CN113987313A

CN113987313A - A method for determining geographic points of interest and a method for training a model for determining geographic points of interest

Info

Publication number: CN113987313A
Application number: CN202111341967.2A
Authority: CN
Inventors: 王禹
Original assignee: Beijing Dajia Internet Information Technology Co Ltd
Current assignee: Beijing Dajia Internet Information Technology Co Ltd
Priority date: 2021-11-12
Filing date: 2021-11-12
Publication date: 2022-01-28
Anticipated expiration: 2041-11-12
Also published as: CN113987313B

Abstract

The present disclosure relates to a method for determining a geographic point of interest, a method for training a model for determining a geographic point of interest, an apparatus, an electronic device, and a computer-readable storage medium. The training method and method for determining the geographic interest point model include: determining the initial geographic interest point matching the account according to the account attribute information of the account; acquiring the geographic interest point attribute information corresponding to the initial geographic interest point; according to the geographic interest point attribute Information and account attribute information to obtain the access probability of the account for the initial geographic POI; according to the access probability, filter out the target geographic POI corresponding to the account from the initial geographic POI. The method for determining a geographic point of interest provided by the present disclosure can combine account attributes and attributes of the geographic point of interest to determine the geographic point of interest, thereby improving the intelligence of the determination of the geographic point of interest.

Description

Method for determining geographical interest points and method for training geographical interest point determination model

Technical Field

The present disclosure relates to the field of internet technologies, and in particular, to a method for determining geographic interest points, a method and an apparatus for training a geographic interest point determination model, an electronic device, and a computer-readable storage medium.

Background

With the development of internet technology, in internet consumption scenes such as short videos or live broadcasts, technology for implementing information recommendation based on geographic positions is also developed, wherein the technology for implementing information recommendation by using geographic interest points is one of conventional means of such information recommendation based on geographic positions, and for example, store information of nearby stores, coupon information and the like can be recommended to an account in the process of recommending short videos to the account.

In the related technology, in the current technology for realizing information recommendation by using geographic interest points, the determination of the geographic interest points mainly adopts a distance-first determination principle, namely, the geographic interest points which are closer to the current position of an account are used as the geographic interest points needing information recommendation, so that the determination of the current geographic interest points is simpler and is not intelligent enough.

Disclosure of Invention

The present disclosure provides a method for determining geographic interest points, a method for training a model for determining geographic interest points, an apparatus, an electronic device, and a computer-readable storage medium, so as to at least solve the problems of simple determination and insufficient intelligence of geographic interest points in the related art. The technical scheme of the disclosure is as follows:

according to a first aspect of the embodiments of the present disclosure, there is provided a method for determining geographic points of interest, including:

according to account attribute information of an account, determining an initial geographic interest point matched with the account;

acquiring geographic interest point attribute information corresponding to the initial geographic interest point;

inputting a trained geographic interest point determination model according to the geographic interest point attribute information and the account attribute information to obtain the access probability of the account aiming at the initial geographic interest point;

and screening out target geographical interest points corresponding to the account from the initial geographical interest points according to the access probability.

In an exemplary embodiment, the obtaining, according to the geographic point of interest attribute information and the account attribute information, an access probability of the account with respect to the initial geographic point of interest includes: inputting the geographic interest point attribute information and the account attribute information into a trained geographic interest point determination model to obtain the access probability; the geographic interest point determination model is obtained by training the geographic interest point determination model to be trained according to the sample account attribute information of the sample account, the sample geographic interest point attribute information of the sample geographic interest point and the actual access probability of the sample account for the sample geographic interest point.

In an exemplary embodiment, the determining an initial geographic point of interest matching the account according to the account attribute information of the account includes: acquiring candidate geographic interest points according to the account attribute information; acquiring candidate interest point attribute information corresponding to the candidate geographic interest points; determining the initial geographic point of interest from the candidate geographic points of interest based on the candidate point of interest attribute information.

In an exemplary embodiment, the account attribute information includes an account location of the account; the obtaining of the candidate geographic interest points according to the account attribute information includes: and taking the geographic interest point with the distance from the account position smaller than a preset distance threshold value as the candidate geographic interest point.

In an exemplary embodiment, after the screening out the target geographic interest points corresponding to the account from the initial geographic interest points, the method further includes: acquiring recommendation information associated with the target geographical interest point according to the access probability of the target geographical interest point; and pushing the recommendation information to the account.

In an exemplary embodiment, the obtaining recommendation information associated with the target geographic interest point according to the access probability of the target geographic interest point includes: acquiring recommendation sequence information of the target geographical interest points according to the access probability of the target geographical interest points; and acquiring recommendation information associated with the target geographic interest points according to the recommendation sequence information.

According to a second aspect of the embodiments of the present disclosure, there is provided a training method of a geographic interest point determination model, including:

obtaining sample account attribute information of a sample account, sample geographical interest point attribute information of a sample geographical interest point and actual access probability of the sample account for the sample geographical interest point;

extracting sample account characteristics corresponding to the sample account attribute information and sample interest point characteristics corresponding to the sample geographical interest point attribute information;

inputting the sample account characteristics and the sample interest point characteristics into a geographic interest point determination model to be trained to obtain a predicted access probability of the sample account for the sample geographic interest point;

and training the geographic interest point determination model to be trained according to the actual access probability and the predicted access probability to obtain the trained geographic interest point determination model.

In an exemplary embodiment, after obtaining the trained geographic point of interest determination model, the method further includes: obtaining verification data for performing model tuning on the trained geographic interest point determination model from the sample account attribute information and the sample geographic interest point attribute information; inputting the verification data into the trained geographic interest point determination model, and performing model tuning on the trained geographic interest point determination model by using the actual access probability corresponding to the verification data.

In an exemplary embodiment, after performing model tuning on the trained geographic interest point determination model, the method further includes: obtaining test data for testing the geographic interest point determination model after model tuning from the sample account attribute information and the sample geographic interest point attribute information; inputting the test data into the geographical interest point determination model after model tuning, and testing the geographical interest point determination model after model tuning by using the actual access probability corresponding to the test data.

According to a third aspect of the embodiments of the present disclosure, there is provided a geographic point of interest determination apparatus, including:

an initial interest point determining unit configured to determine an initial geographic interest point matched with an account according to account attribute information of the account;

an interest point attribute obtaining unit configured to perform obtaining of geographic interest point attribute information corresponding to the initial geographic interest point;

an access probability obtaining unit configured to obtain an access probability of the account for the initial geographic interest point according to the geographic interest point attribute information and the account attribute information;

and the target interest point determining unit is configured to perform screening of target geographic interest points corresponding to the account from the initial geographic interest points according to the access probability.

In an exemplary embodiment, the access probability obtaining unit is further configured to perform inputting the geographic interest point attribute information and the account attribute information into a trained geographic interest point determination model to obtain the access probability; the geographic interest point determination model is obtained by training the geographic interest point determination model to be trained according to the sample account attribute information of the sample account, the sample geographic interest point attribute information of the sample geographic interest point and the actual access probability of the sample account for the sample geographic interest point.

In an exemplary embodiment, the initial point of interest determination unit is further configured to perform obtaining a candidate geographic point of interest according to the account attribute information; acquiring candidate interest point attribute information corresponding to the candidate geographic interest points; determining the initial geographic point of interest from the candidate geographic points of interest based on the candidate point of interest attribute information.

In an exemplary embodiment, the account attribute information includes an account location of the account; the initial point of interest determination unit is further configured to perform, as the candidate geographical point of interest, a geographical point of interest having a distance from the account location that is less than a preset distance threshold.

In an exemplary embodiment, the determining of the geographic point of interest further includes: the recommendation information pushing unit is configured to execute obtaining recommendation information associated with the target geographic interest point according to the access probability of the target geographic interest point; and pushing the recommendation information to the account.

In an exemplary embodiment, the recommendation information pushing unit is further configured to execute obtaining recommendation order information of the target geographic interest point according to the access probability of the target geographic interest point; and acquiring recommendation information associated with the target geographic interest points according to the recommendation sequence information.

According to a fourth aspect of the embodiments of the present disclosure, there is provided a training apparatus for a geographic point of interest determination model, including:

a sample information obtaining unit configured to perform obtaining sample account attribute information of a sample account, sample geographical point of interest attribute information of a sample geographical point of interest, and an actual access probability of the sample account for the sample geographical point of interest;

the sample feature extraction unit is configured to extract sample account features corresponding to the sample account attribute information and sample interest point features corresponding to the sample geographic interest point attribute information;

the prediction probability obtaining unit is configured to input the sample account characteristics and the sample interest point characteristics into a geographic interest point determination model to be trained to obtain a prediction access probability of the sample account for the sample geographic interest point;

and the model training unit is configured to train the geographic interest point determination model to be trained according to the actual access probability and the predicted access probability to obtain the trained geographic interest point determination model.

In an exemplary embodiment, the training device of the geographic interest point determination model further includes: the model tuning module is configured to execute model tuning of the trained geographic interest point determination model from the sample account attribute information and the sample geographic interest point attribute information; inputting the verification data into the trained geographic interest point determination model, and performing model tuning on the trained geographic interest point determination model by using the actual access probability corresponding to the verification data.

In an exemplary embodiment, the training device of the geographic interest point determination model further includes: the method comprises the steps of obtaining test data used for testing a geographical interest point determination model after model tuning from sample account attribute information and sample geographical interest point attribute information; inputting the test data into the geographical interest point determination model after model tuning, and testing the geographical interest point determination model after model tuning by using the actual access probability corresponding to the test data.

According to a fifth aspect of embodiments of the present disclosure, there is provided an electronic apparatus including: a processor; a memory for storing the processor-executable instructions; wherein the processor is configured to execute the instructions to implement the method for determining a geographical point of interest as defined in any one of the embodiments of the first aspect or the method for training a geographical point of interest determination model as defined in any one of the embodiments of the second aspect.

According to a fourth aspect of embodiments of the present disclosure, there is provided a computer-readable storage medium, wherein instructions of the computer-readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the method for determining a geographical point of interest according to any one of the embodiments of the first aspect, or the method for training a geographical point of interest determination model according to any one of the embodiments of the second aspect.

According to a fifth aspect of embodiments of the present disclosure, there is provided a computer program product, which includes instructions that, when executed by a processor of an electronic device, enable the electronic device to perform the method for determining a geographic point of interest according to any one of the embodiments of the first aspect, or the method for training a model for determining a geographic point of interest according to any one of the embodiments of the second aspect.

The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:

determining an initial geographic point of interest matched with the account according to the account attribute information of the account; acquiring geographic interest point attribute information corresponding to the initial geographic interest point; obtaining the access probability of the account aiming at the initial geographical interest point according to the geographical interest point attribute information and the account attribute information; and screening out target geographical interest points corresponding to the account from the initial geographical interest points according to the access probability. According to the method, the account attribute of the account and the attribute of the initial geographic interest point corresponding to the account are used for screening out the target geographic interest point which can be used for information recommendation, and compared with the method that the geographic interest point close to the account is directly used as the geographic interest point for information recommendation, the method for determining the geographic interest point can be used for determining the geographic interest point by combining the account attribute and the attribute of the geographic interest point, so that the intelligence of determining the geographic interest point can be improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure and are not to be construed as limiting the disclosure.

FIG. 1 is a flow chart illustrating a method for geographic point of interest determination in accordance with an exemplary embodiment.

FIG. 2 is a flow diagram illustrating the determination of an initial geographic point of interest that matches an account in accordance with an exemplary embodiment.

FIG. 3 is a flow diagram illustrating a method of training a geographic point of interest determination model in accordance with an exemplary embodiment.

FIG. 4 is a flow diagram illustrating model tuning of a geographic point of interest determination model in accordance with an exemplary embodiment.

FIG. 5 is a flow diagram illustrating model testing of a geographic point of interest determination model in accordance with an exemplary embodiment.

FIG. 6 is a flow diagram illustrating model training in a geographic point of interest recommendation method in accordance with an exemplary embodiment.

FIG. 7 is a flowchart illustrating an online application of a geographic point of interest recommendation method in accordance with an exemplary embodiment.

Fig. 8 is a block diagram illustrating a geographic point of interest determination apparatus in accordance with an example embodiment.

FIG. 9 is a block diagram illustrating a training apparatus of a geographic point of interest determination model in accordance with an exemplary embodiment.

FIG. 10 is a block diagram illustrating an electronic device in accordance with an example embodiment.

Detailed Description

In order to make the technical solutions of the present disclosure better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

It should also be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for presentation, analyzed data, etc.) referred to in the present disclosure are both information and data that are authorized by the user or sufficiently authorized by various parties.

Fig. 1 is a flowchart illustrating a method for determining geographical points of interest, which may be used in a terminal as shown in fig. 1, according to an exemplary embodiment.

In step S101, an initial geographic point of interest matching an account is determined according to account attribute information of the account.

The initial geographic interest point is a geographic interest point determined in advance according to account attributes of the account, the account may be an account for which recommendation information related to the geographic interest point needs to be recommended, the account attribute information refers to account attributes of the account for which recommendation information recommendation needs to be performed, and may include, for example, a current location of the account, what type of geographic interest point the account is prone to browse, a geographic interest point of which the dining type, shopping type or scenic spot type is prone, whether the account has a preference for browsing the geographic interest point, and other related account attributes. Specifically, when the terminal needs to recommend recommendation information carrying information related to geographic interest points to an account, account attribute information of the account may be collected first, and geographic interest points matching account attributes of the account are determined based on the account attribute information of the account and serve as initial geographic interest points.

For example, when an account tends to browse geographical points of interest of a restaurant type, the terminal may determine the geographical points of interest of the restaurant type as initial geographical points of interest from the geographical points of interest, and if an account tends to browse the geographical points of interest of a shopping type, the terminal may screen the geographical points of interest of the shopping type from the geographical points of interest as initial geographical points of interest. In addition, the initial geographic interest point may be determined according to the location of the account, that is, the geographic interest point near the current location of the account may be used as the initial geographic interest point.

Step S102, obtaining geographic interest point attribute information corresponding to the initial geographic interest point.

The geographic interest point attribute information refers to the attribute information corresponding to the initial geographic interest point obtained in step S101, and may include, for example: the location of the point of interest of the initial geographic point of interest, the type of the initial geographic point of interest, the number of visits of the initial geographic point of interest, and the information integrity of the initial geographic point of interest, among others. Specifically, after the terminal determines the initial geographic interest point, the terminal may acquire attribute information corresponding to the initial geographic interest point as geographic interest point attribute information corresponding to the initial geographic interest point.

Step S103, obtaining the access probability of the account aiming at the initial geographical interest point according to the geographical interest point attribute information and the account attribute information;

and step S104, screening out target geographical interest points corresponding to the account from the initial geographical interest points according to the access probability.

The access probability can be used for representing the possibility that the account accesses the initial geographic interest point, the higher the access probability is, the higher the possibility that the account accesses the initial geographic interest point is indicated, and the target geographic interest point refers to the finally determined geographic interest point, and the geographic interest point can be used for providing relevant recommendation information for the account. Specifically, after obtaining the geographic interest point attribute information corresponding to the initial geographic interest point, the terminal may obtain, according to the account attribute information of the account and the geographic interest point attribute information, a probability that the user accesses the initial geographic interest point, for example, by using a trained geographic interest point determination model, process the geographic interest point attribute information and the account attribute information to obtain an access probability that the account accesses the initial geographic interest point; for another example, the access conditions of accounts with different account attributes for the screened initial geographic interest points may be collected in advance, an account matching the account attribute of the account requiring geographic interest point recommendation may be found from the access conditions, and after the access probability of the account is obtained according to the access conditions of the matching account for the initial geographic interest points, a target geographic interest point finally used for information recommendation for the account may be screened from the initial geographic interest points based on the obtained access probability, for example, the initial geographic interest point with the access probability greater than a certain preset access probability threshold may be used as the target geographic interest point.

For example, the access situation of an account a, an account B, an account C, and an account D to a certain initial geographic point of interest a may be collected in advance, where the account attributes of the account a and the account D match with the account attribute information of an account E that needs to make recommendation information related to geographic point of interest to make a recommendation, and the account a accesses the initial geographic point of interest a, and the account D does not access the initial geographic point of interest a, then the access probability of the account E to the initial geographic point of interest a may be obtained as 50%.

In the method for determining the geographical interest points, the initial geographical interest points matched with the account are determined according to the account attribute information of the account; acquiring geographic interest point attribute information corresponding to the initial geographic interest point; obtaining the access probability of the account aiming at the initial geographical interest point according to the geographical interest point attribute information and the account attribute information; and screening out target geographical interest points corresponding to the account from the initial geographical interest points according to the access probability. According to the method, the account attribute of the account and the attribute of the initial geographic interest point corresponding to the account are used for screening out the target geographic interest point which can be used for information recommendation, and compared with the method that the geographic interest point close to the account is directly used as the geographic interest point for information recommendation, the method for determining the geographic interest point can be used for determining the geographic interest point by combining the account attribute and the attribute of the geographic interest point, so that the intelligence of determining the geographic interest point can be improved.

In an exemplary embodiment, step S103 may further include: inputting the attribute information of the geographic interest points and the account attribute information into a trained geographic interest point determination model to obtain access probability; and the geographic interest point determination model is obtained by training the geographic interest point determination model to be trained according to the sample account attribute information of the sample account, the sample geographic interest point attribute information of the sample geographic interest point and the actual access probability of the sample account for the sample geographic interest point.

The geographic point of interest determination model is pre-trained, a network model for determining the probability of access of the account with respect to the initial geographic point of interest, the model is obtained by training pre-collected sample account attribute information, sample geographical interest point attribute information and actual access probability of a sample account aiming at sample geographical interest points, wherein the sample account attribute information refers to account attribute information for an account used to train the geographic point of interest determination model, i.e., account attribute information for the sample account, and the sample geographic point of interest attribute information is attribute information for the geographic point of interest used to train the geographic point of interest determination model, i.e., attribute information of the sample geographic point of interest, the actual access probability is the access probability of the sample account for the sample geographic point of interest, the access probability is used for representing whether the sample user accesses the sample geographic interest point, and the access probability can be shown in the form of a label. Specifically, the terminal may collect in advance sample account attribute information of a sample account, sample geographical interest point attribute information of a sample geographical interest point, and an actual access probability of the sample account for the sample geographical interest point, so that the geographic interest point determination model to be trained may be trained using the information, and a trained geographic interest point determination model may be obtained. Then, when the access probability of the account for the initial geographic interest points needs to be obtained, the account attribute information and the geographic interest point attribute information of the account can be input into a geographic interest point determination model which is trained in advance, and the access probability of the account for accessing each initial geographic interest point is obtained through output of the geographic interest point determination model.

In this embodiment, the access probability of the account accessing each initial geographic interest point may be obtained through a pre-trained geographic interest point determination model, and the geographic interest point determination model is obtained by training according to the sample account attribute information, the sample geographic interest point attribute information, and the actual access probability of the sample account for the sample geographic interest point, so that the accuracy of the obtained access probability of the account for the initial geographic interest point may be improved.

In an exemplary embodiment, as shown in fig. 2, step S101 may further include:

in step S201, candidate geographic interest points are obtained according to the account attribute information.

In this embodiment, in order to prevent the initial geographic interest point attribute information of the initial geographic interest point input to the geographic interest point determination model from being obtained directly according to the account attribute information, and thus the operation speed of the geographic interest point determination model is affected, in this embodiment, the terminal may first perform preliminary screening on the geographic interest points screened by the account attribute information, that is, candidate geographic interest points, and then obtain the initial geographic interest point used for inputting the geographic interest point attribute information to the geographic interest point determination model. Specifically, the terminal may determine, from the plurality of geographic interest points according to the account attribute information of the account, a geographic interest point that is suitable for the account attribute information as a candidate geographic interest point.

In step S202, candidate interest point attribute information corresponding to the candidate geographic interest point is obtained.

The candidate interest point attribute information refers to attribute information of the candidate geographical interest point screened by the terminal in step S201, and after the terminal determines the candidate geographical interest point, the terminal may obtain the attribute information of the candidate geographical interest point as the candidate interest point attribute information.

In step S203, an initial geographic interest point is determined from the candidate geographic interest points based on the candidate interest point attribute information.

Wherein, the determination of the initial geographic interest point from the candidate geographic interest points can be based on the preset geographic interest point screening rule for screening, for example the rule may be a filtering rule that filters by the number of visits to each candidate geographical point of interest, selecting candidate geographical interest points with more visiting times from the obtained candidate geographical interest points as initial geographical interest points, the rule may also be a screening rule for screening according to the account goodness of each candidate geographic interest point, that is, a candidate geographic interest point with a higher goodness of the candidate geographic interest points is screened from the obtained candidate geographic interest points as an initial geographic interest point, or a screening rule for screening according to the position of each candidate geographic interest point, namely, candidate geographical interest points with the positions closer to the current position of the account are screened out from the obtained candidate geographical interest points and are used as initial geographical interest points, and the like.

In this embodiment, the terminal may filter the candidate geographical interest points obtained according to the account attribute information to obtain the initial geographical interest points, and may reduce the number of geographical interest point attribute information of the initial geographical interest points input to the geographical interest point determination model, thereby increasing the operation speed of the geographical interest point determination model.

Further, the account attribute information may include an account location of the account; step S201 may further include: and taking the geographic interest points with the distance from the account position smaller than a preset distance threshold value as candidate geographic interest points.

In this embodiment, the selection of the candidate geographic interest points may be determined according to the current account location of the account in the account attribute information, and the terminal may filter out geographic interest points close to the account location from the multiple geographic interest points as the candidate geographic interest points. Specifically, the terminal may determine the account location where the account is currently located from the account attribute information, and use a preset distance threshold to set, as the candidate geographic interest point, a geographic interest point whose distance from the account location where the account is currently located is smaller than the distance threshold.

In this embodiment, the terminal may use the geographic interest point whose distance from the account location is smaller than the preset distance threshold as the candidate geographic interest point, so that it may be ensured that the location of the target geographic interest point screened by the geographic interest point determination model may be located near the account location where the account is currently located, and thus, the correlation between the determined target geographic interest point and the account may be further improved.

In an exemplary embodiment, after step S103, the method may further include: acquiring recommendation information associated with the target geographical interest point according to the access probability of the target geographical interest point; and pushing the recommendation information to an account.

After the terminal obtains the target geographic interest point, the terminal may further obtain recommendation information associated with the target geographic interest point based on the determined target geographic interest point, for example, the determined target geographic interest point may be a certain restaurant enterprise, the terminal may obtain information associated with the restaurant enterprise, the information may be related offer information of the restaurant enterprise may be pushed to the account as recommendation information associated with the target geographic interest point, and if the determined recommendation information associated with the target geographic interest point is a certain shopping store, the terminal may obtain merchant information of the store as recommendation information associated with the target geographic interest point and push the merchant information to the account.

Specifically, after the terminal obtains the target geographic interest point, the terminal may find out the recommendation information related to the target geographic interest point based on the determined access probability of the target geographic interest point, and may select the recommendation information related to the target geographic interest point with the highest access probability from the recommendation information related to the plurality of target geographic interest points and push the recommendation information to the account.

In this embodiment, the terminal may further obtain, based on the determined access probability corresponding to the target geographic interest point, recommendation information associated with the target geographic interest point and push the recommendation information to the account, so that the recommendation information recommended by the terminal better meets the requirement of the account, and the access probability of the account to the recommendation information is improved.

Further, obtaining recommendation information associated with the target geographic interest point according to the access probability of the target geographic interest point may further include: acquiring recommendation sequence information of the target geographical interest points according to the access probability of the target geographical interest points; and acquiring recommendation information associated with the target geographic interest points according to the recommendation sequence information.

The recommendation sequence information refers to the size sequence information of the access probability of the account for the target geographic interest points, after the terminal obtains the access probability of the target geographic interest points, the size sequence information of the access probability of the target geographic interest points can be sequenced, so that recommendation sequence information corresponding to each target geographic interest point is obtained, and then the terminal can obtain recommendation information associated with the target geographic interest points based on the recommendation sequence information. For example, after obtaining the recommendation sequence information corresponding to the target geographic interest point, the terminal may output the recommendation information by using a recommendation engine, with the recommendation sequence information being one input dimension of the preset recommendation engine.

In this embodiment, the recommendation information may be obtained based on the access probability of the target geographic interest point, and the recommendation information for pushing information to the account may be the recommendation information that best meets the account requirement, so that the recommendation quality of the recommendation information and the traffic conversion rate of the recommendation information are improved.

Fig. 3 is a flowchart illustrating a training method of a geographic interest point determination model according to an exemplary embodiment, and as shown in fig. 3, the training method of the geographic interest point determination model may be used in a terminal, and includes the following steps.

In step S301, sample account attribute information for a sample account, sample geographic point of interest attribute information for a sample geographic point of interest, and an actual access probability for a sample account for a sample geographic point of interest are obtained.

The sample account attribute information refers to account attribute information of an account which is acquired by the terminal in advance and used for training the geographic interest point determination model, namely the account attribute information of the sample account, the sample geographic interest point attribute information refers to attribute information of a geographic interest point which is acquired by the terminal in advance and used for training the geographic interest point determination model, namely the attribute information of the sample geographic interest point, the actual access probability refers to the access probability of the sample account for the sample geographic interest point, and the access probability can be displayed in a form of a label. For example, if a sample account has actually visited the sample geographic point of interest, the actual probability of visiting the sample account for the sample geographic point of interest may be set to 1, and if the sample account has not actually visited the sample geographic point of interest, the actual probability of visiting the sample account for the sample geographic point of interest may be set to 0.

Specifically, the terminal may collect historical access information of a sample account accessing a sample geographic point of interest in advance, and may determine, according to the historical access information, an actual access probability of the sample account for the sample geographic point of interest, and account attribute information of the sample account and attribute information of the sample geographic point of interest as sample account attribute information and sample geographic point of interest attribute information, respectively.

In step S302, sample account features corresponding to the sample account attribute information and sample point of interest features corresponding to the sample geographic point of interest attribute information are extracted.

The sample account characteristics refer to account characteristics obtained after characteristic extraction is carried out on the sample account attribute information, and the sample interest point characteristics refer to interest point characteristics obtained after characteristic extraction is carried out on the sample geographic interest point attribute information. The terminal may perform data cleaning, transformation, and other processing on the sample account attribute information and the sample geographic interest point attribute information obtained in step S301, so as to obtain sample account characteristics and sample interest point characteristics.

In step S303, inputting the sample account characteristics and the sample interest point characteristics into a geographic interest point determination model to be trained, and obtaining a predicted access probability of the sample account for the sample geographic interest point;

in step S304, the geographic interest point determination model to be trained is trained according to the actual access probability and the predicted access probability, so as to obtain a trained geographic interest point determination model.

The geographic interest point determination model to be trained can be a geographic interest point determination model which needs to be trained, and the predicted access probability refers to the probability that the geographic interest point determination model to be trained accesses the sample geographic interest point according to the input sample account characteristics and the sample interest point characteristics. The terminal can input the sample account characteristics and the sample interest point characteristics into a geographic interest point determination model needing to be trained, the geographic interest point determination model predicts the probability of the sample account accessing the sample geographic interest point, namely the predicted access probability of the sample account aiming at the sample geographic interest point is output, and then the terminal can train the geographic interest point determination model according to the actual access probability and the predicted access probability of the sample account aiming at the sample geographic interest point as the difference loss of the model, so that the trained geographic interest point determination model can be obtained.

In the training method of the geographic interest point determination model, sample account attribute information of a sample account, sample geographic interest point attribute information of a sample geographic interest point and actual access probability of the sample account for the sample geographic interest point are obtained; extracting sample account characteristics corresponding to the sample account attribute information and sample interest point characteristics corresponding to the sample geographic interest point attribute information; inputting the sample account characteristics and the sample interest point characteristics into a geographic interest point determination model to be trained to obtain the predicted access probability of the sample account for the sample geographic interest points; and training the geographic interest point determination model to be trained according to the actual access probability and the predicted access probability to obtain the trained geographic interest point determination model. The method and the device can train the geographic interest point determination model through the pre-collected sample account attribute information, the sample geographic interest point attribute information and the actual access probability of the sample account for the sample geographic interest point so as to obtain the trained geographic interest point determination model, and realize the determination of the geographic interest point by utilizing the geographic interest point determination model so as to improve the intelligence of the determined geographic interest point.

As shown in fig. 4, after step S304, the method may further include:

in step S401, test data for testing the geographic point of interest determination model after model tuning is obtained from the sample account attribute information and the sample geographic point of interest attribute information.

In this embodiment, after the training of the geographic interest point determination model is completed, further model tuning may be performed on the trained geographic interest point determination model, and specifically, in this embodiment, after the sample account attribute information and the sample geographic interest point attribute information are obtained, the sample account attribute information and the sample geographic interest point attribute information may be classified according to a certain proportion, and the classified sample account attribute information and the sample geographic interest point attribute information are used as training data for training the geographic interest point determination model and verification data for performing model tuning on the trained geographic interest point determination model. After the training of the geographic interest point determination model is completed, corresponding verification data can be obtained from the pre-classified sample account attribute information and the sample geographic interest point attribute information.

For example, after the terminal finishes collecting the sample account attribute information and the sample geographical interest point attribute information, the collected sample account attribute information and the sample geographical interest point attribute information may be divided into training data and verification data according to a preset proportion, and after the terminal finishes training the geographical interest point determination model by using the training data, the pre-divided verification data may be obtained from the sample account attribute information and the sample geographical interest point attribute information.

In step S402, the verification data is input into the trained geographic interest point determination model, and the trained geographic interest point determination model is optimized by using the actual access probability corresponding to the verification data.

And then, the terminal can input the acquired verification data into the trained geographic interest point determination model, so that the predicted access probability corresponding to the verification data can be obtained through the geographic interest point determination model, and the trained geographic interest point determination model is further subjected to model tuning treatment by using the actual access probability corresponding to the verification data and the predicted access probability corresponding to the verification data.

In this embodiment, after the terminal completes the training of the geographic interest point determination model, the model tuning of the geographic interest point determination model can be further realized by using the verification data, so that the model accuracy of the trained geographic interest point determination model can be further improved, and the accuracy of the determined geographic interest point can be further improved.

Further, as shown in fig. 5, after step S402, the method may further include:

in step S501, test data for testing the geographic interest point determination model after model tuning is obtained from the sample account attribute information and the sample geographic interest point attribute information.

In this embodiment, after the model tuning of the geographic interest point determination model is completed, further model testing may be performed on the geographic interest point determination model, and specifically, in this embodiment, after the sample account attribute information and the sample geographic interest point attribute information are obtained, the sample account attribute information and the sample geographic interest point attribute information are classified according to a certain proportion to obtain training data and verification data, and besides, a part of the sample account attribute information and the sample geographic interest point attribute information may be further classified as test data for testing the model-tuned geographic interest point determination model. After the model tuning of the geographic interest point determination model is completed, corresponding test data can be obtained from the pre-classified sample account attribute information and the sample geographic interest point attribute information.

For example, after the terminal completes the acquisition of the sample account attribute information and the sample geographical interest point attribute information, the acquired sample account attribute information and the sample geographical interest point attribute information may be divided into training data, verification data, and test data according to a preset ratio, which may be a 7:2:1 manner, and after the terminal completes the tuning of the geographical interest point determination model by using the verification data, the pre-divided test data may be further obtained from the sample account attribute information and the sample geographical interest point attribute information.

In step S502, the test data is input into the geographical interest point determination model after model tuning, and the geographical interest point determination model after model tuning is tested by using the actual access probability corresponding to the test data.

After the optimization of the geographical interest point determination model is completed, the terminal can also input the test data into the optimized geographical interest point determination model, so that the predicted access probability corresponding to the test data can be obtained through the geographical interest point determination model, and the actual access probability corresponding to the test data and the predicted access probability corresponding to the test data are utilized to perform test processing on the overall performance of the trained geographical interest point determination model.

In this embodiment, after the terminal completes model tuning of the geographic interest point determination model by using the verification data, the test data may be used to perform further performance test on the geographic interest point determination model, so as to further ensure the model performance of the geographic interest point determination model.

In an exemplary embodiment, a geographical interest point recommendation method based on multidimensional attributes is further provided, where a plurality of POIs suitable for a user are trained through data of two links according to request information when the user uses an app and multidimensional information of the POIs (namely geographical interest points), the POIs are ranked, and a recommendation is performed for the user according to the ranking, so as to improve recommendation quality and recommendation efficiency.

The user-side data may include:

(1) the position of the user: location of the user when using the app;

(2) user local life attributes: whether the user has local life (paying attention to scenes such as catering and shopping) attributes, namely whether the user has the preference of browsing POI (the characteristics are summarized by historical clicking and browsing various POI behaviors);

(3) user local life preference: local life types (such as restaurants, shopping, scenic spots and the like) which users tend to browse are preferred (features are summarized by historical clicking and browsing various POI behaviors).

And the data of the POI side data may include:

(1) POI location: administrative divisions and longitude and latitude positions where POI are located;

(2) type of POI: the type to which the POI belongs (e.g., restaurant, hotel, shopping, scenic spot, etc.);

(3) POI heat: the number of exposed and clicked POIs;

(4) POI attribute richness: such as price, label, hours of operation, etc.

Specifically, the present embodiment may be composed of the following two processes:

1. model training: training user information and a relationship with a POI clicked by the user information through historical data, wherein a model training process can be shown in FIG. 6, and specifically comprises the following steps:

(1) collecting historical data: and reporting information through user history. When a user historically browses videos hung with the POI, information such as the number/proportion of clicks of the user under each POI category, the stay time of a POI detail page and the like describes local life attributes and local life preferences of the user; according to information such as the position of a user when browsing a video hung with a POI in history, the category, the heat degree and the information richness of the POI hung on the video and whether the user finally clicks the POI hung on the video;

(2) characteristic engineering: cleaning and transforming the data, and dividing training data, verification data and test data;

(3) training a supervised model: and (3) constructing a supervised machine learning model through the characteristics and dependent variables (whether the user clicks the POI) obtained in the characteristic engineering, carrying out model tuning on a verification set, and finally obtaining user information when the user browses the video hung with the POI and a relation model of whether the POI information clicks the POI or not.

2. And (3) online prediction: the online prediction is implemented by using the obtained model, and the process of the online prediction may be as shown in fig. 7, and specifically may include the following steps:

(1) POI recall: and finding all POI in a certain range (for example, in a range of 1 kilometer) as a recommendation candidate set according to the position reported when the user uses the APP.

(2) Coarse sorting: and according to the POI categories and the heat, roughly ordering business rules of high value (such as catering, shopping and scenic spot categories), POI heat and the like.

(3) Fine sorting: and selecting top 10% of POI in the coarse sorting result, transmitting the user information and the POI information into a trained model to obtain the probability of the user clicking the POI, sorting the POI obtained in the coarse sorting according to the probability from high to low, wherein the sorting (or the probability value) is used as one dimension in a video recommendation engine, and recommending videos with the POI with the highest prediction probability for the users with local life attributes.

According to the embodiment, the user data and the objective POI data are combined and matched and ranked, so that more effective POIs (i.e. POIs which are more likely to be favored by the user) are obtained, and the POI information which is likely to be preferred by the user is transmitted to the video recommendation engine, so that the conversion and the change from the online traffic to the offline traffic of the user are improved.

It should be understood that although the various steps in the flowcharts of fig. 1-7 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-7 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least some of the other steps.

It is understood that the same/similar parts between the embodiments of the method described above in this specification can be referred to each other, and each embodiment focuses on the differences from the other embodiments, and it is sufficient that the relevant points are referred to the descriptions of the other method embodiments.

Fig. 8 is a block diagram illustrating a geographic point of interest determination apparatus in accordance with an example embodiment. Referring to fig. 8, the apparatus includes an initial point of interest determining unit 801, a point of interest attribute acquiring unit 802, an access probability acquiring unit 803, and a target point of interest determining unit 804.

An initial point of interest determination unit 801 configured to perform determining an initial geographic point of interest matching an account according to account attribute information of the account;

an interest point attribute obtaining unit 802 configured to perform obtaining geographic interest point attribute information corresponding to the initial geographic interest point;

an access probability obtaining unit 803 configured to obtain an access probability of the account with respect to the initial geographic point of interest according to the geographic point of interest attribute information and the account attribute information;

a target interest point determining unit 804 configured to perform screening out a target geographic interest point corresponding to the account from the initial geographic interest points according to the access probability.

In an exemplary embodiment, the access probability obtaining unit 803 is further configured to perform inputting the geographic interest point attribute information and the account attribute information into the trained geographic interest point determination model to obtain an access probability; and the geographic interest point determination model is obtained by training the geographic interest point determination model to be trained according to the sample account attribute information of the sample account, the sample geographic interest point attribute information of the sample geographic interest point and the actual access probability of the sample account aiming at the sample geographic interest point.

In an exemplary embodiment, the initial point of interest determining unit 801 is further configured to perform obtaining a candidate geographic point of interest according to the account attribute information; acquiring candidate interest point attribute information corresponding to the candidate geographic interest points; an initial geographic point of interest is determined from the candidate geographic points of interest based on the candidate point of interest attribute information.

In an exemplary embodiment, the account attribute information includes an account location of the account; the initial point of interest determination unit 801 is further configured to perform a determination of a geographical point of interest having a distance to the account location smaller than a preset distance threshold as a candidate geographical point of interest.

In an exemplary embodiment, the determining apparatus of the geographic point of interest further includes: the recommendation information pushing unit is configured to execute acquiring recommendation information associated with the target geographic interest point according to the access probability of the target geographic interest point; and pushing the recommendation information to an account.

In an exemplary embodiment, the recommendation information pushing unit is further configured to execute obtaining recommendation sequence information of the target geographic interest point according to the access probability of the target geographic interest point; and acquiring recommendation information associated with the target geographic interest points according to the recommendation sequence information.

FIG. 9 is a block diagram illustrating a training apparatus of a geographic point of interest determination model in accordance with an exemplary embodiment. Referring to fig. 9, the apparatus includes a sample information acquisition unit 901, a sample feature extraction unit 902, a prediction probability acquisition unit 903, and a model training unit 904.

A sample information obtaining unit 901 configured to perform obtaining sample account attribute information of a sample account, sample geographical interest point attribute information of a sample geographical interest point, and an actual access probability of the sample account for the sample geographical interest point;

a sample feature extraction unit 902 configured to perform extraction of sample account features corresponding to the sample account attribute information and sample interest point features corresponding to the sample geographic interest point attribute information;

a prediction probability obtaining unit 903, configured to perform inputting the sample account characteristics and the sample interest point characteristics into a geographic interest point determination model to be trained, so as to obtain a prediction access probability of the sample account for the sample geographic interest point;

and the model training unit 904 is configured to perform training on the geographic interest point determination model to be trained according to the actual access probability and the predicted access probability to obtain the trained geographic interest point determination model.

In an exemplary embodiment, the training device for the geographic interest point determination model further includes: the model tuning module is configured to execute model tuning verification data for the trained geographic interest point determination model from the sample account attribute information and the sample geographic interest point attribute information; and inputting verification data into the trained geographical interest point determination model, and performing model tuning on the trained geographical interest point determination model by using the actual access probability corresponding to the verification data.

In an exemplary embodiment, the training device for the geographic interest point determination model further includes: the model testing module is configured to execute the test data for testing the adjusted geographic interest point determination model from the sample account attribute information and the sample geographic interest point attribute information; inputting the test data into the geographical interest point determination model after model tuning, and testing the geographical interest point determination model after model tuning by using the actual access probability corresponding to the test data.

With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

FIG. 10 is a block diagram illustrating an electronic device 1000 for determination of a geographic point of interest, in accordance with an example embodiment. For example, the electronic device 1000 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a gaming console, a tablet device, a medical device, a fitness device, a personal digital assistant, and so forth.

Referring to fig. 8, electronic device 1000 may include one or more of the following components: processing component 1002, memory 1004, power component 1006, multimedia component 1008, audio component 1010, interface to input/output (I/O) 1012, sensor component 1014, and communications component 1016.

The processing component 1002 generally controls the overall operation of the electronic device 1000, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing components 1002 may include one or more processors 1020 to execute instructions to perform all or a portion of the steps of the methods described above. Further, processing component 1002 may include one or more modules that facilitate interaction between processing component 1002 and other components. For example, the processing component 1002 may include a multimedia module to facilitate interaction between the multimedia component 1008 and the processing component 1002.

The memory 1004 is configured to store various types of data to support operations at the electronic device 1000. Examples of such data include instructions for any application or method operating on the electronic device 1000, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 1004 may be implemented by any type or combination of volatile or non-volatile storage devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, optical disk, or graphene memory.

The power supply component 1006 provides power to the various components of the electronic device 1000. The power components 1006 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power for the electronic device 1000.

The multimedia component 1008 includes a screen that provides an output interface between the electronic device 1000 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1008 includes a front facing camera and/or a rear facing camera. The front camera and/or the rear camera may receive external multimedia data when the electronic device 1000 is in an operating mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

The audio component 1010 is configured to output and/or input audio signals. For example, the audio component 1010 includes a Microphone (MIC) configured to receive external audio signals when the electronic device 1000 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signal may further be stored in the memory 1004 or transmitted via the communication component 1016. In some embodiments, audio component 1010 also includes a speaker for outputting audio signals.

I/O interface 1012 provides an interface between processing component 1002 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.

The sensor assembly 1014 includes one or more sensors for providing various aspects of status assessment for the electronic device 1000. For example, the sensor assembly 1014 may detect an open/closed state of the electronic device 1000, the relative positioning of components, such as a display and keypad of the electronic device 1000, the sensor assembly 1014 may also detect a change in the position of the electronic device 1000 or components of the electronic device 1000, the presence or absence of user contact with the electronic device 1000, orientation or acceleration/deceleration of the device 1000, and a change in the temperature of the electronic device 1000. The sensor assembly 1014 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 1014 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1014 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 1016 is configured to facilitate wired or wireless communication between the electronic device 1000 and other devices. The electronic device 1000 may access a wireless network based on a communication standard, such as WiFi, a carrier network (such as 2G, 3G, 4G, or 5G), or a combination thereof. In an exemplary embodiment, the communication component 1016 receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communications component 1016 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

In an exemplary embodiment, the electronic device 1000 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors, or other electronic components for performing the above-described methods.

In an exemplary embodiment, a computer-readable storage medium comprising instructions, such as the memory 1004 comprising instructions, executable by the processor 1020 of the electronic device 1000 to perform the above-described method is also provided. For example, the computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

In an exemplary embodiment, a computer program product is also provided, which includes instructions executable by the processor 1020 of the electronic device 1000 to perform the above-described method.

It should be noted that the descriptions of the above-mentioned apparatus, the electronic device, the computer-readable storage medium, the computer program product, and the like according to the method embodiments may also include other embodiments, and specific implementations may refer to the descriptions of the related method embodiments, which are not described in detail herein.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for determining geographic points of interest, comprising:

obtaining the access probability of the account aiming at the initial geographical interest point according to the geographical interest point attribute information and the account attribute information;

2. The method for determining geographic interest points according to claim 1, wherein the obtaining the access probability of the account with respect to the initial geographic interest point according to the geographic interest point attribute information and the account attribute information comprises:

inputting the geographic interest point attribute information and the account attribute information into a trained geographic interest point determination model to obtain the access probability; the geographic interest point determination model is obtained by training the geographic interest point determination model to be trained according to the sample account attribute information of the sample account, the sample geographic interest point attribute information of the sample geographic interest point and the actual access probability of the sample account for the sample geographic interest point.

3. The method for determining geographic points of interest of claim 2, wherein determining the initial geographic point of interest matching the account according to the account attribute information of the account comprises:

acquiring candidate geographic interest points according to the account attribute information;

acquiring candidate interest point attribute information corresponding to the candidate geographic interest points;

determining the initial geographic point of interest from the candidate geographic points of interest based on the candidate point of interest attribute information.

4. The method of claim 3, wherein the account attribute information comprises an account location of the account;

the obtaining of the candidate geographic interest points according to the account attribute information includes:

and taking the geographic interest point with the distance from the account position smaller than a preset distance threshold value as the candidate geographic interest point.

5. A training method of a geographic interest point determination model is characterized by comprising the following steps:

6. An apparatus for determining geographic points of interest, comprising:

7. A training apparatus for a geographic point of interest determination model, comprising:

8. An electronic device, comprising:

a processor;

a memory for storing the processor-executable instructions;

wherein the processor is configured to execute the instructions to implement the method of determining a geographical point of interest according to any one of claims 1 to 4 or the method of training a geographical point of interest determination model according to claim 5.

9. A computer-readable storage medium, wherein instructions in the computer-readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the method for determining a geographical point of interest of any one of claims 1 to 4, or the method for training a geographical point of interest determination model of claim 5.

10. A computer program product comprising instructions therein, which when executed by a processor of an electronic device, enable the electronic device to perform the method of determining a geographical point of interest of any one of claims 1 to 4, or the method of training a geographical point of interest determination model of claim 5.