Title: A novel binary tree and pattern matching mechanism for building effective question answering system for agriculture domain

Authors: Prashant Y. Niranjan; Vijay S. Rajpurohit

Addresses: Department of Computer Science and Engineering, KLS Gogte Institute of Technology, Belagavi, India ' Department of Computer Science and Engineering, KLS Gogte Institute of Technology, Belagavi, India

Abstract: Agricultural farming has been a main source of income for approximately 58% of India's residents. However, because of the dreadful circumstances in the agriculture industry, around 10,000 farmers yearly die. These situations emerge whenever crop products may not be as planned, resulting in significant indebtedness on the farm owner's shoulders. As a result, we propose a question answering system that aids farmers by delivering answers to major agricultural inquiries. Some of the inquiries include agricultural planting, plant disease, and seasonal difficulties. As a result, farmers will profit from making better crop selections, boosting their output. The question answering system is created with a pattern-matching technology that uses an optimised binary-tree method to categorise various inquiries and deliver a solution to the farmers. The proposed question and answering system for agriculture domain (QAS-AD) model increases the efficiency of overall computation overhead by 28.33%, and reduces the time computation by 32.33% when compared with the existing system. Moreover, the proposed QAS-AD reduces the computation of the tree structure and byte overhead. Experiment outcome shows the proposed question answering system for agricultural domain achieves much higher relevance score with less time to retrieve answer in comparison with conventional QAS.

Keywords: agriculture; pattern matching; processing overhead; ranking; semantic; storage overhead.

DOI: 10.1504/IJRIC.2022.129675

International Journal of Research, Innovation and Commercialisation, 2022 Vol.4 No.2, pp.92 - 109

Received: 21 Jul 2022
Received in revised form: 28 Sep 2022
Accepted: 28 Sep 2022
Published online: 20 Mar 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article