Poi data fusion method, electronic device, and storage medium

Abstract

Embodiments of the present application relate to the technical field of electronic map, and disclose a POI data fusion method, an electronic device and a computer readable storage medium. The POI data fusion method comprises: obtaining a first data set and a second data set to be fused, wherein the first data set and the second data set each comprise a plurality of POI data; searching for POI data in the first data set and the second data set located in a same geohash nine-square grid according to geohash encodings of the POI data in the first data set and the second data set; performing a Cartesian product on the POI data in the first data set and the second data set located in the same geohash nine-square grid to form different data groups; calculating a similarity of two POI data in each data group, and determining whether the similarity exceeds a threshold value. The POI data fusion method, the electronic device and the computer readable storage medium provided by the embodiments of the present application can improve the fusion efficiency of POI data.

Description

Technical Field

[0001] This application relates to the field of electronic map technology, and in particular to a POI data fusion method, electronic device, and storage medium. Background Technology

[0002] With the rapid development of internet technology, users' demand for data is increasing. In geographic information data, POI (point of interest) refers to point-like data on internet electronic maps. It can be a building, a tourist attraction, a school, a bus stop, an ATM, a shop, etc. Generally, a point-like data item typically includes four attributes: name, address, coordinates, and category. It may also include information such as phone number and reviews.

[0003] To enhance the richness of electronic map data, it is necessary to continuously update the information contained in POI data, enabling newly generated POI data to be integrated with existing POI data, thereby providing users with comprehensive and detailed information services. However, with the explosive growth of POI data, improving the efficiency of POI data integration is a crucial issue. Summary of the Invention

[0004] The purpose of this application is to provide a POI data fusion method, electronic device, and storage medium that can improve the fusion efficiency of POI data.

[0005] To address the aforementioned technical problems, embodiments of this application provide a POI data fusion method, which includes:

[0006] Obtain the first and second datasets to be merged. Both the first and second datasets contain multiple POI data.

[0007] Based on the geohash encoding of each POI in the first and second datasets, find the POI data that are located in the same geohash 3x3 grid in the first and second datasets;

[0008] Perform a Cartesian product on the POI data in the first and second datasets that are located in the same geohash 3x3 grid to form different data groups;

[0009] Calculate the similarity between two POI data in each data group and determine whether the similarity exceeds a threshold. If the similarity between two POI data in a certain data group exceeds the threshold, then merge the two POI data in that data group to form fused POI data.

[0010] The embodiments of this application also provide an electronic device, which includes a processor and a memory communicatively connected to the processor. The memory stores instructions that can be executed by the processor, and the instructions are executed by the processor to enable the processor to perform the above-described POI data fusion method.

[0011] The embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described POI data fusion method.

[0012] The POI data fusion method, electronic device, and computer-readable storage medium provided in this application, when fusing POI data, first use the geohash encoding of each POI data to find POI data located within the same geohash 3x3 grid. Then, similarity calculation is performed on any two POI data located within the same geohash 3x3 grid to fuse POI data belonging to the same entity object. This avoids the computational redundancy caused by fully combining all POI data when calculating similarity, thus improving the efficiency of POI data fusion. Attached Figure Description

[0013] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0014] Figure 1 This is a flowchart of a POI data fusion method provided in some embodiments of this application;

[0015] Figure 2 This is a flowchart of a POI data fusion method provided in other embodiments of this application;

[0016] Figure 3 These are schematic diagrams of the structure of electronic devices provided in some embodiments of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been presented in the various embodiments of this application to enable readers to better understand this application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solutions claimed in this application can be implemented. The division of the various embodiments below is for the convenience of description and should not constitute any limitation on the specific implementation of this application. The various embodiments can be combined with and referenced by each other without contradiction.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0019] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0020] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of this application according to the specific circumstances.

[0021] With the rapid development of mobile communication technology, people face massive amounts of data every day, with new data constantly being generated. Therefore, data fusion is becoming increasingly important. Through data fusion, data from different data sources can be combined into new data according to a certain standard. Information is then delivered to users based on this new data to meet their needs. As internet content grows rapidly, users demand more comprehensive and accurate data, which necessitates improved efficiency in data fusion.

[0022] Points of Interest (POI) data is a type of data that users in geographic information systems frequently need, and it can be used to represent various different entities. To provide users with high-quality services, POI data needs to be continuously integrated and updated to offer road condition information and detailed descriptions of surrounding buildings, facilitating users' access to comprehensive information about their desired destinations for selection and planning.

[0023] POI data comes from a wide range of sources and can be obtained instantly from common electronic map software. However, POI data collected from multiple data sources will have some differences. To quickly merge and adjust the information contained in POI data from different sources into a more complete whole, some embodiments of this application provide a POI data fusion method. When fusing POI data from different data sources, the method queries POI data located within the same geohash 3x3 grid based on the geohash encoding of each POI data, and then compares the similarity between two different POI data within the same geohash 3x3 grid. This avoids the inefficiency caused by performing a full match between two POI datasets, i.e., calculating the similarity for every two POI data in both datasets.

[0024] It's important to note that geohash is an address encoding method. In geohash encoding, the Earth can be viewed as a two-dimensional plane, which can be recursively decomposed into smaller sub-blocks. Each sub-block shares the same code within a certain latitude and longitude range. Geohash encoding maps the two-dimensional coordinates of a point on a map to a string code. Each code represents a specific rectangular area, and all points within that area share the same code. The longer the string code, the higher the precision, and the smaller the corresponding rectangular area on the map. Specifically, the geohash encoding rule divides an area into a nine-square grid; the latitude and longitude string code determines the position of a point within that grid.

[0025] The following is combined with Figure 1 This application describes a POI data fusion method provided in some embodiments.

[0026] like Figure 1 As shown, some embodiments of this application provide a POI data fusion method, which includes the following steps:

[0027] Step S100: Obtain the first dataset and the second dataset to be merged. Both the first dataset and the second dataset include at least one POI data.

[0028] Specifically, both the first and second datasets are collections containing Points of Interest (POI) data, which can be obtained from different data providers. POI data can include information such as address, name, and latitude / longitude that represents a specific entity on an electronic map. Through POI data, users can obtain specific information about a specific entity. One of the first and second datasets contains basic POI data, while the other contains external POI data. In other words, one of the first and second datasets can be considered the POI dataset requiring content updates, while the other can be considered the POI dataset obtained externally for information fusion.

[0029] In practice, due to the large volume of POI data, different POI data are typically grouped into a single set table during the fusion process. This set table contains numerous different POI data points. These POIs may be located in the same or nearby areas, or in two geographically distant areas. Furthermore, some POI data points have high similarity; although there are some differences in specific content, they belong to the same entity on the electronic map. Therefore, these POI data points need to be merged to reduce data redundancy and update the data content. However, for datasets containing numerous POIs, arbitrarily selecting two POIs for comparison would significantly increase the workload and hinder the efficiency of POI data fusion.

[0030] Therefore, in some embodiments of the POI data fusion method provided in this application, after obtaining two different POI datasets, the regional information of the location of each POI data is determined. This is to identify POI data located within the same geographical region, and then the similarity of these POI data is calculated.

[0031] Step S200: Based on the geohash encoding of each POI data in the first dataset and the second dataset, find the POI data in the first dataset and the second dataset that are located in the same geohash 3x3 grid.

[0032] Geohash encoding is the latitude and longitude address encoding of POI data. Each POI corresponds to a specific latitude and longitude data information, and the geohash encoding method can convert the latitude and longitude data of these POI data into a string. The character length of the string can be determined according to the positional precision required when merging POI data. The geohash nine-square grid is a grid unfolded from the Earth. Figure 2 The 3D plane is divided into multiple different location regions, and each specific geohash 3x3 grid corresponds to a specific location. Figure 2A specific location region on a 3D plane. Furthermore, each specific geohash 3x3 grid corresponds to the same geohash encoded information, i.e., the same string.

[0033] In this way, by using the geohash encoding information of each POI data, the location information can be determined. Figure 2 POI data within the same location area on the dimensional plane are categorized, thereby pre-determining the POI data that need to be calculated for similarity based on the location information of each POI data.

[0034] Step S300: Perform a Cartesian product on the POI data located in the same geohash 3x3 grid in the first and second datasets to form different data groups.

[0035] After finding the POI data within the same geohash 3x3 grid, these POI data are paired up to calculate the similarity between them.

[0036] For example, the first dataset includes three Points of Interest (POIs) A, B, and C, while the second dataset includes two POIs D and E. By using geohash encoding to find that POIs A, B, and D are located within the same geohash 3x3 grid, we can perform a Cartesian product (pairwise combination) on these three POIs. This results in three data groups: a first group containing A and B, a second group containing B and D, and a third group containing A and D. After forming these data groups, we need to calculate the similarity between the two POIs in each group.

[0037] Step S400: Calculate the similarity between two POI data in each data group and determine whether the similarity exceeds the threshold. If the similarity between two POI data in a certain data group exceeds the threshold, then merge the two POI data in that data group to form POI fused data.

[0038] Specifically, similarity can be determined from two dimensions: text and position. Text similarity matching methods can employ the Lewisonstein distance algorithm, the Jaro distance algorithm, and the Jaccard similarity method. The Lewisonstein distance algorithm calculates the edit distance between two strings and can be used to measure the similarity between them. The Jaro distance algorithm can also be used to calculate the similarity between two strings using a certain matching coefficient. The Jaccard similarity method compares the similarity between two strings based on the statistical probability of their differences and dispersion. Positional similarity can also be measured using certain calculation methods to measure the similarity between two Points of Interest (POI) datasets.

[0039] Furthermore, for each similarity calculation method, a certain threshold can be set to determine whether two POI data sets match. If the similarity calculation result of two POI data sets exceeds the threshold, then it can be determined that the two POI data sets belong to the same entity object in terms of information content. Therefore, the two POI data sets can be merged in terms of content, thereby forming more complete POI fused data.

[0040] It's important to note that the merging of two POI datasets can be performed according to pre-defined weights. When the similarity between two POI datasets exceeds a threshold and merging is necessary, the content to be retained in the merged dataset is selected based on the priority of the two POI datasets. The weights can be set on a dataset-by-dataset basis or by setting the weights for each field type within the POI dataset.

[0041] The POI data fusion method provided in some embodiments of this application pre-identifies POI data within the same geohash 3x3 grid by using the geohash encoding of each POI data. Then, it calculates the similarity between any two POI data within the same geohash 3x3 grid to fuse POI data belonging to the same entity object. This avoids the computational redundancy caused by fully combining all POI data when calculating similarity, thus improving the efficiency of POI data fusion.

[0042] In some embodiments of this application, step S100, obtaining the first dataset and the second dataset to be fused, may include the following steps:

[0043] Step S110: Obtain POI data from different data sources and store them in the Hive data warehouse.

[0044] Different data sources can be common electronic map software or location recommendation software, or data located during actual location exploration. These POI data from different sources differ in information content and emphasis; therefore, they can be categorized according to the completeness of their content. This allows for the subsequent assignment of weights for different POI data during the fusion process. For example... Figure 2 As shown, the acquired POI data can be stored in a Hive data warehouse. Hive is a data warehouse tool based on a distributed platform that can be used for data extraction, transformation, and loading. Using Hive data warehouses can help improve the efficiency of data fusion.

[0045] Step S120: Select POI data with more field types to form the first dataset, and POI data with fewer field types to form the second dataset.

[0046] The field types include at least one of name, address, and latitude / longitude.

[0047] In other words, POI data with more complete content information is placed in the first dataset, while POI data with less complete content information is stored in the second dataset. This allows the first dataset to be assigned a higher weight than the second dataset during the POI data fusion process.

[0048] In some embodiments, the first dataset has a higher priority than the second dataset.

[0049] It should be noted that when field type information is repeated, some field type information can be removed, and only one field type information can be selectively retained. For example, in a POI dataset, if the name in a field type appears multiple times, only one name needs to be kept.

[0050] In some embodiments of this application, similarity may include distance similarity. Step S400, which calculates the similarity between two POI data in each data group and determines whether the similarity exceeds a threshold, may include the following steps:

[0051] i. Calculate the distance d between the locations of the two POI data points.

[0052] Specifically, the distance d between the locations of two POI data points can be calculated based on latitude and longitude data, and the unit of distance d can be meters or kilometers.

[0053] ii. Iterate through the declaration array array(A1, A2, ..., An) to find the distance d between the locations of the two POI data points, and compare the distance d between the locations of the two POI data points with the size relationship of each element in the declaration array. Here, A1, A2, ..., An in the declaration array are all pre-defined distance values ​​arranged in an increasing order.

[0054] In one embodiment, the declared array can be array(25, 50, 100, 200, 300, 500, 550, 600). By iterating through the declared array using the distance d, the range of distances between entity objects in two POI datasets can be determined, thus facilitating the calculation of the similarity between the two POI datasets.

[0055] iii. Calculate the distance similarity D between the two POI data points as follows, and determine whether the distance similarity exceeds the threshold:

[0056] If d < array(i), then .

[0057] For example, if the distance d between the entity objects of two POI data is 45, then after iterating through the array declared by the distance d, we can know that d < array(1), and the distance similarity between the two POI data is... .

[0058] It should be noted that the distance similarity calculation result is within the range of 0 or greater and 1 or less. The closer the distance similarity between two POIs is to 0, the lower the similarity between them. Conversely, the closer the distance similarity is to 1, the higher the similarity between them.

[0059] In some embodiments of this application, the similarity may also include Jaccard similarity. Step S400, which calculates the similarity between two POI data in each data group and determines whether the similarity exceeds a threshold, may include the following steps:

[0060] i. Perform field segmentation on the text in the two POI data according to different types, where the text types include name, address and category.

[0061] In other words, the texts of the two POI datasets are segmented into words based on the different information they refer to. This segmentation process breaks down the POI text into shorter words, which facilitates subsequent similarity calculations between the different texts.

[0062] In one example, one piece of POI data contains the text information of "Huandong Lianhua Supermarket (Huandong Road Store)", and another piece of POI data contains the text information of "Zhangdong Town Huandong Lianhua". Then, during the word segmentation process, the text of the first piece of POI data will be segmented into:

[0063]

[0064] During the word segmentation process, the text of the second piece of POI data will be segmented into:

[0065]

[0066] After the division is completed, duplicate words can be removed.

[0067] ii. Calculate the Jaccard similarity of the name fields in the two pieces of POI data.

[0068] Specifically, the name field of each piece of POI data text directly reflects the entity object corresponding to that piece of POI data. Therefore, by calculating the similarity of the name fields in the two pieces of POI data, it can be accurately determined whether the two pieces of POI data belong to the same entity object.

[0069] For example, there are two text data X and Y. Then the Jaccard similarity value of these two text data is the number of intersection fields between X and Y divided by the number of union fields between X and Y, which can be listed in a mathematical expression as:

[0070]

[0071] Suppose data X includes fields "Chipone, Weizhi, icbc", and data Y includes fields "Juchuangxin, Weizhi". Then the intersection result of data X and data Y is "Weizhi", X∩Y = 1. The union result of data X and data Y is "Chipone, Weizhi, icbc, Juchuangxin", X∪Y = 4. Therefore, the Jaccard similarity value of data X and data Y is 1 / 4.

[0072] In addition, when calculating the Jaccard similarity, the judgment of the similarity of fields in the text can also be added. Suppose the text similarity between "Chipone" and "Juchuangxin" is judged to be two texts expressing the same information. Then the Jaccard similarity value of data X and data Y is 2 / 3. During the specific calculation process, the field texts in data X and data Y can be combined pairwise, and the text similarity of the two combined field texts can be calculated respectively, and finally the Jaccard similarity of data X and data Y can be calculated.

[0073] The specific calculation rules can be referred to the following steps:

[0074] (1) If the number of text fields in data X is greater than the number of text fields in data Y, swap data X and data Y;

[0075] (2) Calculate the similarity matrix between data X and data Y. Each row of the matrix corresponds to the field text of data X, and each column of the matrix corresponds to the field text of data Y.

[0076] (3) Calculate the similarity between the two fields of text in each row. If the similarity is higher than the set value, the result of X∩Y is increased by 1 and the result of X∪Y is decreased by 1.

[0077] In practice, the data text may also include brand information. In this case, the similarity between the two POIs can be directly determined as 0 or 1 based on whether their name fields are in the same brand list. This can replace the process of calculating similarity. If the data text does not contain brand information, similarity can be calculated using edit distance similarity or cosine similarity of word vectors, with word vectors being the preferred method.

[0078] iii. Determine whether the Jaccard similarity of the name field in the two POI data exceeds the threshold.

[0079] Specifically, in practice, since the name field in each POI data may differ in the number of characters, a threshold can be set according to the actual differences.

[0080] In some embodiments of this application, after the step of performing field segmentation on the text in the two POI data according to different types, and before the step of calculating the Jaccard similarity of the name text in the two POI data, the following steps are also included:

[0081] Determine if there is any overlap in the category text of two POI data sets. If there is no overlap in the category text of the two POI data sets, then the similarity between the two POI data sets is 0.

[0082] Specifically, the category text of each POI directly reflects the classification of the entity object corresponding to that POI. For example, each POI might point to an entity object in the category of "supermarket," "school," or "hospital." If the category texts of two POIs do not overlap, meaning they do not contain text indicating the same category, it means that the two POIs do not belong to the same entity object. Therefore, the similarity between two POIs can be initially screened and judged based on the category text.

[0083] In addition, after determining whether there is any overlap in the category text of the two POI data, the following steps are also included:

[0084] Determine whether both POIs contain name text. If one POI contains name text and the other does not, the similarity between the two POIs is 0. If both POIs contain name text, calculate the Jaccard similarity between the name text in the two POIs.

[0085] Specifically, if only one of the two POI data sets being compared contains name text, it means that there is no name information that needs to be merged between the two POI data sets, and therefore the similarity value returned is 0.

[0086] If both POIs include name text, then perform a Jaccard similarity calculation on the name text of the two POIs, and determine the degree of similarity between the two POIs based on the Jaccard similarity calculation result.

[0087] In some embodiments of this application, if two data groups with a similarity exceeding a threshold contain the same POI data, a connected subgraph of the POI data in these two data groups can be created, and a dataframe containing each POI data can be created based on the connected subgraph.

[0088] For example, if three Points of Interest (POIs) F, G, and H are considered to belong to the same entity, a connected subgraph can be generated for F, G, and H to indicate that these three POIs have a similarity relationship. The creation of a connected subgraph can be achieved using graphframes in the Spark distributed data platform. This allows more than two POIs to be linked together through a connected subgraph. Furthermore, a dataframe of these POIs can be created based on the connected subgraph, which is essentially a POI data table including segmented field type information.

[0089] like Figure 2 As shown, by creating a dataframe containing multiple Points of Interest (POIs) with high similarity (i.e., belonging to the same entity), these POI data are merged according to their weighted priorities. The merged POI data can be written to a dataset containing the basic POI data already stored in the Hive data warehouse. Simultaneously, a fusion log, including details of the fusion process, can be stored in the Hive data warehouse.

[0090] Some embodiments of this application also provide an electronic device, which includes a processor 101 and a memory 102 communicatively connected to the processor 101. The memory 102 stores instructions that can be executed by the processor. The instructions are executed by the processor 101 so that the processor 101 can perform the above-described POI data fusion method.

[0091] The memory 102 and processor 101 are connected via a bus, which may include any number of interconnecting buses and bridges. The bus connects various circuits of one or more processors 101 and memory 102 together. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. A bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by processor 101 is transmitted over a wireless medium via an antenna, which further receives data and transmits it to processor 101.

[0092] Processor 101 is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory 102 can be used to store data used by processor 101 during operation.

[0093] Some embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described POI data fusion method.

[0094] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes: a USB flash drive, a portable hard drive, and a read-only memory (ROM). Various media that can store program code, such as only memory, random access memory (RAM), magnetic disks or optical disks.

[0095] Those skilled in the art will understand that the above embodiments are specific implementations of this application, and in practical applications, various changes can be made in form and detail without departing from the spirit and scope of this application.

Claims

1. A POI data fusion method, characterized in that, include: Obtain a first dataset and a second dataset to be merged. Both the first dataset and the second dataset include multiple POI data. One of the first dataset and the second dataset is a dataset containing basic POI data, and the other is a dataset containing external POI data. Based on the geohash encoding of each POI data in the first dataset and the second dataset, the POI data located in the same geohash 3x3 grid in the first dataset and the second dataset are searched. Each specific geohash 3x3 grid corresponds to a specific location area on the two-dimensional plane of the map and corresponds to the same geohash encoding information string. Through the geohash encoding information of each POI data, the POI data whose location information is in the same location area on the two-dimensional plane of the map are classified. Perform a Cartesian product on the POI data in the first dataset and the second dataset that are located in the same geohash 3x3 grid to form different data groups; Calculate the similarity between two POI data in each data group and determine whether the similarity exceeds a threshold. If the similarity between two POI data in a certain data group exceeds the threshold, then merge the two POI data in that data group to form POI fused data. The step of obtaining the first dataset and the second dataset to be merged includes: Acquire POI data from different data sources and store it in a Hive data warehouse; The first dataset is composed of POI data with more field types, and the second dataset is composed of POI data with fewer field types. The field types include at least one of name, address, and latitude and longitude, and the first dataset has a higher priority than the second dataset. If two sets of data contain the same POI data and their similarity exceeds a threshold, a connected subgraph of the POI data in these two sets of data is created, and a dataframe containing each POI data is created based on the connected subgraph. The POI data is then fused according to the weight priority.

2. The POI data fusion method according to claim 1, characterized in that, The similarity includes distance similarity. The calculation of the similarity between two POI data points in each data group, and the determination of whether the similarity exceeds a threshold, includes: Calculate the distance d between the locations of two POI data points; The distance d between the locations of two POI data points is traversed through the declaration array array (A1, A2, ..., An), and the distance d between the locations of the two POI data points is compared with the size relationship of each element in the declaration array. Here, A1, A2, ..., An in the declaration array are all pre-set distance values ​​arranged in an increasing order. The distance similarity D between two POI data points is calculated as follows, and it is then determined whether the distance similarity exceeds a threshold: If d < array(i), then D = 1 - i * 0.

1.

3. The POI data fusion method according to claim 1, characterized in that, The similarity includes Jaccard similarity. The calculation of the similarity between two POIs in each data group, and the determination of whether the similarity exceeds a threshold, includes: The text in the two POI datasets is segmented into words according to different data types, including name, address, and category. Calculate Jaccard similarity for the name field in two POI datasets; Determine if the Jaccard similarity of the name field in two POI datasets exceeds a threshold.

4. The POI data fusion method according to claim 3, characterized in that, After performing field segmentation on the text in the two POI datasets according to different types, and before calculating the Jaccard similarity of the name fields in the two POI datasets, the following steps are also included: Determine if there is any overlap in the category text of two POI data sets. If there is no overlap in the category text of the two POI data sets, then the similarity between the two POI data sets is 0.

5. The POI data fusion method according to claim 4, characterized in that, After determining whether there is an overlap in the category text of the two POI data, the following steps are also included: Determine whether both POIs contain name text. If one POI contains name text and the other does not, the similarity between the two POIs is 0. If both POIs contain name text, calculate the Jaccard similarity between the name text in the two POIs.

6. An electronic device, characterized in that, include: processor; A memory communicatively connected to the processor, the memory storing instructions executable by the processor, the instructions being executed by the processor to enable the processor to perform the POI data fusion method as described in any one of claims 1 to 5.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the POI data fusion method according to any one of claims 1 to 5.

Images