Area naming method, device and electronic equipment

Abstract

The present application provides a method, apparatus, electronic device, and computer-readable storage medium for naming an area; the method includes: obtaining a plurality of first points of interest within a first area to be named, and obtaining a keyword set corresponding to each of the first points of interest; determining a first weight of each keyword based on the number of times each keyword appears in the keyword set corresponding to the plurality of first points of interest; obtaining a plurality of second points of interest within a second area, wherein the second area includes the first area; modifying the first weight of each keyword based on the total number of second points of interest within the second area and the number of second points of interest corresponding to each keyword within the second area to obtain a second weight of each keyword; and determining the name of the first area based on the keyword with the largest second weight. The present application can be applied to fields such as maps for naming areas.

Description

Technical Field

[0001] The present application relates to the technical field of electronic maps, and in particular to a method, device, electronic device, and computer-readable storage medium for naming an area. Background Art

[0002] Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to achieve the best results.

[0003] An electronic map is a system for map production and application. It's a computer-generated map capable of visual display. Landmarks on an electronic map typically belong to different regions. Naming regions facilitates the clear presentation of information on the electronic map and its application to various functional products. However, prior art for naming regions remains at the manual annotation stage, making it difficult to meet the demand for large-scale, flexible region naming. Summary of the Invention

[0004] The embodiments of the present application provide a method, device, electronic device, and computer-readable storage medium for naming an area, which can be applied to fields such as maps for naming areas.

[0005] The technical solution of the embodiment of the present application is implemented as follows:

[0006] This embodiment of the present application provides a method for naming a region, including:

[0007] Acquire multiple first points of interest in a first area to be named, and acquire a keyword set corresponding to each of the first points of interest;

[0008] determining a first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the plurality of first points of interest;

[0009] Acquire a plurality of second points of interest in a second area, wherein the second area includes the first area;

[0010] modifying the first weight of each keyword according to the total number of second points of interest in the second area and the number of second points of interest corresponding to each keyword in the second area to obtain a second weight of each keyword;

[0011] The name of the first area is determined based on the keyword having the largest second weight.

[0012] The present invention provides a device for naming a region, including:

[0013] An interest point module, configured to obtain a plurality of first interest points in a first area to be named, and obtain a keyword set corresponding to each of the first interest points;

[0014] A first weight module, configured to determine a first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the plurality of first interest points;

[0015] The point of interest module is further configured to obtain a plurality of second points of interest in a second area, wherein the second area includes the first area;

[0016] a second weighting module, configured to modify the first weight of each keyword according to the total number of second points of interest in the second area and the number of second points of interest corresponding to each keyword in the second area, so as to obtain a second weight of each keyword;

[0017] The naming module is configured to determine a name of the first area based on the keyword having the largest second weight.

[0018] In the above scheme, the point of interest module is also used to perform any of the following processing to determine the second area before obtaining multiple second points of interest in the second area: taking the upper-level area of ​​the first area as the second area; taking the center of the first area as the center, extending the line between the center and the boundary point of the first area by a set multiple, and taking the end point of the extension processing as the boundary point to obtain the second area defined by the center and the boundary point; taking any area including the first area as the second area.

[0019] In the above scheme, the point of interest module is also used to obtain at least one of the following types of attribute information of each first point of interest: name information, category information, and detail information, wherein the geographical location of the first point of interest is within the first area; extract keywords from the attribute information of each first point of interest to form a keyword set corresponding to each first point of interest.

[0020] In the above scheme, the point of interest module is also used to perform at least one of the following processing: performing word segmentation processing on the attribute information based on a dictionary set to obtain keywords corresponding to the attribute information; performing word segmentation processing on the attribute information based on a sequence generation model to obtain keywords corresponding to the attribute information.

[0021] In the above scheme, the point of interest module is also used to match multiple characters of the attribute information with the character strings in the dictionary set in a specific direction, use the character string with the maximum length in the obtained attribute information as a keyword, and continue to match the remaining characters of the attribute information with the dictionary set from left to right until each character in the attribute information participates in the match; wherein, the specific direction includes: the backward direction starting from the first character of the attribute information, and the forward direction starting from the last character of the attribute information; the remaining characters are the characters in the attribute information that are not matched as the keywords.

[0022] In the above scheme, the point of interest module is also used to perform sequence labeling processing on the attribute information through the sequence generation model to obtain the character type corresponding to each character in the attribute information, so as to generate a keyword sequence corresponding to the attribute information according to the character type of each character; wherein, the sequence generation model is trained based on attribute information samples and pre-labeled keyword sequences.

[0023] In the above scheme, the first weight module is further used to sum the number of times each keyword appears in multiple keyword sets to obtain the total number of keywords corresponding to the first area; and determine the first weight of each keyword based on a first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords.

[0024] In the above scheme, the first weight module is also used to determine the first weight of each of the keywords by performing any one of the following processes: determining the first weight of each keyword by a first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords; scoring each keyword through a first word scoring model to obtain a first score representing the importance of the keyword, and determining the product of the first score of each keyword and the first ratio as the first weight of each keyword; wherein, the first word scoring model is trained based on keyword samples and the first scores of pre-labeled keywords.

[0025] In the above scheme, the second weight module is also used to determine, for each of the keywords, a correction parameter of the keyword that is positively correlated with the total number of points of interest in the second area and negatively correlated with the number of points of interest in the second area that include the keyword; and the first weight of each of the keywords is corrected according to the correction parameter to obtain the second weight of each of the keywords.

[0026] In the above scheme, the second weight module is also used to determine the correction parameter of the keyword by performing any one of the following: determining a second ratio of the total number of second points of interest in the second area to the number of second points of interest including the keyword in the second area, as the correction parameter of the keyword; scoring the keyword through a second word scoring model to obtain a second score characterizing the uniqueness of the keyword, multiplying the second score by the number of points of interest including the keyword in the second area, and determining a third ratio of the total number of second points of interest in the second area to the multiplication result, as the correction parameter of the keyword; wherein the second word scoring model is trained based on keyword samples and the second scores of pre-annotated keywords.

[0027] In the above scheme, the second weight module is also used to score the multiple correction operators that correct the first weight through the correction parameters through the operator scoring model to obtain third scores corresponding to the multiple correction operators; through the correction operator with the largest third score, the correction parameters and the first weight of each keyword are fused to obtain the second weight of each keyword; wherein, the operator scoring model is trained based on the third score including operator samples and pre-labeled operator samples, and the second weight of the keyword is positively correlated with the correction parameter and positively correlated with the weight.

[0028] In the above scheme, the naming module is also used to determine the second highest-weighted keyword as the name of the first area, and then display the name of the first area on the human-computer interaction interface; in response to negative feedback information received regarding the name, present a page for editing the name; in response to an editing operation received on the page, collect edited names from multiple editing operations, and update the name of the first area according to the edited name.

[0029] An embodiment of the present application provides an electronic device, including:

[0030] a memory for storing executable instructions;

[0031] The processor is configured to implement the area naming method provided in the embodiment of the present application when executing the executable instructions stored in the memory.

[0032] An embodiment of the present application provides a computer-readable storage medium storing executable instructions for implementing the area naming method provided in the embodiment of the present application when executed by a processor.

[0033] The embodiments of the present application have the following beneficial effects:

[0034] The keywords of the points of interest in the first area can be used to stably characterize the characteristics of the area. The weight of the criticality of the keywords is characterized by extracting the number of times the keywords of the points of interest in the first area, ensuring the stability of weight reasoning, and combining the statistical data for the keywords in the buffer area including the first area to correct the weight, effectively removing the influence of irrelevant words on the naming results, thereby improving the accuracy and efficiency of the naming results. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figures 1A-1B This is a schematic diagram of the structure of the regional naming system provided in an embodiment of the present application;

[0036] Figure 2 is a structural diagram of an electronic device provided in an embodiment of the present application;

[0037] Figures 3A-3D Schematic diagram of the process of the region naming method provided in the embodiment of the present application;

[0038] Figure 4 This is a flowchart of the region naming method provided in an embodiment of the present application;

[0039] Figure 5 Schematic diagram of the buffer area of ​​the area naming method provided in the embodiment of the present application;

[0040] Figure 6 This is a schematic diagram of the interface of the area naming method provided in the embodiment of the present application;

[0041] Figures 7A-7C This is a regional schematic diagram of the regional naming method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0043] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0044] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0046] Before further describing the embodiments of the present application in detail, the nouns and terms involved in the embodiments of the present application are explained. The nouns and terms involved in the embodiments of the present application are subject to the following interpretations.

[0047] 1) Points of Interest (POIs) are landmarks on electronic maps. These include scenic spots, government departments, commercial establishments (such as gas stations, department stores, supermarkets, restaurants, hotels, convenience stores, and hospitals), tourist attractions (such as parks and public restrooms), historical sites, and transportation facilities (such as bus stops, parking lots, speed cameras, and speed limit signs). POIs must have associated data such as names, categories, coordinates, and elevation to appear on electronic maps.

[0048] 2) Cognitive area: an area with the same public recognition name, for example, a business district (Xidan in Beijing, Coastal City in Shenzhen), a plot of land with a common public historical memory (Dashilan in Beijing, Shenzhen). Cognitive area can also be an artificially delineated area with certain similar functional areas.

[0049] In related technologies, naming geographic areas requires accurate geographic boundaries and map information of the target area, which is then distributed to experts, who then name the area according to a certain pattern. For example, the area can be named according to the orientation of mountains, rivers, or according to representative landmarks in the area. With the advent of the information age, a large number of social media texts and images with geographic identifiers have appeared on the Internet. Related technologies use geotagged photo data to develop a clustering method based on triangulation to construct the range of different place names. Tagged photos can be used to estimate the names of specific areas in the city using kernel density estimation. Social networks can be used to determine the probability that different areas belong to a certain name, which in turn can be used to name specific areas. House rental and sale information can also be used to extract geographic cognitive names. Related technologies extract online house rental and sale related posts, their coordinates, and text content to obtain the name of the target area.

[0050] However, in the examples of this application, it was found that geographical naming is greatly influenced by subjective factors, and naming experts are limited to familiar areas, which makes it difficult to promote on a large scale. Extracting names through the Internet (for example, extracting names from dynamics tagged with geographical information posted on social media) has the following two disadvantages:

[0051] 1. Social media data sources are unstable, especially as different social media platforms grow and decline. For example, the types of active users on social network A today are different from those when it was first published; while social network B's website has been permanently closed. The instability of social media data sources will lead to unstable naming results, and the place names cannot be consistent over a long period of time.

[0052] 2. When extracting names from social media, clustering and density discrimination are used. The selection of clustering parameters and thresholds is subjective and cannot be accurately and objectively judged.

[0053] The embodiments of the present application provide a method, device, electronic device, and computer-readable storage medium for naming an area, which can objectively characterize the weight representing the criticality of each keyword by extracting the point of interest data in the target area, and modify the weight in combination with the statistical data for the keyword in the buffer area, thereby effectively improving the accuracy of area naming. The following describes an exemplary application of the electronic device provided by the embodiment of the present application. The device provided by the embodiment of the present application can be implemented as various types of user terminals such as laptop computers, tablet computers, desktop computers, set-top boxes, mobile devices (for example, mobile phones, portable music players, personal digital assistants, dedicated messaging devices, portable gaming devices), and vehicle-mounted devices, and can also be implemented as a server. Below, an exemplary application when the device is implemented as a terminal will be described.

[0054] See also Figure 1A , Figure 1A This is a schematic diagram of the structure of the area naming system provided in an embodiment of the present application. The terminal 400 is connected to the server 200 via the network 300. The network 300 can be a wide area network or a local area network, or a combination of the two. The terminal 400 presents the target area to be named. The terminal 400 obtains from the server 200 a plurality of first points of interest in the first area to be named, a set of keywords corresponding to each first point of interest, the total number of a plurality of second points of interest in the second area, and the number of second points of interest corresponding to each keyword in the second area; the terminal 400 determines the first weight of each keyword based on the information obtained from the server, and corrects the first weight of each keyword to obtain the second weight of each keyword. The keyword with the largest second weight is determined as the name of the first area and presented on the terminal 400.

[0055] In some embodiments, the terminal 400 is connected to the server 200 via the network 300, and the terminal 400 presents the target area to be named and sends the target area to be named to the server 200, so that the server 200 determines the first weight of each keyword based on the multiple first points of interest in the first area to be named, the keyword set corresponding to each first point of interest, the total number of multiple second points of interest in the second area, and the number of second points of interest corresponding to each keyword in the second area, and corrects the first weight of each keyword to obtain the second weight of each keyword, determines the keyword with the largest second weight as the name of the first area, and returns the name of the first area to the terminal 400 for presentation.

[0056] See also Figure 1B , Figure 1B This is a structural diagram of the regional naming system provided by an embodiment of the present application. The nodes in the blockchain network need to go to the authentication center 700 for identity authentication. The terminal 400 can be a node in the blockchain network. The data related to the first point of interest and the second point of interest (for example, a keyword set) are stored in the form of key-value pairs on the blockchain network. When the smart contract of the node (terminal 400, terminal 500-1, terminal 500-2 and terminal 500-3) is called, each node executes the regional naming method provided by the embodiment of the present application to determine the name of the first area. The name that passes the consensus (for example, confirmed by more than half of the nodes) is used as the name of the first area. The name of the first area will be stored on the chain, and the name query operation for the first area is supported.

[0057] The following describes an exemplary application scenario of the area naming method provided in the embodiments of the present application.

[0058] In some embodiments, the regional naming method provided in the embodiments of the present application can be applied to the takeaway order delivery scenario. Based on multiple takeaway merchants in a certain geographical area, the names of the regions to which the multiple takeaway merchants belong are determined, so that when the takeaway platform allocates takeaway delivery business, the delivery range can be divided into different regions and their corresponding names can be determined, which facilitates the allocation of orders to delivery personnel in different regions.

[0059] In some embodiments, the area naming method provided in the embodiments of the present application can be applied to electronic map browsing scenarios. When a user demarcates an area on the map, the name of the demarcated area can be determined, and the user can be supported to personalize the name of the area to facilitate the user's subsequent use in the corresponding area. By providing personalized area naming services, the efficiency of electronic map retrieval can be improved.

[0060] In some embodiments, the server 200 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The terminal 400 can be a smart phone, tablet computer, laptop computer, desktop computer, smart speaker, smart watch, car-mounted device, etc., but is not limited to this. The terminal and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiments of the present application.

[0061] See also Figure 2 , Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, and is described by taking the electronic device as a terminal as an example. Figure 2 The terminal 400 shown includes: at least one processor 410, a memory 450, at least one network interface 420, and a user interface 430. The various components in the terminal 400 are coupled together via a bus system 440. It is understood that the bus system 440 is used to achieve connection and communication between these components. In addition to including a data bus, the bus system 440 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, the bus system 440 is not shown in FIG. Figure 2 Various buses are labeled as bus system 440 .

[0062] The processor 410 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., where the general-purpose processor can be a microprocessor or any conventional processor, etc.

[0063] The user interface 430 includes one or more output devices 431 that enable presentation of media content, including one or more speakers and / or one or more visual display screens. The user interface 430 also includes one or more input devices 432, including user interface components that facilitate user input, such as a keyboard, mouse, microphone, touch screen display, camera, other input buttons and controls.

[0064] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard drives, optical drives, etc. The memory 450 may optionally include one or more storage devices that are physically remote from the processor 410.

[0065] The memory 450 includes volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory may be a read-only memory (ROM), and the volatile memory may be a random access memory (RAM). The memory 450 described in the embodiments of the present application is intended to include any suitable type of memory.

[0066] In some embodiments, the memory 450 can store data to support various operations, examples of which include programs, modules, and data structures, or a subset or superset thereof, as exemplified below.

[0067] The operating system 451 includes system programs for handling various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic businesses and handle hardware-based tasks; a network communication module 452 is used to reach other computing devices via one or more (wired or wireless) network interfaces 420. Exemplary network interfaces 420 include: Bluetooth, Wireless Compatibility Certification (WiFi), and Universal Serial Bus (USB); a presentation module 453 is used to enable information to be presented (for example, a user interface for operating peripheral devices and displaying content and information) via one or more output devices 431 associated with the user interface 430 (for example, a display screen, a speaker, etc.); an input processing module 454 is used to detect one or more user inputs or interactions from one of the one or more input devices 432 and translate the detected inputs or interactions.

[0068] In some embodiments, the area naming device provided in the embodiments of the present application can be implemented in software. Figure 2 An area naming device 455 stored in the memory 450 is shown, which can be software in the form of a program and plug-in, etc., including the following software modules: a point of interest module 4551, a first weight module 4552, a second weight module 4553 and a naming module 4554. These modules are logical, and therefore can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be explained below.

[0069] The area naming method provided in the embodiment of the present application will be described in conjunction with the exemplary application and implementation of the terminal provided in the embodiment of the present application.

[0070] See also Figure 3A , Figure 3A This is a flow chart of the region naming method provided in the embodiment of the present application, which will be combined with Figure 3A The steps shown are explained.

[0071] In step 101, a plurality of first points of interest in a first area to be named are obtained, and a keyword set corresponding to each first point of interest is obtained.

[0072] For example, see Figure 5 , Figure 5 This is a schematic diagram of the buffer area of ​​the area naming method provided in the embodiment of the present application. The first area is Figure 5 In area A, the first area is the target area to be named. The first area can be determined according to set coordinates or captured according to a calibration operation on an electronic map. Point of interest data of the first area is obtained from the data on the electronic map. The first point of interest can be any landmark on the electronic map, such as a tourist attraction, a gas station, a department store, a supermarket, a restaurant, a hotel, a convenience store, a hospital, a transportation facility, etc.

[0073] In some embodiments, obtaining the keyword set corresponding to each first point of interest in step 101 can be achieved through the following technical solution: obtaining at least one of the following types of attribute information of each first point of interest from the point of interest data of the first area: name information, category information, and detail information, and extracting multiple keywords from the attribute information of each first point of interest to form a keyword set corresponding to each first point of interest.

[0074] As an example, when the first point of interest is a restaurant in the first area, the attribute information of the restaurant includes the name information of the restaurant, for example, Mcxlaw (R&F Plaza Store), the attribute information also includes category information, for example, catering, the attribute information also includes detailed information of the restaurant (for example, the restaurant's main cuisine, the restaurant's cooperative brands, the restaurant's franchisees, etc.), the attribute information of the first point of interest includes at least one of the above information, and keywords are extracted from the attribute information. For example, if the attribute information of the restaurant is "Mcxlaw (R&F Plaza Store)", multiple keywords are extracted from "Mcxlaw (R&F Plaza Store)" to obtain a keyword set corresponding to the first point of interest.

[0075] In some embodiments, extracting keywords from the attribute information of each first point of interest can be achieved by at least one of the following processes: performing word segmentation processing on the attribute information based on a dictionary set to obtain keywords corresponding to the attribute information; performing word segmentation processing on the attribute information based on a sequence generation model to obtain keywords corresponding to the attribute information.

[0076] As an example, the process of extracting keywords from attribute information is achieved through word segmentation. The word segmentation methods can include the following two: the first is dictionary-based word segmentation, which divides sentences into words according to the dictionary; the second is character-based word segmentation, which divides sentences into multiple characters and then combines the characters into words to find the optimal segmentation strategy.

[0077] In some embodiments, the above-mentioned word segmentation processing of attribute information based on the dictionary set to obtain keywords corresponding to the attribute information can be achieved through the following technical solution: matching multiple characters of the attribute information with the character strings in the dictionary set in a specific direction, using the character string with the maximum length in the obtained attribute information as the keyword, and continuing to match the remaining characters of the attribute information with the dictionary set from left to right until each character in the attribute information participates in the match; wherein, the specific direction includes: the backward direction starting from the first character of the attribute information, and the forward direction starting from the last character of the attribute information; the remaining characters are the characters in the attribute information that are not matched as the keywords.

[0078] As an example, the attribute information is treated as a sentence and matched with the character strings in the dictionary set. When the character string with the maximum length in the attribute information is matched, the matched character string is used as a keyword. For example, the attribute information "McX-Lao (R&F Plaza Store)" includes multiple characters "McX-Lao / R&F Plaza / Store". Starting from the first character (Mc) of the attribute information, it is matched backward with the character strings in the dictionary set. In the dictionary set, "Mc", "McX" and "McX-Lao" are matched in turn. "McX-Lao Fu" is not matched at that time. Therefore, "McX-Lao" is the character string with the maximum length and is extracted as the keyword. The remaining characters of the attribute information "R&F Plaza / Store" are matched with the character strings in the dictionary set from left to right to obtain the character string with the maximum length "R&F Plaza" until every character in the attribute information participates in the match. After every character in the attribute information participates in the match, all keywords can be extracted.

[0079] In some embodiments, the above-mentioned word segmentation processing of attribute information based on the sequence generation model to obtain keywords corresponding to the attribute information can be achieved through the following technical solution: sequence labeling processing is performed on the attribute information through the sequence generation model to obtain the character type of each character in the corresponding attribute information, so as to generate a keyword sequence corresponding to the attribute information according to the character type of each of the characters; wherein the sequence generation model is trained based on attribute information samples and pre-labeled keyword sequences.

[0080] As an example, unlike word segmentation based on a dictionary set, character-based word segmentation does not match words in sentences in advance, but regards word segmentation as a sequence labeling problem. The attributes of each character in the attribute information are first predicted through a sequence generation model. The predicted candidates include: Begin, Inside, Outside, End, and Single. These candidates represent the position of the character. The attribute information is sequence-labeled based on the predicted character attributes to obtain a keyword sequence corresponding to the attribute information. The input is the features composed of each character and its preceding and following characters, and the output is the word segmentation result (keyword sequence) obtained based on the predicted attributes of each character. The sequence generation model includes hidden Markov model, naive Bayes classification, etc.

[0081] As an example of a training sequence generation model, it is obtained by training based on attribute information samples and pre-labeled keyword sequences. The sequence generation model is a model that predicts the keyword sequence of attribute information. During the training phase of the model, the attribute information samples are input into the sequence generation model to predict the keyword sequence of the attribute information samples. The error between the predicted keyword sequence and the pre-labeled keyword sequence of the keyword sample is back-propagated between the sequence generation model to update the parameters of the sequence generation model.

[0082] In step 102, a first weight of each keyword is determined according to the number of times each keyword appears in a keyword set corresponding to a plurality of first interest points.

[0083] As an example, each first point of interest corresponds to a keyword set, and multiple first points of interest correspond to multiple keyword sets. For example, landmark A has a keyword set {McX-Laurels, R&F Plaza, store}, and landmark B has a keyword set {KFC, R&F Plaza, store}. Then the keyword "R&F Plaza" appears 2 times in the keyword sets corresponding to multiple first points of interest, and the keyword "McX-Laurels" appears 1 time in the keyword sets corresponding to multiple first points of interest.

[0084] See also Figure 3B , Figure 3B This is a flow chart of the region naming method provided in the embodiment of the present application, based on Figure 3A In step 102 , determining the first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the plurality of first interest points can be achieved through steps 1021 - 1022 .

[0085] In step 1021 , the number of times each keyword appears in multiple keyword sets is summed to obtain the total number of keywords corresponding to the first region.

[0086] As an example, for the keyword set A corresponding to coordinate A (i.e., the first point of interest A) and the keyword set A corresponding to coordinate B (i.e., the first point of interest B), there are the following four keywords: "McX-Lao", "KFC", "R&F Plaza", and "Store". Among them, the number of times the keyword "McX-Lao" appears in multiple keyword sets refers to the number of times the keyword "McX-Lao" appears in keyword set A and keyword set B, which is 1. The number of times "KFC" appears in keyword set A and keyword set B is 1, the number of times the keyword "R&F Plaza" appears in keyword set A and keyword set B is 2, and the number of times the keyword "Store" appears in keyword set A and keyword set B is 2. The number of times each keyword appears in multiple keyword sets is summed, that is, the above multiple times are added together, and the total number of keywords corresponding to the first area is 6.

[0087] In step 1022, a first weight of each keyword is determined based on a first ratio of the number of times each keyword appears in the plurality of keyword sets to the total number of keywords.

[0088] In some embodiments, the above-mentioned determination of the first weight of each keyword based on the first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords can be achieved by: performing any one of the following processes: determining the first weight of each keyword by the first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords; scoring each keyword through a first word scoring model to obtain a first score representing the importance of the keyword, and determining the product of the first score of each keyword and the first ratio as the first weight of each keyword.

[0089] As an example of training a first word scoring model, training is performed based on keyword samples and pre-labeled first scores of keywords. The first word scoring model is a model that predicts a first score representing the importance of the keyword. During the model training phase, keyword samples are input into the first word model to predict a score representing the importance of the keyword (the first score). The error between the predicted score and the pre-labeled first score of the keyword sample is back-propagated through the first word model to update the parameters of the first word scoring model.

[0090] As an example, the first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords is the word frequency of each keyword, and the word frequency is determined as the first weight of each keyword. For example, for the keyword "McX Law", the word frequency of the keyword "McX Law" is 1 / 6, and for the keyword "R&F Plaza", the word frequency of the keyword "R&F Plaza" is 1 / 3. The word frequency of each keyword is used as the first weight to objectively represent the importance of the keyword to the target area, which is conducive to improving the objectivity of area naming. Since keywords are stable words that can represent the first point of interest, that is, keywords such as "McX Law" and "R&F Plaza" will not be replaced at any time (words such as "R&F Plaza" and "McDonald's" are words with long-term recognition), which is conducive to improving the stability of area naming.

[0091] As an example, each keyword is scored by the first word scoring model to obtain a first score that represents the importance of the keyword. For example, the first score output by the first word scoring model for the keyword "store" is lower than the first score output by the first word scoring model for the keyword "McX-Lao", and the first score output by the first word scoring model for the keyword "McX-Lao" is lower than the first score output by the first word scoring model for the keyword "Fuli Plaza". The first ratio is adjusted by multiplying the first score of each keyword by the first ratio, and the adjusted weight is determined as the first weight of each keyword. By combining the deep learning model with objective statistical data, the core meaning of each keyword is learned on the basis of ensuring the objectivity of the first weight, so that the first weight can more effectively represent the criticality of the keyword to the first area (target area).

[0092] In step 103, a plurality of second points of interest in a second area are acquired, wherein the second area includes the first area.

[0093] As an example, the second area is formed by expanding the first area by a specific area, radius (if the first area is an area within a circle), and diagonal length (if the first area is a rectangular area), see Figure 7A , Figure 7A This is a regional schematic diagram of the regional naming method provided in an embodiment of the present application. Figure 7A The first region is a rectangular region 701A, and the second region 702A is a region having the same center as the first region, partially overlapping with the diagonal line, and having a longer diagonal line (longer than the diagonal line of the first region).

[0094] As an example, the second region is a complete upper-level region including the first region. There are n levels between the second region and the first region, where n is a non-negative integer. When divided by administrative level, if the second region is the administrative region above the first region, n is zero. Figure 7B , Figure 7B This is a regional schematic diagram of the regional naming method provided in an embodiment of the present application. Figure 7B The first area is community A, and the second area is street B, the upper-level administrative area of ​​the first area, wherein the second area is the complete area of ​​street B.

[0095] As an example, the second region is an incomplete superior region that includes a portion of the first region that exceeds a set ratio (e.g., 50%). It is only necessary to ensure that the proportion of the area of ​​the second region exceeding the first region to the area of ​​the first region exceeds the set ratio to ensure that there are enough names of the second region to eliminate the influence of irrelevant high-frequency words on the naming and effectively control the statistics and calculation amount in the naming process. See Figure 7C , Figure 7C This is a regional schematic diagram of the regional naming method provided in an embodiment of the present application. Figure 7C The first area is community A, and the second area is part of street B, which is the upper-level administrative area of ​​the first area. The second area is an incomplete area of ​​street B, and the area of ​​the second area exceeding the first area accounts for more than a set ratio (for example, 50%) to the area of ​​the first area.

[0096] For example, the ratio of the area of ​​the second region exceeding the first region to the area of ​​the first region is related to the specific application scenario. The larger the area of ​​the second region, the more second region names can be obtained to eliminate the influence of irrelevant high-frequency words on the naming, thereby making the determined first region name more representative and the first region naming more accurate. Different business scenarios have different requirements for the representativeness of the first region naming. For different representativeness requirements, there are second region area settings corresponding to multiple representativeness requirements. For example, assuming that the representativeness requirement for region naming in an electronic map browsing scenario is lower than that in a food delivery business, the representativeness requirement for region naming can be quantified by a parameter. The area of ​​the second region involved in the naming is determined based on the representativeness requirement of the first region name in the business scenario, where the representativeness requirement is positively correlated with the area of ​​the second region. In response to the user's selection operation for the representativeness requirement, the area of ​​the second region corresponding to the representativeness requirement is determined, and a second region that meets the required area is obtained based on the first region. The subsequent steps 104-105 are performed using the second point of interest data and associated keyword data in the second region. The user's selection operation can be triggered at any time before step 104.

[0097] In step 104, the first weight of each keyword is modified according to the total number of second points of interest in the second area and the number of second points of interest corresponding to each keyword in the second area to obtain a second weight of each keyword.

[0098] See also Figure 3D , Figure 3D This is a flow chart of the region naming method provided in the embodiment of the present application, based on Figure 3A In step 104, the first weight of each keyword is corrected according to the total number of second points of interest in the second area and the number of second points of interest corresponding to each keyword in the second area, and the second weight of each keyword can be obtained through steps 1041-1042.

[0099] In step 1041 , for each keyword, a correction parameter of the keyword is determined that is positively correlated with the total number of second points of interest in the second area and negatively correlated with the number of second points of interest including the keyword in the second area.

[0100] In some embodiments, the above-mentioned determination of the correction parameter of the keyword that is positively correlated with the total number of second points of interest in the second area and negatively correlated with the number of second points of interest including the keyword in the second area can be achieved by the following technical solution: performing any one of the following processes: determining a second ratio of the total number of second points of interest in the second area to the number of second points of interest including the keyword in the second area as a correction parameter for the keyword; scoring the keyword through a second word scoring model to obtain a second score characterizing the uniqueness of the keyword, multiplying the second score by the number of second points of interest including the keyword in the second area, and determining a third ratio of the total number of second points of interest in the second area to the multiplication result as a correction parameter for the keyword.

[0101] As an example of training a second word scoring model, training is performed based on keyword samples and pre-labeled second scores of keywords. The second word scoring model is a model that predicts a second score that represents the uniqueness of each keyword. During the model training phase, keyword samples are input into the second word scoring model to predict the second score that represents the uniqueness of the keyword. The error between the predicted second score and the pre-labeled second score of the keyword sample is back-propagated through the second word scoring model to update the parameters of the second word scoring model.

[0102] As an example, a second ratio of the total number of second points of interest in the second area to the number of second points of interest including the keyword in the second area is determined as a correction parameter for the keyword. The second ratio between the number of second points of interest including a certain keyword in the second area and the total number of second points of interest in the second area can be used as a correction parameter for the keyword.

[0103] For example, the second area includes three second points of interest. In addition to the above-mentioned coordinates A and B, it also includes a coordinate C that is in the second area but not in the first area. The coordinate C has the keyword set {Burger King, Shuangjing, store}. Therefore, it can be determined that the number of second points of interest including the keyword "Fuli Plaza" in the second area is 2, and the number of second points of interest in the second area is 3. The correction parameter is 1.5 (that is, the ratio is directly used as the correction parameter), or the second ratio is taken as the logarithm with a base of 10 (when taking the logarithm, the denominator is added with 1 to prevent the denominator from being 0), to obtain the keyword correction parameter. The correction parameter can also be lg1.5, see formula (1):

[0104]

[0105] Among them, IDF i is the correction parameter of keyword i, |P'| refers to the total number of second interest points in the second area, |{t∈P':i∈t}| is the number of second interest points containing keyword i in the second area. In order to prevent the denominator in formula (1) from being 0, 1 is added to the denominator.

[0106] As an example, each keyword is scored using the second word scoring model to obtain a second score that characterizes the uniqueness of each keyword. For example, the second score output by the second word scoring model for the keyword "shop" is lower than the second score output by the second word scoring model for the keyword "McX-Lao." The second score of each keyword is multiplied by the second ratio (to obtain a third ratio) as a correction parameter. By combining a deep learning model with objective statistical data, the core meaning of each keyword is learned while ensuring the objectivity of the correction parameter, so that the correction parameter can more effectively correct the first weight, so that the obtained second weight can truly represent the first area.

[0107] In step 1042, the first weight of each keyword is modified according to the modification parameter to obtain the second weight of each keyword.

[0108] As an example, the correction parameter of each keyword is multiplied by the first weight of the keyword to obtain the second weight of each keyword. For example, the first weight of the keyword "store" is higher, thereby indicating that the keyword "store" has a higher criticality. The correction parameter of the keyword "store" is lower, thereby indicating that the keyword "store" has a lower uniqueness. The first weight of the keyword "R&F Plaza" is higher, thereby indicating that the keyword "R&F Plaza" has a higher criticality. The correction parameter of the keyword "R&F Plaza" is higher, higher than the correction parameter of the keyword "store", thereby indicating that the keyword "R&F Plaza" has a higher uniqueness. After correction processing (multiplication), the second weight of the keyword "R&F Plaza" obtained is higher than the second weight of the keyword "store", that is, the keyword "R&F Plaza" is more representative of the first area and is therefore suitable to be the name of the first area.

[0109] In some embodiments, the above-mentioned correction processing of the first weight of each keyword according to the correction parameter to obtain the second weight of each keyword can be achieved through the following technical solution: multiple correction operators that correct the first weight according to the correction parameter are scored through the operator scoring model to obtain third scores corresponding to the multiple correction operators; the correction parameter and the first weight of each keyword are fused through the correction operator with the largest third score to obtain the second weight of each keyword; wherein the operator scoring model is trained based on the third score including operator samples and pre-labeled operator samples, and the second weight of the keyword is positively correlated with the correction parameter and positively correlated with the weight.

[0110] As an example, the operator scoring model is a model that predicts a third score that characterizes the regional adaptability of each operator. During the training phase of the model, the operator sample is input into the operator scoring model to predict the third score that characterizes the regional adaptability of the operator. The error between the predicted third score and the pre-labeled third score of the operator sample is back-propagated between the operator scoring models to update the parameters of the operator scoring model. Each operator is scored by the operator scoring model to obtain a third score that characterizes the regional adaptability of each operator. For example, for a certain first region and the corresponding second region, the third score output by the operator scoring model for the operator "correction parameter multiplied by the first weight" is different from the third score output by the operator scoring model for the operator "square of the correction parameter multiplied by the first weight". For different The first area and different second areas have different densities of second interest points, which means that the same operator is not applicable to all areas. It is necessary to score each operator through the operator scoring model in a certain scenario (a combination of the first area and the second area), and use the operator with the highest third score as the operator to perform correction processing in this scenario. The correction processing is essentially a fusion processing, which fuses the correction parameters with the first weight, and uses the fusion result as the second weight, and names it based on the second weight. When scoring each operator through the operator scoring model, the object processed by the operator scoring model is not only the operator, but also the data related to the second interest points of the first area and the second area, such as the number of second interest points, the distribution of second interest points, keywords for each second interest point, etc.

[0111] In step 105 , the name of the first area is determined based on the keyword having the largest second weight.

[0112] As an example, the keyword having the largest second weight is determined as the name of the first region.

[0113] In some embodiments, see Figure 3C , Figure 3C This is a flow chart of the region naming method provided in the embodiment of the present application, based on Figure 3A Before obtaining a plurality of second points of interest in the second area in step 103, any one of steps 106-108 may be performed to determine the second area.

[0114] In step 106, the upper-level region of the first region is used as the second region.

[0115] As an example, the first area is expanded to obtain the second area, and the second area is the superior area of ​​the first area. In terms of administrative dimension, the second area is the superior administrative area of ​​the first area. For example, the second area is Street A, and the first area is Community B within Street A. In terms of geographical dimension, the second area is the superior geographical area of ​​the first area. For example, the second area is the area within the Third Ring Road of Beijing, and the first area is the area within the Second Ring Road of Beijing. According to multi-level grid division, for example, an area is divided into nine even grids. The second area is the area represented by the four grids in the upper left corner, and the first area is the area represented by the one grid in the upper left corner.

[0116] In step 107, with the center of the first area as the center, the line connecting the center and the boundary point of the first area is extended by a set multiple, and the end point of the extension is used as the boundary point to obtain a second area defined by the center and the boundary point.

[0117] As an example, the first area is expanded to obtain the second area, the center of the first area is reused as the center of the second area, and the line between the center and the boundary point of the first area is extended away from the center. The extension distance can be set and adjusted according to the specific situation. The end point of the extension process is used as the boundary point of the second area, thereby determining the second area with the center and the boundary point as constraints. When the second area is a circle, it can be defined by the radius formed by the center to the boundary point, the center is the center of the circle, and the boundary point is the point on the circumference. When the second area is a rectangle, the boundary point is the vertex of the second area or the boundary point is the midpoint of the side of the second area.

[0118] As an example, when the line connecting the center and the boundary point of the first area is extended by a set multiple, all the boundary points of the first area can be traversed. For each boundary point, the line connecting the center and the boundary point is extended by a set ratio, and the end point of the extension is used as the boundary point of the second area. When extending the line connecting different boundary points and the center, different ratios can be extended (to obtain a second area of ​​any shape including the first area) or the same ratio (to obtain an enlarged version of the first area), and the second area is composed of multiple boundary points obtained by extension.

[0119] In step 108 , an arbitrary region including the first region is acquired as a second region.

[0120] For example, any area including the first area is used as the second area, and the second area acts as a buffer zone for the first area. This is to reduce the impact of keywords with high keyness in all areas. For example, the keyword "store" is universal in a large range, and even all food-related points of interest (first or second points of interest) will be associated with this keyword.

[0121] In some embodiments, step 105 determines the name of the first area based on the keyword with the largest second weight, and then displays the name of the first area on the human-computer interaction interface; in response to negative feedback information received regarding the name, presents a page for editing the name; in response to the editing operation received in the page, collects edited names from multiple editing operations, and updates the name of the first area according to the edited name.

[0122] As an example, after the second highest weighted keyword is determined as the name of the first area, the name of the first area is displayed on the human-computer interaction interface. Figure 6 , Figure 6 This is an interface diagram of the area naming method provided in an embodiment of the present application. After the first area name 602 is displayed on the human-computer interaction interface 601, negative feedback information regarding the name is received from the user. The negative feedback information is information indicating that the user does not approve of the displayed name of the first area. The negative feedback information can be carried in the user's interactive operation on the human-computer interaction interface 601. For example, a click operation on a certain control presents a page 603 (editing area) for editing the name in the human-computer interaction interface. In response to the editing operation received on page 603, the edit names from multiple editing operations are collected, and the name of the first area is updated according to the edit name. For example, if the edit names of multiple editing operations are all Lecheng Plaza, and the name of the first area displayed in the human-computer interaction interface 601 is R&F Plaza, Lecheng Plaza will be used as the name of the first area. If multiple editing operations display different edit names, one of the edit names will be used as the name of the first area according to the set rules.

[0123] Below, an exemplary application of the embodiment of the present application in an electronic map browsing scenario will be described.

[0124] In some embodiments, when a user defines an area on a map as a target area to be named (first area), the store names within a target area to be named usually carry the regional name of the target area that is recognized by the public, for example, XXX Xidan branch, XXX Coast City branch, etc. These contents are important information for automatic naming of regions. However, not all such information is for the target area, but may be for the upper-level area, for example, XXX Beijing branch, XXX Shenzhen branch. Therefore, the regional naming method provided in the embodiment of the present application accurately extracts the names within the target area and excludes the upper-level names.

[0125] In some embodiments, see Figure 4 , Figure 4It is a process structure diagram of the area naming method provided in an embodiment of the present application, which mainly includes the following four main processes: buffer area creation (determined based on the target area to be named), point of interest extraction (obtained through the store point of interest data set), store name segmentation, keyword extraction based on word frequency-reverse file frequency, and obtaining area naming results based on the keyword extraction results.

[0126] In some embodiments, see Figure 5 , Figure 5 This is a schematic diagram of the buffer area of ​​the area naming method provided in an embodiment of the present application. The buffer area refers to an area extending a certain range outside the target area. The area of ​​the buffer area can be set to 10 to 20 times the area of ​​the target area to be named, or directly set to the superior area, for example, the superior administrative area. The creation of the buffer area is mainly for comparison with the target area to be named to exclude the influence of the name of the superior area. The buffer area construction can be achieved through geographic information system software. Below, the target area to be named is set as target area A (area to be named A), and the buffer area is set as buffer area A'.

[0127] In some embodiments, a store point of interest refers to a store landmark with a store name and coordinate information. The coordinate information can be used to determine whether a store is located in the target area, and the store points of interest in the target area A (set as set P, i.e., the set of first points of interest) and the store points of interest in the buffer area A' (set as set P', i.e., the set of second points of interest) are extracted respectively, where set P is included in set P'.

[0128] In some embodiments, the Chinese place names are composed of phrases including multiple Chinese characters, and phrases (multiple keywords) are extracted from the store names (attribute information) of the store points of interest (first point of interest and second point of interest), and word segmentation processing is performed on each store name in each set.

[0129] Store Name Word segmentation results xx Lao (Coastal City Store) {xx Law, Coast City, Store} Niuwang Temple (East Building of Coastal City) {Bull King, Temple, Coastal City, East Block} Li x (Poly Culture Plaza Store) {Li x, Poly, Culture, Plaza, Store} Saixwei (Shenzhen Poly Culture Plaza) {Sports, Shenzhen, Poly, Culture, Plaza}

[0130] Table 1: Word segmentation results

[0131] In some embodiments, Term Frequency-Inverse Document Frequency (TF-IDF) can be used to evaluate the importance of a word to a set of documents or one of the documents in a corpus. The importance of a word increases in direct proportion to the number of times it appears in the document, but decreases in inverse proportion to the frequency of its appearance in the corpus. In the regional naming method provided in the embodiment of the present application, the frequency of occurrence of a keyword in the P set is calculated, and the ratio of the total number of stores in the P' set to the number of stores containing the keyword in the P' set is compared to evaluate whether the keyword is the name of the target area referred to by the P set. The term frequency refers to the frequency (first weight) of a given word i in the P set, see formula (2):

[0132]

[0133] Among them, n i,P Refers to the total number of occurrences of word i in set P (for example, the total number of occurrences of "Coastal City" in set P), Refers to the total number of words that appear in set P, TF i,P is the frequency of a given word i in the set P.

[0134] The reverse file frequency is used to measure the general importance of a word. The reverse file frequency of a specific word can be obtained by dividing the total number of files by the number of files containing the word, and then taking the logarithm of the obtained ratio to the base 10. In the regional naming method provided in the embodiment of the present application, the reverse file frequency of a keyword can be obtained by dividing the total number of stores in the P' set by the number of stores containing the keyword, and then taking the logarithm of the obtained quotient to the base 10, see formula (3):

[0135]

[0136] Among them, IDF i is the correction parameter of keyword i (the reverse file frequency of keyword i), |P'| refers to the number of stores in the P' set, |{t∈P':i∈t}| is the number of stores in the P' set that contain keyword i. In order to prevent the denominator in the calculation formula of the reverse file frequency from being 0, 1 is added to the denominator. Finally, the second weight of each keyword i is TF i,P ×IDF i , traverse all keywords and take the word with the second highest weight as the name of the area to be named A.

[0137] After determining the name of the area to be named, since the name of the area to be named excludes the naming of the superior area, it is effectively distinguished from the superior area and can accurately characterize the geographical characteristics of the area to be named. Therefore, the user can use the determined name to perform personalized marking on the electronic map, which can achieve precise marking and facilitate the user to search based on the precise marking during subsequent searches through the electronic map, thereby improving the efficiency of electronic map retrieval while ensuring retrieval accuracy.

[0138] The following continues to describe the exemplary structure of the region naming device 455 provided in the embodiment of the present application as a software module. In some embodiments, such as Figure 2 As shown, the software modules stored in the area naming device 455 of the memory 450 may include: an interest point module 4551, which is used to obtain multiple first interest points in the first area to be named, and obtain a keyword set corresponding to each first interest point; a first weight module 4552, which is used to determine the first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the multiple first interest points; the interest point module 4551 is also used to obtain multiple second interest points in the second area, wherein the second area includes the first area; a second weight module 4553 is used to correct the first weight of each keyword according to the total number of second interest points in the second area and the number of second interest points corresponding to each keyword in the second area to obtain the second weight of each keyword; a naming module 4554 is used to determine the name of the first area based on the keyword with the largest second weight.

[0139] In some embodiments, the point of interest module 4551 is also used to perform any of the following processing to determine the second area before obtaining multiple second points of interest in the second area: taking the upper-level area of ​​the first area as the second area; taking the center of the first area as the center, extending the line between the center and the boundary point of the first area by a set multiple, and taking the end point of the extension processing as the boundary point to obtain the second area defined by the center and the boundary point; taking any area including the first area as the second area.

[0140] In some embodiments, the point of interest module 4551 is also used to obtain at least one of the following types of attribute information for each first point of interest: name information, category information, and detail information, wherein the geographical location of the point of interest is within the first area; and extract keywords from the attribute information of each first point of interest to form a keyword set corresponding to each first point of interest.

[0141] In some embodiments, the point of interest module 4551 is also used to perform at least one of the following processing: performing word segmentation processing on the attribute information based on a dictionary set to obtain keywords corresponding to the attribute information; performing word segmentation processing on the attribute information based on a sequence generation model to obtain keywords corresponding to the attribute information.

[0142] In some embodiments, the point of interest module 4551 is also used to match multiple characters of the attribute information with the character strings in the dictionary set in a specific direction, use the character string with the maximum length in the obtained attribute information as the keyword, and continue to match the remaining characters of the attribute information with the dictionary set from left to right until each character in the attribute information participates in the match; wherein the specific direction includes: the backward direction starting from the first character of the attribute information, and the forward direction starting from the last character of the attribute information; the remaining characters are the characters in the attribute information that have not been matched as the keyword.

[0143] In some embodiments, the point of interest module 4551 is also used to perform sequence labeling processing on the attribute information through a sequence generation model to obtain the character type of each character in the corresponding attribute information, so as to generate a keyword sequence corresponding to the attribute information according to the character type of each character; wherein the sequence generation model is trained based on attribute information samples and pre-labeled keyword sequences.

[0144] In some embodiments, the first weight module 4552 is also used to sum the number of times each keyword appears in multiple keyword sets to obtain the total number of keywords corresponding to the first area; and determine the first weight of each keyword based on the first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords.

[0145] In some embodiments, the first weight module 4552 is further used to determine the first weight of each of the keywords by performing any one of the following processes: determining the first weight of each keyword by a first ratio of the number of times each keyword appears in multiple keyword sets to the total number of keywords; scoring each keyword through a first word scoring model to obtain a first score representing the importance of the keyword, and determining the product of the first score of each keyword and the first ratio as the first weight of each keyword; wherein the first word scoring model is trained based on keyword samples and the first scores of pre-labeled keywords.

[0146] In some embodiments, the second weight module 4553 is also used to determine, for each keyword, a correction parameter of the keyword that is positively correlated with the total number of points of interest in the second area and negatively correlated with the number of points of interest in the second area that include the keyword; and the first weight of each keyword is corrected according to the correction parameter to obtain the second weight of each keyword.

[0147] In some embodiments, the second weight module 4553 is also used to determine the correction parameter of the keyword by performing any one of the following: determining a second ratio of the total number of second points of interest in the second area to the number of second points of interest including the keyword in the second area, as the correction parameter of the keyword; scoring the keyword through a second word scoring model to obtain a second score characterizing the uniqueness of the keyword, multiplying the second score by the number of second points of interest including the keyword in the second area, and determining a third ratio of the total number of second points of interest in the second area to the multiplication result, as the correction parameter of the keyword; wherein the second word scoring model is trained based on keyword samples and the second scores of pre-labeled keywords.

[0148] In some embodiments, the second weight module 4553 is also used to score multiple correction operators that correct the first weight through correction parameters through an operator scoring model to obtain third scores corresponding to the multiple correction operators; the correction parameters and the first weight of each keyword are fused through the correction operator with the largest third score to obtain the second weight of each keyword; wherein the operator scoring model is trained based on the third scores of operator samples and pre-labeled operator samples, and the second weight of the keyword is positively correlated with the correction parameter and positively correlated with the weight.

[0149] In some embodiments, the naming module 4554 is also used to display the name of the first area on the human-computer interaction interface after determining the second highest-weighted keyword as the name of the first area; in response to negative feedback information received regarding the name, present a page for editing the name; in response to an editing operation received on the page, collect edited names from multiple editing operations, and update the name of the first area according to the edited name.

[0150] The present invention provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the region naming method described in the present invention.

[0151] The embodiment of the present application provides a computer-readable storage medium storing executable instructions, wherein the executable instructions are stored. When the executable instructions are executed by a processor, the processor will execute the region naming method provided by the embodiment of the present application, for example, Figures 3A-3D Shown is the area naming method.

[0152] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface storage, optical disk, or CD-ROM; or various devices including one or any combination of the above memories.

[0153] In some embodiments, executable instructions may be in the form of a program, software, software module, script, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0154] As an example, executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinating files (e.g., files storing one or more modules, subroutines, or code portions).

[0155] By way of example, executable instructions may be deployed to be executed on one computing device, or on multiple computing devices at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

[0156] In summary, through the embodiments of the present application, keywords of points of interest in the target area are extracted, and weights representing the criticality of each keyword are characterized. Since the keywords of points of interest in the target area are long-term valid information, the stability of weight reasoning can be improved. Combined with the statistical data for keywords in the buffer area including the target area, the weights are corrected to overcome the influence of irrelevant words on the naming results, thereby improving the accuracy and efficiency of the naming results without human intervention.

[0157] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent replacements, and improvements made within the spirit and scope of the present application are included in the scope of protection of the present application.

Claims

1. A method for naming an area, characterized in that: The method comprises: Acquire multiple first points of interest in a first area to be named, and acquire a keyword set corresponding to each of the first points of interest; determining a first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the plurality of first points of interest; Taking the center of the first region as the center, extending the line between the center and a boundary point of the first region by a set multiple, and using the end point of the extension as the boundary point, to obtain a second region defined by the center and the boundary point; The ratio of the area of ​​the second region exceeding the first region to the area of ​​the first region exceeds a set ratio; Acquire a plurality of second points of interest within the second area; Predicting the keyword sample using a second word scoring model to obtain a second score representing the uniqueness of the keyword; updating the second word scoring model based on an error between the second score and the pre-labeled second score of the keyword sample; Scoring the keyword using the second word scoring model to obtain the second score; multiplying the second score by the number of second points of interest that include the keyword in the second area, and determining a third ratio of the total number of second points of interest in the second area to the multiplication result as a correction parameter for the keyword; Modify the first weight of each keyword according to the modification parameter to obtain a second weight of each keyword; The name of the first area is determined based on the keyword having the largest second weight.

2. The method according to claim 1, characterized in that The obtaining of a keyword set corresponding to each first point of interest includes: Acquire at least one of the following types of attribute information of each of the first points of interest: name information, category information, and detail information, wherein the geographical location of the first point of interest is within the first area; Keywords are extracted from the attribute information of each of the first interest points to form a keyword set corresponding to each of the first interest points.

3. The method according to claim 2, characterized in that The extracting keywords from the attribute information of each first point of interest includes: Perform at least one of the following actions: Perform word segmentation processing on the attribute information based on a dictionary set to obtain keywords corresponding to the attribute information; The attribute information is segmented based on a sequence generation model to obtain keywords corresponding to the attribute information.

4. The method according to claim 3, characterized in that The word segmentation processing of the attribute information based on the dictionary set to obtain keywords corresponding to the attribute information includes: Matching multiple characters of the attribute information with the character strings in the dictionary set in a specific direction, using the character string with the longest length in the attribute information as a keyword, and continuing to match the remaining characters of the attribute information with the dictionary set from left to right until every character in the attribute information participates in the match; The specific direction includes: a backward direction starting from the first character of the attribute information, and a forward direction starting from the last character of the attribute information; the remaining characters are characters in the attribute information that are not matched as the keyword.

5. The method according to claim 3, characterized in that The segmentation processing of the attribute information based on the sequence generation model to obtain keywords corresponding to the attribute information includes: Performing sequence labeling processing on the attribute information using the sequence generation model to obtain a character type corresponding to each character in the attribute information, and generating a keyword sequence corresponding to the attribute information according to the character type of each character; The sequence generation model is trained based on attribute information samples and pre-labeled keyword sequences.

6. The method according to claim 1, wherein Determining the first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the plurality of first points of interest includes: summing the number of times each keyword appears in the plurality of keyword sets to obtain a total number of keywords corresponding to the first region; A first weight of each keyword is determined according to a first ratio of the number of times each keyword appears in the plurality of keyword sets to the total number of keywords.

7. The method according to claim 6, characterized in that Determining a first weight of each keyword according to a first ratio of the number of times each keyword appears in the plurality of keyword sets to the total number of keywords includes: The first weight of each keyword is determined by performing any one of the following processes: determining a first ratio of the number of times each keyword appears in the plurality of keyword sets to the total number of keywords as a first weight for each keyword; Scoring each keyword using a first word scoring model to obtain a first score representing the importance of the keyword, and multiplying the first score of each keyword by the first ratio to determine a first weight for each keyword; The first word scoring model is trained based on keyword samples and first scores of pre-labeled keywords.

8. The method according to claim 1, characterized in that The modifying the first weight of each keyword according to the modification parameter to obtain the second weight of each keyword includes: Scoring the plurality of correction operators that correct the first weight using the correction parameter using an operator scoring model to obtain third scores corresponding to the plurality of correction operators; fusing the correction parameter of each keyword and the first weight using a correction operator having the largest third score to obtain a second weight of each keyword; The operator scoring model is trained based on a third score including operator samples and pre-labeled operator samples, and the second weight of the keyword is positively correlated with the correction parameter and is positively correlated with the weight.

9. The method according to any one of claims 1 to 8, characterized in that After determining the name of the first area based on the keyword with the largest second weight, the method further includes: Displaying the name of the first area on the human-computer interaction interface; In response to receiving negative feedback information for the name, presenting a page for editing the name; In response to an edit operation received in the page, edit names resulting from a plurality of the edit operations are collected, and the name of the first area is updated according to the edit names.

10. A device for naming an area, characterized in that: The device comprises: An interest point module, configured to obtain a plurality of first interest points in a first area to be named, and obtain a keyword set corresponding to each of the first interest points; A first weight module, configured to determine a first weight of each keyword according to the number of times each keyword appears in the keyword set corresponding to the plurality of first interest points; The point of interest module is further configured to extend a line connecting the center of the first area and a boundary point of the first area by a set multiple, with the center of the first area as the center, and use the end point of the extension as the boundary point, to obtain a second area defined by the center and the boundary point; wherein the area of ​​the second area exceeding the first area accounts for an area of ​​the first area exceeding a set ratio; and obtain a plurality of second points of interest within the second area; a second weighting module, configured to predict the keyword sample using a second word scoring model to obtain a second score representing the uniqueness of the keyword; update the second word scoring model based on an error between the second score and the pre-labeled second score of the keyword sample; score the keyword using the second word scoring model to obtain the second score; multiply the second score by the number of second points of interest in the second area that include the keyword, and determine a third ratio of the total number of second points of interest in the second area to the multiplication result as a correction parameter for the keyword; and correct the first weight of each keyword according to the correction parameter to obtain a second weight for each keyword; The naming module is configured to determine the keyword with the largest second weight as the name of the first area.

11. An electronic device, characterized in that: include: a memory for storing executable instructions; A processor, configured to implement the area naming method according to any one of claims 1 to 9 when executing the executable instructions stored in the memory.

12. A computer-readable storage medium, characterized in that Executable instructions are stored, and when executed by a processor, the area naming method according to any one of claims 1 to 9 is implemented.

13. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the area naming method according to any one of claims 1 to 9 is implemented.

Images