Research On Continuous Skyline Query Processing Algorithms Based On Incremental Motion Models

Continuous Skyline queries returns the subset of data that has an advantage in multi attributes,which plays an important role in database query processing area.In recent years,mobile applications are rapid advancing,which puts forward new challenges to continuous query processing on large-scale data;Besides,it promotes massive data associated with geographic information which brings a new demand of analyzing and processing on the geo-textual data.In this thesis,continuous Skyline query processing technology based on incremental motion models is studied and a solution of continuous Skyline query on geo-textual data is proposed.The main work and contributions are listed as follows:(1)We worked on the modeling of query point movement for continuous Skyline query processing.Practical and matched incremental motion models are utilized to substitute the traditional distance function,which relaxed the constraint on the movement of query objects.According to the basic attributes of different query objects,such as the maximum moving speed,moving direction and so on,this thesis puts forward two motion models: the directional incremental motion model and the extended motion model.(2)We looked into the continuous Skyline query processing technology based on the directional incremental motion model.First,the directional incremental motion model can be applied to estimate the query point's position in the near future.Then,the datasets will be preprocessed with two kinds of pruning strategies.At the meanwhile,according to the change of the distance between data points and the query point,three kinds of events are formulated.Finally,at each moment,events are processed based on their types,while new events for maintaining the future Skyline results will be generated.(3)We studied the continuous Skyline query processing on geo-textual data,and a solution based on an extended incremental motion model is proposed.First,the extended incremental motion model can be used to estimate the query point's position in the near future.After that,according to the characteristics of extended incremental motion model,a subset of data points which may affect the Skyline over the coming moments can be picked out.Besides,a textual relevance function is applied to evaluate the similarity between the query keyword and the ones of geo-textual objects.Finally,specific data objects are accessed at each moment so that the Skyline results are maintained.