Gate address anomaly detection method and device, equipment and storage medium

Abstract

The invention discloses a gate address anomaly detection method and device, equipment and a storage medium. The gate address anomaly detection method comprises the steps of obtaining a to-be-processed gate address data set; performing structured processing on any one of the gate address data in the gate address data set according to the address component so as to perform aggregation processing on each structured address data; performing abnormal point identification on the structured address data in any aggregation result according to the structured address data in a preset range associated with any aggregation result to obtain an initial abnormal gate address; and verifying the initial abnormal gate address to obtain a target abnormal gate address. According to the method, structured processing is performed on the gate address data, the address is decomposed into clear components, the address can be analyzed and matched more accurately, abnormal point identification is performed based on the gate address data in the preset range associated with the aggregation result, misjudgment caused by a single data source can be reduced by using multi-dimensional gate address information, and the accuracy of abnormal point identification is improved. Therefore, the initial abnormal gate address is verified, and the accuracy of abnormal point identification is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular, to a method, device, equipment and storage medium for detecting abnormal door addresses. Background Art

[0002] Currently, the detection process of abnormal door address points mainly relies on the retrieval of address texts. Specifically, it is necessary to search and verify the address text through a map retrieval service. The effect of this process depends on the capabilities and accuracy of the map retrieval service. If the performance of the map retrieval service is poor or there are errors, it will directly affect the recall rate and accuracy of the detection of abnormal door address points. In addition, when correcting coordinates, the current solutions mainly rely on the relevant data of the target address itself, and this single data source limits the comprehensiveness and accuracy of the correction. Summary of the Invention

[0003] Based on this, it is necessary to provide a method, device, equipment and storage medium for detecting abnormal door addresses to solve at least one of the problems existing in the above technical problems.

[0004] The present invention provides a method for detecting abnormal door addresses, including:

[0005] Obtaining a set of door address data to be processed;

[0006] Structuring any door address data in the set of door address data according to address components to perform aggregation processing on each structured address data;

[0007] Identifying abnormal points in the structured address data in any aggregation result based on the structured address data within a preset range associated with any aggregation result to obtain an initial abnormal door address;

[0008] Verifying the initial abnormal door address to obtain a target abnormal door address.

[0009] Optionally, according to a method for detecting abnormal door addresses provided by the present invention, the structuring any door address data in the set of door address data according to address components includes:

[0010] For any of the door address data:

[0011] Disassembling the door address data according to address components to obtain the address entity information in the door address data and the component type corresponding to the address entity information;

[0012] Based on the address entity information and the component type, performing normalization processing on the door address data and performing structuring processing on the normalized door address data.

[0013] Optionally, according to a method for detecting abnormal door addresses provided by the present invention, the structured address data includes address entity information and a component type corresponding to the address entity information;

[0014] The aggregating process for each structured address data includes:

[0015] Correcting each of the address entity information and its corresponding component type;

[0016] Filtering the corrected structured address data;

[0017] Aggregating the filtered structured address data.

[0018] Optionally, according to a method for detecting abnormal door addresses provided by the present invention, the identifying of abnormal points for the structured address data in any one aggregation result to obtain an initial abnormal door address based on the structured address data within a preset range associated with any one aggregation result includes:

[0019] For any one of the aggregation results:

[0020] Performing redundancy processing on the structured address data in the aggregation result to obtain a target aggregation result;

[0021] Determining the first number of door addresses corresponding to the target aggregation result;

[0022] If the first number of door addresses is less than a preset number threshold, then based on each structured address data in the target aggregation result, determining the second number of door addresses corresponding to the structured address data located within the preset range;

[0023] Determining the initial abnormal door address based on the second number of door addresses associated with each structured address data.

[0024] Optionally, according to a method for detecting abnormal door addresses provided by the present invention, after determining the first number of door addresses corresponding to the target aggregation result, further including:

[0025] If the first number of door addresses is greater than or equal to the preset number threshold, then clustering the structured address data in the target aggregation result;

[0026] Determining the initial abnormal door address according to the clustering result.

[0027] Optionally, according to a method for detecting abnormal door addresses provided by the present invention, the verifying of the initial abnormal door address to obtain a target abnormal door address includes:

[0028] For the initial abnormal door address in any one aggregation result:

[0029] Determine the first shortest distance between the initial abnormal address and the normal address data in the aggregation result, and count the number and proportion of abnormal points of the initial abnormal address in the aggregation result, where the normal address data refers to the address data that does not belong to the initial abnormal address among all structured address data;

[0030] Verify the initial abnormal address according to the first shortest distance, the number of abnormal points, and the proportion, to obtain the target abnormal address; and / or,

[0031] Determine the distances between the various structured address data in the aggregation result, and calculate the standard deviation according to the distances between the various structured address data;

[0032] Verify the initial abnormal address according to the standard deviation and the second shortest distance between the initial abnormal address and the normal address data, to obtain the target abnormal address; and / or,

[0033] Determine the entity information of the initial abnormal address and the normal address data belonging to the road component type according to the component type;

[0034] Judge whether the entity information of the initial abnormal address and the normal address data belonging to the road component type is the same, and determine the target abnormal address according to the first judgment result; and / or,

[0035] Determine the contour area corresponding to all the normal address data in the aggregation result;

[0036] Judge whether the initial abnormal address is within the contour area, and determine the target abnormal address according to the second judgment result.

[0037] Optionally, according to a method for detecting address anomalies provided by the present invention, after verifying the initial abnormal address to obtain the target abnormal address, it further includes:

[0038] Use the structured address data redundant in the redundancy processing process as the address data to be processed;

[0039] Determine the third shortest distance between each piece of address data to be processed and the normal address data, and the fourth shortest distance from the target abnormal address;

[0040] Compare the third shortest distance and the fourth shortest distance, and determine whether the address data to be processed is an abnormal address according to the comparison result.

[0041] The present invention also provides an address anomaly detection device, including:

[0042] An acquisition module for acquiring a set of address data to be processed;

[0043] An aggregation module for structurally processing any address data in the set of address data according to address components to perform aggregation processing on each structured address data;

[0044] An identification module for identifying abnormal points in the structured address data in any aggregation result based on the structured address data within a preset range associated with any aggregation result to obtain an initial abnormal address;

[0045] A verification module for verifying the initial abnormal address to obtain a target abnormal address.

[0046] The present invention also provides a computer device, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the above-mentioned address abnormality detection method is implemented.

[0047] The present invention also provides one or more readable storage media storing computer-readable instructions. When the computer-readable instructions are executed by a processor, the above-mentioned address abnormality detection method is implemented.

[0048] The above-mentioned address abnormality detection method, device, equipment, and storage medium include: obtaining a set of address data to be processed; structurally processing any address data in the set of address data according to address components to perform aggregation processing on each structured address data; identifying abnormal points in the structured address data in any aggregation result based on the structured address data within a preset range associated with any aggregation result to obtain an initial abnormal address; verifying the initial abnormal address to obtain a target abnormal address. By structurally processing the address data and decomposing the address into clear components, the address can be parsed and matched more accurately. Furthermore, abnormal points are identified based on the address data within the preset range associated with the aggregation result, and multi-dimensional address information can reduce misjudgments caused by a single data source. Then, the initial abnormal address is verified to improve the accuracy of abnormal point identification. Description of the Drawings

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments of the present invention. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0050] Figure 1 It is a flowchart of an address abnormality detection method in an embodiment of the present invention;

[0051] Figure 2It is one of the schematic diagrams of the scene map provided by an embodiment of the present invention;

[0052] Figure 3 It is the second of the schematic diagrams of the scene map provided by an embodiment of the present invention;

[0053] Figure 4 It is the third of the schematic diagrams of the scene map provided by an embodiment of the present invention;

[0054] Figure 5 It is a schematic structural diagram of a door address abnormality detection device in an embodiment of the present invention;

[0055] Figure 6 It is a schematic diagram of a computer device in an embodiment of the present invention. Detailed implementation manners

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0057] The terms used in one or more embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present invention. The singular forms "a", "the" and "said" used in one or more embodiments of the present invention are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of the present invention refers to and includes any or all possible combinations of one or more of the associated listed items.

[0058] In one embodiment, specifically, as Figure 1 shown, Figure 1 It is a schematic flowchart of a door address abnormality detection method in an embodiment of the present invention. The embodiment of the present invention provides a door address abnormality detection method, including the following steps:

[0059] Step S11, obtaining a set of door address data to be processed;

[0060] It should be noted that the set of door address data includes multiple door address data. Among them, the door address data (Address Data) refers to address data containing geographical location information, and is usually used to describe the detailed address of a specific location.

[0061] Step S12, performing structured processing on any door address data in the set of door address data according to address components to perform aggregation processing on each structured address data;

[0062] Specifically, each door address data is decomposed into multiple address components. Among them, common address components include country, province / state, city / district, street, house number, building number, unit number, floor number, house number, etc., so as to obtain the address entity information in the door address data and the component type corresponding to the address entity information. Further, based on the address entity information and its component type, the door address data is structurally processed to obtain the structured address data corresponding to each door address data. For example: No. 9, Fuxing Road, Haidian District, Beijing. The structured information is as follows: Beijing, 2|Haidian District, 4|Fuxing Road, 8|No. 9, 21. Among them, the numbers used in the example represent entity categories, 2 represents city, 4 represents region, 8 represents road, and 21 represents road number. In other embodiments, before the structural processing, the door address data is normalized first. For example, digital categories such as house numbers are uniformly converted into Arabic numerals, etc., to ensure the consistency and normalization of the door address data. Further, the structured address data is aggregated. For example, algorithms such as Jaccard similarity, cosine similarity, K-means clustering, DBSCAN clustering, etc. are used for aggregation processing.

[0063] Step S13, according to the structured address data within a preset range associated with any aggregation result, identify abnormal points in the structured address data in any aggregation result to obtain initial abnormal door addresses;

[0064] Specifically, the following steps are performed for each structured address data in the aggregation result: Determine other structured address data within the preset range of this structured address data. Other structured address data refers to the remaining data in the aggregation result that does not include this structured address data. For example, query each structured address data within 50 meters near this structured address data, and determine the number of door addresses corresponding to each structured address data located within the preset range. Further, based on the second door address number associated with each structured address data, determine whether the structured address data is an abnormal point, so as to identify the initial abnormal door address.

[0065] In other embodiments, in order to improve the accuracy of abnormal address recognition, before the abnormal point recognition operation, the structured address data in each aggregation result can be preprocessed for redundancy. Specifically: score the structured address data according to data source, data confidence, update time, associated data volume, etc., so as to sort each structured address data according to the scores. According to the sorting result, perform deduplication processing on each structured address data. Optionally, in one embodiment, if the source of the structured address data already exists in the processed data, then filter the structured address data, and then continue to check the next structured address data; in another embodiment, if the source of a certain structured address data is considered to have low confidence (that is, less reliable or inaccurate), but there already exists address data that is the same as this address data and comes from a high-confidence data source, then filter the structured address data.

[0066] In other embodiments, different abnormal point recognition schemes can also be selected in combination with the number of addresses in the structured address data in the aggregation result. More specifically: determine the number of addresses corresponding to the structured address data in the aggregation result. If the number of addresses corresponding to the structured address data is less than the preset number threshold, then perform the above steps of determining other structured address data within the preset range of the structured address data, so as to implement the operation of recognizing abnormal points based on the address data within the preset range associated with the aggregation result; if the first number of addresses is greater than or equal to the preset number threshold, then cluster each structured address data in the target aggregation result; for example, use the DBSCAN clustering algorithm. Then, based on the clustering result, the initial abnormal address is recognized.

[0067] Step S14, verify the initial abnormal address to obtain the target abnormal address.

[0068] To further improve the accuracy of gateway anomaly detection, secondary verification is performed on the initially identified abnormal gateways. Specifically, in one embodiment, the shortest distance between the initial abnormal gateway and the normal gateway data in the aggregation result is calculated, where the normal gateway data refers to the gateway data that does not belong to the initial abnormal gateway among all structured address data. If the shortest distance exceeds the preset distance, it proves that the initial abnormal gateway is valid, and then the initial abnormal gateway is used as the target abnormal gateway. In addition, in other embodiments, information such as the number and / or proportion of the initial abnormal gateways corresponding to the aggregation result can also be combined for verification. In one embodiment, even if the initial abnormal gateway is very close to the normal gateway data, but if there is a situation of crossing roads between the initial abnormal gateway and the normal gateway data in the aggregation result, the initial abnormal gateway is still determined to be valid. Optionally, according to the entity information of the road component types in the initial abnormal gateway and the normal gateway data, it is determined whether there is a road crossing relationship between the initial abnormal gateway and the normal gateway data. If there is a road crossing relationship, it proves that the initial abnormal gateway is valid, and the initial abnormal gateway is used as the target abnormal gateway. In one embodiment, even if the initial abnormal gateway is very close to the normal gateway data, but if the initial abnormal gateway is not located in the contour area formed by each normal gateway data in the aggregation result, the initial abnormal gateway is still determined to be valid. The specific verification process will be elaborated in the following embodiments and will not be repeated here.

[0069] The embodiment of the present invention adopts the above solution, including: obtaining a set of gateway data to be processed; performing structured processing on any gateway data in the set of gateway data according to address components to perform aggregation processing on each structured address data; identifying abnormal points in the structured address data in any aggregation result according to the structured address data within a preset range associated with any aggregation result to obtain an initial abnormal gateway; verifying the initial abnormal gateway to obtain a target abnormal gateway. By performing structured processing on the gateway data and decomposing the address into clear components, the address can be parsed and matched more accurately. Furthermore, abnormal points are identified based on the gateway data within a preset range associated with the aggregation result, and multi-dimensional gateway information can reduce misjudgment caused by a single data source. Then, the initial abnormal gateway is verified to improve the accuracy of abnormal point identification.

[0070] In one embodiment of the present invention, performing structured processing on any gateway data in the set of gateway data according to address components includes:

[0071] For any of the gateway data: disassembling the gateway data according to address components to obtain the address entity information in the gateway data and the component type corresponding to the address entity information; based on the address entity information and the component type, performing normalization processing on the gateway data and performing structured processing on the normalized gateway data.

[0072] Specifically, the following steps are performed for any of the above-mentioned door address data: disassemble the door address data according to the address components to analyze and identify each address component in the door address data (i.e., the address entity information in this embodiment), and determine the component type corresponding to each component; for example, for the door address data: No. 1, Zhongguancun Street, Haidian District, Beijing, the identified address entity information and the corresponding component types are as follows: "Beijing" corresponds to "city", "Haidian District" corresponds to "district", "Zhongguancun Street" corresponds to "road", and "No. 1" corresponds to "road number". Further, based on the address entity information and the component type, the door address data is normalized. For example, the normalization process includes converting uppercase English characters to lowercase, converting traditional Chinese characters to simplified Chinese, converting Chinese numerals to Arabic numerals, and aligning units. Then, the structured processing is performed on the normalized door address data. It can be understood that the complete address data is split according to spatial geographical entities, and information such as administrative divisions, roads, and house numbers in the address is identified.

[0073] Through the above solution, the embodiment of the present invention realizes the structured processing of each door address data according to the address components, ensuring the consistency and standardization of the address data, and reducing data redundancy and errors.

[0074] In an embodiment of the present invention, the aggregation processing of each structured address data includes:

[0075] Correct each address entity information and its corresponding component type; perform filtering processing on the corrected structured address data; perform aggregation processing on the filtered structured address data.

[0076] It should be noted that when there are errors in the components split by address structuring, the category and content of the components are corrected. Specifically, each address entity information and its corresponding component type are corrected. The component correction is mainly inferred by using the coherence between the component contents and the coherence of the categories. For example, the structured address data is as follows: Shanghai, 2|Huangpu District, 4|Huaihai Middle Road, 1000 Lane, 9|Building 9, 22; among them, Huaihai Middle Road, 1000 Lane is of the road type, and the label should be changed to 8. The corrected structured address data is as follows: Shanghai, 2|Huangpu District, 4|Huaihai Middle Road, 8|1000 Lane, 8|Building 9, 22.

[0077] Further, filter the corrected structured address data. For example, invalid data to be filtered is preset. The invalid data refers to invalid components in the structured address data, such as single digits (20) in the road component, English letters (za), and those with only "lane" but no "road" in Shanghai data. Optionally, the filtering operation is implemented through a preset regular expression. Keywords to be filtered can also be preset. For example, filter according to a manually summarized keyword dictionary. If the structured address data hits the preset keywords, then filter the structured address data. In addition, address suffixes to be filtered can be preset, such as "University", "Development Zone", etc. If the structured address data has such suffixes, then filter the structured address data. Further, perform aggregation processing on the filtered structured address data.

[0078] Through the above solution, the embodiment of the present invention realizes improving the quality and usability of data by correcting, filtering, and aggregating address data, and further improves the accuracy of abnormal door address detection.

[0079] In an embodiment of the present invention, for the structured address data within a preset range associated with any aggregation result, identify abnormal points in the structured address data in any aggregation result to obtain an initial abnormal door address, including:

[0080] For any one of the aggregation results: perform redundancy processing on the structured address data in the aggregation result to obtain a target aggregation result; determine the number of the first door addresses corresponding to the target aggregation result; if the number of the first door addresses is less than a preset number threshold, then based on each structured address data in the target aggregation result, determine the number of the second door addresses corresponding to the structured address data located within the preset range; based on the number of the second door addresses associated with each structured address data, determine the initial abnormal door address.

[0081] It should be noted that, in order to reduce the misleading of subsequent aggregation caused by data redundancy, in this embodiment, it is necessary to perform redundancy filtering processing on the structured address data in the aggregation result. Specifically, for any one of the aggregation results, the following steps are executed:

[0082] Step 1, score the structured address data according to data source, data confidence level, update time, associated data volume, etc.

[0083] Step 2: Use the H3 spatial index to convert the geographical coordinates of each structured address data to obtain the H3 index. It should be noted that H3 is a hexagonal grid system that can convert geographical coordinates into the index of a hexagonal grid. Data within the same H3 index are regarded as data in the same group. Optionally, select each structured address data corresponding to a preset H3 index from each group of data. For example, select data with a smaller H3 index level for subsequent processing. Selecting a smaller level can ensure the deletion of completely similar coordinates while not losing the distribution characteristics of the data.

[0084] Step 3: Sort each selected structured address data according to the scores calculated in Step 1.

[0085] Step 4: According to the sorting result, perform deduplication processing on each structured address data. For example, if the source of a structured address data already exists in the processed data, then filter out this structured address data and continue to check the next structured address data; if the source of a certain structured address data is considered to have low confidence, but there already exists an address data with the same content as this address data and from a high-confidence data source, then filter out this structured address data.

[0086] Perform the above operations on all data until all structured address data are traversed, so as to obtain the final target aggregation result.

[0087] Furthermore, determine the number of first door addresses corresponding to the target aggregation result; then compare the number of first door addresses with a preset quantity threshold, where the preset quantity threshold can be set according to the actual situation. For example, the preset quantity threshold is set to 3. If the number of first door addresses is less than the preset quantity threshold, then for each structured address data in the target aggregation result, perform the following steps: determine each structured address data within a preset range of the structured address data. For example, query the address data within 50 meters of the structured address data, and determine the number of second door addresses corresponding to each structured address data located within the preset range.

[0088] Even further, based on the number of second door addresses associated with each structured address data, determine whether the structured address data is an outlier, so as to identify the initial abnormal door addresses. For example, set a quantity threshold. When the number of second door addresses is less than the set quantity threshold, then determine that this structured address data is an initial abnormal door address. For example, at No. 100 Yishan Road, Xuhui District, Shanghai, after redundancy removal, there are 2 structured address data (A and B); recall the address data within 50 meters of the coordinates of point A and point B respectively. Among them, there are 20 address data containing Yishan Road near point A, and only 1 address data containing Yishan Road near point B; then it is proved that the data at point A is normal and the data at point B is abnormal.

[0089] In addition, in an embodiment of the present invention, after determining the number of first addresses corresponding to the target aggregation result, the following steps are further included:

[0090] If the number of first addresses is greater than or equal to a preset number threshold, cluster each structured address data in the target aggregation result; and determine the initial abnormal address according to the clustering result.

[0091] Specifically, if the number of first addresses is greater than or equal to a preset number threshold, cluster the structured address data in the target aggregation result. For example, the DBSCAN clustering algorithm is used. It should be noted that DBSCAN is a density-based clustering algorithm that can well measure the aggregation between coordinates and is more suitable for the current coordinate error correction scenario. DBSCAN has two core parameters, the minimum distance e for forming clusters and the minimum number of samples n for forming clusters. For the minimum distance e for forming clusters, a set of default values is specified from small to large, and the parameters are used in turn for clustering. If the clustering stop condition is met, the loop ends and the abnormal value is returned. The process of aggregation error correction is as follows:

[0092] Step 1: Specify a set of default parameters E: [e1, e2, e3], such as [50, 100, 200];

[0093] Step 2: Use the DBSCAN algorithm to cluster each structured address data in the target aggregation result. Optionally, the group labels that meet the clustering are marked as non-negative integers, and the group labels that do not form clusters are marked as -1;

[0094] Step 3: Determine whether the clustering is completed. For the group that does not form a cluster (marked as an abnormal address), end the clustering and return the abnormal address; otherwise, go to Step 1 to traverse the default parameters E until the clustering meets the stop condition.

[0095] Through the above solution, the embodiment of the present invention realizes redundant processing of the structured address data in the aggregation result, effectively improving the accuracy of abnormal point recognition. In addition, by combining the address data within the preset range associated with the aggregation result for abnormal point recognition, the probability of misrecognition caused by a single data source can be reduced.

[0096] In an embodiment of the present invention, verifying the initial abnormal address to obtain the target abnormal address includes:

[0097] For any initial abnormal address in an aggregation result:

[0098] Determine the first shortest distance between the initial abnormal address and the normal address data in the aggregation result, and count the number of abnormal points and the proportion of the initial abnormal address in the aggregation result, where the normal address data refers to the address data in all structured address data that does not belong to the initial abnormal address; according to the first shortest distance, the number of abnormal points and the proportion, verify the initial abnormal address to obtain the target abnormal address; and / or,

[0099] Determine the distances between the structured address data in the aggregation result, and calculate the standard deviation based on the distances between the structured address data; according to the standard deviation and the second shortest distance between the initial abnormal address and the normal address data, verify the initial abnormal address to obtain the target abnormal address; and / or,

[0100] Determine the entity information of the initial abnormal address and the normal address data belonging to the road component type according to the component type; judge whether the entity information of the initial abnormal address and the normal address data belonging to the road component type is the same, and determine the target abnormal address according to the first judgment result; and / or,

[0101] Determine the contour area corresponding to all the normal address data in the aggregation result; judge whether the initial abnormal address is within the contour area, and determine the target abnormal address according to the second judgment result.

[0102] It should be noted that in order to improve the accuracy of data error correction, the initially identified initial abnormal address is verified twice. Specifically, for any initial abnormal address in the aggregation result, the following steps are performed:

[0103] In an embodiment, determine the first shortest distance between the initial abnormal address and the normal address data in the aggregation result, and count the number of abnormal points and the proportion of the initial abnormal address in the aggregation result, where the proportion is the proportion of the number of abnormal points of the initial abnormal address and the total number of all structured address data in the aggregation result. Compare the first shortest distance with a preset distance threshold, compare the number of abnormal points with a preset abnormal point number threshold, and compare the proportion with a preset proportion threshold. If the first shortest distance is greater than the preset distance threshold, the number of abnormal points is less than the preset abnormal point number threshold, and the proportion is less than the preset proportion threshold, it is determined that the initial abnormal address is valid, and the initial abnormal address is used as the target abnormal address. For example, if the first shortest distance is greater than 5 kilometers, the number of abnormal points is less than 2, and the abnormal point proportion is less than 0.1, it is determined that the initial abnormal address is valid.

[0104] In one embodiment, the distances between the respective structured gateway data in the calculated aggregation result are calculated, and then, based on the distances between the respective structured gateway data, the standard deviation is calculated. In addition, the second shortest distance between the initial abnormal gateway and the normal gateway data is calculated; further, if the second shortest distance is greater than a preset multiple of the standard deviation, it is determined that the initial abnormal gateway is valid, and the initial abnormal gateway is used as the target abnormal gateway. If the second shortest distance is not greater than the preset multiple of the standard deviation, it is determined that the initial abnormal gateway is invalid, that is, it is determined that the initial abnormal gateway is not an abnormal point. The preset multiple can be set according to the actual situation. Optionally, it is set to 3. It can be understood that with reference to Figure 2 , Figure 2 is one of the schematic diagrams of the scene map provided by an embodiment of the present invention. For the judgment of the short-distance case, the normal point clusters are relatively close, the abnormal points are relatively sparse, the standard deviation of the central cluster is 8m, and if the distance from the outer points to the central cluster exceeds 24m, it is considered that the initial abnormal gateway is valid.

[0105] In one embodiment, the respective normal gateway data within a preset range from the initial abnormal gateway are determined, and then, according to the component type, the entity information of the initial abnormal gateway and each normal gateway data belonging to the road component type is determined; for example, the structured address data is as follows: Shanghai, 2|Huangpu District, 4|Huaihai Middle Road, 9|Building 9, 22; Huaihai Middle Road is the entity information of the road component type. Further, multiple normal gateway data with the same entity information of the road component type are selected, and it is judged whether the entity information of the initial abnormal gateway and the selected multiple normal gateway data belonging to the road component type is the same. If the first judgment result is that the entity information is different, it is proved that the initial abnormal gateway and the normal gateway data do not belong to the same road, and it is determined that the initial abnormal gateway is valid, and the initial abnormal gateway is used as the target abnormal gateway. For example, with reference to Figure 3 , Figure 3 is the second schematic diagram of the scene map provided by an embodiment of the present invention. For the judgment of the short-distance case, if the initial abnormal gateway is very close to the normal gateway data (for example, within 50m), but there is a cross-road relationship between the initial abnormal gateway and 3 normal gateway data, it is considered that the initial abnormal gateway is valid.

[0106] In one embodiment, the contour area corresponding to all the normal gateway data in the aggregation result is determined; it is judged whether the initial abnormal gateway is within the contour area. If the second judgment result is that the initial abnormal gateway is not within the contour area, it is determined that the initial abnormal gateway is valid, and the initial abnormal gateway is used as the target abnormal gateway. If the second judgment result is that the initial abnormal gateway is within the contour area, it is determined that the initial abnormal gateway is invalid, that is, the initial abnormal gateway is a normal gateway. For example, with reference to Figure 4 , Figure 4 is the third schematic diagram of the scene map provided by an embodiment of the present invention.Figure 4 There is a certain distance between the initial abnormal gate address below and the cluster above, but they are within the same AOI contour area (the square contour of the building), then it is considered that the initial abnormal gate address is invalid, that is, the initial abnormal gate address is not an abnormal point.

[0107] Through the above solution, the embodiment of the present invention realizes the secondary verification of the initially identified abnormal gate address after the abnormal point is identified from the structured address data, effectively improving the accuracy of the gate address abnormality detection.

[0108] In an embodiment of the present invention, after verifying the initial abnormal gate address to obtain the target abnormal gate address, it further includes:

[0109] Taking the structured address data redundant during the redundancy process as the address data to be processed; determining the third shortest distance between each piece of the address data to be processed and the normal gate address data, and the fourth shortest distance from the target abnormal gate address; comparing the third shortest distance and the fourth shortest distance, so as to determine whether the address data to be processed is an abnormal gate address according to the comparison result.

[0110] It should be noted that during the above redundancy process, some address data are redundantly filtered, resulting in some address data not participating in the abnormal point identification process. Therefore, it is necessary to re-mark the address data with unknown tags.

[0111] Specifically, taking the structured address data redundant during the redundancy process as the address data to be processed, the following steps are performed for each piece of address data to be processed: calculating the third shortest distance between the address data to be processed and the normal gate address data, and calculating the fourth shortest distance between the address data to be processed and the target abnormal gate address. Further, comparing the third shortest distance and the fourth shortest distance, if the comparison result is that the third shortest distance is greater than the fourth shortest distance, it is determined that the address data to be processed is not an abnormal gate address. If the comparison result is that the third shortest distance is not greater than the fourth shortest distance, it is determined that the address data to be processed is an abnormal gate address.

[0112] Through the above solution, the embodiment of the present invention realizes determining whether the address data to be processed with unknown tags belongs to an abnormal gate address according to the shortest distance between the address data to be processed with unknown tags and the normal gate address data, and the shortest distance from the target abnormal gate address, improving the efficiency of abnormal detection of the address data with unknown tags.

[0113] It should be understood that the magnitudes of the sequence numbers of the above steps in the embodiments do not mean the order of execution. The order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0114] In one embodiment, a device for detecting abnormal door addresses is provided. The device for detecting abnormal door addresses corresponds one-to-one with the method for detecting abnormal door addresses in the above embodiment. As Figure 5 shown, Figure 5 FIG. 4 is a schematic structural diagram of a device for detecting abnormal door addresses according to an embodiment of the present invention. The device for detecting abnormal door addresses includes:

[0115] An acquisition module 21, configured to acquire a set of door address data to be processed;

[0116] An aggregation module 22, configured to perform structured processing on any door address data in the set of door address data according to address components, so as to perform aggregation processing on each structured address data;

[0117] An identification module 23, configured to identify abnormal points in the structured address data in any aggregation result according to the structured address data within a preset range associated with any aggregation result, so as to obtain an initial abnormal door address;

[0118] A verification module 24, configured to verify the initial abnormal door address to obtain a target abnormal door address.

[0119] The aggregation module 22 is further configured to:

[0120] For any of the door address data:

[0121] Disassemble the door address data according to address components to obtain address entity information in the door address data and a component type corresponding to the address entity information;

[0122] Based on the address entity information and the component type, perform normalization processing on the door address data, and perform structured processing on the normalized door address data.

[0123] The aggregation module 22 is further configured to:

[0124] Correct each address entity information and its corresponding component type;

[0125] Perform filtering processing on each corrected structured address data;

[0126] Perform aggregation processing on each filtered structured address data.

[0127] The identification module 23 is further configured to:

[0128] For any of the aggregation results:

[0129] Perform redundancy processing on the structured address data in the aggregation result to obtain a target aggregation result;

[0130] Determine a first door address quantity corresponding to the target aggregation result;

[0131] If the number of first door addresses is less than the preset quantity threshold, based on each structured address data in the target aggregation result, determine the number of second door addresses corresponding to the structured address data located within the preset range;

[0132] Based on the number of second door addresses associated with each structured address data, determine the initial abnormal door address.

[0133] The door address anomaly detection device further includes:

[0134] If the number of first door addresses is greater than or equal to the preset quantity threshold, perform clustering on each structured address data in the target aggregation result;

[0135] Based on the clustering result, determine the initial abnormal door address.

[0136] The door address anomaly detection device further includes:

[0137] Use the structured address data redundant during the redundancy processing as the to-be-processed address data;

[0138] Determine the third shortest distance between each to-be-processed address data and the normal door address data, and the fourth shortest distance from the target abnormal door address;

[0139] Compare the third shortest distance and the fourth shortest distance to determine whether the to-be-processed address data is an abnormal door address according to the comparison result.

[0140] The verification module 24 is further configured to:

[0141] For any initial abnormal door address in the aggregation result:

[0142] Determine the first shortest distance between the initial abnormal door address and the normal door address data in the aggregation result, and count the number of abnormal points and the proportion of the number of the initial abnormal door address in the aggregation result, where the normal door address data refers to the door address data that does not belong to the initial abnormal door addresses among all structured address data;

[0143] Verify the initial abnormal door address according to the first shortest distance, the number of abnormal points and the proportion of the number to obtain the target abnormal door address; and / or,

[0144] Determine the distances between each structured door address data in the aggregation result, and calculate the standard deviation according to the distances between each structured door address data;

[0145] Verify the initial abnormal door address according to the standard deviation and the second shortest distance between the initial abnormal door address and the normal door address data to obtain the target abnormal door address; and / or,

[0146] Determine the entity information of the road component type to which the initial abnormal address and the normal address data belong according to the component type;

[0147] Judge whether the entity information of the road component type to which the initial abnormal address and the normal address data belong is the same, and determine the target abnormal address according to the first judgment result; and / or,

[0148] Determine the contour areas corresponding to all normal address data in the aggregation result;

[0149] Judge whether the initial abnormal address is within the contour area, and determine the target abnormal address according to the second judgment result.

[0150] For the specific definition of the address anomaly detection device, reference can be made to the definition of the address anomaly detection method in the foregoing text, which will not be elaborated here. Each module in the above address anomaly detection device can be implemented in whole or in part by software, hardware and their combination. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory of the computer device in the form of software, so as to facilitate the processor to call and execute the operations corresponding to the above modules.

[0151] In one embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 6 shown Figure 6 is a schematic diagram of a computer device in an embodiment of the present invention. The computer device includes a processor, a memory, a network interface and a database connected through a device bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium and an internal memory. The readable storage medium stores an operating device, computer-readable instructions and a database. The internal memory provides an environment for the operation of the operating device and computer-readable instructions in the readable storage medium. The database of the computer device is used to store the data involved in the address anomaly detection method. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer-readable instructions are executed by the processor, an address anomaly detection method is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0152] In one embodiment, a computer device is provided. The computer device can be a terminal device, and its internal structure diagram can be as Figure 6As shown. The computer device includes a processor, a memory, and a network interface connected by a device bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium. The readable storage medium stores computer-readable instructions. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer-readable instructions are executed by the processor, a gateway address anomaly detection method is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0153] In one embodiment, a computer device is provided, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the gateway address anomaly detection method as described above are implemented.

[0154] In one embodiment, a readable storage medium is provided. The readable storage medium stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps of the gateway address anomaly detection method as described above are implemented. Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through computer-readable instructions. The computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When the computer-readable instructions are executed, they can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in this application can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or an external cache. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0155] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is used as an example. In actual applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0156] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for detecting anomaly of a door address, characterized in that: include: Obtaining a set of door address data to be processed; Performing structured processing on any gate address data in the gate address data set according to the address components, so as to perform aggregation processing on each structured address data; According to the structured address data within a preset range associated with any aggregation result, an abnormal point identification is performed on the structured address data in any aggregation result to obtain an initial abnormal gate address; The initial abnormal gate address is verified to obtain a target abnormal gate address.

2. The door address anomaly detection method according to claim 1, characterized in that: The step of performing structural processing on any gate address data in the gate address data set according to the address component comprises: For any of the above mentioned portal address data: Decomposing the portal address data according to the address components to obtain the address entity information in the portal address data and the component type corresponding to the address entity information; Based on the address entity information and the component type, the gate address data is normalized, and the normalized gate address data is structured.

3. The door address anomaly detection method according to claim 1, characterized in that: The structured address data includes address entity information and component types corresponding to the address entity information; The aggregating processing of each structured address data includes: Modifying each of the address entity information and its corresponding component type; Filtering the corrected structured address data; Aggregate the structured address data after filtering.

4. The method for detecting anomaly of a door address according to claim 1, characterized in that: The step of identifying abnormal points of the structured address data in any aggregation result according to the structured address data in a preset range associated with any aggregation result to obtain an initial abnormal gate address includes: For any of the above aggregation results: Performing redundancy processing on the structured address data in the aggregation result to obtain a target aggregation result; Determine the number of first gate addresses corresponding to the target aggregation result; If the number of first gate addresses is less than a preset number threshold, determining the number of second gate addresses corresponding to the structured address data within a preset range based on each structured address data in the target aggregation result; The initial abnormal gate address is determined based on the number of second gate addresses associated with each structured address data.

5. The door address anomaly detection method according to claim 4, characterized in that: After determining the number of first gate addresses corresponding to the target aggregation result, the method further includes: If the number of first gate addresses is greater than or equal to a preset number threshold, clustering each structured address data in the target aggregation result; According to the clustering result, the initial abnormal gate address is determined.

6. The method for detecting anomaly of a door address according to claim 1, characterized in that: The checking of the initial abnormal gate address to obtain the target abnormal gate address includes: For the initial abnormal gate address in any aggregation result: Determine the first shortest distance between the initial abnormal gate address and the normal gate address data in the aggregation result, and count the number and proportion of abnormal points of the initial abnormal gate address in the aggregation result, wherein the normal gate address data refers to the gate address data that does not belong to the initial abnormal gate address in all structured address data; According to the first shortest distance, the number of abnormal points and the number ratio, the initial abnormal gate address is verified to obtain the target abnormal gate address; and / or, Determine the distance between each structured portal address data in the aggregation result, so as to calculate the standard deviation according to the distance between each structured portal address data; According to the standard deviation and the second shortest distance between the initial abnormal gate address and the normal gate address data, the initial abnormal gate address is verified to obtain the target abnormal gate address; and / or, According to the component type, determining that the initial abnormal gate address and the normal gate address data belong to entity information of the road component type; Determine whether the initial abnormal gate address and the normal gate address data belong to the same entity information of the road component type, so as to determine the target abnormal gate address according to the first judgment result; and / or, Determine the contour area corresponding to all normal gate address data in the aggregation result; It is determined whether the initial abnormal gate address is in the contour area, and the target abnormal gate address is determined according to the second determination result.

7. The door address anomaly detection method according to claim 4, characterized in that: After verifying the initial abnormal gate address and obtaining the target abnormal gate address, the method further includes: Using the redundant structured address data in the redundancy processing process as the address data to be processed; Determine the third shortest distance between each of the to-be-processed address data and the normal gate address data, and the fourth shortest distance to the target abnormal gate address; The third shortest distance is compared with the fourth shortest distance to determine whether the address data to be processed is an abnormal gate address according to the comparison result.

8. A door address anomaly detection device, characterized in that: include: An acquisition module, used to acquire a set of door address data to be processed; An aggregation module, used for performing structured processing on any gate address data in the gate address data set according to the address components, so as to perform aggregation processing on each structured address data; An identification module, used to identify abnormal points of structured address data in any aggregation result according to structured address data within a preset range associated with any aggregation result, and obtain an initial abnormal gate address; The verification module is used to verify the initial abnormal gate address to obtain a target abnormal gate address.

9. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executed on the processor, characterized in that: When the processor executes the computer-readable instructions, the gate address anomaly detection method as described in any one of claims 1 to 7 is implemented.

10. A readable storage medium having computer readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by a processor, the gate address anomaly detection method as described in any one of claims 1 to 7 is implemented.

Images