Predicting amount of fertilizer substances for gardening by using Rule Based Technique / Farah Dalila Rodzuan

This project is developed to predict the amount of fertilizer substances for gardening by implementing Rule Based Technique. Rimbun Ventures Nursery is chosen as stakeholder for this project, which is located at Taman Pulai Flora, Johor Bahru, Johor where it’s near to Pulai Spring Resort and Technol...

Full description

Saved in:
Bibliographic Details
Main Author: Rodzuan, Farah Dalila
Format: Thesis
Language:English
Published: 2017
Subjects:
Online Access:https://ir.uitm.edu.my/id/eprint/18759/1/TD_FARAH%20DALILA%20RODZUAN%20CS%2017_5.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This project is developed to predict the amount of fertilizer substances for gardening by implementing Rule Based Technique. Rimbun Ventures Nursery is chosen as stakeholder for this project, which is located at Taman Pulai Flora, Johor Bahru, Johor where it’s near to Pulai Spring Resort and Technology Park University Technology Malaysia (UTM). There are few users of this system which are the owner of the nursery itself, the gardener and the other users in related area of knowledge. Currently, there is very few system that predict the amount of fertilizer substances to be used, this lead to several problems such as flower plant stunted growth because of lacking nutrient, misuse type of fertilizer and fertilizer amount used over or understated. In order to minimize this problem, the objectives had been identified for conducting and monitoring this research. This prediction system helps the nursery by providing the suggestion on the fertilizer amount needs for flower plant, according to the uncertainty attributes such as the age of flower plant, temperature (Celcius), water requirement and soil pH value. The Rule Based technique was chosen to analyze and produce the rules from the data that have been gathered from the expert knowledge. Beforehand, the data have to undergo data pre-processing phase in order to clean, transform, reduce and smooth the raw data. Based on the data that have been pre-processed, the uncertainty attributes were divided into several categories to ease the rules produced such age of flower plant that has been categorized into threes according to its height, which are small, medium and large. From the rules, the system can suggest a few suggested percentages of the fertilizer amount needed for flower plants. The rules embedded into the prototype system to demonstrate on how the rules functioning. Therefore, the type of fertilizer to be used also being suggested according to soil pH values. In order to control the entire process involved, a waterfall methodology has been chosen. Even though this project have been completed, there is a lot of enhancement that can be done to improve such as within the scope of this research on covering the other type of plants such as herb, fruit and vegetables. Furthermore, this prototype system platform also can be changed into mobile applications for more easy and quick uses as well as can address the need of the testing phase. In conclusion, this research proved to be beneficial not only for the stakeholder but also for the other users.