A method, system, electronic device and storage medium for generating a point of interest portrait

Abstract

The embodiments of the present invention provide a method, system, electronic device and storage medium for generating portraits of points of interest, and relate to the field of positioning technology. The method includes: obtaining multiple Wi-Fi lists, each Wi-Fi list includes the MAC address of one or more Wi-Fis; searching for Wi-Fis with similar relationships in multiple Wi-Fi lists according to the MAC addresses of the Wi-Fis in the Wi-Fi lists, and assigning them to a Wi-Fi cluster; determining the binding relationship between other Wi-Fis and POIs in each Wi-Fi cluster according to the target Wi-Fis included in each Wi-Fi cluster, and the binding relationship between the target Wi-Fis and points of interest POIs that are predetermined and verified for authenticity. The solution of the present invention can accurately bind POIs even in the face of semanticless Wi-Fis, thereby improving the portrait coverage rate and making the subsequent perception algorithm that strongly relies on the portrait more accurate.

Description

Technical Field

[0001] The present invention relates to the field of positioning technology, and in particular to a method, system, electronic device and storage medium for generating an image of a point of interest. Background Art

[0002] User store perception is a common scenario in sales business and can be used in a variety of recommendation systems. For example, when it is detected that a user has arrived at a store, group coupons, discount information, etc. of the store can be recommended to enable customers to have a better service experience. User store perception is also called POI (Point of Interest) perception. After access to POI perception, sales performance can be significantly improved.

[0003] Since GPS (Global Positioning System) signals are weak indoors and drift errors are large, existing indoor positioning solutions use the surrounding Wi-Fi and Bluetooth information that users connect to or scan to determine whether the user has reached the POI, which is called "network positioning." Wi-Fi-POI portraits refer to the binding relationship between Wi-Fi and POI. The accuracy of POI perception strongly depends on the construction of Wi-Fi-POI portraits.

[0004] The construction of existing Wi-Fi-POI portraits usually involves comparing the semantic similarity between Wi-Fi names and POI names, as well as comparing the GPS overlap between Wi-Fi and POI. However, due to the large indoor GPS positioning drift error, the Wi-Fi-POI portraits based on GPS overlap are likely to be biased, making it difficult to ensure accuracy.

[0005] Moreover, since many Wi-Fi names have no semantics, such as the semantic-free Wi-Fi "CMCC-urjk", and the semantic similarity-based solution cannot solve the binding problem of non-semantic relationships, the number of Wi-Fi-POI portrait binding relationships constructed by this method is limited, the coverage rate is low, and it is difficult to use in positioning based on Wi-Fi-POI portraits. Summary of the invention

[0006] The present invention provides a method, system, electronic device and storage medium for generating a point of interest portrait, which solve the problem in the prior art that Wi-Fi-POI portraits cannot be bound without semantic relationships and have low coverage.

[0007] In a first aspect of the present invention, a method for generating a point of interest portrait is provided, comprising:

[0008] Acquire multiple Wi-Fi lists reported by the user device through a positioning request, each Wi-Fi list includes one or more media access control addresses (MAC addresses) of Wi-Fi;

[0009] According to the MAC addresses of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar relationships in multiple Wi-Fi lists, and classify the Wi-Fi with similar relationships into a Wi-Fi cluster;

[0010] According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the point of interest POI that is predetermined and authenticity-verified, the binding relationship between other Wi-Fi and the POI in each Wi-Fi cluster except the target Wi-Fi is determined.

[0011] Optionally, the step of searching for Wi-Fis with similar relationships in multiple Wi-Fi lists according to the MAC addresses of the Wi-Fis in the Wi-Fi list, and assigning the Wi-Fis with similar relationships to a Wi-Fi cluster includes:

[0012] According to the MAC addresses of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar MAC addresses in each Wi-Fi list, and group the Wi-Fi with similar MAC addresses in each Wi-Fi list into one Wi-Fi set;

[0013] In the Wi-Fi sets in the multiple Wi-Fi lists, a Wi-Fi set including the same Wi-Fi is searched, and all Wi-Fis in the Wi-Fi set including the same Wi-Fi are classified into one Wi-Fi cluster.

[0014] Optionally, according to the MAC address of the Wi-Fi in the Wi-Fi list, the step of searching for a Wi-Fi with a similar MAC address in each Wi-Fi list includes:

[0015] Compare the MAC addresses between different Wi-Fi in each Wi-Fi list;

[0016] If the MAC addresses of two Wi-Fi networks are identical except for one or two last bytes, it is determined that the MAC addresses of the two Wi-Fi networks are similar.

[0017] Optionally, in the Wi-Fi sets in the multiple Wi-Fi lists, the step of searching for a Wi-Fi set including the same Wi-Fi includes:

[0018] The multiple Wi-Fi lists are traversed in units of preset areas, and a Wi-Fi set including the same Wi-Fi is searched in the Wi-Fi sets of the multiple Wi-Fi lists traversed in each unit.

[0019] Optionally, according to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the point of interest POI that is predetermined and authenticity-verified, the step of determining the binding relationship between other Wi-Fis except the target Wi-Fi in each Wi-Fi cluster and the POI includes:

[0020] If the Wi-Fi cluster contains a target Wi-Fi, the POI bound to the target Wi-Fi is determined to be the POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi;

[0021] If the Wi-Fi cluster includes multiple target Wi-Fis, and there are different POIs among the POIs bound to the multiple target Wi-Fis, a target POI is selected from multiple candidate POIs bound to the multiple target Wi-Fis, and the target POI is determined to be a POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi.

[0022] Optionally, the step of selecting a target POI from a plurality of candidate POIs bound to the plurality of target Wi-Fis includes:

[0023] The other Wi-Fi in the Wi-Fi cluster except the target Wi-Fi is used as the Wi-Fi to be recalled;

[0024] Scoring the multiple candidate POIs according to one or more of the characteristic information of the Wi-Fi cluster, the characteristic information of the candidate POI, the similar characteristics between the Wi-Fi to be recalled and the candidate POI, and the similar characteristics between the Wi-Fi to be recalled and the target Wi-Fi;

[0025] A POI with the highest score and a score exceeding a preset threshold is selected from the multiple candidate POIs as the target POI.

[0026] Optionally, the characteristic information of the Wi-Fi cluster includes: the number of Wi-Fis included in the Wi-Fi cluster;

[0027] The characteristic information of the candidate POI includes: one or more of the number of candidate POIs and Wi-Fi number thresholds corresponding to the types of the candidate POIs;

[0028] The similar features between the Wi-Fi to be recalled and the candidate POI include: the similarity between the names of the Wi-Fi to be recalled and the candidate POI;

[0029] The similarity features between the Wi-Fi to be recalled and the target Wi-Fi include: one or more of: MAC address similarity between the Wi-Fi to be recalled and the target Wi-Fi, name similarity between the Wi-Fi to be recalled and the target Wi-Fi, signal strength difference between the Wi-Fi to be recalled and the target Wi-Fi, and the number of times the Wi-Fi to be recalled and the target Wi-Fi appear together in the same Wi-Fi list.

[0030] Optionally, after determining the binding relationship between other Wi-Fis except the target Wi-Fi in each Wi-Fi cluster and the POI according to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that is predetermined and authenticity-verified, the method further includes:

[0031] Determine whether the number of Wi-Fi networks bound to each POI is greater than the Wi-Fi number threshold corresponding to the type of the POI;

[0032] If it is greater, the binding relationship between the POI and Wi-Fi will be discarded.

[0033] Optionally, after the Wi-Fis with similar relationships are grouped into one Wi-Fi cluster, the method further includes:

[0034] Obtain identification information corresponding to each Wi-Fi in the Wi-Fi cluster, take the identification information of each Wi-Fi as a vertex, connect the vertices corresponding to different Wi-Fi in the Wi-Fi cluster, and construct a connectivity graph;

[0035] According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that is predetermined and authenticity-verified, the steps of determining the binding relationship between other Wi-Fis except the target Wi-Fi in each Wi-Fi cluster and the POI include:

[0036] According to the connectivity graph, a target Wi-Fi included in each Wi-Fi cluster is determined.

[0037] In a second aspect of the present invention, a system for generating a portrait of a point of interest is provided, comprising:

[0038] A first acquisition module is used to acquire multiple Wi-Fi lists reported by the user equipment through a positioning request, each Wi-Fi list includes one or more media access control addresses (MAC addresses) of Wi-Fi;

[0039] A cluster building module is used to search for Wi-Fis with similar relationships in multiple Wi-Fi lists according to the MAC addresses of the Wi-Fis in the Wi-Fi list, and classify the Wi-Fis with similar relationships into a Wi-Fi cluster;

[0040] The binding module is used to determine the binding relationship between other Wi-Fis except the target Wi-Fi in each Wi-Fi cluster and the POI according to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that is predetermined and authenticity-verified.

[0041] In a third aspect of the present invention, there is also provided an electronic device, comprising: a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0042] Memory, used to store computer programs;

[0043] The processor is used to implement the steps in the above-mentioned method for generating a portrait of a point of interest when executing the program stored in the memory.

[0044] In a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method for generating a portrait of a point of interest as described above is implemented.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] In an embodiment of the present invention, a large number of Wi-Fis are classified by building Wi-Fi clusters, and POIs can be accurately bound even for Wi-Fis without semantic relationships. When performing POI profiling, the embodiment of the present invention first obtains multiple Wi-Fi lists reported by the user device through a positioning request, and then searches for Wi-Fis with similar relationships in multiple Wi-Fi lists based on the MAC (Media Access Control Address) address of the Wi-Fi in the Wi-Fi list, and assigns the Wi-Fis with similar relationships to a Wi-Fi cluster. Since multiple Wi-Fis of the same POI are usually deployed by the same manufacturer, the MAC of the Wi-Fi is similar. Therefore, based on the MAC address of the Wi-Fi, multiple similar Wi-Fis deployed by the same manufacturer can be accurately found, and these similar Wi-Fis can be built into a Wi-Fi cluster. The Wi-Fis in the same cluster can be bound to the same POI, thereby achieving classification processing for a large number of Wi-Fis. Next, just find the POI bound to the Wi-Fi cluster. The embodiment of the present invention predetermines a group of authentic and reliable binding relationships between target Wi-Fi and POI. According to the target Wi-Fi contained in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and POI, the POI bound to the Wi-Fi cluster is found, and the binding relationship between other Wi-Fi and POI in each Wi-Fi cluster except the target Wi-Fi can be accurately determined, thereby completing the POI portrait based on the Wi-Fi cluster. The embodiment of the present invention searches for similar Wi-Fi by MAC address, and clusters similar Wi-Fi to obtain a cluster of Wi-Fi corresponding to a POI, thereby realizing the classification processing of a large number of Wi-Fis, and can accurately bind POIs even in the face of non-semantic Wi-Fis, solving the problem that Wi-Fi-POI portraits cannot achieve non-semantic relationship binding, improving the portrait coverage, making the subsequent perception algorithm that strongly relies on the portrait more accurate, and avoiding the risk of portrait binding errors caused by a single GPS drift error, thereby improving accuracy.

[0047] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for describing the embodiments are briefly introduced below.

[0049] Figure 1 A schematic diagram of a process for generating a point of interest portrait according to an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of the sub-steps of the method for generating a point of interest portrait provided by an embodiment of the present invention;

[0051] Figure 3 A flowchart of another sub-step of the method for generating a point of interest portrait provided by an embodiment of the present invention;

[0052] Figure 4 Another schematic diagram of the process of generating a point of interest portrait provided by an embodiment of the present invention;

[0053] Figure 5 A schematic block diagram of a system for generating a point of interest portrait according to an embodiment of the present invention;

[0054] Figure 6 A schematic block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can also be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to enable the scope of the present invention to be fully communicated to those skilled in the art.

[0056] Figure 1 is a flow chart of a method for generating a point of interest portrait provided by an embodiment of the present invention. Figure 1 As shown, the interest point portrait generation method includes:

[0057] Step 101: Acquire multiple Wi-Fi lists reported by a user device through a positioning request, each Wi-Fi list includes one or more Wi-Fi media access control addresses (MAC addresses).

[0058] Here, firstly, multiple Wi-Fi lists reported by the user equipment through the positioning request are obtained, each Wi-Fi list includes one or more Wi-Fi MAC addresses, so as to perform similar Wi-Fi clustering based on the MAC addresses of the Wi-Fi in the Wi-Fi list.

[0059] When the user device sends a positioning request, it also reports the Wi-Fi list searched at the current location. The Wi-Fi list includes the MAC address of the Wi-Fi and may also include the name of the Wi-Fi, signal strength and other information.

[0060] The multiple Wi-Fi lists obtained in step 101 may be Wi-Fi lists reported by any multiple user devices through positioning requests.

[0061] Step 102: According to the MAC address of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar relationship in multiple Wi-Fi lists, and classify the Wi-Fi with similar relationship into a Wi-Fi cluster.

[0062] Here, the MAC address can be used as the unique identifier of Wi-Fi due to its uniqueness, and since multiple Wi-Fis at the same POI are usually deployed by the same manufacturer, the MAC addresses of Wi-Fis are similar. Therefore, based on the MAC address of Wi-Fis, Wi-Fis with similar relationships can be accurately found in multiple Wi-Fi lists, and Wi-Fis with similar relationships can be built into Wi-Fi clusters. Since Wi-Fis in the same cluster are composed of similar Wi-Fis, they can be processed in the same way and can be bound to the same POI, thereby realizing the classification of a large number of Wi-Fis. By classifying a large number of Wi-Fis, the calculation process can be simplified, and the binding calculation of a single Wi-Fi and a POI can be realized, and the binding calculation of a Wi-Fi cluster constructed by multiple Wi-Fis and a POI can be realized. The binding of multiple Wi-Fis and POIs can be processed through one calculation, which improves the processing efficiency, and even semantic-free Wi-Fis can be accurately bound to POIs.

[0063] Step 103: According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that is predetermined and authenticity-verified, the binding relationship between other Wi-Fis except the target Wi-Fi in each Wi-Fi cluster and the POI is determined.

[0064] Here, a batch of authentic and reliable binding relationships between target Wi-Fi and POI, namely "seed" portraits, are determined in advance. According to the target Wi-Fi contained in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and POI, the POI bound to the Wi-Fi cluster is found, and the binding relationship between other Wi-Fi and POI in each Wi-Fi cluster except the target Wi-Fi can be accurately determined, thereby completing the POI portrait based on the Wi-Fi cluster. Even in the face of semanticless Wi-Fi, the POI can be accurately bound, which improves the portrait coverage rate, makes the subsequent perception algorithm that strongly relies on the portrait more accurate, and avoids the risk of portrait binding errors caused by a single GPS drift error, thereby improving accuracy.

[0065] The embodiment of the present invention does not limit how to determine the binding relationship between the target Wi-Fi and the POI, and any method that can determine the authentic binding relationship between the target Wi-Fi and the POI can be applied to the embodiment of the present invention. For example, the embodiment of the present invention can obtain a list of Wi-Fi locations requested near the POI, and obtain a highly credible binding relationship between the target Wi-Fi and the POI, i.e., a "seed" portrait, by comparing semantic similarity or mining the true value of a service source, for subsequent expansion, but is not limited thereto.

[0066] The method for generating a point of interest portrait in an embodiment of the present invention searches for similar Wi-Fi by MAC address, and clusters similar Wi-Fi to obtain a cluster of Wi-Fi corresponding to a POI. Even in the face of non-semantic Wi-Fi, the POI can be accurately bound, solving the problem that Wi-Fi-POI portraits cannot achieve non-semantic relationship binding, improving the portrait coverage, and making the subsequent perception algorithm that strongly relies on the portrait more accurate. And by clustering multiple Wi-Fi, the risk of portrait binding errors caused by a single GPS drift error is avoided, and the accuracy is improved.

[0067] Preferably, Figure 2 As shown, the above step 102 includes:

[0068] Step 1021: According to the MAC address of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar MAC address in each Wi-Fi list, and assign the Wi-Fi with similar MAC address in each Wi-Fi list to one Wi-Fi set.

[0069] Here, firstly, based on the MAC address of the Wi-Fi, the Wi-Fi with similar MAC address is searched in each Wi-Fi list, and the Wi-Fi with similar MAC address in each list is classified into a Wi-Fi set, so as to build a Wi-Fi cluster based on the Wi-Fi sets between the lists.

[0070] For example, based on the MAC address of the Wi-Fi in List A, Wi-Fi1 and Wi-Fi2 with similar MAC addresses are found in List A, and Wi-Fi1 and Wi-Fi2 are classified into one Wi-Fi set; similarly, based on the MAC address of the Wi-Fi in List B, Wi-Fi2 and Wi-Fi3 with similar MAC addresses are found in List B, and Wi-Fi2 and Wi-Fi3 are classified into one Wi-Fi set.

[0071] Generally, the first three bytes of the MAC address are fixed codes of the manufacturer, and the last three bytes are allocated by the manufacturer. Multiple Wi-Fis of the same POI are usually deployed by the same manufacturer, and the MACs are similar. Based on this feature, preferably, the above step 1021 includes:

[0072] Compare the MAC addresses of different Wi-Fis in each Wi-Fi list; if the MAC addresses of two Wi-Fis are the same except for the last one or two bytes, then it is determined that the MAC addresses of the two Wi-Fis are similar.

[0073] At this time, by comparing the last bytes of the MAC addresses between different Wi-Fis, you can accurately find the Wi-Fis with similar MAC addresses in the Wi-Fi list.

[0074] For example, a group of Wi-Fis at the same POI "Haidilao Hotpot" is: 00:16:78:00:40:50@Haidilao, 00:16:78:00:40:52@.0, 00:16:78:00:40:64@xxbtest. The MAC addresses of these three Wi-Fis are: 00:16:78:00:40:50, 00:16:78:00:40:52, 00:16:78:00:40:64, the first five bytes are "00:16:78:00:40", and only the last byte is different. When these three Wi-Fis appear in the same Wi-Fi list, by comparing the MAC addresses of different Wi-Fis in the Wi-Fi list, these three Wi-Fis can be determined as Wi-Fis with similar MAC addresses, forming a Wi-Fi set, which is used to cluster with Wi-Fi sets in other lists to build a Wi-Fi cluster.

[0075] Step 1022: Search for a Wi-Fi set including the same Wi-Fi in the Wi-Fi sets in the multiple Wi-Fi lists, and assign all Wi-Fi in the Wi-Fi set including the same Wi-Fi to one Wi-Fi cluster.

[0076] Here, after obtaining a Wi-Fi set in a single list based on the MAC address, Wi-Fi clustering is performed between multiple lists. That is, in the Wi-Fi sets in multiple lists, check whether there is a Wi-Fi set containing the same Wi-Fi, and assign the Wi-Fi in the Wi-Fi set containing the same Wi-Fi to the same cluster.

[0077] For example, it is known from list A that Wi-Fi1 and Wi-Fi2 belong to one Wi-Fi set, and it is known from list B that Wi-Fi2 and Wi-Fi3 belong to one Wi-Fi set. The Wi-Fi sets in list A and list B contain the same Wi-Fi2, so Wi-Fi1, Wi-Fi2, and Wi-Fi3 are classified into the same cluster.

[0078] By clustering a large number of Wi-Fis and building Wi-Fi clusters, Wi-Fis in the same cluster can be treated the same way and bound to the same POI, thus achieving classification of a large number of Wi-Fis. Classification of Wi-Fis can simplify the calculation process, and the binding calculation of a single Wi-Fi and a POI can be upgraded to the binding calculation of a Wi-Fi cluster built by multiple Wi-Fis and a POI. Multiple Wi-Fis can be bound to POIs in one calculation, which improves processing efficiency. It can also accurately bind non-semantic Wi-Fis to POIs.

[0079] Preferably, the above step 1022 includes:

[0080] The multiple Wi-Fi lists are traversed in units of preset areas, and a Wi-Fi set including the same Wi-Fi is searched in the Wi-Fi sets of the multiple Wi-Fi lists traversed in each unit.

[0081] When the user equipment sends a positioning request, it reports the Wi-Fi list and the location information at the same time, so the Wi-Fi list can be matched with the location.

[0082] Here, for a large number of Wi-Fi lists reported from different locations, the Wi-Fi set containing the same Wi-Fi can be searched in the Wi-Fi set of the Wi-Fi list traversed by each unit by traversing in the preset area to build a Wi-Fi cluster. The Wi-Fi clusters obtained by clustering belong to similar Wi-Fi reported in the same preset area. These similar Wi-Fi are bound to the same POI according to the binding relationship between the target Wi-Fi and the POI, which can ensure accuracy and avoid the clustering range being too large, resulting in similar Wi-Fi deployed by the same manufacturer in different chain stores being incorrectly clustered. It can also ensure coverage and avoid the clustering range being too small, resulting in Wi-Fi set up in schools, hospitals, etc. across large regions not being clustered into the same Wi-Fi cluster.

[0083] The size of the preset area can be set according to demand, such as being set to a 6-bit Geohash block. In this case, the Wi-Fi lists carried in all positioning requests are traversed in units of 6-bit Geohash blocks, and Wi-Fi clusters are constructed between multiple lists.

[0084] Among them, Geohash converts the two-dimensional longitude and latitude representing the location into a string. For example, the GeoHash strings of a certain area are WX4ER, WX4G2, WX4G3, etc. Each string represents a certain rectangular area. In other words, all points (longitude and latitude coordinates) in this rectangular area share the same GeoHash string, which can protect privacy (only indicate the approximate area location instead of the specific point) and is easier to cache.

[0085] Preferably, after the above step 102, the method further includes:

[0086] Obtain identification information corresponding to each Wi-Fi in the Wi-Fi cluster, take the identification information of each Wi-Fi as a vertex, connect the vertices corresponding to different Wi-Fi in the Wi-Fi cluster, and construct a connectivity graph;

[0087] The above step 103 includes:

[0088] According to the connectivity graph, a target Wi-Fi included in each Wi-Fi cluster is determined.

[0089] At this time, a connectivity graph can be established between multiple lists to intuitively represent the clustered Wi-Fi clusters. Then, based on the connectivity graph, the target Wi-Fi included in each Wi-Fi cluster can be determined. Then, based on the binding relationship between the target Wi-Fi and the POI, the binding relationship between other Wi-Fi and the POI in the Wi-Fi cluster can be determined.

[0090] The identification information of each Wi-Fi is used to uniquely identify the Wi-Fi. The identification information may be a hash code formed by combining Geohash6 string information, MAC address information, SSID (Service Set Identifier) ​​information, etc., but is not limited thereto.

[0091] Among them, the identification information of each Wi-Fi is taken as a vertex. After the vertices corresponding to different Wi-Fi in the Wi-Fi cluster are connected, the weight of the edge can represent the MAC similarity relationship. Since the Wi-Fi in the same Wi-Fi cluster are similar Wi-Fi, the weight of the edge in the same Wi-Fi cluster can be set to 1.

[0092] Of course, the above-mentioned method of constructing a connectivity graph to save Wi-Fi clusters is only a preferred implementation mode. The embodiment of the present invention may also use other methods to save Wi-Fi clusters, such as through an array, etc., which will not be described one by one here.

[0093] Preferably, Figure 2 As shown, the above step 103 includes:

[0094] Step 1031: If the Wi-Fi cluster includes a target Wi-Fi, the POI bound to the target Wi-Fi is determined to be the POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi.

[0095] Here, if the Wi-Fi cluster contains only one target Wi-Fi, the POI bound to the target Wi-Fi can be directly determined as the POI bound to other Wi-Fi in the Wi-Fi cluster.

[0096] For example, Wi-Fi1, Wi-Fi2, and Wi-Fi3 belong to the same cluster. It is known that Wi-Fi1 has a binding relationship with POI1. Therefore, both Wi-Fi2 and Wi-Fi3 establish a binding relationship with POI1.

[0097] Step 1032: If the Wi-Fi cluster includes multiple target Wi-Fis, and there are different POIs among the POIs bound to the multiple target Wi-Fis, a target POI is selected from the multiple candidate POIs bound to the multiple target Wi-Fis, and the target POI is determined to be a POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi.

[0098] Here, if there are multiple target Wi-Fis in the Wi-Fi cluster and they conflict with each other (mostly due to POI redundancy and coincidence of similar MAC addresses), it is necessary to select a target POI from multiple candidate POIs for other Wi-Fis in the Wi-Fi cluster to be recalled.

[0099] For example, it is known that Wi-Fi1, Wi-Fi2, and Wi-Fi3 belong to the same cluster, and Wi-Fi1-POI1 and Wi-Fi2-POI2 are both acquired "seed" images. Then Wi-Fi3 needs to select a target POI from POI1 and POI2.

[0100] Preferably, Figure 3 As shown, in the above step 1032, the step of selecting a target POI from a plurality of candidate POIs bound to the plurality of target Wi-Fis includes:

[0101] Step 10321: other Wi-Fi in the Wi-Fi cluster except the target Wi-Fi are used as Wi-Fi to be recalled.

[0102] Step 10322: score the multiple candidate POIs according to one or more of the characteristic information of the Wi-Fi cluster, the characteristic information of the candidate POI, the similar characteristics between the Wi-Fi to be recalled and the candidate POI, and the similar characteristics between the Wi-Fi to be recalled and the target Wi-Fi.

[0103] Here, multiple candidate POIs are scored based on various information such as characteristic information of the Wi-Fi cluster, characteristic information of the candidate POIs, similar characteristics between the Wi-Fi to be recalled and the candidate POIs, and similar characteristics between the Wi-Fi to be recalled and the target Wi-Fi. The final bound POI is determined based on the scoring, which can further improve the accuracy of binding and find the most matching POI for the Wi-Fi to be recalled.

[0104] Preferably, the characteristic information of the Wi-Fi cluster includes: the number of Wi-Fis included in the Wi-Fi cluster, but is not limited thereto;

[0105] The characteristic information of the candidate POI includes: one or more of the number of candidate POIs and Wi-Fi number thresholds corresponding to the types of the candidate POIs, but is not limited thereto;

[0106] The similar features between the Wi-Fi to be recalled and the candidate POI include: similarity between the names of the Wi-Fi to be recalled and the candidate POI, but not limited thereto;

[0107] The similarity features between the Wi-Fi to be recalled and the target Wi-Fi include: one or more of: MAC address similarity between the Wi-Fi to be recalled and the target Wi-Fi, name similarity between the Wi-Fi to be recalled and the target Wi-Fi, signal strength difference between the Wi-Fi to be recalled and the target Wi-Fi, and the number of times the Wi-Fi to be recalled and the target Wi-Fi appear together in the same Wi-Fi list, but are not limited thereto.

[0108] The embodiment of the present invention can score by a machine learning scoring model, but is not limited thereto. When scoring by a model, step 10322 can combine the above feature information into a one-dimensional input vector, and then input it into the trained machine learning scoring model, and output the score of each candidate POI by the machine learning scoring model.

[0109] Step 10323: Select a POI with the highest score and a score exceeding a preset threshold from the multiple candidate POIs as the target POI.

[0110] At this time, based on the scoring situation, the POI with the highest score and a score exceeding a preset threshold is selected as the target POI, which further improves the accuracy of binding and finds the most matching POI for the Wi-Fi to be recalled.

[0111] The preset threshold can be set according to the requirements and can be set to any value. By setting the preset threshold, the probability of false matching is avoided. When the score is low, the group of data can be directly discarded and re-matched.

[0112] Due to differences in POI types, Wi-Fi deployment characteristics are also different. For example, for large-category POIs, including companies, schools, hospitals, and hotels, the internal Wi-Fi deployment has significant characteristics such as density and similarity. Such POIs should correspond to large clusters in Wi-Fi clustering. For small-category POIs, including ordinary businesses, community buildings, and other non-campus types, there are usually fewer internal Wi-Fis. Such POIs should correspond to small clusters in Wi-Fi clustering.

[0113] In the embodiment of the present invention, based on the above POI type differences and Wi-Fi deployment characteristics, after the Wi-Fi cluster is bound to the POI, verification is performed again according to the POI type differences to further ensure accuracy. The verification steps are as follows.

[0114] Preferably, after the above step 103, the method further includes:

[0115] Determine whether the number of Wi-Fi networks bound to each POI is greater than the Wi-Fi number threshold corresponding to the type of the POI;

[0116] If it is greater, the binding relationship between the POI and Wi-Fi will be discarded.

[0117] At this time, different types of POIs are treated in layers, and corresponding Wi-Fi quantity thresholds are set for different types of POIs. If the number of Wi-Fis bound to the POI is significantly greater than its corresponding Wi-Fi quantity threshold, it means that a binding error has occurred. The group of data can be directly discarded and re-matched to further ensure the accuracy of the binding.

[0118] In the embodiment of the present invention, by discussing POIs in layers, the Wi-Fi-POI portrait construction errors caused by the accidental MAC similarity are reduced. For example, if it is known that the POI is an ordinary merchant, but corresponds to a Wi-Fi cluster of size 100, it is obviously wrong. It may be due to coincidence that the Wi-Fi MAC address of the POI is similar to the MAC address of the large category POI, causing clustering errors, which can be filtered by POI layers.

[0119] The following is an example of a specific application of the method for generating a point of interest portrait according to an embodiment of the present invention.

[0120] like Figure 4 As shown, the method for generating a point of interest portrait in an embodiment of the present invention includes:

[0121] Step 401: Determine the binding relationship between the target Wi-Fi and the POI, and build a highly reliable "seed" portrait.

[0122] Step 402: Compare MAC addresses between different Wi-Fis in each Wi-Fi list, find Wi-Fis with similar MAC addresses, and assign Wi-Fis with similar MAC addresses in each Wi-Fi list to one Wi-Fi set.

[0123] Step 403: construct a Wi-Fi connectivity graph for multiple historical Wi-Fi lists, and cluster them according to MAC address similarity, that is, traverse multiple Wi-Fi lists in units of preset areas, and search for Wi-Fi sets containing the same Wi-Fi in the Wi-Fi sets of the multiple Wi-Fi lists traversed by each unit, and classify the Wi-Fi sets containing the same Wi-Fi into one Wi-Fi cluster to implement Wi-Fi clustering.

[0124] Step 404: Using the Wi-Fi clustering results, for Wi-Fi clusters without “seed” portrait conflicts, use the Wi-Fi-POI relationship to expand and recall the portraits. That is, if the Wi-Fi cluster contains a target Wi-Fi, the POI bound to the target Wi-Fi is determined to be the POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi.

[0125] Step 405: Using the Wi-Fi clustering results, for Wi-Fi clusters with conflicting “seed” portraits, construct a feature relationship between the Wi-Fi to be recalled and the candidate POIs, and use a machine learning scoring model to select the optimal POI as the binding portrait. That is, if the Wi-Fi cluster contains multiple target Wi-Fis, and the multiple target Wi-Fis are bound to different POIs, then according to one or more of the feature information of the Wi-Fi cluster, the feature information of the candidate POIs, the similar features between the Wi-Fi to be recalled and the candidate POIs, and the similar features between the Wi-Fi to be recalled and the target Wi-Fi, use a machine learning scoring model to score the multiple candidate POIs, and select the POI with the highest score and a score exceeding a preset threshold from the multiple candidate POIs as the binding target POI.

[0126] Step 406: Output the Wi-Fi-POI portrait table.

[0127] The method for generating a point of interest portrait in an embodiment of the present invention searches for similar Wi-Fi by MAC address, and clusters similar Wi-Fi to obtain a cluster of Wi-Fi corresponding to a POI. Even in the face of non-semantic Wi-Fi, the POI can be accurately bound, solving the problem that Wi-Fi-POI portraits cannot achieve non-semantic relationship binding, improving the portrait coverage, and making the subsequent perception algorithm that strongly relies on the portrait more accurate. And by clustering multiple Wi-Fi, the risk of portrait binding errors caused by a single GPS drift error is avoided, and the accuracy is improved.

[0128] See also Figure 5 As shown, the embodiment of the present invention further provides a system 500 for generating a portrait of an interest point, including:

[0129] A first acquisition module 501 is used to acquire multiple Wi-Fi lists reported by a user equipment through a positioning request, each Wi-Fi list includes a media access control address MAC address of one or more Wi-Fi;

[0130] A cluster building module 502 is used to search for Wi-Fis with similar relationships in multiple Wi-Fi lists according to the MAC addresses of the Wi-Fis in the Wi-Fi list, and classify the Wi-Fis with similar relationships into a Wi-Fi cluster;

[0131] The binding module 503 is used to determine the binding relationship between other Wi-Fis except the target Wi-Fi in each Wi-Fi cluster and the POI according to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that is predetermined and authenticity-verified.

[0132] The point of interest portrait generation system 500 of the embodiment of the present invention searches for similar Wi-Fi through MAC addresses, and clusters similar Wi-Fi to obtain a cluster of Wi-Fi corresponding to a POI. Even in the face of non-semantic Wi-Fi, the POI can be accurately bound, solving the problem that Wi-Fi-POI portraits cannot achieve non-semantic relationship binding, improving the portrait coverage, and making the subsequent perception algorithm that strongly relies on the portrait more accurate. And through multiple Wi-Fi clusters, the risk of portrait binding errors caused by a single GPS drift error is avoided, and the accuracy is improved.

[0133] Preferably, the cluster building module 502 includes:

[0134] The first search submodule is used to search for Wi-Fi with similar MAC addresses in each Wi-Fi list according to the MAC addresses of the Wi-Fi in the Wi-Fi list, and classify the Wi-Fi with similar MAC addresses in each Wi-Fi list into one Wi-Fi set;

[0135] The construction submodule is used to search for a Wi-Fi set including the same Wi-Fi in the Wi-Fi sets in the multiple Wi-Fi lists, and assign all Wi-Fi in the Wi-Fi set including the same Wi-Fi to one Wi-Fi cluster.

[0136] Preferably, the first search submodule includes:

[0137] A comparison unit, used for comparing MAC addresses between different Wi-Fi in each Wi-Fi list;

[0138] The determining unit is configured to determine that the MAC addresses of the two Wi-Fi networks are similar if the MAC addresses of the two Wi-Fi networks are identical except for one or two last bytes.

[0139] Preferably, the construction submodule includes:

[0140] The searching unit is used to traverse the multiple Wi-Fi lists in units of preset areas, and search for a Wi-Fi set containing the same Wi-Fi in the Wi-Fi sets of the multiple Wi-Fi lists traversed by each unit.

[0141] Preferably, the binding module 503 includes:

[0142] The first determination submodule is configured to determine the POI bound to the target Wi-Fi as the POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi if the Wi-Fi cluster includes a target Wi-Fi;

[0143] The second determination submodule is used to select a target POI from multiple candidate POIs bound to the multiple target Wi-Fis if the Wi-Fi cluster includes multiple target Wi-Fis and there are different POIs among the POIs bound to the multiple target Wi-Fis, and determine the target POI as a POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi.

[0144] Preferably, the second determining submodule includes:

[0145] A to-be-recalled determining unit, configured to determine other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi as the to-be-recalled Wi-Fis;

[0146] A scoring unit, configured to score the multiple candidate POIs according to one or more of the characteristic information of the Wi-Fi cluster, the characteristic information of the candidate POI, the similar characteristics between the Wi-Fi to be recalled and the candidate POI, and the similar characteristics between the Wi-Fi to be recalled and the target Wi-Fi;

[0147] The selection unit is used to select a POI with the highest score and a score exceeding a preset threshold from the multiple candidate POIs as a target POI.

[0148] Preferably, the characteristic information of the Wi-Fi cluster includes: the number of Wi-Fis included in the Wi-Fi cluster;

[0149] The characteristic information of the candidate POI includes: one or more of the number of candidate POIs and Wi-Fi number thresholds corresponding to the types of the candidate POIs;

[0150] The similar features between the Wi-Fi to be recalled and the candidate POI include: the similarity between the names of the Wi-Fi to be recalled and the candidate POI;

[0151] The similarity features between the Wi-Fi to be recalled and the target Wi-Fi include: one or more of: MAC address similarity between the Wi-Fi to be recalled and the target Wi-Fi, name similarity between the Wi-Fi to be recalled and the target Wi-Fi, signal strength difference between the Wi-Fi to be recalled and the target Wi-Fi, and the number of times the Wi-Fi to be recalled and the target Wi-Fi appear together in the same Wi-Fi list.

[0152] Preferably, the system further comprises:

[0153] A judgment module is used to judge whether the number of Wi-Fi networks bound to each POI is greater than a threshold of the number of Wi-Fi networks corresponding to the type of the POI;

[0154] The discarding module is used to discard the binding relationship between the POI and the Wi-Fi if it is greater than.

[0155] Preferably, the system further comprises:

[0156] A graph construction module is used to obtain identification information corresponding to each Wi-Fi in the Wi-Fi cluster, take the identification information of each Wi-Fi as a vertex, connect the vertices corresponding to different Wi-Fi in the Wi-Fi cluster, and construct a connected graph;

[0157] The binding module 503 includes:

[0158] The third determination submodule is configured to determine a target Wi-Fi included in each Wi-Fi cluster according to the connectivity graph.

[0159] The point of interest portrait generation system 500 of the embodiment of the present invention searches for similar Wi-Fi through MAC addresses, and clusters similar Wi-Fi to obtain a cluster of Wi-Fi corresponding to a POI. Even in the face of non-semantic Wi-Fi, the POI can be accurately bound, solving the problem that Wi-Fi-POI portraits cannot achieve non-semantic relationship binding, improving the portrait coverage, and making the subsequent perception algorithm that strongly relies on the portrait more accurate. And through multiple Wi-Fi clusters, the risk of portrait binding errors caused by a single GPS drift error is avoided, and the accuracy is improved.

[0160] As for the above-mentioned system embodiment, since it is basically similar to the method embodiment, the relevant parts may refer to the partial description of the method embodiment.

[0161] An embodiment of the present invention further provides an electronic device, which may be a server. Figure 6As shown, it includes a processor 601 , a communication interface 602 , a memory 603 and a communication bus 604 , wherein the processor 601 , the communication interface 602 , and the memory 603 communicate with each other via the communication bus 604 .

[0162] The memory 603 is used to store computer programs.

[0163] When the processor 601 is used to execute the program stored in the memory 603, the following steps are implemented:

[0164] Acquire multiple Wi-Fi lists reported by the user device through a positioning request, each Wi-Fi list includes one or more media access control addresses (MAC addresses) of Wi-Fi;

[0165] According to the MAC addresses of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar relationships in multiple Wi-Fi lists, and classify the Wi-Fi with similar relationships into a Wi-Fi cluster;

[0166] According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the point of interest POI that is predetermined and authenticity-verified, the binding relationship between other Wi-Fi and the POI in each Wi-Fi cluster except the target Wi-Fi is determined.

[0167] The communication bus mentioned in the above electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0168] The communication interface is used for communication between the above electronic device and other devices.

[0169] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0170] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0171] In another embodiment provided by the present invention, a computer-readable storage medium is also provided, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the method for generating a point of interest portrait described in the above embodiment.

[0172] In another embodiment provided by the present invention, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute the method for generating a portrait of a point of interest described in the above embodiment.

[0173] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk Solid State Disk (SSD)), etc.

[0174] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0175] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0176] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A method for generating an interest point portrait, characterized in that: include: Obtain multiple Wi-Fi lists reported by the user device through a positioning request, each Wi-Fi list includes one or more media access control addresses (MAC addresses) of Wi-Fi; According to the MAC address of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar relationship in multiple Wi-Fi lists, and classify the Wi-Fi with similar relationship into a Wi-Fi cluster; According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that has been predetermined and verified for authenticity, the binding relationship between other Wi-Fi and the POI in each Wi-Fi cluster except the target Wi-Fi is determined; According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the point of interest POI that is predetermined and authenticity-verified, the steps of determining the binding relationship between other Wi-Fi and the POI except the target Wi-Fi in each Wi-Fi cluster include: If the Wi-Fi cluster contains a target Wi-Fi, the POI bound to the target Wi-Fi is determined to be the POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi; If the Wi-Fi cluster includes multiple target Wi-Fis, and there are different POIs among the POIs bound to the multiple target Wi-Fis, a target POI is selected from multiple candidate POIs bound to the multiple target Wi-Fis, and the target POI is determined to be a POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi; The step of selecting a target POI from a plurality of candidate POIs bound to the plurality of target Wi-Fi networks comprises: The other Wi-Fi in the Wi-Fi cluster except the target Wi-Fi is used as the Wi-Fi to be recalled; Scoring the multiple candidate POIs according to one or more of the characteristic information of the Wi-Fi cluster, the characteristic information of the candidate POI, the similar characteristics between the Wi-Fi to be recalled and the candidate POI, and the similar characteristics between the Wi-Fi to be recalled and the target Wi-Fi; Selecting a POI with the highest score and a score exceeding a preset threshold from the multiple candidate POIs as a target POI; The characteristic information of the Wi-Fi cluster includes: the number of Wi-Fis included in the Wi-Fi cluster; The characteristic information of the candidate POI includes: one or more of the number of candidate POIs and the Wi-Fi number threshold corresponding to the type of the candidate POI; The similar features between the Wi-Fi to be recalled and the candidate POI include: the similarity between the names of the Wi-Fi to be recalled and the candidate POI; The similar features between the Wi-Fi to be recalled and the target Wi-Fi include: one or more of: MAC address similarity between the Wi-Fi to be recalled and the target Wi-Fi, name similarity between the Wi-Fi to be recalled and the target Wi-Fi, signal strength difference between the Wi-Fi to be recalled and the target Wi-Fi, and the number of times the Wi-Fi to be recalled and the target Wi-Fi appear together in the same Wi-Fi list.

2. The method for generating a point of interest portrait according to claim 1, characterized in that: The steps of searching for Wi-Fis with similar relationships in multiple Wi-Fi lists according to the MAC addresses of the Wi-Fis in the Wi-Fi list, and assigning the Wi-Fis with similar relationships to a Wi-Fi cluster include: According to the MAC address of the Wi-Fi in the Wi-Fi list, search for Wi-Fi with similar MAC address in each Wi-Fi list, and classify the Wi-Fi with similar MAC address in each Wi-Fi list into one Wi-Fi set; In the Wi-Fi sets of the multiple Wi-Fi lists, a Wi-Fi set including the same Wi-Fi is searched, and all Wi-Fis in the Wi-Fi set including the same Wi-Fi are classified into one Wi-Fi cluster.

3. The method for generating a point of interest portrait according to claim 2, characterized in that: Based on the MAC address of the Wi-Fi in the Wi-Fi list, the steps for finding the Wi-Fi with a similar MAC address in each Wi-Fi list include: Compare the MAC addresses of different Wi-Fi in each Wi-Fi list; If the MAC addresses of two Wi-Fi networks are identical in bytes except for the last one or two bytes, it is determined that the MAC addresses of the two Wi-Fi networks are similar.

4. The method for generating a point of interest portrait according to claim 2, characterized in that: Among the Wi-Fi sets in multiple Wi-Fi lists, the steps of searching for a Wi-Fi set that includes the same Wi-Fi include: The multiple Wi-Fi lists are traversed in units of preset areas, and a Wi-Fi set including the same Wi-Fi is searched in the Wi-Fi sets of the multiple Wi-Fi lists traversed in each unit.

5. The method for generating a point of interest portrait according to claim 1, characterized in that: After determining the binding relationship between other Wi-Fi and POI except the target Wi-Fi in each Wi-Fi cluster according to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and POI that is predetermined and authenticity-verified, the method further includes: Determine whether the number of Wi-Fi networks bound to each POI is greater than the Wi-Fi number threshold corresponding to the type of the POI; If it is greater, the binding relationship between the POI and Wi-Fi will be discarded.

6. The method for generating a point of interest portrait according to claim 1, characterized in that: After the Wi-Fis with similar relationships are grouped into one Wi-Fi cluster, the method further includes: Obtain identification information corresponding to each Wi-Fi in the Wi-Fi cluster, take the identification information of each Wi-Fi as a vertex, connect the vertices corresponding to different Wi-Fi in the Wi-Fi cluster, and construct a connectivity graph; According to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the point of interest POI that is predetermined and authenticity-verified, the steps of determining the binding relationship between other Wi-Fi and the POI except the target Wi-Fi in each Wi-Fi cluster include: According to the connectivity graph, a target Wi-Fi included in each Wi-Fi cluster is determined.

7. A point of interest portrait generation system, characterized in that: include: A first acquisition module is used to acquire multiple Wi-Fi lists reported by the user equipment through a positioning request, each Wi-Fi list includes one or more media access control addresses (MAC addresses) of Wi-Fi; A clustering module is constructed, which is used to search for Wi-Fis with similar relationships in multiple Wi-Fi lists according to the MAC addresses of the Wi-Fis in the Wi-Fi list, and classify the Wi-Fis with similar relationships into a Wi-Fi cluster; A binding module, used to determine the binding relationship between other Wi-Fi and POI except the target Wi-Fi in each Wi-Fi cluster according to the target Wi-Fi included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi and the POI that is predetermined and authenticity-verified; The step of determining the binding relationship between other Wi-Fi networks except the target Wi-Fi network in each Wi-Fi cluster and the POI according to the target Wi-Fi network included in each Wi-Fi cluster and the binding relationship between the target Wi-Fi network and the POI that is predetermined and authenticity-verified includes: If the Wi-Fi cluster contains a target Wi-Fi, the POI bound to the target Wi-Fi is determined to be the POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi; If the Wi-Fi cluster includes multiple target Wi-Fis, and there are different POIs among the POIs bound to the multiple target Wi-Fis, a target POI is selected from multiple candidate POIs bound to the multiple target Wi-Fis, and the target POI is determined to be a POI bound to other Wi-Fis in the Wi-Fi cluster except the target Wi-Fi; The step of selecting a target POI from a plurality of candidate POIs bound to the plurality of target Wi-Fi networks comprises: The other Wi-Fi in the Wi-Fi cluster except the target Wi-Fi is used as the Wi-Fi to be recalled; Scoring the multiple candidate POIs according to one or more of the characteristic information of the Wi-Fi cluster, the characteristic information of the candidate POI, the similar characteristics between the Wi-Fi to be recalled and the candidate POI, and the similar characteristics between the Wi-Fi to be recalled and the target Wi-Fi; Selecting a POI with the highest score and a score exceeding a preset threshold from the multiple candidate POIs as a target POI; The characteristic information of the Wi-Fi cluster includes: the number of Wi-Fis included in the Wi-Fi cluster; The characteristic information of the candidate POI includes: one or more of the number of candidate POIs and the Wi-Fi number threshold corresponding to the type of the candidate POI; The similar features between the Wi-Fi to be recalled and the candidate POI include: the similarity between the names of the Wi-Fi to be recalled and the candidate POI; The similar features between the Wi-Fi to be recalled and the target Wi-Fi include: one or more of: MAC address similarity between the Wi-Fi to be recalled and the target Wi-Fi, name similarity between the Wi-Fi to be recalled and the target Wi-Fi, signal strength difference between the Wi-Fi to be recalled and the target Wi-Fi, and the number of times the Wi-Fi to be recalled and the target Wi-Fi appear together in the same Wi-Fi list.

8. An electronic device, characterized in that: include: A processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the steps in the method for generating a point of interest portrait as described in any one of claims 1 to 6 when executing the program stored in the memory.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for generating a point of interest portrait as described in any one of claims 1 to 6 is implemented.

Images