Archive Information Management System Based on K-Means Clustering with Elbow Method

From the past few years, an effective archive information management system is important for preserving and organizing records which ensures easy retrieval and maintaining data integrity. Traditional approaches for archive information management system had faced several challenges which include manual file handling and lack of security access control. Therefore, this research proposes K-Means clustering with elbow method for archive information management system. Initially, data is collected from Internal Document Management System (IDMS) dataset consists of document records with attributes such as document type, department, and archival date. Then, the input data is preprocessed using K-nearest neighbors (KNN) imputation to handle the missing values by replacing with similar attributes. Also, one-hot encoding is used to transforming categorical data into numeral and finally it normalized with help of min max normalization. Finally, proposed K-Means clustering with elbow method is applied to group similar documents into structured clusters, optimizing accessibility and retrieval while reducing manual effort in document management. From the results, the proposed K-Means clustering with elbow method provided outstanding results in terms of precision (98.00<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>), recall (93.00<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>), and Fl-score (94.00<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>) when compared to existing Support Vector Machine (SVM).