Web Based Decision Support System for Smartphone Selection Using the K-Nearest Neighbor (KNN) Method

Abstract

The development of smartphones has been remarkably rapid, from their early generations to the present day. This evolution has led to an increasing number of vendors offering a wide array of products with highly modern features. The diversity of brands, specifications, and price points often fails to align with the specific needs of users; furthermore, many users lack an understanding of smartphone specifications, frequently choosing devices based solely on their appearance rather than their actual functionality. Given these challenges, a decision-support system is needed to assist users in making informed choices. The K-Nearest Neighbor (KNN) algorithm can serve as a decision-support tool; it functions by calculating the proximity between data variables, allowing users to select a device that matches their desired functionality. To address this issue, the researchers developed a "Web-Based Decision Support System for Smartphone Selection Using the K-Nearest Neighbor (KNN) Method." Black-box testing yielded satisfactory results regarding system and algorithm functionality, though a limitation was identified in the processor selection component the established parameters were unable to encompass the full range of available processor data. In a series of 14 test runs, the system produced 8 correct classifications, resulting in an overall accuracy of 57%. Ultimately, a smartphone decision-support system utilizing the K-Nearest Neighbor algorithm can be effectively implemented.