EFFICIENT KEYWORD SET SEARCH WITH PMHR

ABSTRACT Multi-dimensional datasets are datasets in which each data point consists of set of keywords in abstract space that concede to develop a new mechanisms for querying and explore these multi-dimensional dataset. In this proposal, we learn nearest keyword set search queries on text-rich, multi-dimensional datasets with ranking functions. A peculiar method called PMHR (Projection and Multi-scale Hashing with Ranking) that uses random projection, hash-based index structure, and ranking. Ranking is done based on repentance of keywords. Ranking is done by using tf-idf technique and provide efficient results. Keyword-based search is done on text-rich multi-dimensional datasets which facilitates many peculiar applications and mechanisms. We considered objects that are appended with keywords and are impacted in a vector space. From these datasets, we are going to study queries that are from the tightest groups of points by gratifying a given set of keywords.