Data Indexing in High Dimensional Space

Indexing of the data is an essential part of the search problem. Whilst a well studied set of efficient algorithms of indexing and search exist for a data in the spaces of the dimensionality less then 5, these algorithms are inefficient or not applicable for high dimensional spaces. We review existing studies of the problems related to indexing in a high dimensional space for a nearest neighbour search problem. Possible methods of the designated problems solution that are applicable in the areas of data analysis such as clustering and hidden structure elicitation are regarded. The question on the applicability of the various dimensional reduction methods for an indexing and nearest neighbour search problem.