KNNE: A Novel Searching Schema

Abstract

The spatial keywords search problem has a lot of  buzz recently and rightly so, In spatial database usually the objects (Hospitals / Hotels / ATM ) are associated with keywords. In spatial keyword search, Best Keyword Cover (BKC) query is new version of mCK query. The baseline algorithm is one implementation of  BKC query, Which is based on indexing the objects with an R*-tree like index, Called Keyword Rating (KRR*-tree). We observed when the number of query keywords increases, The performance of the baseline algorithm drop  dramatically  as  result of massive candidate keywords covers generated. To overcome this drawback, We proposed an alternative algorithm of  Best Keyword Cover (BKC) query, We develop a much more scalable algorithm called keywords nearest neighbor expansion (KNNE) algorithm, The KNNE algorithm is supported by indexing keyword rating. This algorithm focus on searching the nearest neighbor by combining  both keywords search and nearest neighbor search. The new algorithm introduce a new concept of keywords rating. The keyword rating helps in decision making. The keyword  nearest  neighbor expansion algorithm  compared  with the Baseline algorithm, It's  significant to reduces the number of candidate keywords cover generated.