This thesis contributes a novel visual analysis approach for multi-dimensional data by enhancing the utility of scatterplot matrices (SPLOMs), focusing on automatic cluster detection and trend/pattern/functional dependency identification utilizing suitable color linking strategies among views. Traditional SPLOMs effectively perform analysis tasks, including identifying dimensionality within or between-cluster structures and trend/spatial correlation across 2D subspaces. Despite being effective for the tasks, traditional SPLOMs often require extensive manual exploration throughbrushing and linking among the views to uncover patterns across multiple dimensions. To address these limitations, this bachelor thesis studies several automatic assigning color strategies for view linking. In particular, colors from a 2D color map will be assigned to points of the scatterplot of interest, taking advantage of two specific clustering techniques: Mean Shift and Self-Organizing Maps (SOM). We utilize Mean Shift to assign colors to scatterplot points to observe spatial information between cluster structures. Meanwhile, SOM is used and adapted to maximize the number of colors to be used, at the same time preserving the spatial information to support observing trend/correlation patterns within-cluster structures among 2D subspaces. We proposed a visual analysis approach with multiple link views from the two strategies. A main and comparison view are created which allows users to directly compare clusters and patterns between different projections without repeatedly scanning through the entire SPLOM, thus significantly reducing cognitive load and manual effort. The Master View allows users to select a specific scatterplot for detailed analysis, automatically assigning colors based on cluster detection results. The Comparison View enables users to juxtapose another scatterplot against the Master View, facilitating quick comparative analysis between different dimensional combinations. Interactions in our visual analysis approach occur through dynamic linking among these views. When clusters are automatically identified and colored in one view (the Main view), these colors propagate consistently across all other views, including the Comparison view and SPLOM. Users can interactively adjust clustering parameters, select subsets of clusters for further refinement, and apply clustering algorithms again on these subsets. Such interactions enable users to iteratively drill down into specific parts of their data, exploring cluster structures at varying granularities. The primary goal of these interactions and workflows is to facilitate rapid identification of meaningful clusters and functional dependencies across multiple dimensions, while minimizing manual selection efforts and enhancing exploratory efficiency. An instantiated interface is also implemented, integrating the Mean Shift algorithm for automatic cluster detection and Self-Organizing Maps (SOMs) for topology-preserving color assignments. This enhances the SPLOM by two side-by-side views of SPLOM entries for quick comparisons. Additionally, the user is able to control the color diversity interactively and adjust the display of the data in various ways. This allows for an intuitive and highly customizable workflow, which can adjust views, color diversity, and clustering techniques. The interface also supports interactive refinement, allowing users to isolate subsets of clusters and reapply clustering algorithms for more granular analysis. The thesis showcases several case studies from several synthetic datasets; the respective computational performance evaluation is carried out for each strategy. Through the qualitative and quantitative evaluations, the visual analysis approach demonstrates significant improvements in cluster detection accuracy, functional dependency identification, and cognitive load reduction compared to traditional SPLOM interfaces. A robust tool for effective multi-dimensional data analysis within Jupyter Notebooks is also provided (with code) and attached to the thesis submission.