My project this year is about Multidimensional Indexing, by looking into a indexing method called "Pyramid Technique" which claim itself more efficient with higher dimensionality thus towards breaking the CURSE OF DIMENSIONALITY. The complete paper can be found HERE!

My main objective has changed since mid-year, I was trying to generate datasets that would "break" the Pyramid Technique, but after take an insight look, the objective changed to compare the performance of the Pyramid Technique to the Sequential Scan, because generally all the indexing methods are bad when the dimension gets too high, and the Pyramid Technique is better than the other methods because it is not recursive, and it only partitions the space once and the indexing is handled by a B+ tree. So when the dimensionality gets high enough, it becomes several linear scans.

The implemented source codes are in my Articles.