When decision makers are required to query terabytes or petabytes of datasets, narrowing the search space helps in shortening the time for querying. We investigated the use of a feature space multidimensional index to narrow the search space. When studying different indexing techniques we found that each fits certain data spaces. We identified a hybrid approach that adapts to different types of indexed data with the goal to optimize both insertion and query performance. We also investigated how the multidimensional index could be used to define a complex feature space from different data types, which we call a `composite' Data View. Such an index is independent of the data location and distribution. We believe our approach could address the problem of dealing with massive dynamic datasets for optimum query performance.