Date of Award

5-2023

Document Type

Project

Degree Name

Master of Science in Information Systems and Technology

Department

Information and Decision Sciences

First Reader/Committee Chair

Shayo, Conrad

Abstract

Agroforestry farming is one of the challenging sectors to grow the crops or farming or variety of trees from the ancient days due to erosion and desertification. This Culminating Experience Project explored how recommendation system can be developed and used in agroforestry. The research questions are Q1. What methods can be used to improve the accuracy and reliability of soil-based agroforestry tree species recommendation systems? Q2. How can agroforestry tree species recommendation systems be tailored to the needs of different stakeholders, such as smallholder farmers or agribusinesses? Q3. What will be the top three, tree recommendations using natural language processing based on varying soil content? Data was collected from two datasets the Agroforestry Database and the European Commission's extension of the periodic Land Use/Land Cover Area Frame Survey. The findings are: 1) Various Natural language processing techniques such as cosine similarity, count vectorization, and TF-IDF can significantly enhance the system's ability to analyze and process large amounts of Data collection, validation, and monitoring to improve the accuracy and reliability of soil-based agroforestry tree species recommendation systems. 2) Cosine similarity achieve to recommend tree species based on soil test report data collected by the European Commission's extension of the periodic Land Use/Land Cover Area Frame Survey and tailored based on various soil properties helps the smallholders, stake holders, farmers to best decisions to increase their growth. 3) Natural language processing techniques such as cosine similarity, count vectorization, and TF-IDF can be employed to analyze soil data and identify the tree species that are most appropriate for different soil types.The conclusions are: 1) The system's ability to analyze and process large volumes of data accurately, and the recommendations provided by the system can become more effective and reliable. 2) The system's recommendations can become more relevant, practical, and acceptable, leading to higher adoption rates and better outcomes.3) Develop The proposed agroforestry tree species recommendation system provides top three trees recommendations using cosine similarity, TFIDF and Count vectorization techniques. Furthermore, areas for future research that emerged from this study include the need to improve the sustainability and productivity of agroforestry practices, enhance ecosystem services, and promote economic, social benefits and identify additional strategies for improving the accuracy and reliability by getting additional feedback about the trees recommendation from the stakeholders and farmers directly in design and development.

Download

  Contact Author

Share

COinS