Continuous Mutual Nearest Neighbour Processing on Moving Objects in Spatiotemporal Datasets

Abstract

Abstract-This paper proposed a new algorithm for answering a novel kind of nearest neighbour search, that is, continuous mutual nearest neighbour (CMNN) search. In this kind of query, by providing a set of objects O and a query object q, CMNN continuously returns the set of objects from O, which is among the k 1 nearest neighbours of q; meanwhile, q is one of their k 2 nearest neighbours. CMNN queries are important in many applications such as decision making, pattern recognition and although it is useful in service providing systems, such as police patrol, taxi drivers, mobile car repairs and so forth. In this paper, we have proposed the first work for handling CMNN queries efficiently, without any assumption on object movements. The most important feature of this work is incremental evaluation and scalability. Utilizing an incremental evaluation technique led to a significant decrease in processing time.Index Terms-Moving objects, nearest neighbor, query processing, spatio-temporal.

Mobile Location Indexing Based On Synthetic Moving Objects

“…For example, a well-known technique such as k the nearest neighbor is a structureless technique and it is very easy to implement. The general work of structureless technique is that distance is calculated from all nodes to the service node of a query and the node with the closest distance is regarded as the nearest neighbor [4][5]. These techniques are very simple but the value of k affects the result.…”