A skyline query and update method based on time-varying information in a road network environment

Abstract

This invention discloses a skyline query and update method based on time-varying information in a road network environment. The method first preprocesses data based on skyline extension properties, then utilizes the road network Voronoi data structure for data refinement and pruning, and stores road network information via a grid index. Secondly, a grid index extension tree is proposed as the primary concept for querying. To better handle the constantly changing information in the road network, two data structures are designed to maintain road network information. Theoretical research and experiments demonstrate that the proposed method can effectively handle skyline queries based on time-varying information in a road network environment and effectively resolve the issue of inconsistent query results caused by the constantly changing object attribute information and road conditions in this environment.

Description

Technical Field

[0001] The present invention relates to the field of time-varying information skyline query, and in particular to a skyline query and update method based on time-varying information in a road network environment. Background Art

[0002] Skyline querying based on time-varying information in a road network environment is a key research topic in the field of spatial databases. In recent years, with the proliferation of mobile devices and the widespread use of wireless network services, location query services have become an essential user need, with skyline query serving as one of the most important enabling technologies for location-based services. With the development of the mobile internet, numerous location-based applications and services, such as finding nearby restaurants and booking hotels, have emerged. Skyline querying can be used to address real-world needs such as ranking products that match user interests and providing them with restaurants or hotels of interest. Therefore, skyline query technology holds significant research value in search-based services.

[0003] Existing research on skyline query algorithms in road network environments primarily focuses on static road networks, where object attributes and road conditions remain unchanged. However, in real-world applications, object attributes and road network conditions constantly change over time, severely limiting the practicality of this research. Summary of the Invention

[0004] In response to the above defects and improvement needs of the prior art, the present invention provides a skyline query and update method based on time-varying information in a road network environment.

[0005] The technical solution of the present invention is: a skyline query and update method based on time-varying information in a road network environment, which mainly includes:

[0006] 1. Traditional skyline query methods in road network environments fail to account for the presence of redundant data. This paper proposes a strategy for handling redundant data. This strategy leverages the dominance relationship of the skyline to extend the relationships between skyline levels, filtering out useless data, effectively addressing the problem of low query efficiency and improving computational performance.

[0007] 2. In response to the low query efficiency of traditional skyline query methods in road network environments, two data structures are proposed to maintain road network information, which effectively solve the problem of low query efficiency caused by the continuous changes in object attribute information and road conditions in the road network.

[0008] 3. Aiming at the problem of low efficiency in maintaining road network information in traditional skyline query methods under road network environments, a skyline object update algorithm is proposed, and combined with two data structures, it effectively solves the problem of low query efficiency caused by time-varying information.

[0009] Furthermore, the two data structures refer to TADA and RODL data structures, where TADA represents the time-varying attributes in the query, and an attribute dominance array (TADA) is designed to represent the dominance relationship between object attributes. When processing skyline queries in a road network environment, TADA stores objects whose distance to the query point is less than or equal to the user-defined distance range. When the attributes change, the objects retained in TADA are not affected by the change of time-varying attributes. If the dominance relationship of the object changes due to the change of the time-varying attribute object, the query result set may be affected; the ascending linked list of road network distances (RDOL, if the distances are equal, the position of the object in RODL is determined according to the dominance relationship in TADA), proposed by the RODL data structure, is similar to TADA. Only objects whose road network distance does not exceed the user-defined distance will be stored in the ascending linked list. If an object o i Position in RDOL in another object o j Before the position, then o i There is a chance to dominate o j , because its road network distance is smaller than the query point. o 2, o 3) = 1, in RODL o 2 positions in o 3 before. Therefore o 2. Dominate in terms of "price", "grade" and "attribute" o 3. Although o 3 is not a node n 1 SP, but due to o The distance of 3 is within the user-defined range. o 3 needs to be kept in RODL.

[0010] Furthermore, the skyline object update algorithm is expressed as o i If it does not belong to RODL, after updating the road network distance, the object o i To Node n Road network distance d ( o i， n )<d Ɛ , then the object o i It needs to be added to TADA and RODL to compare the dominance relationship and update the result set. o i , o j ) = -1, and the position in RODL o i exist o j After that, then at this time o i Cannot be updated to the result set SP, on the contrary, o i ∈SP. When there are some pairs o j Belongs to the result set, only when o The properties of the distance are better than o j hour, o j It can be removed from the result set, at which point the result set is updated.

[0011] When the object o i Belongs to RODL, but o i When it does not belong to SP, the side length is updated after the distance d' ( o i， n )< d Ɛ , so the object o i The location in RODL is o j Before, it was possible to be o Domination, if present in TADA ( o i , o j ) = 1 (Description o i Non-distance attribute ratio o j Good), at this time you can o j Remove from the result set and update the result set SP. In addition, due to the update of the edge length, the o i The previous object may no longer dominate o i ,at this time o iCan be a candidate SP. When it does not exist in TADA ( o i , o j ) = -1 and does not exist o j The distance property is better than o i ,at this time o i It can be used as a global SP, and the result set SP is updated at this time.

[0012] When the object o i Belong to RODL, and o i When it belongs to SP, due to the long update distance d' ( o i， n )< d Ɛ , the distance attribute is better, at this time o i Still belongs to the result set SP. If there are any objects in the result set o j , while meeting the conditions in TADA ( o i , o j ) = 1 and in RODL o i The location is o j Before, then o j Will be removed from the result set and the result set updated.

[0013] The beneficial effect of the present invention is that the present invention further protects user location information query by adopting the K-dominant combined with semantic similarity algorithm, which not only reduces the time overhead of query, but also further protects the user's query privacy. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of a skyline query and update method based on time-varying information in a road network environment according to the present invention.

[0015] Figure 2 It is a schematic diagram of the initial data structure TADA of the present invention.

[0016] Figure 3 It is a schematic diagram of TADA after the property changes of the present invention.

[0017] Figure 4Schematic diagram of an example of road closure according to the present invention.

[0018] Figure 5 It is a schematic diagram of the data structure RODL of the present invention.

[0019] Figure 6 This is a schematic diagram of the skyline expansion tree of the node of the present invention Implementation Method

[0020] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the examples described are only some implementation examples of the present invention, rather than all embodiments. In addition, it should be understood that after reading this invention, those skilled in the art can make various modifications or changes to the present invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0021] The present invention discloses a skyline query and update method based on time-varying information in a road network environment. The specific operation process includes:

[0022] Step 1: Since all data objects are located in the same road network and the database contains a large amount of information that needs to be maintained, a grid index can effectively manage this data and provide the data structure required for query when storing data. The grid index extension tree further improves query efficiency. Based on this, the query algorithm 1 for the skyline query in the road network is proposed as shown in Algorithm 1:

[0023] Algorithm 1: Obtaining candidate skyline sets

[0024] Input: Preferred distance area d Ɛ ,side e and the two nodes of the edge i , j

[0025] Output: Global candidate SP points

[0026] 1. First, the node i is the center of the circle, the preferred distance d Ɛ Make a circle for the radius C ;

[0027] 2. Store the object information based on the area where the circular area formed intersects with the grid unit;

[0028] 3. Calculation Object o i The Euclidean distance to the node i;

[0029] 4. If the Euclidean distance is greater than the preference distance, further pruning is performed.

[0030] 5. Calculate road network distance through grid index;

[0031] 6. If the road network distance is within the distance range, o i If it is not dominated by other static properties, o i As a candidate SP point, it is added to the candidate skyline set.

[0032] Step 2: Due to the large amount of data and the complexity of the original data set in the road network environment, in order to further improve the query efficiency, a storage array based on time-varying objects is used to maintain object attribute information, avoid a large number of unnecessary road network operations, and reduce computing overhead. When the attributes of an object change, the object's dominant array TADA needs to be updated in real time. The main idea when dealing with changes in the attributes of time-varying objects is: first, by checking whether the object with time-varying attributes appears in TADA and RDOL to determine whether the change in the attributes of the time-varying object affects the final result set SP, then check whether the value in TADA and the position of the object in RDOL have changed. If the object o The properties from o . pro becomes o . pro ', o If it does not appear in TADA and RDOL, then the result set SP does not need to be updated at this time, and there is no need to pay attention to the new attributes of the object at this time, so it can be ignored. o If the object appears in TADA and RDOL, then the object o i and another object o j The dominance relationship may change, which means that the result set SP needs to be further verified. Depending on whether the object is included in the result set, the result set update can be divided into four cases: (1) o i Belong to SP and o j Belongs to SP, (2) o i Belong to SP and o j Not belonging to SP, (3) o i Not part of SP and o j Belongs to SP, (4) o i Not part of SP and oj Does not belong to SP.

[0033] Algorithm 2 Pruning algorithm based on time-varying attributes in road networks

[0034] Input: object, SP corresponding to the node

[0035] Output: SP* after pruning

[0036] 1. Create the attribute dominance array TADA of the object and create the ascending linked list RDOL of the road network distance.

[0037] 2. Store the object information into the data structure according to the generated object attribute dominance array and the ascending linked list based on the road network distance.

[0038] 3. Judgment object o i The relationship with SP, if o i Belong to SP and o j Belongs to SP.

[0039] 4. When satisfied o i Belong to SP and o j When it belongs to SP, after the object attribute changes, if in the control array TADA ( o i , o j ) value is -1, and at this time o j The distance property is better than o i , then the object o i Prune; if the attribute changes in TADA ( o i , o j ) is 1, and o i The distance property is better than o j , then the object o j Pruning.

[0040] 5. Meet the conditions o i Belong to SP and o j If it does not belong to SP, when the object attributes change, if in TADA ( o i , oj ) changes from 0 to -1, and in RODL o j The location is o i Before, this time o i Pruning; if in TADA ( o i , o j ) changes from 1 to -1, and no other object dominates o j , in RODL o j The location is o i Before, at this time o i Being pruned.

[0041] 6. Meet the conditions o i Not part of SP and o j When it belongs to SP, if in TADA ( o i , o j ) changes from -1 to 0, and o i is dominated by other objects, at which point the object o i Pruned; if in TADA ( o i , o j ) changes from -1 to 1, and o i In RODL o j Afterwards, the object o i Being pruned.

[0042] 7. Update SP* based on the pruning situation.

[0043] Changes in time-varying edge lengths affect not only the position and number of objects in RODL but also the number of objects in TADA. This means that an object will be added or removed from both RDOL and TADA as the edge length increases or decreases. Therefore, when the edge length changes, the values in TADA and RODL are updated accordingly. We will then discuss whether this affects the update of the result set.

[0044] First consider the connection nodes n i and nj The length of the edge becomes "×" (that is, the road is closed). At this time, the grid index expansion tree is combined to determine whether the result set is affected. If there is no parent-child relationship in the tree with this node as the parent node, n i and n j Node, the time-varying edge length change cannot affect the objects in TADA and RODL, and the result set is not affected. n i and n j Nodes contain n i and n j The subtree of the node will be affected and deleted from the tree. At this time, it is necessary to traverse the nodes in the remaining expanded tree to update the road distance between the objects.

[0045] Step 3: Assume the connection node n 1 and node n Road 5 is temporarily closed, traverse the node n 1's skyline expansion tree shows that n 1 and n 5 There is a parent-child relationship. n The subtree of 1 is removed from the expanded tree. Therefore o 2 and o The road network distance of 4 needs to be recalculated to "infinity" and 120 respectively. o 2 and o 4. Delete from TADA and RODL. Therefore, if the road status between two random nodes changes from open to closed, the subtree of the node with the parent-child relationship with the node in the skyline expansion tree can be deleted to reduce repeated calculations during road network updates. Prune a closed road in the road network.

[0046] Step 4: Discuss the update problem of the result set after pruning. That is, when the object o i To Node n The road network distance from d ( o i， n ) becomes d' ( o i， n ). At this time, the update process of the result set can be divided into the following three cases: (1) Object o i Not part of RODL; (2) Object o i Belongs to RODL, buto i Not belonging to SP; (3) Object o i Belong to RODL, and o i Belongs to SP.

[0047] When the object o i If it does not belong to RODL, after updating the road network distance, the object o i To Node n Road network distance d ( o i, n )> d Ɛ , at this time the object o i Does not appear in RODL, so the result set does not need to be updated.

[0048] When the object o i Belongs to RODL, but o i When the object does not belong to SP o i To Node n The road network distance exceeds d Ɛ , directly change the node n Remove from TADA and RODL, then the result set does not need to be updated. d ( o i, n )< d Ɛ When the position of the object in RODL is based on d' ( o i, n ) for dynamic adjustment. For the result set SP, it is not affected by o i No update is required.

[0049] When the object o i Belong to RODL, and o i When it belongs to SP. If the edge length is updated after the distance d' ( o i, n )> d Ɛ , you need to delete the object from TADA and RODL oi , and the result set SP changes; secondly, if there is an object o j ∈RODL, o j Belongs to SP, if o j Only o i Dominate, and o i The location in RODL is o j Before, it can be used as a candidate SP. If it does not exist in TADA ( o k , o j ) = 1 and in RODL o j exist o k The previous object, the result set can now be added o j For SP. If the edge length is updated after the distance d' ( o i, n )< d Ɛ , and the object o i Still in TADA and RODL, but at this time o i The position in RODL is higher than before, so the result is concentrated o i will be deleted; otherwise, if o i The dominant object is no longer dominated, so it can be used as a candidate SP. o i Is it deleted? First check whether TADA contains o i The rows and columns of o j The corresponding relationship ( o i , o j ) = -1, then o i The updated position in RODL is the same as o j Comparison, only when o j exist o i When the object is in front of o iWill be completely deleted. On the contrary, in order to determine whether the candidate SP can be used as the global SP, it is necessary to check all the SPs in TADA that have ( o i , o j ) = 1 and in RODL the position is o i Each object after o j ,once o j The location is o i Before, and o j is not dominated by any object, then o j Can be added to the result set as a global SP point, otherwise as a candidate SP point o j Cannot be added to the result set as a global SP point.

[0050] When connecting nodes n i and n j The length of the side changes from "×" to e (That is, the road changes from closed to open). If the grid index expansion tree does not contain n i and n j Nodes, which means that the edge length is not in the distance d Ɛ In this case, the result set SP does not need to be updated. On the contrary, if the tree contains n i and n j These two nodes, the result set may be updated. Also assume that the object o i To Node n The road network distance from d ( o i, n ) becomes d' ( o i, n ). At this time, the update process of the result set is divided into the following three cases: (1) Object o i Not part of RODL; (2) Object o i Belongs to RODL, but o i Not belonging to SP; (3) Object oi Belong to RODL, and o i Belongs to SP.

[0051] When the object o i If it does not belong to RODL, after updating the road network distance, the object o i Road network distance to node i d ( o i, n )< d Ɛ , then the object o i It needs to be added to TADA and RODL to compare the dominance relationship and update the result set. o i , o j ) = -1, and the position in RODL o i exist o j After that, then at this time o i Cannot be updated to the result set SP, on the contrary, o i ∈SP. When there are some objects o j Belongs to the result set, only when o The properties of the distance are better than o j hour, o j It can be removed from the result set, at which point the result set is updated.

[0052] When the object o i Belongs to RODL, but o j When it does not belong to SP, the edge length update distance d' ( o i, n )< d Ɛ , so the object o i The location in RODL is o j Before, it was possible to be o Domination, if present in TADA ( o i , o j ) = 1 (Descriptiono i Non-distance attribute ratio o j Good), at this time you can o j Remove from the result set and update the result set SP. In addition, due to the update of the edge length, the o i The previous object may no longer dominate o i ,at this time o i Can be a candidate SP. When it does not exist in TADA ( o i , o j ) = -1 and does not exist o j The distance property is better than o i ,at this time o i It can be used as a global SP, and the result set SP is updated at this time.

[0053] When the object o i Belong to RODL, and o i When it belongs to SP, due to the long update distance d' ( o i, n )< d Ɛ , the distance attribute is better, at this time o i Still belongs to the result set SP. If there are any objects in the result set o j , while meeting the conditions in TADA ( o i , o j ) = 1 and in RODL o i The location is o j Before, then o j Will be removed from the result set and the result set updated.

[0054] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be fully applied to various fields suitable for the present invention. For those skilled in the art, further modifications can be easily realized. Therefore, the present invention is not limited to the specific details and illustrations shown and described herein without departing from the general concept defined by the claims and their equivalents.

Claims

1. A skyline query and update method based on time-varying information in a road network environment, characterized by The following steps are involved: Step 1: Use the dominance relationship of the skyline to extend the relationship between skyline levels, filter out useless data, effectively solve the problem of low query efficiency, and improve computing performance; Step 2: Maintain road network information through two data structures to effectively solve the problem of low query efficiency caused by the continuous changes in object attribute information and road conditions in the road network; Step 3: Propose a skyline object update algorithm, which is expressed as when object o i If it does not belong to RODL, after updating the road network distance, object o i The road network distance d(o i， n) <d ε , then object o i Need to be added to TADA and RODL to compare the dominance relationship and update the result set; when there is (o i ,o j )=-1, and position o in RODL i In o j After that, then at this time o i Cannot update to the result set SP, on the contrary, o i Belongs to SP; when there are some objects o j Belong to the result set only when the attribute distance of o is better than o j When, o j It is removed from the result set, at which point the result set is updated. In addition, the following two situations need to be judged when updating data: (1) When object o i Belongs to RODL, but o i When it does not belong to SP, the edge length is updated after the distance d'(o i， n) <d ε , so the object o i The position in RODL is o j Before, it was possible to be dominated by o, if there was (o i ,o j )=1, then o j Remove from the result set and update the result set SP; In addition, due to the update of the edge length, the previously i The previous object may no longer dominate o i , at this time o i Can be used as a candidate SP; when it does not exist in TADA (o i ,o j )=-1 and there is no o j The distance property is better than o i , at this time o i It can be used as a global SP, and the result set SP is updated at this time; (2) When object o i Belongs to RODL, and o i When it belongs to SP, due to the long update distance d'(o i， n) <d ε , the distance attribute is better, then o i Still belongs to the result set SP; if there is any object o in the result set j , and meet the conditions in TADA (o i ,o j )=1 and in RODL o i The location is o j Before, then o j It will be removed from the result set, the result set will be updated, and the two data structures will be combined to effectively solve the problem of low query efficiency caused by time-varying information; among them, TADA is the attribute dominance array, and RODL is the ascending linked list of road network distances.

2. The method for skyline query and update based on time-varying information in a road network environment according to claim 1, characterized in that: When filtering the dataset, the skyline layer is divided according to the distance attribute, and the custom distance for the query is d ε , in the distance attribute d ε A perpendicular line is drawn at the layer where the perpendicular line intersects with the highest layer in the skyline preprocessing layer as the boundary, and the skyline points of the layers higher than the skyline layer are pruned.

3. The method for skyline query and update based on time-varying information in a road network environment according to claim 1, characterized in that: If there is a data point p, with the query point q as the center, d ε Make a circle with radius C q , if the network Voronoi neighboring points of a data point p are pruned, then the point p is also pruned.

Images