A Novel Hybrid Big Data Handling Optimized Architecture for Document Clustering

Document clustering is an important aspect of big data. Data growth has been directed to time for many days, and storing the data in a structured manner is an issue. Clustering is a strategy and procedure for automatically grouping together relevant documents. The clustering of documents is a solution for organizing important database items. Our paper's major purpose is to bring together pertinent records. This work proposes a hybrid big data handling architecture as a clustering technique that effectively clusters documents. This project uses the K-Means and Gaussian Firefly Algorithms to cluster documents and search for beginning values, as well as gauze to optimize and increase the accuracy of square and time errors. Finally, to update the cluster's initial values or centroid, we add the squared errors. The cluster's efficiency improves as a result of this. The experimental data is sufficient, and the simulation results are extremely trustworthy. With a low squared error sum, the high-quality solutions look quite promising.