Industry needs for fast visualization techniques to view the large datasets. To add more insights out of data dimensionality, volume, variety, and complexity cannot be covered by traditional types of visualization. Hence, such techniques represent vital elements of contemporary visualization methods, synergizing computational efficiency, interpretability, and decision-making (DMD) in big data visualization applications. Some of the key dimensionality reduction methods are principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), and uniform manifold approximation and projection (UMAP). Though PCA is a computationally cheap linear method, it will struggle in case the data is not linear. However, t-SNE and UMAP are nonlinear approaches with higher quality of visualization, most useful in complex structures. t-SNE works great to preserve local relationships, but it is computationally expensive, which makes it a poor fit for very large datasets. Compared to t-SNE, UMAP offers the best of both worlds, providing a low-computation, scalable, and high-quality method at the same time, making UMAP frequently a prime candidate in big data analytics. Clustering and sampling techniques are used for data summarization which helps in improving data management as well as data interpretation. Some of the bests are supervised algorithms like regression, and unsupervised algorithms like k-means and DBSCAN to cluster groups of data into useful classes. This allows for faster processing time and ensures the integrity of dataset. This study compares methods through visualization clarity, computation efficiency, and user feedback. The findings of this study enhance big data visualization approaches and provide empirical justification within the professional field of organizational public relations ultimately leads to enhanced data-driven communication and decision-making.