Nearest Neighbor Browsing and Search with Keywords in Spatial Databases

Abstract

Users may search for different type of things from anywhere. But Search results depend on the user entered query which has to satisfy their searched properties that is stored in the spatial database. Due to rapid growth of users it becomes essential to optimize search results based on nearest neighbour property in spatial databases. Conventional spatial queries, such as range search and nearest neighbour retrieval, involve only geometric properties of objects which satisfies condition on geometric objects. Nowadays many modern applications aim to find objects satisfying both a spatial condition and a condition on their associated texts which is known as Spatial keyword search. For example, instead of considering all the hotels, a nearest neighbor query would instead ask for the hotel that is closest to among those who provide services such as pool, internet at the same time. For this type of query a variant of inverted index is used that is effective for multidimensional points and comes with an R-tree which is built on every inverted list, and uses the algorithm of minimum bounding method that can answer the nearest neighbor queries with keywords in real time.