Point of interest state determination method, model training method, and computer storage medium

Abstract

Embodiments of the present application provide a point of interest state determination method, a model training method and a computer storage medium, wherein the point of interest state determination method comprises: determining first navigation data of a first navigation of a navigation object and second navigation data of a second navigation corresponding to the first navigation according to historical navigation session data of the navigation object, wherein a distance between a starting point in the second navigation data and an ending point in the first navigation data of the first navigation is less than a preset distance threshold; obtaining first navigation data of a first navigation corresponding to a same ending point according to the ending point in the first navigation data of the first navigation of different navigation objects; determining a second navigation feature of a point of interest corresponding to the ending point based on the first navigation data of the first navigation of the same ending point and the second navigation data of the second navigation corresponding to the first navigation; and determining a state of the point of interest corresponding to the ending point based on the second navigation feature of the point of interest corresponding to the ending point.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of data mining, and particularly relate to a method for determining a state of a point of interest and a computer storage medium. BACKGROUND

[0002] With the development of the Internet, electronic maps provide convenience for users' travel. Among them, a point of interest (POI) is an important part of an electronic map. Generally, a POI corresponds to a geographical object in the real world that can perform a specific activity, including institutions, stores, facilities, or places, etc.

[0003] With the passage of time, the state of a geographical object in the real world will change, such as closing down, demolition, suspension of business, relocation, etc., and these changes will cause the corresponding POI to be invalid, i.e., the geographical object in the real world corresponding to the POI no longer exists or stops operating. Taking a store as an example, if the store has closed down or relocated in the real world, the POI needs to be marked as invalid or deleted in the electronic map, so that users will not search for the POI in the electronic map or even if the user searches for the POI, the user can know that the POI is in an invalid state. If the POI is not marked as invalid or deleted in the electronic map, when the user travels to the geographical object corresponding to the POI, the user will find that the store has closed down or relocated, which will greatly damage the user experience. Therefore, an electronic map provider needs to timely dig out the POIs that have become invalid and update the POIs in the electronic map in a timely manner, so that the POIs in the electronic map are consistent with the real situation of the geographical objects in the real world, and the accuracy and effectiveness of the data provided by the electronic map to users are ensured. SUMMARY

[0004] Therefore, embodiments of the present application provide a state determination scheme for a point of interest to at least partially solve the above problems.

[0005] According to a first aspect of embodiments of the present application, a method for determining a state of a point of interest is provided, including: determining first navigation data of a first navigation of a navigation object and second navigation data of a second navigation corresponding to the first navigation according to historical navigation session data of the navigation object, wherein a distance between a starting point in the second navigation data and an ending point in the first navigation data is less than a preset distance threshold; obtaining first navigation data of a first navigation corresponding to a same ending point according to the ending point in the first navigation data of the first navigation of different navigation objects; determining a second navigation feature of a point of interest corresponding to the same ending point based on the first navigation data of the first navigation of the same ending point and the second navigation data of the second navigation corresponding to the first navigation; and determining a state of the point of interest corresponding to the same ending point based on the second navigation feature of the point of interest corresponding to the same ending point.

[0006] According to a second aspect of embodiments of the present application, a method for training a machine learning model is provided, including: obtaining a point of interest that has failed and a point of interest that has not failed as sample points of interest, and sample second navigation features of the sample points of interest, wherein the sample second navigation features are determined based on first navigation data of a first navigation of a sample point of interest and second navigation data of a second navigation corresponding to the first navigation, and a distance between a starting point in the second navigation data and the sample point of interest in the first navigation data is less than a preset distance threshold; and training the machine learning model by taking the sample second navigation features as input, so that the machine learning model outputs state information of the sample points of interest.

[0007] According to a third aspect of embodiments of the present application, an electronic device is provided, including: a processor, a memory, a communication interface and a communication bus, the processor, the memory and the communication interface complete communication with each other through the communication bus; the memory is used to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the method according to the first aspect or the second aspect.

[0008] According to a fourth aspect of embodiments of the present application, a computer storage medium is provided, and the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method according to the first aspect or the second aspect.

[0009] According to the state determination scheme of the interest point provided in the embodiments of the present application, the scene that the navigation object initiates a navigation request again, i.e., performs secondary navigation, to the target interest point after navigating to the interest point is used as the basis for mining the invalid interest point. Based on this, the embodiments of the present application obtain the first navigation data and the second navigation data corresponding to the same terminal point according to the historical navigation session data of the navigation object, to determine the secondary navigation feature of the interest point corresponding to the terminal point, and further determine the state of the interest point corresponding to the terminal point. In this way, the phenomenon that the state of the interest point cannot be effectively determined in the prior art when the data such as shipping data is missing on some interest points (such as gas stations, automatic teller machines, etc.) is remedied. Because the data related to navigation in the historical navigation session data of the navigation object is more accurate and timely, the state of the interest point can be more effectively and accurately determined. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0011] Figure 1 A flowchart of a state determination method of an interest point in the embodiments of the present application;

[0012] Figure 2 A flowchart of another state determination method of an interest point in the embodiments of the present application;

[0013] Figure 3 A flowchart of a training method of a machine learning model in the embodiments of the present application;

[0014] Figure 4 A structural schematic diagram of an electronic device in the embodiments of the present application. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments in the present application should belong to the scope of protection of the present application.

[0016] The specific implementation of the embodiments of the present application will be further described below with reference to the drawings of the embodiments of the present application.

[0017] Referring to Figure 1 , Figure 1 is a flowchart of a point of interest state determination method in an embodiment of the present application.

[0018] The point of interest state determination method of the present embodiment can include the following steps:

[0019] In step S110, first navigation data of a first navigation and second navigation data of a second navigation corresponding to the first navigation of the navigated object are determined according to historical navigation session data of the navigated object.

[0020] The navigation session data is session data generated by the navigated object in the process of using the map software for navigation, which can include the start time, end time, starting point, ending point and navigation route of the navigation, etc. In the present embodiment, the historical navigation session data is session data generated by multiple navigated objects in the navigation process in a historical time period, for example, navigation session data in the last 1 month, 2 months or 3 months, etc.

[0021] In actual life, there are cases where a certain location fails due to various reasons, resulting in the navigated object being unable to achieve the intended purpose when reaching the location through the first navigation, and needing to perform a second navigation. At this time, the previous navigation can be considered as the first navigation, and the subsequent navigation can be considered as the second navigation corresponding to the first navigation. The two navigations can correspond to different navigation session data, or can correspond to the same navigation session data. The first navigation data corresponding to the first navigation and the second navigation data corresponding to the second navigation can be obtained from the navigation session data.

[0022] The obtained navigation data at least includes information of the endpoint of the first navigation, information of the starting point and endpoint of the second navigation, and information of the point of interest corresponding to the endpoint. Optionally, it can also include information such as the type of the point of interest corresponding to the endpoint of the first navigation, the type of the point of interest corresponding to the endpoint of the second navigation, and the start time and end time corresponding to the two navigations, etc.

[0023] Among the above data, the distance between the starting point of the second navigation data requiring the second navigation and the endpoint of the first navigation data is less than a preset distance threshold. Since the second navigation is the navigation of the navigated object to another point of interest soon (for example, within 1 hour or 2 hours) after navigating to a point of interest in a trip, the preset distance threshold can be set by those skilled in the art according to actual needs. Since the two navigations are usually very close, the preset distance threshold can be set within a small distance range. For example, the preset distance threshold can be 20 meters, 50 meters, etc., which is not limited in the present embodiment.

[0024] It should be noted that in the embodiments of the present application, "a plurality of", "a plurality of" and "a plurality of" related to "a plurality of" are intended to mean two and more than two.

[0025] In step S120, the first navigation data of the first navigation corresponding to the same terminal point is obtained according to the terminal point in the first navigation data of the first navigation of different navigated objects.

[0026] As described above, in the embodiments of the present application, the historical navigation session data is the data of a plurality of navigated objects. Therefore, the corresponding first navigation data and the corresponding terminal point can be extracted from the historical navigation session data of each navigated object. However, for different first navigation data, the terminal points may be different, and the same terminal point and the first navigation data corresponding to the terminal point need to be selected.

[0027] In one possible way, this step can be implemented as follows: clustering the terminal points in the first navigation data of the first navigation of different navigated objects according to the terminal point identifiers of the terminal points in the first navigation data, obtaining clustered terminal points; and obtaining the first navigation data of the first navigation corresponding to the same clustered terminal point for different clustered terminal points. By clustering the terminal points in the first navigation data, the terminal points (with the same terminal point identifier) that may have different literal expressions but are actually the same place are aggregated, which on the one hand facilitates the subsequent data processing, and on the other hand makes the determination of the terminal point of the first navigation more accurate.

[0028] In step S130, the secondary navigation feature of the interest point corresponding to the terminal point is determined based on the first navigation data of the first navigation of the same terminal point and the second navigation data of the second navigation corresponding to the first navigation.

[0029] The secondary navigation feature of the interest point corresponding to the terminal point is used to indicate the feature of the navigated object based on the interest point corresponding to the terminal point for re-navigation, including but not limited to the feature of the interest point corresponding to the terminal point. It should be noted that for a navigation, the terminal point may itself be an interest point; and if the terminal point is not an interest point, the interest point closest to the terminal point can be taken as the interest point corresponding to the terminal point.

[0030] In one feasible approach, this step can be implemented as follows: based on the first navigation data of the first navigation to the same destination, determine the type of point of interest corresponding to that destination, the number of navigations, and the first time of navigation to that destination; based on the second navigation data of the second navigation, determine the type of point of interest corresponding to the destination of the second navigation and the second time of navigation to that destination; based on the number of navigations, the type of point of interest corresponding to the destination of the first navigation, the type of point of interest corresponding to the destination of the second navigation, the first time, and the second time, determine the secondary navigation characteristics of the destination of the first navigation. This information effectively characterizes and determines the secondary navigation characteristics of the destination of the first navigation.

[0031] Optionally, determining the secondary navigation characteristics of the endpoint of the first navigation based on the navigation count, the type of the point of interest corresponding to the endpoint of the first navigation, the type of the point of interest corresponding to the endpoint of the second navigation, the first time, and the second time can be achieved by: determining the secondary navigation characteristics of the endpoint of the first navigation based on the navigation count, the number of times the type of the point of interest corresponding to the endpoint of the second navigation is the same as the type of the point of interest corresponding to the endpoint of the first navigation, and the time difference between the second time and the first time. The more navigation counts, the more times the type is the same, and the shorter the time difference, the more likely it is that the second navigation was performed when the first navigation was unsuccessful; that is, the point of interest corresponding to the endpoint of the first navigation may be invalid. Therefore, this method allows for a more objective and accurate determination of the secondary navigation characteristics.

[0032] Step S140: Determine the state of the point of interest corresponding to the endpoint based on the secondary navigation features of the point of interest corresponding to the endpoint.

[0033] As mentioned earlier, the secondary navigation features of the endpoint include features specific to the point of interest (POI) corresponding to that endpoint. Therefore, based on the secondary navigation features of the POI corresponding to the endpoint, the state of the POI can be determined, such as whether it is in a failed state or still in a valid state.

[0034] In one feasible approach, this step can be implemented as follows: obtaining the number of times that the type of the point of interest corresponding to the endpoint of the second navigation is the same as the type of the point of interest corresponding to the endpoint of the first navigation; obtaining a first ratio of the number of times the type of the point of interest is the same as ...

[0035] If the interest point is invalid, the types of the interest points corresponding to the end points of the two navigations should be the same; the number of times of the same should also be proportional to the corresponding number of times of navigation, and the time difference between the two navigations (the time difference between the end time of the first navigation and the start time of the second navigation) should be within a short preset time difference, and the number of times that the time difference is less than the preset time difference should also be proportional to the corresponding number of times of navigation. Based on this, the state of the interest point corresponding to the end point can be determined relatively accurately.

[0036] For example, if the first ratio is greater than a first threshold (for example, it can be 0.6, 0.7, etc.), and the second ratio is greater than a second threshold (for example, it can be 0.8, 0.9, etc.), it can be determined that the interest point corresponding to the end point is an invalid interest point, otherwise, it can be determined that the interest point corresponding to the end point is not an invalid interest point. Wherein, the preset time difference can be set by those skilled in the art according to the actual situation, for example, it can be 1 hour, 2 hours, or a few minutes, a few tens of minutes, etc., and the embodiments of the present application do not limit this.

[0037] Optionally, the number of times that the type of the interest point corresponding to the end point of the second navigation is the same as the type of the interest point corresponding to the end point of the first navigation can include: obtaining the number of times that the type of the interest point corresponding to the end point of the second navigation and the type of the interest point corresponding to the end point of the first navigation are the same in the same level in the preset type hierarchical relationship. Through the form of the preset type hierarchical relationship, the number of times that the types are the same can be more accurately determined.

[0038] Wherein, the type of the interest point can include multiple levels, for example, there can be a large class of interest points, and multiple middle classes can be included under the large class, and each middle class can include multiple small classes, and so on. Of course, the large class, the middle class and the small class are exemplary, and more levels can be included in actual application. Alternatively, in a feasible manner, the hierarchical relationship of the type of the interest point can be implemented as a tree structure. Except for the root node, each level in other levels can include multiple nodes, i.e. types.

[0039] For example, for the catering class, the corresponding preset type hierarchical relationship can be: the large class is the catering class, the middle class is Chinese food, western food, etc., the small class corresponding to Chinese food can include: Sichuan cuisine, Cantonese cuisine, Shanghai cuisine, etc., and the small class corresponding to western food includes: French cuisine, English cuisine, Italian cuisine, etc. Assuming that the name of a certain interest point is **Sichuan Restaurant, the large class of the interest point type of the interest point is the catering class, the middle class is Chinese food, and the small class is Sichuan cuisine.

[0040] Correspondingly, when the number of times that the end point of the second navigation corresponds to the same type of the interest point and the end point of the first navigation corresponds to the same type of the interest point in the preset type hierarchical relationship is determined, the determination can be made for a specific level, or the determination can be made level by level. For example, in the case where the preset type hierarchical relationship includes three levels, the number of times that the types are the same can include the number of times that the large categories are the same, the number of times that the medium categories are the same, and the number of times that the small categories are the same.

[0041] Through the above information and process, the state of the interest point corresponding to the end point of the first navigation can be determined, such as invalid or valid.

[0042] According to the interest point state determination method provided in the embodiments of the present application, based on the fact that the navigation object will usually initiate a navigation request again, that is, perform secondary navigation, to the target interest point after navigating to the interest point that has expired, the secondary navigation scenario is used as a basis for mining invalid interest points. Based on this, the embodiments of the present application obtain first navigation data and second navigation data corresponding to the same end point according to the historical navigation session data of the navigation object, to determine the secondary navigation characteristics of the interest point corresponding to the end point, and further determine the state of the interest point corresponding to the end point. In this way, the phenomenon that the state of the interest point cannot be effectively determined in the prior art when some data such as shipping data is missing on some interest points (such as gas stations, automatic teller machines, etc.) is remedied. Because the data related to navigation in the historical navigation session data of the navigation object is more accurate and timely, the state of the interest point can be more effectively and accurately determined.

[0043] Referring to Figure 2 , Figure 2 is a flowchart of another interest point state determination method in the embodiments of the present application.

[0044] In this embodiment, the interest point state determination method provided in the embodiments of the present application is described by taking determination of the state of the interest point through a machine learning model as an example.

[0045] The interest point state determination method in this embodiment can include the following steps:

[0046] In step S210, first navigation data of the first navigation of the navigation object and second navigation data of the second navigation corresponding to the first navigation are determined according to the historical navigation session data of the navigation object.

[0047] In the second navigation data of the second navigation, the distance between the start point and the end point in the first navigation data of the first navigation is less than a preset distance threshold.

[0048] In step S220, first navigation data corresponding to the same end point is obtained according to the end point in the first navigation data of the first navigation of different navigation objects.

[0049] In step S230, the second navigation feature of the interest point corresponding to the terminal point is determined based on the first navigation data of the first navigation of the same terminal point and the second navigation data of the second navigation corresponding to the first navigation.

[0050] The specific implementation of steps S210-S230 can refer to the description of the corresponding part in the foregoing embodiments, which will not be described here again.

[0051] In step S240, the second navigation feature of the interest point corresponding to the terminal point is taken as the input of the machine learning model pre-trained for state judgment of the interest point, and the state of the interest point corresponding to the terminal point is predicted by the machine learning model.

[0052] In the embodiments of the present application, the machine learning model is used for state judgment of the interest point. In a feasible manner, the machine learning model can implement the state judgment as a binary classification problem, and thus various algorithms for binary classification can be used, such as logistic regression, SVM (Support Vector Machines), random forest, XGBoost (eXtreme Gradient Boosting), CNN (Convolutional Neural Network), etc.

[0053] In addition, the output of the machine learning model can adopt a suitable form according to actual needs, such as a form of probability or Boolean value, etc. Preferably, the probability is adopted to accurately express the state of the interest point that may be invalid or valid. Based on this, in a feasible manner, predicting the state of the interest point corresponding to the terminal point by the machine learning model can be implemented as follows: obtaining the state vector of the interest point corresponding to the terminal point of the first navigation output by the machine learning model, the state vector at least including the invalid state probability; determining the state of the interest point corresponding to the terminal point according to the comparison result of the invalid state probability and the first preset threshold. The first preset threshold can be set by those skilled in the art according to actual needs, and the embodiments of the present application do not limit this.

[0054] For example, if the state vector output by the machine learning model is (0.8, 0.2), it means that the probability of the interest point being an invalid interest point is 0.8 and the probability of the interest point not being an invalid interest point is 0.2. If the first preset threshold is 0.6, it can be determined that the interest point is an invalid interest point. If the invalid vector of the interest point corresponding to the terminal point is (0.3, 0.7), it can be determined that the interest point is not an invalid interest point.

[0055] Although it can be roughly determined according to the state vector whether the interest point is a failed interest point, in order to further improve the accuracy of the judgment, in one possible way, further comprising: if it is determined that the interest point corresponding to the terminal point is in a failed state according to the comparison result of the failed state probability and the first preset threshold, judging the relationship between the failed state probability and a second preset threshold; if the failed state probability is greater than the second preset threshold, deleting the interest point corresponding to the terminal point, wherein the second preset threshold is greater than the first preset threshold; or if the failed state probability is not greater than the second preset threshold, displaying the interest point corresponding to the terminal point and a deletion option for the interest point corresponding to the terminal point, and deleting the interest point corresponding to the terminal point after receiving an operation for the deletion option.

[0056] For example, in some cases, according to the failed state probability in the state vector, it can not be determined whether the interest point is a failed interest point, for example, the state vector of the interest point corresponding to the terminal point is (0.65, 0.35), then it cannot be accurately determined that the interest point is necessarily a failed interest point. In this case, the interest point can be further filtered to determine whether its failed state probability is greater than a second preset threshold (the specific value of the second preset threshold can be flexibly set by those skilled in the art according to actual needs, which is greater than the first preset threshold).

[0057] If the failed state probability of the interest point corresponding to the terminal point is greater than the second preset threshold, the interest point can be directly deleted; and if the failed state probability is not greater than the second preset threshold, the information of the interest point can be displayed, and the interest point is deleted after receiving a deletion instruction of the interest point from the user. Thus, with the help of manual judgment, the state of the interest point can be more accurately determined.

[0058] For example, if the first preset threshold is 0.6 and the second preset threshold is 0.8, assuming that the state vector corresponding to the interest point corresponding to a certain terminal point is (0.9, 0.1), since its failed state probability 0.9 is greater than the second preset threshold 0.8, the interest point can be directly deleted. If the state vector corresponding to the interest point is (0.75, 0.25), since 0.75 is less than 0.8, the information of the interest point can be displayed to the platform maintenance personnel, and the operation is performed in a manual manner to determine whether the interest point is a failed interest point and whether to delete it. If the manual selection is a deletion option, the interest point can be deleted.

[0059] Through the embodiment, based on the scenario that the navigation object initiates a navigation request again, i.e., performs secondary navigation, to the target interest point after navigating to the interest point that expires, the expired interest point is mined. Based on this, the embodiment of the application obtains first navigation data and second navigation data corresponding to the same terminal point according to historical navigation session data of the navigation object, to determine secondary navigation features of the interest point corresponding to the terminal point, and further determine the state of the interest point corresponding to the terminal point. Thus, the phenomenon that the state of the interest point cannot be effectively determined in the prior art in the case of missing data such as shipping data on some interest points (such as gas stations, automatic teller machines, etc.) is remedied. Because the data related to navigation in the historical navigation session data of the navigation object is more accurate and timely, the state of the interest point can be more effectively and accurately determined.

[0060] And by means of the machine learning model, the state of the interest point can be more accurately and reasonably determined.

[0061] Reference is made to Figure 3 , Figure 3 A flowchart of a training method of a machine learning model of the embodiment of the application.

[0062] In the embodiment, the training process of the machine learning model is mainly described, especially the sample data acquisition process in the training process.

[0063] The training method of the machine learning model of the embodiment can include the following steps:

[0064] In step S310, expired and unexpired interest points are obtained as sample interest points, and sample secondary navigation features of the sample interest points are obtained.

[0065] The sample secondary navigation features are determined based on first navigation data of the first navigation of the sample interest point and second navigation data of the second navigation corresponding to the first navigation. The distance between the starting point in the second navigation data of the second navigation and the sample interest point in the first navigation data of the first navigation is less than a preset distance threshold. The preset distance threshold can be flexibly set by those skilled in the art according to actual needs, and the embodiment of the application does not limit this.

[0066] In the embodiment of the application, the expiration of the interest point includes expired and unexpired, and the expired interest point can be selected as a positive sample and the unexpired interest point can be selected as a negative sample. The positive sample can be an expired interest point manually selected in a recent historical period, and the negative sample can also be an interest point that is verified and is currently in use. In this way, both the positive sample and the negative sample can have a high accuracy. In order to make the trained machine learning model have high accuracy, the ratio of the positive sample to the negative sample can be equal to or close to 1:1.

[0067] The sample secondary navigation feature of the sample interest point in the embodiment is used to indicate the feature of the object being navigated based on the sample interest point for re-navigation.

[0068] In an available manner, the sample secondary navigation feature of the sample interest point can be implemented as follows: obtaining the first navigation data of the first navigation of the sample interest point and the second navigation data of the second navigation corresponding to the first navigation; determining the type corresponding to the sample interest point, the navigation times and the first time of navigation to the sample interest point based on the first navigation data of the first navigation; determining the type of the interest point corresponding to the end point of the second navigation and the second time of navigation to the end point based on the second navigation data of the second navigation; and determining the sample secondary navigation feature of the sample interest point according to the navigation times, the type of the sample interest point, the type of the interest point corresponding to the end point of the second navigation, the first time and the second time.

[0069] Further optionally, the determination of the sample secondary navigation feature of the sample interest point according to the navigation times, the type of the sample interest point, the type of the interest point corresponding to the end point of the second navigation, the first time and the second time can be implemented as follows: determining the sample secondary navigation feature of the sample interest point according to the navigation times, the times of the type of the interest point corresponding to the end point of the second navigation being the same as the type of the sample interest point, and the time difference between the second time and the first time.

[0070] It should be noted that the above process is described briefly, and in the specific implementation, the description of the generation of the secondary navigation feature for the interest point corresponding to the end point in the foregoing embodiments can be referred to, and will not be described herein.

[0071] In step S320, the sample secondary navigation feature is taken as the input to train the machine learning model, so that the machine learning model outputs the state information of the sample interest point.

[0072] In an available manner, the output of the machine learning model can be a state vector.

[0073] Based on this, the machine learning model can be trained with the sample secondary navigation feature as the input and the state vector corresponding to the sample interest point as the output, wherein the state vector includes the probability that the sample interest point is a failure interest point and the probability that the sample interest point is not a failure interest point.

[0074] For example, if the state vector is (1, 0), it means that the corresponding sample interest point is a failure interest point, that is, the probability that it is a failure interest point is 1, and (0, 1) means that the sample interest point is not a failure interest point, that is, the probability that it is a failure interest point is 0. Conversely, (0, 1) can also mean a failure interest point, and (1, 0) means a non-failure interest point.

[0075] After the machine learning model is trained, it can be applied to the interest point state determination method described in the foregoing embodiments to output accurate interest point state information.

[0076] Referring to Figure 4 , Figure 4 The electronic device is a structure schematic diagram in the embodiments of the present application, and the specific implementation of the electronic device is not limited in the embodiments of the present application.

[0077] As Figure 4 shown, the electronic device can include a processor 402, a communications interface 404, a memory 406, and a communications bus 408.

[0078] Among them:

[0079] The processor 402, the communications interface 404, and the memory 406 complete mutual communication through the communications bus 408.

[0080] The communications interface 404 is configured to communicate with other electronic devices or servers.

[0081] The processor 402 is configured to execute the program 410, and specifically can execute the related steps described in any of the method embodiments.

[0082] Specifically, the program 410 can include program code including computer operation instructions.

[0083] The processor 402 can be a 4CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or an integrated circuit configured to implement one or more embodiments of the present application. One or more processors included in the smart device can be the same type of processor, such as one or more CPUs; or can be different types of processors, such as one or more CPUs and one or more ASICs.

[0084] The memory 406 is configured to store the program 410. The memory 406 can include a high-speed RAM memory, and can also include a non-volatile memory such as at least one disk memory.

[0085] The specific implementation of each step in the program 410 can refer to the corresponding description in the corresponding steps and units of the method described in any of the above method embodiments, and has corresponding beneficial effects, which will not be described here. Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described devices and modules can refer to the corresponding process description in the foregoing method embodiments, which will not be described here.

[0086] It should be noted that, according to the needs of implementation, each component / step described in the embodiments of the present application can be split into more components / steps, or two or more components / steps or part of the operation of the components / steps can be combined into a new component / steps, to achieve the purpose of the embodiments of the present application.

[0087] The above method according to the embodiments of the present application can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium such as a CD ROM, a RAM, a floppy disk, a hard disk or an optical disk, or be implemented by computer code originally stored in a remote recording medium or a non-transitory machine readable medium downloaded through a network and stored in a local recording medium, so that the method described herein can be processed by such software on a recording medium using a general computer, a special processor or programmable or special hardware such as ASIC or FPGA. It can be understood that the computer, processor, microprocessor controller or programmable hardware includes a storage component (for example, RAM, ROM, flash memory, etc.) that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, processor or hardware, the method described herein is implemented. In addition, when a general computer accesses the code for implementing the method shown herein, the execution of the code will convert the general computer into a special computer for executing the method shown herein.

[0088] Those of ordinary skill in the art can realize that the units and method steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of the present application.

[0089] The above implementation is only used to illustrate the embodiments of the present application, and is not limited to the embodiments of the present application. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present application, therefore all equivalent technical solutions also belong to the scope of the embodiments of the present application, and the patent protection scope of the embodiments of the present application should be defined by the claims.

Claims

1. A method for determining a state of a point of interest, comprising: determining first navigation data of a first navigation of a navigation object and second navigation data of a second navigation corresponding to the first navigation according to historical navigation session data of the navigation object, wherein a distance between a start point in the second navigation data and an end point in the first navigation data is less than a preset distance threshold; obtaining first navigation data of a first navigation corresponding to a same end point according to end points in the first navigation data of the first navigation of different navigation objects; determining a second navigation feature of a point of interest corresponding to the end point based on the first navigation data of the first navigation of the same end point and the second navigation data of the second navigation corresponding to the first navigation, wherein the second navigation feature of the point of interest corresponding to the end point is used to indicate a feature of the navigation object for re-navigation based on the point of interest corresponding to the end point; and determining a state of the point of interest corresponding to the end point based on the second navigation feature of the point of interest corresponding to the end point.

2. The method of claim 1, wherein, The determining of the second navigation feature of the point of interest corresponding to the end point based on the first navigation data of the first navigation of the same end point and the second navigation data of the second navigation corresponding to the first navigation comprises: determining a type of the point of interest corresponding to the end point, a navigation frequency and a first time of navigation to the end point based on the first navigation data of the first navigation of the same end point; determining a type of the point of interest corresponding to the end point of the second navigation and a second time of navigation to the end point based on the second navigation data of the second navigation; determining the second navigation feature of the end point of the first navigation according to the navigation frequency, the type of the point of interest corresponding to the end point of the first navigation, the type of the point of interest corresponding to the end point of the second navigation, the first time and the second time.

3. The method of claim 2, wherein, The determining of the second navigation feature of the end point of the first navigation according to the navigation frequency, the type of the point of interest corresponding to the end point of the first navigation, the type of the point of interest corresponding to the end point of the second navigation, the first time and the second time comprises: determining the second navigation feature of the end point of the first navigation according to the navigation frequency, a frequency of the type of the point of interest corresponding to the end point of the second navigation being same as the type of the point of interest corresponding to the end point of the first navigation, and a time difference between the second time and the first time.

4. The method of claim 3, wherein, The determining of the state of the point of interest corresponding to the end point based on the second navigation feature of the point of interest corresponding to the end point comprises: obtaining a frequency of the type of the point of interest corresponding to the end point of the second navigation being same as the type of the point of interest corresponding to the end point of the first navigation; obtaining a first ratio of the frequency to a navigation frequency of the point of interest corresponding to the end point of the first navigation; obtaining a second ratio of a frequency of the time difference being less than a preset time difference to the navigation frequency of the point of interest corresponding to the end point of the first navigation; and determining the state of the point of interest corresponding to the end point according to the first ratio and the second ratio.

5. The method of claim 4, wherein, The obtaining of the frequency of the type of the point of interest corresponding to the end point of the second navigation being same as the type of the point of interest corresponding to the end point of the first navigation comprises: The number of times that the second end point corresponding to the interest point and the first end point corresponding to the interest point are of the same type in the preset type hierarchical relationship and are in the same level.

6. The method of any one of claims 1-3, wherein, The first navigation data corresponding to the same end point is obtained according to the end point in the first navigation data of the first navigation of different navigated objects, and the method comprises the steps of: According to the end point identifier of the end point in the first navigation data, the end points in the first navigation data of the first navigation of different navigated objects are clustered to obtain clustered end points; For different clustered end points, the first navigation data corresponding to the same clustered end point is obtained.

7. The method of any one of claims 1-3, wherein, The state of the interest point corresponding to the end point is determined based on the secondary navigation feature of the end point, and the method comprises the steps of: The secondary navigation feature of the interest point corresponding to the end point is taken as the input of the machine learning model which is pre-trained to judge the state of the interest point, and the state of the interest point corresponding to the end point is predicted by the machine learning model.

8. The method of claim 7, wherein, The state of the interest point corresponding to the end point is predicted by the machine learning model, and the method comprises the steps of: An interest point state vector corresponding to the end point of the first navigation is output by the machine learning model, and the state vector at least includes a failure state probability; According to the comparison result of the failure state probability and the first preset threshold, the state of the interest point corresponding to the end point is determined.

9. The method of claim 8, wherein, The method further comprises the steps of: If it is determined that the interest point corresponding to the end point is in a failure state according to the comparison result of the failure state probability and the first preset threshold, the relationship between the failure state probability and a second preset threshold is determined; If the failure state probability is greater than the second preset threshold, the interest point corresponding to the end point is deleted, wherein the second preset threshold is greater than the first preset threshold; Or, if the failure state probability is not greater than the second preset threshold, the interest point corresponding to the end point and a deletion option for the interest point corresponding to the end point are displayed, and the interest point corresponding to the end point is deleted after receiving an operation for the deletion option.

10. A machine learning model training method, wherein, The method comprises the steps of: Obtaining the interest points that have failed and have not failed as sample interest points, and sample secondary navigation features of the sample interest points, wherein the sample secondary navigation features are determined based on the first navigation data of the first navigation of the sample interest points and the second navigation data of the second navigation corresponding to the first navigation, and the distance between the start point in the second navigation data and the sample interest point in the first navigation data is less than a preset distance threshold; The sample secondary navigation features are taken as the input to train the machine learning model, so that the machine learning model outputs the state information of the sample interest points, wherein the secondary navigation feature of the interest point corresponding to the end point is used to indicate the feature of the navigated object based on the interest point corresponding to the end point for re-navigation.

11. The method of claim 10, wherein, The sample secondary navigation features of the sample interest points are obtained, and the method comprises the steps of: The first navigation data of the first navigation of the sample interest points and the second navigation data of the second navigation corresponding to the first navigation are obtained; determine a type of the sample interest point, a navigation frequency and a first time of navigation to the sample interest point based on first navigation data of the first navigation; determine a type of an interest point corresponding to a terminal point of the second navigation and a second time of navigation to the terminal point based on second navigation data of the second navigation; determine a sample secondary navigation feature of the sample interest point according to the navigation frequency, the type of the sample interest point, the type of the interest point corresponding to the terminal point of the second navigation, the first time and the second time.

12. The method of claim 11, wherein, The determining of the sample secondary navigation feature of the sample interest point according to the navigation frequency, the type of the sample interest point, the type of the interest point corresponding to the terminal point of the second navigation, the first time and the second time comprises: determine the sample secondary navigation feature of the sample interest point according to the navigation frequency, the number of times that the type of the interest point corresponding to the terminal point of the second navigation is the same as the type of the sample interest point, and the time difference between the second time and the first time. 13.A computer storage medium having stored thereon a computer program, the program being executed by a processor to implement the method according to any one of claims 1 to 12.

Images