6 V May 2018

http://doi.org/10.22214/ijraset.2018.5183
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at www.ijraset.com

Crop Recommendation System for Precision

Agriculture
Lakshmi. N1, Priya. M2, Mrs. Sahana Shetty3, Mr. Manjunath C.R 4
1, 2
Final year Btech student, Department of Computer Science and Engineering, Jain University, Bangalore, Karnataka, India
3, 4
Professor, Department of Computer Science and Engineering, Jain University, Bangalore, Karnataka, India

Abstract: To develop a recommendation system for modern agriculture Big Data is a vastly evolved technology in computer
engineering that can be used in various domains like agriculture, medicines and retails etc. By using this big data more number
of previous data can be found and from that a new system can be developed which can act as a predictive system. Agriculture is
the important factor of economy in India and till now many of the methods followed by agriculture is outdated ,by using newest
rends IOT, data mining and big data which plays an important in developing a new technology to improvise the agriculture field
and make the usage simpler and economic
Keywords: Agriculture, Big Data, IOT, Data Mining

I. INTRODUCTION
In current times the agriculture methods has radically improved due to globalization. A range of causes of agriculture strength is still
impacted in India. Several advanced equipment consume to improve the regain the strength. The method is Precision Agriculture.
Precision Agriculture is growing in India. Precision agriculture is the data of “site-specific” farming. It provides various inputs and
suitable output which outcomes in the agriculture. Though precision agriculture has deliver better improvement there exists some
issue. Many system exists which helps in farm land systems. System suggest crops, manures and even agricultural
methods. Recommendation of crops is one main province in precision agriculture. Recommendation of crops is needful relative on
a variety of parameter. Precision agriculture main purposes in identify these limitations in a site-specific way in demand to decide
issues concerning crop selection. The “site-specific” method of such system need to be supervise to get better results. Only some of
the results in precision agriculture provides particular results. However in farming the situation is vital that the references are made
precise and accurate since if any mistakes occurs it might lead to serious measureable and assets damage. Large number of enquiry
work is existence done, in order to acquire a precise and also well-organized crop prediction model.

II. CASE REPORT

Paper [1] which makes a relative study of categorization algorithms and their performance which helps to know the yield and
predict it in precision agriculture. These algorithms are implemented in the various data set which are collected from various fields
and form long back years which helps in yield prediction on ragi crop.
Paper [2] it concludes the requirement for crop yield prediction and its major usage and the role in a nation’s planned guiding
principle which are made in agriculture development field .A Framework is developed called Extensible Crop Yield Prediction
Framework (XCYPF) is developed. It provides facilities to bendable addition of an assortment of techniques and methods towards
crop yield prediction tool called XCYPF is developed in order to help people to predict more varieties of crops and their yield
prediction by using independent and dependent variables.
Paper [3] it states the procedure of agricultural data through data mining and visual data mining techniques are predicted. It helps to
reduce high dimensional agricultural data into smaller size helps to acquire the useful knowledge of the data which are related to the
yield, by taking input attributes like fertilizers.
Paper [4] states that the recommendation system to be developed for the different types of soil using big data analytics and other
various factors.
Paper [5] acknowledges the significance of crop selection and some of the factors which decides the selection of crop like market
price, production rate and policies of the government that are discussed. In this paper it also proposes Crop Selection Method (CSM)
that is used to solve the yield of crop and selecting the crop and requires to improve the net yield rate of the crop. It predicts the
series of crops which needs to be selected according to the seasons including the factors like soil type, weather crop type, water type
which are all used in estimation of crop yield for crops like cereal in some major districts of India.

© IJRASET: All Rights are Reserved 1132

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at www.ijraset.com

Paper [7] it focuses to solve the complex problem of choosing the classifiers which are used for assembling learning knowledge. A
technique to select an accurate classifiers set which are present in pool of classifiers it has been discovered. This development helps
in achieving the aims like performance and high rate of accuracy.
Paper [8] aims in solving the problem related in food problems in the country of Egypt. A framework has been proposed that
predicts the production, and it also imports of that meticulous year. Artificial Neural Network and Multi Layer Prescription is
proposed for perception WEKA to build the prediction.
Paper [9] yield is analyzed and category are predicted. Crop yield is identified as Classification rule which are taken from prediction
which depicts Naïve Byes and k-Nearest Neighbor algorithms that are used in prediction for yield prediction of soil.

A. Problem Definition
To develop a recommendation system using big data analytics for crop growth by taking all the factors like weather, geographical
features ,water utility and land and to build a simpler mechanism to predict the types of soil and the crops that are suitable to be
grown in that soil.

III. METHODOLOGY
By the details in the laboratory and the results acquired are important sources of data for various types of test and for researchers.
The tests which are conducted are water, soil analysis, compost, plant manure, green roof media, biosolids, and green house media
testing etc .Information on agriculture is provided by website that provides statics and real time data that are related to agriculture
economic entities, some of the expert discussions are provided by particular articles on specific problems. Information on
agricultural website aims to provide information to everyone at any place or any time which empowers farmer to accept the recent
technology. To download WebPages from information agricultural website flume are used which helps to download the pages and
save them on Hadoop distributed file system. This framework is used to fulfill the needs and help agriculture officers and
agricultural researchers to analysis and come up with recommendation solution for the growth of crop by taking the evidences from
historic data.

Table 1 types of soil and crop growth

TYPE OF SOIL CROPS GROWN

1 Alluvial Soil Tobacco, Cotton ,Rice
2 Black Soil Rice, Sugarcane, Wheat
3 Red and Yellow Soil Ragi, Potato, Pulses
4 Late rite Soil Rubber , Coffee
5 Arid Soil Corn, Millers, Barley
6 Forest and Mountain
Soil
Tea, Coffee, Tropical
Fruits

© IJRASET: All Rights are Reserved 1133

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at www.ijraset.com

Figure 1Estimation of agriculture in 2016.

For conjecture taking the effects of high temperature on the paddy and their measures to be taken in how to control the problem in
the particular district. Feature selection is defined has the process o f selecting features and relevant information in the form of
subset. The selection process of feature selection takes place at the data preprocessing on model learning. Feature extraction is the
process that needs to be done to normalization of data for removal of technical variation through replacing the missing values since
these missing values they distracts the process of decision making.

Figure1.1.Methodology of system.

A. Data Preparation
In this paper, weather data are used to collect in Agriculture Meteorological Department (AMD) from 2016 up to now there are 28
weather stations, contain various agro atmospheric index, control and pass strict feature examination for example, the statistical
values are especially the extreme parts are verified and proved. AMD are also proved to be time consistent data. The related indexes
has an accuracy more than 99.99% and other absolute humidity have an accuracy of 99.4%. From the figure 1.3 the distribution
weather station can be found.

Figure 1.3 District wise rainfall pattern data

B. Map Reduce Weather Data Processing Structure

Weather data have been depending on static technique established on several traditional resources. The examples for weather
processing structure are usually form the model from mean values were absolute if 2years have comparable weather patterns.

© IJRASET: All Rights are Reserved 1134

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at www.ijraset.com

Though, the results of the actual weather patterns cannot replicate the growth of the little changes in the weather conditions. Hence,
the big data brings us to figure the models for all micro difference which is used to figure the whole datasets that were not using
before. The new weather data processing structure divide the computing process into 2 phase, which can be define as Map and
Reduce In the Map phase the function succeeds all the input values/key and produces the median key/value. in the Reduce phase the
function combines all the values which contains same value of median key and finally produces the result. Firstly weather data
subdivided into multiple fragments. Moreover, the Map function is executed and the data is categorized according to the certain
instructions and then written in local hard drive. After the Map phase is performed, the Reduce phase is executed where the values
will be consolidation and shuffled the intermediate data having the same year and the output is written to distributed file systems.
The final result can be obtained by merging all the Reduce phase output at last.

C. Dataset Collection
In India the dataset which are collected from the soil specific attributes from particular part of Karnataka. In accumulation, the
general crop data is used in the similar online source. Some of the crops that are used as a sample in the model are groundnut,
paddy, pulses, coriander, millet, sugarcane, cotton, banana and different types of vegetables. The training data set the number of
instances of each crop is depicted. Some of the main attributes of the soil which are taken from the sample are texture, color,
drainage, depth, ph, water holding, erosion and permeability.

Fig 1.4 Pie chart of dataset

D. Similarity of the weather – Nearest Neighbors

Similarity of weather is defined by distances between the weather, which is the multivariable analysis and usage varied weather
indexes. Each climatic index is treated as one dimensional space. So for m number of indexes, m dimensional spaces are found. The
distance between two years can be calculated through the valuation of similarity between the two years and the system that can be
seen as the spot in m dimensional space. The specify two years can be similar if the distances is small. Hence the nearest neighbor
of the particular year can be found through the similarity between any two years.
A P A1 A2 ... An P1 P2 ... Pm 11 12 1 21 22 2 1 2 ... ... ...... ... n n m m mn XX X XX X XX X
One specific year having different properties is described as A and P stands for each year. The measurement between the distance of
any two points like Ps and Pt basically uses Euclidean distance that is shown as:

IV. CONCLUSION AND FUTURE SCOPE

Big data has transformed agriculture into an intellectual world which provides more detail information. Decision making is
achieved through a variety of technique in big data analytics. Agriculture plays a primary role in India. Improvement in
agriculture helps the prosperity of nation and the farmers. Through this technique we can help farmers to sow the right seed based
on the requirements which helps in the increase productivity and profit will be acquired from the technique. Overall productivity
of the nation can be increased by planting the right crop and increasing the yield. This can be further improved by improving the
data set with more number of attributes in yield prediction. The current framework is essential for officers and researchers and
recommending the desire solution based on historical data and evidence. Since massive data set already exist we can take its
advantage by using it in the proper way. Our system contains two major parts. Map reduce weather is the first part used in data

© IJRASET: All Rights are Reserved 1135

 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at www.ijraset.com

processing structure which calculate huge data set on a group of computer. Nearest neighbour technique is used between similar
years which help to calculate the weather distances. Taking an example of the forecast from the year 2016 the yield can be
calculated by using the data from March 2015 to February 2016, from that we can acquire the result from five to six months in
advance. Food security can be improved since the accuracy is beyond the average and methods can be use for decision making.

To conclude the topic, from fast accumulation of the data the data processing from the weather can be qualified enough to support
these data . Mainly if any complex comparative of the data is required the similarity of the weather can be indulged in the data
processing sector to control the computation time.

REFERENCES
[1] Anshal Savla, Parul Dhawan, Himtanaya Bhadada, Nivedita Israni, Alisha Mandholia , Sanya Bhardwaj (2015), ‘Survey of classification algorithms for
formulating yield prediction accuracy in precision agriculture', Innovations in Information, Embedded and Communication systems (ICIIECS).
[2] Aakunuri Manjula, Dr.G.Narsimha (2015), ‘XCYPF: A Flexible and Extensible Framework for Agricultural Crop Yield Prediction’ , Conference on
Intelligent Systems and Control (ISCO
[3] Yash Sanghvi, Harsh Gupta, Harmish Doshi, Divya Koli, Amogh Ansh Divya Koli, Umang Gupta (2015), ‘Comparison of Self Organizing Maps and
Sammon’s Mapping on agricultural datasets for precision agriculture’, International Conference on Innovations in Information, Embedded and
Communication systems (ICIIECS)
[4] Ramanujam, R.Harine Rajashreeń, C.Kavyań, T.Kiruthikań, J.Nishań, S.Pudumalar(2016), ‘Crop Recommendation System for Precision Agriculture’
(ICoAC) .
[5] Paul Monali K. Vishwakarma Santosh Ashok Verma(2015) "Analysis of Soil Behaviour and Prediction of Crop Yield using Data Mining Approach"
International Conference on Computational Intelligence and Communication Network.
[6] Saso Karakatic, Marjan Hericko and Vili Podgorelec (2015), ‘Weighting and sampling data for individual classifiers and bagging with genetic algorithms’
International Joint Conference and Computational Intelligence(IJCCI)
[7] Yue Zhang, Jianxia Chen, Oin Fang, Zhiwei Ye (2016), ‘Fault analysis and Prediction of transmission line based on Fuzzy KNearest Neighbor algorithm’
International Conference on Natural Computation , Fuzzy systems and Knowledge discovery(ICNCFSKD)
[8] Rakesh Kumar, M.P. Singh, Prabhat Kumar and J.P. Singh (2015), ’Crop Selection Method to Maximize Crop Yield Rate using Machine Learning
Technique’, International Conference on Smart Technologies and Management for Computing, Communication, Controls, Energy and Materials (ICSTM)
[9] Xu Y, Zhou W, Cui B, Lu L. Research on performance optimization and visualization tool of hadoop. Proceedings of 10th International Conference on
Computer Science and Education; Cambridge University, USA. 2015 Jul
[10] Lu Q, Li Z, Kihl M, Zhu L, Zhang W. A conceptual framework for big data analytics applications in the cloud. IEEE Access. 2015 Oct; 3(1):944–52.

© IJRASET: All Rights are Reserved 1136