Agriculture Recommendation System

Project Title

Agriculture Recommendation System

Project Area of Specialization Software EngineeringProject Summary

Agriculture plays a very important role in the economy of a country. Agriculture is managed by farmers where farmers grow crops. As the farmer is the most important stakeholder in agriculture production. So, he must know about the production and consumption trends of Agri products. Conventionally, the farmers grow crops based on the previous year's record, keeping in mind which crops were needed the most and were expensive in the last season, which leads to the same problem where almost all farmers think so and again grow the most needed crops which cause to increase the productivity of those reduced crops and decrease the productivity of other crops. In past, many agriculture systems have been developed and available online but they mostly focused on buying and selling, recording and storage of farm data such as monitoring and analyzing farm activities, consumption, and tracking business expenses and farm budgets. To the best of our knowledge, there is no web-based system that provides us with recommendation functionalities. Therefore, we are going to develop a web-based Agriculture Recommendation System based on a Spatio-temporal context where farmers will get recommendations like: what they need to grow, what is the demand for a particular crop in the next season, what are the treatments for a particular crop disease, which fertilizer is good for their crops and which area is good for which kind of crop, based on water availability? In the end, the final product will be the end-to-end web-based agriculture recommendation system which will be based on recommendation rule-based inferencing. The proposed system will make life easier for farmers to easily take decisions regarding agriculture.

Project Objectives

The objective of the proposed system is to develop an Agriculture Recommendation System based on MERN Stack and Semantic Web technology, which will recommend to the farmers: which crop to grow based on production and consumption trends, which fertilizers are best suited for their crops, which pesticides are best for disease if happened to their crops and which area is good for which crop based on water availability.

Project Implementation Method Proposed Project Methodology

For this application, we will prefer to adopt an agile methodology. Agile is a process in which a project is broken down into several stages and the stakeholders are constantly involved with iterations and progression at each stage.

Application Development Architecture   

For the development of this application, we will use the MERN stack because it is fast and efficient. MERN stands for MongoDB, Express JS, React JS, Node JS. MongoDB is a document database. Express JS is a Node JS web framework used for Web Applications. React JS is a JavaScript library used for client-side development.

We have divided the development into three tiers namely, the Presentation tier, Business logic tier, and Data storage tier. 

Presentation Tier

For the proposed application we will use React JS for front-end (client-side) development. In MERN React JS is on the top of the stack. React is a declarative framework by which dynamic client-side applications can be built in HTML. React allows the building of complex interfaces with simple components.

Business Tier

For the server-side or the back-end, we will partially implement the proposed system with Node JS framework Express JS and Semantic Web. Node JS is a runtime environment where developers can build any kind of server-side application in JavaScript. Express JS is the most popular, fast, and flexible Node framework for the web. Moreover, the Semantic Web refers to the "Web of Linked Data" which means allows people to use the web to create data stores, develop vocabulary and write rules to handle the data. The business logic will be developed in OWL and the rules will be defined in Semantic Web Rule Language and querying of that Semantic Web Rule Language will be done using SPARQL. Lastly, we will develop our APIs (application programming interfaces) for the proposed system and we will also utilize Google Geolocation API. An API is an intermediatory software that makes two applications talk to one another [9]. Our APIs will be mainly responsible for communication among the frontend, backend, and Semantic Web. Postman [2] will be used to test the APIs.

Data Storage Tier

In this application MongoDB will be used for managing data in a database. MongoDB is NoSQL document database. NoSQL stands for "not only SQL" which means non-tabular database. SQL stands for structured query language which is a relational database that used tables to store data. MongoDB stores JSON documents in it. Collections are used which are equivalent to tables in the relational database. These collections can have more than one JSON document. A document is equal to a record or row in SQL. Fields are similar to columns in the SQL table. Mongoose will be used to manage the relationship between schema validation, and data and used to work as a translator between objects in MongoDB objects and code objects.

The proposed system, System Level Architecture is shown below.

Benefits of the Project

The proposed system will be helpful for the farmers. As we discussed earlier, the farmers need to know about different trends including what to grow based on consumption and production trends, which fertilizers are good for their crops, and which pesticides are good for their crops in case of the disease to the crops. Moreover, if a farmer is acquainted with that type of information, then he will be able to produce quality crops keeping in mind the supply and demand. This will eventually lead to impact the country's economy in a positive way where there will be no shortage of any kind of crop as the farmer will know what to grow and what not to grow and also will have the knowledge to use the best fertilizers and pesticides for his crops, which will lead to a high quality of the crops.

Technical Details of Final Deliverable

After the successful completion of the project, we should have a complete end-to-end web-based Agriculture Recommendation System where a farmer will be able to sign up to the system and will be able to enter queries like, which crops to grow, which fertilizers are good for his crops, which pesticides are good for his crops according to the disease that occurred to the crops and which crop is good to grow for his area of cultivation based on water availability. On the other hand, firstly the model will recommend to the farmer according to his query. If the model is unable to respond then the advisor will give advice and the farmer will get the best advice related to his submitted problem. As the advisor will give the advice, the model will get trained for that kind of query. Hence this way the model will get mature with time and a time will come when there will be very less interference required from the advisors to respond to the queries.

Final Deliverable of the Project Software SystemCore Industry AgricultureOther IndustriesCore Technology OthersOther TechnologiesSustainable Development Goals Decent Work and Economic Growth, Responsible Consumption and ProductionRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	32000
Hosting	Equipment	1	19000	19000
Domain	Equipment	1	3000	3000
Meeting farmers, pesticides and fertilizers companies stakeholders	Miscellaneous	2	3500	7000
Documentation Printing	Miscellaneous	6	500	3000