A spatial dataset range ordering search method based on grid correlation degree

Abstract

The application belongs to the technical field of data processing, and discloses a spatial data set range ordering search method based on grid correlation, which comprises a data processing stage and a query processing stage.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, specifically relating to a method for supporting range-based sorting and searching of spatial datasets using grid correlation. Background Technology

[0002] With the increasing number of open datasets from governments, businesses, and non-profit organizations, people can access vast amounts of data for analysis through various online systems. These open datasets can be used for decision-making and application development, as well as for training, validating, and enhancing machine learning and artificial intelligence. Most of these datasets contain spatial information, making spatial dataset search an important research direction in the field of datasets. As one of the most fundamental search methods, range-ranked search is supported by most spatial dataset search systems. Therefore, how to efficiently handle range-ranked search on large amounts of spatial datasets is one of the important problems in spatial dataset search.

[0003] Range-ranked searches of spatial datasets require calculating the relevance between the search range and the spatial dataset. However, existing spatial dataset search systems and their prototypes rarely discuss the calculation of this relevance. In similar range-ranked searches, existing literature typically focuses on the number of locations within the search range. In some cases, using the number of locations within the search range as the relevance can be effective, especially when the dataset user needs a large number of spatial data items within the search range. However, this method only considers the local distribution of locations within the spatial dataset, ignoring the impact of the overall distribution of the dataset on the relevance. For example, when using the number of locations as the relevance, a spatial dataset with a larger number of locations is more likely to be more relevance to the search range than a spatial dataset with fewer locations, even if the latter's locations are relatively more concentrated within the search range. Depending on the actual needs of the dataset user, datasets with locations concentrated within the search range are usually the truly desired search results. Therefore, proposing a method that considers both the overall distribution of locations within the spatial dataset and their distribution within the search range when calculating the relevance between the search range and the spatial dataset would improve the usability of range-ranked search results. Summary of the Invention

[0004] To calculate the relevance between a spatial dataset and a search range, a novel calculation method is proposed. This invention provides a spatial dataset range sorting search method based on grid relevance. This method considers both the number of spatial dataset points belonging to the grid and the number of spatial dataset points belonging to the grid in the grid-based relevance calculation, and provides two adjustable parameters: k is a saturation parameter of the spatial dataset, controlling the influence of the number of spatial dataset points belonging to the grid on the relevance; b is a normalization parameter of the spatial dataset, controlling the influence of the total number of spatial dataset points on the relevance.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution:

[0006] This invention is a spatial dataset range sorting search method based on grid relevance. The spatial dataset range sorting search method includes two stages: a first stage, a data processing stage; and a second stage, a query processing stage, specifically including the following processes:

[0007] The data processing stage includes the following steps:

[0008] Step 1.1: Given a spatial database D = {D1, D2, ... D...} n}, where D i A spatial dataset containing several two-dimensional location points is represented as follows: Where m i D represents i The number of location points included determines the division of the global spatial region consisting of all location points in the spatial database D into 2 based on the partitioning parameter θ. θ ×2 θ A set of grids of equal size G = {g 1,1 ,g 1,2 ,…,g u,u-1 ,g u,u}, where g r,c This represents the grid in the r-th row and c-th column of grid G, where u = 2. θ ;

[0009] Step 1.2: Calculate the average number of location points contained in the spatial dataset of spatial database D. For each grid g in the grid set G r,c Calculate the number of location points in the grid (lg) r,c Construct a location point belonging to mesh g. r,c Spatial dataset collection DS(g) r,c )={D i |D i There exists a location point in grid g. r,c If the spatial dataset set DS(g)r,c If not empty, then grid g will be... r,c Add to the Global Non-Empty Mesh Set (GNG);

[0010] Step 1.3: Based on the grid set G, extract each spatial dataset D from the spatial database D. i Convert to grid relevance set A i This ultimately forms a spatial database A = {A1, A2, ..., A} represented by a set of grid correlations. n};

[0011] Step 1.4: Based on the spatial database A represented by the grid correlation set formed in Step 1.3, perform a search on each global non-empty grid g in the global non-empty grid set GNG. r,c Construct a non-empty grid-spatial dataset mapping tuple (g r,c ,DRList r,c ), forming the non-empty grid-spatial dataset mapping index GSI-index={(g r,c ,DRList r,c )|g r,c ∈GNG};

[0012] Assuming the search range for spatial database D is Q, the query processing stage includes the following steps:

[0013] Step 2.1: Based on the grid set G, generate a minimum search grid set G that covers the search range Q. Q ={g r,c |g r,c There is an overlapping region with Q;

[0014] Step 2.2: Utilize the search grid set G Q The non-empty grid-spatial dataset mapping index GSI-index is used to calculate the correlation between the search range Q and each spatial dataset in the spatial database D, thereby determining the final search result Result.

[0015] A further improvement of the present invention is that, in step 1.3, the spatial dataset D is calculated based on the grid set G. i Convert to grid relevance set A i Specifically, it includes the following steps:

[0016] Step 1.3.1: For each mesh g in the mesh set G... r,c If the spatial dataset D i The location point exists in grid g. r,c In the middle, the grid g will be r,c Add to spatial dataset D i Non-empty grid set NG imiddle;

[0017] Step 1.3.2: For the spatial dataset D i For each non-empty grid g in the grid set G r,c Compute spatial dataset D i With grid g r,c Grid correlation

[0018] Step 1.3.3, Spatial Dataset D i The mesh correlations with all non-empty meshes constitute the mesh correlation set.

[0019] A further improvement of the present invention is that, in step 1.3.2, the spatial dataset D is calculated. i With grid g r,c Grid correlation The calculation method is as follows:

[0020]

[0021] Among them, l r,c Representation spatial dataset D i The number of location points contained in the grid set G, where 'a' is the spatial dataset D. i The saturation parameter, b, is the spatial dataset D. i The normalized parameters.

[0022] A further improvement of the present invention is that, in step 1.4, the database A is represented by the relevance set as a globally non-empty grid g. r,c Construct a non-empty grid-spatial dataset mapping tuple (g r,c ,DRList r,c Specifically, it includes the following steps:

[0023] Step 1.4.1: For the spatial dataset set DS(g) r,c Each spatial dataset D in ) i From spatial dataset D i The set of grid correlations A i Extracting the spatial dataset D from the data i With grid g r,c Grid correlation Create a binary tuple and the binary pair Add to grid g r,c Spatial Dataset Mapping List DRList r,c middle.

[0024] Step 1.4.2: Construct a non-empty grid-spatial dataset mapping tuple (g r,c,DRList r,c ).

[0025] A further improvement of the present invention is that: in step 2.2, based on the search grid set G... Q The final search result Result is determined by the non-empty grid-spatial dataset mapping index GSI-index, which includes the following steps:

[0026] Step 2.2.1: For each spatial dataset D in the spatial database D... i Initialize a relevance variable R e l i =0;

[0027] Step 2.2.2: For the search grid set G Q Each search grid g r,c The g is obtained based on the non-empty grid-spatial dataset mapping index GSI-index. r,c The corresponding spatial dataset mapping list DRList r,c For spatial dataset mapping list DRList r,c Each pair Computational spatial dataset D i Corresponding correlation variables

[0028] Step 2.2.3, the corresponding correlation variable R e l i The set of the top k spatial datasets with the highest values ​​is the final search result.

[0029] The beneficial effects of this invention are:

[0030] Specifically, this method transforms a spatial dataset into a set of grid relevances based on grid relevances, and uses the spatial database represented by the set of grid relevances to create a non-empty grid-spatial dataset mapping index. Based on the minimum search grid set covering the search range, the relevance between the search range and the spatial dataset is calculated to obtain the top-k ranking results.

[0031] This invention proposes a grid-based correlation calculation method for spatial datasets, with adjustable parameters based on actual needs. A larger parameter 'a' indicates a greater influence of the number of location points in the spatial dataset within the grid on the correlation. Conversely, a larger parameter 'b' indicates a smaller influence of the number of location points in the spatial dataset on the correlation. For different spatial databases, the methods can be adapted to specific requirements by adjusting parameters 'a' and 'b'.

[0032] This invention constructs an index based on grid relevance, containing mappings between non-empty grids and associated datasets. By using this mapping index, a large amount of location point information can be omitted when storing information, and spatial datasets can be refined into relevance representations, saving storage space. Furthermore, when searching the grid, not only can empty grids be skipped, but associated datasets can also be quickly located through the grid, thereby improving search efficiency.

[0033] The query scheme of this invention has a simple structure and is easy to implement and apply. Attached Figure Description

[0034] Figure 1 This is a flowchart example of the data processing stage and query processing stage of the present invention.

[0035] Figure 2 This is an example of how the present invention generates a grid correlation set based on a spatial dataset.

[0036] Figure 3 This is a mapping index diagram generated by the present invention based on the grid correlation set. Detailed Implementation

[0037] The embodiments of the present invention will be disclosed below with reference to the drawings. For clarity, many practical details will be described in the following description. However, it should be understood that these practical details are not intended to limit the invention. That is, in some embodiments of the invention, these practical details are not essential.

[0038] For ease of description, the relevant symbols are defined as follows: The spatial database is D = {D1, D2, ... D...} n}, where D i A spatial dataset containing several two-dimensional location points is represented as follows: Where m i D represents i The number of location points included, the grid set G = {g 1,1 ,g 1,2 ,…,g u,u-1 ,g u,u}, where g r,c This represents the grid in the r-th row and c-th column of the grid set G, where u = 2. θ The spatial database represented by the grid correlation set is A = {A = {A1, A2, ..., A}. n}}, where A i It is a spatial dataset D i Grid correlation set NG i Representation spatial dataset D i A non-empty grid set, Representation spatial dataset D i With gr,c The grid relevance; the non-empty grid-spatial dataset mapping index is GSI-index = {(g r,c ,DRList r,c )|g r,c ∈GNG}, where GNG is the globally non-empty set of grids. It is g r,c Spatial dataset mapping list, DS(g) r,c ) indicates that there exists a location point belonging to grid g. r,c A collection of spatial datasets.

[0039] This invention is a spatial dataset range ranking search method based on grid relevance. The method comprises two stages: a first stage, data processing; and a second stage, query processing. Specifically, it includes the following processes:

[0040] The data processing stage mainly consists of four steps:

[0041] Step 1.1: Given a spatial database D = {D1, D2, ... D...} n}, where D i A spatial dataset containing several two-dimensional location points is represented as follows: Where m i D represents i The number of location points included determines the division of the global spatial region consisting of all location points in the spatial database D into 2 based on the partitioning parameter θ. θ ×2 θ A set of grids of equal size G = {g 1,1 ,g 1,2 ,…,g u,u-1 ,g u,u}, where g r,c This represents the grid in the r-th row and c-th column of grid G, where u = 2. θ ;

[0042] Step 1.2: Calculate the average number of location points contained in the spatial dataset of spatial database D. For each grid g in the grid set G r,c Calculate the number of location points in the grid (lg) r,c Construct a location point belonging to mesh g. r,c Spatial dataset collection DS(g) r,c )={D i |D i There exists a location point in grid g. r,c If the spatial dataset set DS(g) r,c If not empty, then grid g will be... r,c Add to the Global Non-Empty Mesh Set (GNG);

[0043] Step 1.3: Based on the grid set G, extract each spatial dataset D from the spatial database D. i Convert to grid relevance set A i This ultimately forms a spatial database A = {A1, A2, ..., A} represented by a set of grid correlations. n};

[0044] In this step, the spatial dataset D is determined based on the grid set G. i Convert to grid relevance set A i Specifically, it includes the following steps:

[0045] Step 1.3.1: For each mesh g in the mesh set G... r,c If the spatial dataset D i The location point exists in grid g. r,c In the middle, the grid g will be r,c Add to spatial dataset D i Non-empty grid set NG i middle;

[0046] Step 1.3.2: For the spatial dataset D i For each non-empty grid g in the grid set G r,c Compute spatial dataset D i With grid g r,c Grid correlation

[0047]

[0048] Among them l r,c Representation spatial dataset D i The number of location points contained in the grid set G, where a is the saturation parameter of the spatial dataset and b is the normalization parameter of the spatial dataset.

[0049] Step 1.3.3, Spatial Dataset D i The mesh correlations with all non-empty meshes constitute the mesh correlation set.

[0050] Step 1.4: Based on the spatial database A represented by the grid correlation set formed in Step 1.3, perform a search on each global non-empty grid g in the global non-empty grid set GNG. r,c Construct a non-empty grid-spatial dataset mapping tuple (g r,c ,DRList r,c ), forming the non-empty grid-spatial dataset mapping index GSI-index={(g r,c ,DRList r,c )|g r,c ∈GNG};

[0051] Based on the relevance set, database A is represented as a globally non-empty grid g. r,c Construct a non-empty grid-spatial dataset mapping tuple (g r,c ,DRList r,c Specifically, it includes the following steps:

[0052] Step 1.4.1: For the spatial dataset set DS(g) r,c Each spatial dataset D in ) i From spatial dataset D i The set of grid correlations A i Extracting the spatial dataset D from the data i With grid g r,c Grid correlation Create a binary tuple and the binary pair Add to grid g r,c Spatial Dataset Mapping List DRList r,c middle.

[0053] Step 1.4.2: Construct a non-empty grid-spatial dataset mapping tuple (g r,c ,DRList r,c ).

[0054] Assuming the search range for spatial database D is Q, the query processing stage includes the following steps:

[0055] Step 2.1: Based on the grid set G, generate a minimum search grid set G that covers the search range Q. Q ={g r,c |g r,c There is an overlapping region with Q;

[0056] Step 2.2: Utilize the search grid set G Q The non-empty grid-spatial dataset mapping index GSI-index is used to calculate the correlation between the search range Q and each spatial dataset in the spatial database D, thereby determining the final search result Result.

[0057] In this step, based on the search grid set G Q The final search result Result is determined by the non-empty grid-spatial dataset mapping index GSI-index, which includes the following steps:

[0058] Step 2.2.1: For each spatial dataset D in the spatial database D... i Initialize a relevance variable R e l i =0;

[0059] Step 2.2.2: For the search grid set G Q Each search grid g r,c The g is obtained based on the non-empty grid-spatial dataset mapping index GSI-index. r,c The corresponding spatial dataset mapping list DRList r,c For spatial dataset mapping list DRList r,c Each pair Computational spatial dataset D i Corresponding correlation variables

[0060] Step 2.2.3, the corresponding correlation variable R e l i The set of the top k spatial datasets with the highest values ​​is the final search result.

[0061] Figure 2 The paper shows the distribution of four spatial datasets in a grid set, and transforms them into a grid correlation set representation with parameters a=1.2 and b=0.75.

[0062] Figure 3 The content shown is based on Figure 2 A mapping index constructed from the grid correlation set of the medium spatial dataset.

[0063] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A spatial dataset range sorting and search method based on grid relevance, characterized in that: The spatial dataset range sorting search method includes two stages: the first stage, the data processing stage; The second stage is the query processing stage, in which... The data processing stage is as follows: First, the given spatial database is processed. A global spatial region is constructed based on the distribution of all spatial dataset locations in the spatial database and divided into grids of equal size. Then, the information of the spatial database is statistically analyzed to prepare parameters and sets for correlation calculation and index construction. After that, using these parameters and sets, for each spatial dataset, a grid correlation set is generated based on the number of locations within the corresponding grid. Finally, a non-empty grid-spatial dataset mapping index is constructed based on the spatial database represented by the grid correlation set. The query processing stage is as follows: First, a set of minimal search grids covering the search range is generated based on the grid division. Then, the relevance between each spatial dataset and the search range is obtained through a non-empty grid-spatial dataset mapping index based on the search grid set. Finally, the dataset with the highest relevance is selected. Using spatial datasets as query results, assuming a spatial database. The search scope is The query processing stage includes the following steps: Step 2.1: Based on the mesh set Generate a coverage search range Minimize search grid set ={ | and There are overlapping regions. Step 2.2: Utilize the search grid set Calculate the search range using the non-empty grid-spatial dataset mapping index GSI-index. and spatial database The correlation between each spatial dataset is used to determine the final search results. Specifically, it includes the following steps: Step 2.2.1: Spatial Database Each spatial dataset Initialize a relevance variable =0; Step 2.2.2: For the search grid set Each search grid Obtained from the non-empty grid-spatial dataset mapping index GSI-index Corresponding spatial dataset mapping list Mapping list for spatial datasets Each pair of tuples ( , Computational spatial dataset Corresponding correlation variables = in, For grid relevance; Step 2.2.3, corresponding correlation variables The highest value The final search results are the collection of spatial datasets. .

2. The spatial dataset range sorting and search method based on grid relevance according to claim 1, characterized in that: The data processing stage specifically includes the following steps: Step 1.1: Given a spatial database ,in A spatial dataset containing several two-dimensional location points is represented as follows: ,in express The number of location points included will affect the spatial database. The global spatial region consisting of all location points is determined according to the partitioning parameters. Divided into 2 θ ×2 θ A set of grids of equal size ,in Represents a grid The first in Line number Column grid, = 2 θ ; Step 1.2: Calculate the spatial database The spatial dataset contains an average number of location points. For grid sets Each grid Calculate the number of location points in the grid. And construct the location points belonging to the grid. Spatial dataset collection If the spatial dataset set If not empty, then Add to global non-empty grid set middle; Step 1.3: Based on the mesh set Spatial database Each spatial dataset Convert to grid relevance set This ultimately forms a spatial database represented by a set of grid correlations. ; Step 1.4: Spatial database represented by the grid correlation set formed in Step 1.

3. For a globally non-empty grid set In each globally non-empty grid Construct a non-empty grid-spatial dataset mapping tuple ( ), forming the non-empty grid-spatial dataset mapping index GSI-index = { ( )| } 3. The spatial dataset range sorting and search method based on grid relevance according to claim 2, characterized in that: In step 1.3, based on the mesh set Spatial dataset Convert to grid relevance set Specifically, it includes the following steps: Step 1.3.1, for the mesh set Each grid If spatial dataset The location point exists within the grid. In the middle, the grid will be Add to spatial dataset Non-empty grid set middle; Step 1.3.2, for the spatial dataset In grid collection Each non-empty grid in Computational spatial dataset With grid Grid correlation ; Step 1.3.3, Spatial Dataset The mesh correlations with all non-empty meshes constitute the mesh correlation set. = {( , ) | } 4. The spatial dataset range sorting and search method based on grid relevance according to claim 3, characterized in that: In step 1.3.2, the spatial dataset is calculated. With grid Grid correlation The calculation method is as follows: in, Representation space dataset In grid collection The number of location points included It is a spatial dataset The saturation parameter, It is a spatial dataset The normalized parameters.

5. The spatial dataset range sorting and search method based on grid relevance according to claim 2, characterized in that: In step 1.4, the database is represented by the relevance set. Global non-empty grid Construct a non-empty grid-spatial dataset mapping tuple ( Specifically, it includes the following steps: Step 1.4.1: For the spatial dataset set Each spatial dataset From spatial datasets Grid correlation set Extracting spatial datasets With grid Grid correlation Create a tuple ( , ), and the binary pair ( , Add to grid Spatial dataset mapping list middle; Step 1.4.2: Construct non-empty grid-spatial dataset mapping tuples ( ).

Images