristics are not. 2.3 Project Definition and
Scope To implement an intelligent crop recommendation system using Machine Learning algorithms so that the Indian farmers can make a more informed decision about which crop to grow. • Training Set: District-wise pre-historic data of 5 major and 15 minor crops. • Prediction Parameters: Temperature, Rainfall and soil attributes • Input to the system: Farmer's geographical location and soil attributes • Output: List of crops suitable for growth Department of Information Technology, DJSCE Page 6 Crop Suitability Predictor Chapter 3 Proposed
Solution This chapter contains details about the proposed system architecture, along with the system features and the benefits of the system. 3.1 Proposed Architecture T he figure below illustrates the system architecture of Crop Suitability Predictor: Fig. 1 System Architecture of Crop Suitability Predictor Department of Information Technology, DJSCE Page 7 Crop Suitability Predictor 1) Sub-system 1: Crop Suitability
Predictor This sub-system is fundamentally concerned with performing the primary function of the model, which is, providing crop recommendations to farmers. The steps involved in this sub- system are: I. Acquisition of Training Dataset The accuracy of any machine learning algorithm depends on the number of parameters and the correctness of the training dataset. For the first sub-system, we have made use of 'India Agriculture and Climate Data Set' [7]. This dataset encompasses historical records of soil and meteorological parameters, which were accumulated over the thirty-year period (from 1957-58 to 1986-87). It covers more than two hundred and seventy Indian districts, thereby constituting 13 major states of the country. The aforementioned parameters are provided for five major (bajra, jowar, maize, rice and wheat) and fifteen minor (barley, cotton, groundnut, gram, jute, other pulses, potato, ragi, tur, rapeseed and mustard, sesame, soybean, sugarcane, sunflower, tobacco) crops. The schema of the training dataset is as follows: • S oil Type: DMS01 (not used) , DMS02 (laterite), DMS03 (red and yellow), DMS04 (shallow black), DMS05 (medium black), DMS06 (deep black), DMS07 (mixed red and black), DMS08 (coastal alluvial), DMS09 (deltaic alluvial), DMS10 (calcerous),
DMS11 (gray brown), DMS12 (desert), DMS13 (tarai), DMS14 (black), DMS15 (saline and alkaline), DMS16 (alluvial river), DMS17 (skeletal), DMS18 (saline and deltaic), DMS19 (red), DMS20 (red and gravely). • A quifer thickness: DMAQ3 (value=1 if aquifer is >150 meters thick), DMAQ2 (value=1 if aquifer is 100-150 meters thick), DMAQ1 (value=1 if aquifer is <100 meters thick). • S oil pH: DMPH5 (4.5<pH<5.5), DMPH6 (5.5<pH<6.5), DMPH7 (6.5<pH<7.5), DMPH8 (7.5<pH<8.5). • Thickness of topsoil: DMTS1 (value= 1 if topsoil is 0 - 25 cm. thick), DMTS2 (value= 1 if topsoil is 25-50 cm. thick), DMTS3 (value= 1 if topsoil is 50 - 100 cm. thick), DMTS4 (value = 1 if topsoil is 100 - 300cm. thick), DMTS5 (value = 1 if topsoil is > 300 cm. thick). • Precipitation: M onth-wise rainfall (in mm). • Temperature: M onth-wise temperature (in °C). Department of Information Technology, DJSCE Page 8 Crop Suitability Predictor • L ocation parameters: I ncludes the latitude, longitude, altitude and distance from the sea of the farm. II. Data Preprocessing This is a two-step process. The first step is to remove the missing values which were represented by a dot (‘.’) in the original dataset. The presence of these missing values
deteriorates the value of the data and subsequently hampers the performance of machine learning models. Hence, in order to deal with these missing values, we replace them with large negative values, which the trained model can easily treat as outliers. The second step before the data is ready to be applied to machine learning algorithms is to generate class labels. Since we intend to use supervised learning, class labels are necessary. The original dataset did not come with labels, and hence we had to create them during the data preprocessing phase. The required labels were generated using production (in tons) and area under cultivation (in hectares) for each crop. Those whose production ÷ area value was greater than 0 were given label 1. In all other cases, a class label of 0 was assigned. III. Machine Learning Algorithms Since in the proposed model, more than one class can be assigned to a single instance, Multi- label classification (MLC) would be the ideal choice. Decision Tree, K Nearest Neighbor (K- NN), Random Forest and Neural Network are four machine learning algorithms that have in- built support for MLC. Decision Tree It is a supervised learning algorithm where attributes and class labels are represented using a
tree. Here, root attributes are compared with the record's attribute and subsequently, depending upon the comparison, a new node is reached. This comparison is continued until a leaf node with a predicted class value is reached. Therefore, a modeled decision tree is very efficient for prediction purposes. [8] K-NN It is a non-parametric method used for making predictions. In this, the predicted value is a class membership. The first step of the K-NN algorithm is to identify the k nearest neighbors for each incoming new instance. The instance is classified by a majority vote of these neighbors. In the second step, depending on the label sets of the k neighbors, a label is predicted for the new instance. [9] Department of Information Technology, DJSCE Page 9 Crop Suitability Predictor Random Forest It is an ensemble method of learning that is commonly used for both classification and regression. In order to train the model to perform prediction using this algorithm, the test features must be passed through the rules of each randomly created tree. As a result of this, a different target will be predicted by each random forest for the same test feature. Then, votes are calculated on the basis of each predicted target. The final prediction of the algorithm is the
highest votes predicted target. The fact that random forest algorithm can efficiently handle missing values and that the classifier can never over-fit the model are huge benefits for using this algorithm. [10] Neural Network Neural Network systems progressively improve their performance by learning from examples. They are based on a collection of connected nodes called neurons. Signals are then be transmitted between these neurons using connections. The neurons and connections have a weight associated with them, which is updated and adjusted as learning proceeds. [11] In order to ensure that the model has the highest possible accuracy, we implemented all the four above-mentioned algorithms individually. The performances of the four were then compared, and the one with the highest accuracy was selected for the model. IV. Trained Model and Crop Recommendations After applying the data to different machine learning algorithms, we obtain trained models of the crop recommendation system. The weights of this model can then be saved, and the farmers can easily avail crop recommendations by giving their farm’s soil type, aquifer characteristics, top soil thickness and pH as the input to the system. The rainfall predicted by sub-system 2 is also given as the input to this trained
model. V. Map Visualization A particular crop may be the most suitable for given soil and weather conditions. When all the farmers of one region use the model for the same season, they are bound to get the same recommendations. However, we know that if all the farmers from the same region will grow the same crop, it will result in surplus. To avoid such a condition, we present the Map View feature, where the farmers can view the sow decisions made by his neighboring farmers using a pop-up marker on the map. Accordingly, he can make decisions about his own crops. Department of Information Technology, DJSCE Page 10 Crop Suitability Predictor To implement the Map Visualization feature, we make use of a JavaScript library called Leaflet.js [12]. It is used to produce and display interactive maps on HTML webpages. The advantage of using this library is that it creates maps of any desired tile type, enables us to zoom in and out as well as pan across the map to reach a desired location. We also make use of Flask [13], which is a powerful Python micro-framework that allows building efficient web applications by providing various libraries, tools and technologies. It is a great way of running light-weight web services on hosts.
2) Sub-system 2: Rainfall Predictor Each and every crop has its own rainfall requirement. If this requirement is not met, the crop yield will suffer. On the other hand, if surplus rainfall is available, the yield may again undergo negative consequences. Hence, rainfall is a very important parameter for the growth of any crop. However, farmers cannot be expected to predict the expected rainfall during the months between the sow and harvest season. For this reason, we decided to implement this sub-system, which predicts the rainfall (in mm) for each of the 12 months of the year, depending on the location of the user’s farm. The predicted output of this sub-system can then be fed to sub-system 1 for prediction of crop suitability. The steps involved are: I. Acquisition of Training Dataset For this sub-system, we used the meteorological dataset provided by [14]. This training dataset consists of 117 years (from 1900 to 2017) of month-wise rainfall of all the 29 states in India. II. Data Preprocessing Similar to the data pre-processing step done for sub-system 1, here the missing values are eliminated by replaced with large negative values (-9999). III. Linear Regression Algorithm Linear Regression is a supervised learning approach that is used to predict a quantitative
response (y) from a predictor variable (x) by making use of statistical approach [15]. Given the nature of the training dataset, such a linear relation can be easily predicted between Indian states and the monthly precipitation values. IV. Trained Model and Rainfall Prediction Once the training dataset is fitted to the linear regression algorithm, we get a trained rainfall predictor model. When an Indian state is given as the input, this model gives 12 float values, corresponding to the rainfall (in mm) of the twelve months in that state. Department of Information Technology, DJSCE Page 11 Crop Suitability Predictor 3.2 System Features 3.2.1 USER PORTAL 1. Login/Registration Page: A ll first time users need to create their account on the portal by entering their details such as name, username, contact number, farm size, farm location and email ID. For all the subsequent times that he wants to access the portal, he can log in using that particular username and password. 2. User Profile: O n this page, the users can view and update his/her personal information like password, name, state, and profile picture and so on.
3. Predictor Model: The user can avail the crop predictions by entering his soil attributes and current location. 4. User Dashboard: This page is essentially divided into two sections. The first section gives information about those crops that the user has shown interest in growing for that particular season. The user can show interest by selecting those crops from a list of favorable crops given by the predictor model. The information includes details about the type of crop, its fertilizer requirements, minimum support price, farm harvest price, possible diseases for the crops and potential pests/insects. The second section consists of the farmer’s yield history. It shows a list of all the crops the farmer has grown in the past. Here, information about the crop grown, area of field sown, amount of yield production, profit earned, start date and stop date are displayed in a tabular format. 5. Map View: Using this feature, the user can see what his neighboring farmers are planning to grow for that particular season. If a mainstream of his neighbors are growing the same crop, then he can take a call about not growing that particular crop, despite recommended by the predictor model, in order to avoid surplus in the region. This is because the surplus will only affect him adversely when he tries to sell them in the retail market. 6. Farmers Market: T his feature allows the farmers to market their produce after the
harvest season is over. It also allows them to get a fair and profitable monetary quotation, which is in lines with the current market’s minimum support and retail prices, for their crop. 7. Notifications: Announcements from the admin are displayed in the notifications tab. Admin’s response to the user’s queries or complaints can also be seen here. 8. Submit Query/Complaints: T he users can submit their queries and complaints regarding the portal and/or crop growth to the admin. Department of Information Technology, DJSCE Page 12 Crop Suitability Predictor 9. Online Learning: At the end of the harvest season, we encourage all the users to enter what they grew, the area sown, crop output and so on. This data is used by us to re-train the machine learning model so as to adjust the weights. This feature will help the predictor model reach closer and closer to a cent percent accuracy. 10. Multi-Lingual Feature: T his feature allows the Indian farmers to avail the entire portal in their preferred language (Gujarati, Hindi, Marathi, Punjabi, Tamil and so on). This ensures that the portal is convenient for use by even the non-English speaking farmers. 3.2.2 ADMIN PANEL
The production statistics for the different crops are displayed on the admin’s dashboard in the form of a line graph. Apart from this, the dashboard also displays state-wise production of the different crops using pie charts. 3.3 Benefits 1. The primary benefit of our project is to help farmers gauge the changing environment and market conditions to correctly choose the crop for the season to optimize their profits. 2. Our emphasis is on providing easy access to crucial information regarding different crops in form of datasheets. 3. The system aggregates region wise data of all farmers which can be used for future policy making. 4. The admin and authorities concerned get up-to-date region wise yield statistics which can be used in research on the supply-demand characteristics of that region. 5. We aim to streamline the communication between farmers and authorities and provide a platform to voice complaints or to ask queries through our grievances and announcements portal. 6. The entire web application will act as a basis to creating an online farming community which will aid farmers in making better choices and avoid losses, thus maximizing
the agricultural economic output. Department of Information Technology, DJSCE Page 13 Crop Suitability Predictor Chapter 4 Project Management This chapter looks into project scheduling. It also includes feasibility study and a study of required project resources. In addition to this, chapter 4 also covers project estimation using COCOMO and Function Point (FP) models, and also Risk Management Monitoring and Mitigation (RMMM) plan. 4.1 Project Schedule
Fig. 2 and 3 represent the project schedule and the corresponding Gantt chart respectively. Fig. 2 Project Schedule Department of Information Technology, DJSCE Page 14 Crop Suitability Predictor
Fig. 3 Gantt Chart 4.2 Feasibility Study A feasibility study is used to determine the viability of an idea, such as ensuring a project is legally and technically feasible as well as economically justifiable. It evaluates the project's potential for success. A well-designed study should offer a historical background of the project, such as a description of the product or service, accounting statements, details of operations and management, financial data and legal requirements. The areas of feasibility study for this project are: 4.2.1 Technical Feasibility The datasets of various parameters such as temperature, humidity and economic values are
taken into consideration to cover different array for prediction of the crop. All the system requirements must be stated in the initial phase of the project. Various machine learning algorithms and techniques which provide more precise and high probability will be used. Python will be used as the primary programming language. Data used must be in a consistent format to avoid erroneous results. 4.2.2 Economic Feasibility The hardware requirements of the project are quite minimal which helps in the cost- effectiveness. Also the software requirements are very minimal. Our project will help a great deal in predicting the most suitable crop which would prevent heavy losses to the farmer. Its flexibility and easy maintainability makes it an economically viable model. The project has a great scope in the domain and could be used on a large scale in some future undertaking. Department of Information Technology, DJSCE Page 15 Crop Suitability Predictor 4.2.3 Operational Feasibility Various environmental and manmade factors are taken into consideration while predicting the most suitable crop. In water-limited environments, rainfall variability represents the main factor determining crop production variability and environmental risk. However, other
factors such as soil moisture, soil type, soil fertility, temperature, planting dates, rainfall intensity, and timeliness of rainfall are particularly important when operational seasonal forecasting systems are applied in practical farming system management. Along with this the demand-supply of the particular crop is taken into account so that the farmers do not face any loss and gain maximum benefits. 4.2.4 Legal Feasibility There is no legal feasibility attached with the project. The choice whether to grow the crop or not relies totally on the farmer and he select what suits him best. 4.3 Project Resources The essential project requirements are as follows: 4.3.1 Hardware Requirements: • Minimum 2GB RAM • Processor: Intel Core (2n d G eneration) • System Bus: 64-bit • Nvidia Geforce Graphic Card (minimum 2GB) • Input Devices: Keyboard, Mouse compatible with OS • LAN / Ethernet Connection 4.3.2 Software Requirements
• Python IDE v3.3 and above • XAMPP Server Department of Information Technology, DJSCE Page 16 Crop Suitability Predictor 4.4 Project Estimation 4.4.1 COCOMO Model Basic COCOMO computes software development effort (and cost) as a function of program size. Program size is expressed in estimated thousands of source lines of code. COCOMO applies to three classes of software projects: 1) O rganic projects - \"small\" teams with \"good\" experience working with \"less than rigid\" requirements 2) S emi-detached projects - \"medium\" teams with mixed experience working with a mix of rigid and less than rigid requirements 3) Embedded projects - developed within a set of \"tight\" constraints. It is also combination of
organic and semi-detached projects.(hardware, software, operational) Historical data used for estimation While making estimation about various factors like efforts involved, the number of people required to finish the project and also the duration required for the completion of project following factors are taken into consideration: 1) Determining the characteristics of the project depending on the factors like the size of project, innovations, development environment etc. 2) D etermining the development mode of the project it will help us in identifying various constants and constraints, which will be further used during estimation. 3) Approximated the number of kilo lines of code (KLOC) which is also useful during estimation. Estimation Technique The COCOMO cost estimation model is used by thousands of software project managers, and is based on a study of hundreds of software projects. It is applied to three classes of software projects: Organic, Semi-Detached and Embedded Systems. Unlike other cost estimation models, COCOMO is an open model, so all of the details are published
including: • T he underlying cost estimation equations • Every assumption made in the model (e.g. \"the project will enjoy good management\") • Every definition (e.g. the precise definition of the Product Design phase of a project) Department of Information Technology, DJSCE Page 17 Crop Suitability Predictor • The costs included in an estimate are explicitly stated (e.g. project managers are included, secretaries aren't). The COCOMO model can be applied to three different categories of software projects, namely, organic, semi-detached and embedded. Table 2 gives the values of ab, bb and the exponents cb and db for each of these categories. project Software
Department of Information Technology, DJSCE Page 18 Organic 2.4 1.05 2.5 0.38 Semi-detached 3.0 1.12 2.5 0.35
Embedded 3.6 1.20 2.5 0.32 Table 2 COCOMO Model Advantages of COCOMO estimating model are: • C OCOMO is factual and easy to interpret. One can clearly understand how it works. • Accounts for various factors that affect cost of the project. • Works on historical data and hence is more predictable and accurate. Thus we choose COCOMO Model. Estimate for Effort, Cost and Duration: Project Type: Organic Class KLOC (Approx): 9 Therefore, ab = 2.4 and bb =1.05, cb = 2.5 and db =0.38 Effort Applied (E) = 2.4*9^1.05 = 24.68 person-months Development Time (D) = 2.5*24.69^0.38 = 8.45 months People required (P) = E/D = 24.69/8.45 = 3 people (approx.) Crop Suitability Predictor
4.4.2 Function Point (FP) Analysis The data collected for the domain characteristics are illustrated in Table 3. Domain Characteristics Count Simple Weighting A verage F actor C omplex Count Number of User Input 6 3 4 6 24 Number of User Output 1 4 5 7 5 Number of User Inquiries 5 3 4 6 20 Number of Files 1 7 10 15 10 Number of External Interfaces 1 5 7 10 7 Count Total 66 Table 3. FP Characteristics Table 4 lists the parameter significance ranking required for Function Point (FP) analysis: Questions Significance Fi Does the system need a reliable Backup and Recovery 5 Are data communications required 4 Are there distributed processing functions 0 Is Performance of System Critical 5 Will the system run in an existing, heavily utilized operational environment 3 Does the system require online data entry 0 Are the master files updated online 0 Are the inputs, outputs , files or the inquires complex 4 Is the code which is being deigned being reusable 4 Are conversion and installation included in the design 2 Is the system designed for multiple installation in different organizations 1 Is the application designed to facilitate change and ease of use by the user 5 Total 33 Table 4. FP Parameter Significance Ranking Function Points (FP) = Count Total *(0.65+ (0.01*SUM (Fi ))) = 66*(0.65+ (0.01*33)) Department of Information Technology, DJSCE Page 19 Crop Suitability Predictor
= 6*(0.65+0.33)) = 66*0.98 = 64.68 Considering 4 hours per day the work is done. Project days = (FP * 4) = 258.72 days ~ 8.6 months 4.5 Risk Mitigation, Monitoring and Management The risk impact table is illustrated in Table 5, which classifies the potential risk into different categories and also identifies its probability of occurrence and impact. Risk Category Probability Impact Application crashes TE High 1 1 Prediction Model doesn’t work
Department of Information Technology, DJSCE Page 20 Authentication Issues DE Medium 2 Table 5. Risk Impact Table Impact values: 1-Catastrophic 2- Critical 3-Marginal 4- Negligible
The risks that can be faced by the project team are specified in the first columns which are followed by the category of the risks. • Technical Risks (TE): T hese risks deal with the engineering, testing phases where the execution of the processes maybe affected due to uncertainty. The resources, time are affected. • Business Risks (BU): T hese risks deal with the economic aspect as well as the usage of the product. These risks are generated due to uncertainty. • D evelopment Risks (DE): T hese risks deal with the estimation, productivity as well as compromising on the designs • Project Risks (PE): T hese risks interfere with the completion of the project The likelihood of the risk is specified using qualitative measures, i.e., high, low, and medium. The impact of the risks is specified using quantitative measures where the measures are Crop Suitability Predictor further mapped to the qualitative measures. Tables 6, 7 and 8 are risk information sheets for Crop Suitability Predictor. Risk Information Sheet
Risk ID: RK01 Date: 19/10/2017 Prob.: 40% Impact: High Description: The project describes an efficient crop suitability prediction model for the Farmers. Authenticating and handling large no. of users involves the risk of Application crash. Mitigation: Ensure that application crash is prevented and overall authentication is simplified Monitoring: Continuously monitor the application usage statistics Management: The developers much take into account the issues and solve it. Originator: Z eel Doshi A ssigned: P rof. Neepa Shah Table 6. Risk Information Sheet 1 Risk Information Sheet Risk ID: RK02 Date: 19/10/2017 Prob.: 30% Impact: High Description: The project describes an efficient solution that uses a prediction model using the pre-processed data and user input to forecast which crop would be most suitable for growth depending on environmental conditions like temperature and rainfall. Any form of mishap can cause wrong suggestions to the farmers. Mitigation: Ensure that the software provides proper prediction model according to the data fed by the farmers
Monitoring: M onitor the meteorological website for any predictions of anomaly weather Management: The developers must look into the report and solve the issue. Originator: R ashi Agrawal Assigned: Subhash Nadkarni Table 7. Risk Information Sheet 2 Department of Information Technology, DJSCE Page 21 Crop Suitability Predictor Risk Information Sheet Risk ID: RK03 Date: 19/10/2017 Prob.: 30% Impact: Low Description: M ultiple dependency support – installation of multiple dependencies in the same machine will result into port issues. Mitigation: Before installation check the port number availability Monitoring: M onitor for any such errors during execution Management: If port number issue occurs then we provide the service to change the port number dynamically. Originator: S ubhash Nadkarni Assigned: Z eel Doshi
Table 8. Risk Information Sheet 3 Department of Information Technology, DJSCE Page 22 Crop Suitability Predictor
Chapter 5 Project Design Chapter 5 covers the data design diagram, data flow diagram (DFD), use case diagram, and sequence diagram and interface design of the proposed system. 5.1 Data Design Diagram Fig 4 represents the entity-relationship diagram of the data to be used during the implementation of Crop Suitability Predictor.
Fig. 4 Entity-Relationship Diagram Department of Information Technology, DJSCE Page 23 Crop Suitability Predictor 5.2 Data Flow Diagram A data flow diagram (DFD) is a graphical representation of the flow of data through an information system. A DFD is often used as a preliminary step to create an overview of the system, which can later be elaborated. DFDs can also be used for the visualization of data processing (structured design). Fig 5, 6 and 7 represent Levels 0, 1 and 2 of the Data Flow Diagram respectively.
DFD Level 0 Fig. 5 DFD Level 0 DFD Level 1 Fig. 6 DFD Level 1 Department of Information Technology, DJSCE Page 24
Crop Suitability Predictor DFD Level 2 Fig. 7 DFD Level 2
Department of Information Technology, DJSCE Page 25 Crop Suitability Predictor 5.3 Use Case Diagram The use-case diagram of the proposed work is illustrated in Fig. 8.
Search
