PROJECT TITLE:

Nearest Keyword Set Search in Multi-dimensional Datasets - 2016

ABSTRACT:

Keyword-based mostly search in text-made multi-dimensional datasets facilitates many novel applications and tools. In this paper, we tend to consider objects that are tagged with keywords and are embedded in an exceedingly vector space. For these datasets, we have a tendency to study queries that ask for the tightest groups of points satisfying a given set of keywords. We tend to propose a completely unique methodology referred to as ProMiSH (Projection and Multi Scale Hashing) that uses random projection and hash-primarily based index structures, and achieves high scalability and speedup. We have a tendency to gift an actual and an approximate version of the algorithm. Our experimental results on real and artificial datasets show that ProMiSH has up to 60 times of speedup over state-of-the-art tree-primarily based techniques.