Study On The Managemenrt Technology For Uncertain Data Of Marine Stations

Nowadays the importance of uncertain data has got more and more people’sattention in many areas. However, traditional database are deterministic, the uncertaindata could not be processed in the traditional database. Therefore, the managementtechnology for uncertain data has become a focus of database research.In this paper, we first introduce the possible world model, which is the core datamodel for the uncertain data. And then, the extended model on the relational model forthe uncertain data is introduced. After introducing the data model for uncertain data, weanalyze query processing operations for uncertain data. Then we describe thedifferences of aggregation operations further between uncertain data and traditionalrelational data. At the same time, we propose a variety of aggregation query definitionsof uncertain data.Next, we analyze how the uncertain data of marine stations occurs, and we alsostudy the characteristics of the uncertain data of marine stations. We introduce animproved method for the uncertain data for the marine stations based on the analysis ofTrio and MayBMS which are two typical examples of uncertain database managementsystem. The experiment shows that this method eliminates redundant information andreduces the total number of tables effectively, while the original method may generateredundant information and tables. And in the last part of the chapter4, we studytheimplementation for the representation method for the uncertain data of the marinestations.Because of the aggregation function on the possible world model is directlyapplied into each possible world instance，it can not be calculated in linear time．In thispaper, we elaborate the aggregation query method of uncertain data from possible worldmodel aspect. We apply the aggregation function to the original uncertain relationdirectly, and we make the aggregation operations on possible world instancescorresponding to each tuple operation through transformation and store proceduremethods. This not only guarantees the original semantic of aggregation operations, butalso makes sure every operation calculate in the linear time.Here we discuss the three parts of every aggregation operation, so we avoided theoperation for each possible world instance, on the other hand, we could analyze theresult of aggregation operations from many aspects and angles. Finally in the fifthchapter, we explained the aggregation functions we defined before for uncertain datafrom the possible world semantics perspective, and we discuss the implementation forall the aggregation operations in the aspect of the representation method for the uncertain data of the marine stations we defined.The main innovation of this paper is that we analyze the generation andcharacteristics on uncertain data of marine stations, we also study the representationmethod for uncertain data, and we discuss the implementation for all the aggregationoperations in the aspect of the representation method we defined.