Route generation method and device, medium and program product

Abstract

The invention discloses a route generation method and device, a medium and a program product, and the method comprises the steps: obtaining a route label corresponding to a target user, the route label indicating the feature content of a to-be-generated route, and the feature content being the content interested by the target user; the route label is input into a generation model, at least one POI sequence is generated through the generation model, the POI sequence comprises the characteristic content indicated by the route label, and each POI sequence comprises the transfer relation between different POIs. And for any group of POI sequences, generating a travel route corresponding to the group of POI sequences according to the transfer relationship among different POIs indicated by the group of POI sequences. Namely, through the technical scheme provided by the invention, the generated travel route passes through the POI in which the user is interested, so that the requirement of the user for non-efficient travel is met, and the travel experience of the user is improved.

Description

Technical Field

[0001] This application relates to the field of navigation technology, specifically to a route generation method, apparatus, medium, and program product. Background Technology

[0002] As people's living standards improve, more and more users are demanding non-efficient travel options. Efficient travel refers to travel that prioritizes speed in reaching the destination. Non-efficient travel, on the other hand, focuses more on the scenery, shops, and other aspects of the journey, such as food exploration trips or city exploration trips.

[0003] However, currently, routes are often generated based on the user's origin and destination, and are the shortest routes in terms of time or distance, which fails to provide users with inefficient travel routes.

[0004] Therefore, how to meet users' needs for inefficient travel is an urgent technical problem to be solved. Summary of the Invention

[0005] In view of this, this application provides a route generation method, apparatus, medium, and program product to generate travel routes with distinctive content for users and improve user experience.

[0006] To solve the above problems, the technical solution provided in this application is as follows:

[0007] In a first aspect of this application, a route generation method is proposed, the method comprising:

[0008] Obtain the route tags of the target user, which are used to indicate the features of the route to be generated;

[0009] The route labels are input into the generative model, and at least one set of Point of Interest (POI) sequences output by the generative model are obtained. Each set of POI sequences includes the transition relationship between different POIs. The generative model is pre-trained based on training samples, which include route labels and the POI sequences corresponding to the route labels.

[0010] For any set of POI sequences, generate the corresponding route based on the transition relationships between different POIs indicated by the set of POI sequences.

[0011] In a second aspect of this application, a route generation apparatus is provided, the apparatus comprising:

[0012] The first acquisition unit is used to acquire the route tags of the target user, wherein the route tags are used to indicate the special features of the route to be generated;

[0013] The second acquisition unit is used to input the route label into the generation model and acquire at least one set of point of interest (POI) sequences output by the generation model. Each set of POI sequences includes the transition relationship between different POIs. The generation model is pre-trained based on training samples, which include route labels and POI sequences corresponding to the route labels.

[0014] The generation unit is used to generate a route corresponding to any set of POI sequences based on the transition relationships between different POIs indicated by the set of POI sequences.

[0015] In a third aspect of this application, an electronic device is provided, comprising: one or more processors; and a storage device having one or more programs stored thereon.

[0016] When the one or more programs are executed by the one or more processors, the one or more processors implement the route generation method described in the first aspect.

[0017] In a fourth aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the route generation method described in the first aspect.

[0018] In a fifth aspect of this application, a computer program product is provided that, when the computer program product is run on a computer, causes the computer to implement the route generation method described in the first aspect.

[0019] Therefore, this application has the following beneficial effects:

[0020] To generate routes that meet users' needs for inefficient travel, the method first obtains route tags matching the target user. These tags indicate the unique features of the route to be generated, which are of interest to the target user. The route tags are then input into a generation model, which generates at least one set of Points of Interest (POI) sequences. Each POI sequence contains the unique features indicated by the route tags, and each POI sequence includes transition relationships between different POIs. For any given set of POI sequences, a corresponding route is generated based on the transition relationships between the different POIs indicated by that set. In other words, the technical solution provided in this application generates travel routes that traverse features of interest to the user, meeting their needs for inefficient travel and improving the user experience. Attached Figure Description

[0021] Figure 1 A flowchart of a generative model training method provided in an embodiment of this application;

[0022] Figure 2a A generative model training framework diagram provided in this application embodiment;

[0023] Figure 2b A generative model structure diagram provided in an embodiment of this application;

[0024] Figure 3 A flowchart of a route generation method provided in an embodiment of this application;

[0025] Figure 4 This is a schematic diagram of an application scenario provided by an embodiment of this application;

[0026] Figure 5 This is a schematic diagram of a route generation device provided in an embodiment of this application;

[0027] Figure 6 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation

[0028] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the embodiments of this application will be further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0029] Route planning is one of the main functions offered by map software, and it's also a pressing need for users seeking efficient travel. In other words, map software often focuses on generating efficient travel routes. However, more and more users are generating increasingly diverse route requirements, such as routes along the coast, routes across grasslands, routes to explore nearby museums, routes to explore city food streets, romantic routes for evening walks, and so on.

[0030] To meet users' diverse travel needs, this application provides a route generation method. When providing travel routes to users, the method obtains route tags matching the target user, which reflect the content of interest to the target user. These route tags are then input into a generation model, which uses the route tags to generate at least one set of Points of Interest (POI) sequences matching the route tags. These POI sequences include transition relationships between different POIs. For any given set of POI sequences, a corresponding travel route is generated based on the transition relationships between the different POIs indicated by the sequence. In other words, this technical solution can generate travel routes of interest to users, meeting their travel needs and improving the user experience.

[0031] For ease of understanding, the technical terms involved in this application will be explained below. Unless otherwise specified, the definitions of the technical terms in this application shall prevail.

[0032] Semantic information refers to meaningful information that eliminates uncertainty and can be understood and interpreted using natural language. Specifically, large language models (LLMs) can be used to extract semantic information. LLMs are deep learning models trained on massive amounts of text data, capable not only of generating natural language text but also of deeply understanding its meaning. With the rapid development of LLMs, they can handle various natural language tasks, such as text classification, question answering, and dialogue, making them an important pathway to artificial intelligence.

[0033] As can be seen, LLM can understand users' travel intentions, making it possible to generate routes with semantic information. Therefore, this application can combine the understanding capabilities of LLM with users' route preferences to generate routes containing semantic information, thereby meeting users' diverse travel needs.

[0034] Points of Interest (POI) are a term in Geographic Information Systems (GIS) that refers to any geographic object that can be abstracted as a point, especially geographic entities closely related to people's lives, such as schools, banks, restaurants, gas stations, hospitals, and supermarkets. The main purpose of POIs is to describe the address of things or events, greatly enhancing the ability to describe and query the location of things or events, and improving the accuracy and speed of geographic positioning.

[0035] To facilitate understanding of the technical solution of this application, the training process of the generative model of this application will be explained below with reference to the accompanying drawings.

[0036] See Figure 1 The figure is a flowchart of a generative model training method provided in an embodiment of this application, as shown below. Figure 1 As shown, the method includes:

[0037] S101: Obtain multiple sets of POI sequences and the route label corresponding to each set of POI sequences.

[0038] In this application, to train the generative model, a large number of training samples can be obtained. These training samples include multiple sets of POI sequences actually searched and / or arrived at by users, along with route labels corresponding to each POI sequence. Specifically, these multiple sets of POI sequences include user search records and / or arrival records for consecutive POIs, which contain the user's transfers between different POIs. Here, "user" refers to a user who uses map software to query routes and plan travel routes and has authorized the map software to obtain relevant query information; it does not refer to any specific user.

[0039] It should be noted that a set of POI sequences in this embodiment includes multiple POI sequences and the transition order between multiple POIs. For example, a set of POI sequences including {milk tea shop a1, exhibition hall a2, celebrity residence a3} indicates that the user first arrives at milk tea shop a1, then at exhibition hall a2, and finally at celebrity residence a3. Continuous POIs refer to sequentially using each POI in the POI sequence as a starting point and the next adjacent POI as the destination for searching and / or visiting. For example, if the user's starting position is A, and they search and arrive at position B; then starting from B, they search and arrive at position C; and starting from C, they search and arrive at position D, the resulting POI sequence is A—>B—>C—>D.

[0040] In this embodiment, for each POI, relevant data can be obtained, such as POI type, search popularity, user reviews, and geographical location. The POI type can include coffee shops, milk tea shops, restaurants, shopping malls, exhibition halls, and celebrity residences, etc., which can be determined manually or automatically by the model. This relevant data can be used as part of the input data to train the initial generative model. The trained model, during use, can not only output a set of POI sequences but also the relevant data corresponding to each POI sequence, making it easier for users to understand each POI without requiring them to search for it separately, thus improving the user experience.

[0041] The route labels for each POI sequence can be obtained as follows: For a set of POI sequences, the POI sequences and prompt words are used as input to the semantic alignment module. The semantic alignment module outputs the route labels for the set of POI sequences, and the type of the route label belongs to the type indicated by the prompt words. That is, the semantic alignment module can be used to generate corresponding route labels for a large number of POI sequences, thereby forming a sufficiently large training sample. For example, a set of POI sequences includes {Shop 1: Roast Duck, Shop 2: Noodles with Soybean Paste, Shop 3: Hot Pot Meat, Shop 4: Snacks}, and prompt words include {Scenery, Shopping, Food, City Scenery}. The above POI sequences and prompt words are input into the semantic alignment module. The semantic alignment module performs semantic analysis on the POI sequences and outputs the route label "This is a route for exploring food" within the type specified by the prompt words.

[0042] The semantic alignment module's functionality can be implemented using an LLM (Local Management Model). Specifically, after receiving the input POI sequence and prompt words, the LLM performs semantic analysis on the POI sequence within the type range defined by the prompt words, obtaining the semantic analysis result, i.e., the route label corresponding to the POI sequence. This route label indicates the distinctive features of the route formed by the POI sequence. In other words, the route label describes the route formed by the POI sequence. For example, if the POI sequence is {A, B, C, D}, the corresponding route label would be "This is a scenic route through a distant city."

[0043] Prompts, in this context, are a method of using natural language to guide or stimulate artificial intelligence models to complete specific tasks. Their function is to provide the AI ​​model with contextual information about the input and the parameters it receives, helping the model better understand the intent of the input and respond accordingly. In this embodiment, the preset props can be manually limited; for example, the LLM's description of the route can only include: scenery, shopping, food, cityscapes, culture, nearby, and distant places.

[0044] In this embodiment, by setting prompt words, the route labels generated for different POI sequences are limited to a controllable range to ensure that the determined route labels are within a unified descriptive framework, such as "This is a route of xxxxx", thus avoiding the problem of model training non-convergence caused by diverse descriptions and reducing the training difficulty.

[0045] S102: Train the initial generative model using route labels as input and the POI sequences corresponding to the route labels as ground truth.

[0046] After obtaining a large number of training samples, the route labels are used as input to the initial generative model to obtain the POI sequence output by the initial generative model. The output of the initial generative model is adjusted using the ground truth (the POI sequence corresponding to the route labels) until a preset condition is met, thus completing the training of the generative model. The preset condition can be that the loss function value of the generative model fluctuates within a certain range and no longer decreases significantly, at which point the generative model can be considered to have converged, and training can stop; or, the preset condition can be that the number of iterations reaches a preset number of iterations.

[0047] As can be seen, this embodiment can obtain multiple sets of POI sequences by using data from users' actual search / arrival at POIs, and use each set of POI sequences and the corresponding route labels to train a generative model. This generative model can output POI sequences that match the route labels, so as to provide users with routes that meet their travel needs using these POI sequences.

[0048] For a better understanding of the technology in this application, please refer to [link / reference]. Figure 2a The aforementioned generative model training framework diagram. (See diagram below.) Figure 2a As shown, the semantic alignment module leverages the powerful world knowledge and inductive reasoning capabilities of LLM to describe routes formed by POI sequences within a defined range of prompts, thereby obtaining route labels. In practical use, the POI sequences and prompts from the collected training samples are input into the semantic alignment module, which then obtains the route labels corresponding to each POI sequence. These route labels are used as input, and the corresponding POI sequences are used as ground truth values ​​to train the initial generative model.

[0049] The structure of the generative model is as follows: Figure 2b As shown, the model includes a text encoder module, a condition embedding module, and a Transformer module. The text encoder module encodes the input text to obtain a corresponding vector. The condition embedding module converts user-input prompts, route lengths, and other conditions into vectors. The Transformer module predicts matching POI sequences based on the vectors output by the previous two modules. For the POI sequence output by the Transformer module, the error between it and the ground truth POI sequence is calculated. The parameters of the Transformer module are adjusted based on this error, and the model is retrained until the error is less than a preset threshold, thus completing the training of the generative model.

[0050] After training the generative model, it can be used to generate routes that meet the user's travel intentions. This will be explained below with reference to the accompanying diagram.

[0051] See Figure 3 The figure is a flowchart of a route generation method provided in an embodiment of this application, as shown below. Figure 3 As shown, the method includes:

[0052] S301: Obtain the route tags of the target user.

[0053] In this embodiment, to recommend routes that meet the travel needs of a target user, a route tag corresponding to that user is obtained. This route tag indicates the unique features of the route to be generated. Specifically, the route tag is a simple description of the route, describing its unique features. The route tag can be presented as "This is a route of xxxx," where "xxxx" describes the unique features of the route. For example, it could be a route for exploring food, a shopping route, or a route rich in history and culture.

[0054] The route tags of the target user can be obtained in the following ways:

[0055] The first method involves obtaining the target user's target location information; determining at least one initial route label based on the target location information and a preset correspondence; displaying at least one initial route label; and confirming the initial route label selected by the target user as the target user's route label. The correspondence includes the relationship between location information and initial route labels; one piece of location information can correspond to one or more initial route labels.

[0056] In this embodiment, the correspondence between different location information or location ranges and initial route tags can be pre-stored. After obtaining the target location information corresponding to the target user, one or more initial route tags are determined based on the target location information and the aforementioned correspondence, and these one or more initial route tags are displayed to the target user. In this way, the target user can select according to their own travel needs, and the selected initial route tag will be used as the target user's route tag.

[0057] The target location information can be the target user's current location or the location of interest corresponding to the target user's drag-and-zoom actions on a preset page. A preset page refers to a specific page within the map software, on which the user can trigger drag-and-zoom actions. For example, the preset page could be the homepage of the map software.

[0058] Specifically, when the target user does not trigger any operation on the preset page, one or more initial route labels can be determined based on the target user's current location and the preset correspondence.

[0059] When a target user triggers an action on a preset page, such as a drag or zoom action, the location information that the user is interested in will be determined based on the triggering action, and one or more initial route labels will be determined based on the location information and the preset correspondence.

[0060] If a drag operation is triggered, the location of interest corresponding to the drag operation can be determined in the following ways:

[0061] One approach is to obtain the map displayed on the preset page after the drag operation and use the center point of that map as the location of interest corresponding to the drag operation. Another approach is to determine the location of interest based on the target user's current location before the drag operation and the offset corresponding to the drag operation. Specifically, based on the target user's current location before the drag operation is triggered, the current location information is adjusted according to the drag direction and offset to obtain the location of interest corresponding to the drag operation.

[0062] The second approach involves obtaining the target user's input information; this input information, along with prompts, is then used as input to a semantic alignment module to obtain the route labels output by that module. The route labels are of the type indicated by the prompts, and the input information reflects the target user's travel needs.

[0063] In this embodiment, when a target user triggers an input operation in map software, the input information entered by the target user is acquired. This input information and prompt words are then input into a semantic alignment module. The semantic alignment module analyzes the input information to obtain the target user's travel needs and outputs route tags corresponding to those needs within the type range indicated by the prompt words. In short, the semantic alignment module performs semantic analysis on the input information, obtains semantic analysis results, and determines route tags that match the target user's travel needs based on the semantic analysis results and pre-set prompt words.

[0064] The input information can be voice information entered by the user through map software, which is then converted into text describing the travel needs through voice recognition and text conversion; or it can be text information entered by the user, which is not limited in this embodiment.

[0065] It should be noted that, in this embodiment, the prompt words used when obtaining the route tags of the target user using the semantic alignment module are the same as the prompt words used when obtaining the route tags of the POI sequence using the semantic alignment module during training the generative model.

[0066] S302: Input the route labels into the generative model and obtain at least one set of POI sequences output by the generative model.

[0067] After determining the route tag corresponding to the target user, the route tag is input into the generation model, which then generates at least one set of POI sequences based on the route tag. Each set of POI sequences includes the transition relationships between different POIs, indicating the order of travel between them. For example, the first set of POI sequences includes four POIs: {A, B, C, D}; the second set of POI sequences includes five POI sequences: {A, E, F, D, H}. The travel order indicated by the first set of POI sequences is A->B->C->D; the travel order indicated by the second set of POI sequences is A->E->F->D->H.

[0068] The generative model is pre-trained based on training samples, which include route labels and corresponding POI sequences. For details on the training process of the generative model, please refer to [link to training documentation]. Figure 1 The illustrated embodiment.

[0069] S303: For any set of POI sequences, generate the corresponding route for the set of POI sequences based on the transition relationships between different POIs indicated by the set of POI sequences.

[0070] For at least one set of POI sequences output by the generative model, a travel route will be determined for each set of POI sequences according to the transition relationships between the different POIs indicated by the set of POI sequences.

[0071] S303 can be implemented as follows:

[0072] One approach is to directly concatenate the POIs in a set of POI sequences according to their transfer order to generate the corresponding route for that set of POIs. For example, if a set of POI sequences includes five POI sequences {A, E, F, D, H}, then the corresponding route for that set of POIs would be A->E->F->D->H.

[0073] Another approach is to determine the target user's starting point; determine the location of the last POI in the POI sequence as the destination; and generate a route containing multiple POIs from the POI sequence that matches the route label, based on the starting point, destination, and the transition relationships between the different POIs indicated by the POI sequence. The target user's starting point can be their current location or a starting point entered by the user.

[0074] In short, starting from the target user's current location and ending at the location of the last POI in a sequence of POIs, a route is generated that passes through multiple POIs in that sequence and matches the route tags. Because this route passes through POIs that the target user is interested in, it meets the user's travel needs and improves the user's travel experience.

[0075] For example, if the target user's current location is in Beiyuan, and the identified POI sequence includes {China World Mall, TV Headquarters Building, CITIC Tower}, then the generated travel route would be Beiyuan → China World Mall → TV Headquarters Building → CITIC Tower.

[0076] When the generative model outputs multiple sets of POI sequences, it can generate a route for each set of POI sequences, thereby recommending multiple travel routes that meet the travel needs of the target user, who can then choose from the recommended routes.

[0077] Alternatively, historical travel information for different users can be pre-stored. When multiple sets of POI sequences are obtained, these sequences can be filtered based on the target user's historical travel information before generating a travel route. This avoids recommending routes that users have already visited, improving the user experience.

[0078] For example, if the target user's route label is "This is a route full of delicious food," the model will output three sets of POI sequences based on this route label: POI sequence 1 is {shop a, shop c, shop d, shop f}, POI sequence 2 is {shop b, shop e, shop m, shop n}, and POI sequence 3 is {shop p, shop q, shop k, shop s}. If the target user has visited shops a, d, and f, then when generating subsequent routes, only one route needs to be generated based on POI sequence 2 and one based on POI sequence 3, without needing to generate a route based on POI sequence 1 again.

[0079] It should be noted that when generating a route based on the starting point, ending point, and the transfer relationships between different POIs indicated by a set of POI sequences, the generated route can include all POI sequences in the set, or it can include only some of the POI sequences. For example, a set of POI sequences includes five POI sequences {A, E, F, D, H}. If the starting point is near point A, the generated route can include starting point -> A -> E -> F -> D -> H. If the starting point is between points E and F, and the generated route includes all POIs in the set, the route could be starting point -> E -> A -> F -> D -> H, or starting point -> A -> E -> F -> D -> H, resulting in a longer route and increased travel costs. To reduce detours and lower travel costs, points A and E can be excluded, and the generated route can be starting point -> F -> D -> H. That is, when generating a route, the POIs included in the route can be determined based on the relative position of the starting point and the POIs in the POI sequence, so as to provide the user with the optimal travel route.

[0080] It is evident that the travel routes generated by the above scheme incorporate features that users are interested in, meet users' needs for inefficient travel, and improve the user experience.

[0081] For ease of understanding, see [link to relevant documentation]. Figure 4 The diagram illustrates the route generation framework. In this scenario, the user's input is "First time in Beijing, what are some good places to visit?" Based on this input and prompts, the semantic alignment module generates a route label as "A city sightseeing route." This route label is then input into the generation model, which outputs one or more sets of Points of Interest (POI) sequences, and generates a travel route based on each POI sequence.

[0082] When no user input is triggered, at least one initial route label can be determined based on the user's location information and a preset correspondence. This initial route label will then be displayed to the user, who can choose one that suits their travel needs. In response to the user's selection, the chosen initial route label is input into the generation model to obtain one or more sets of POI sequences, and a travel route is generated based on each set of POI sequences.

[0083] Based on the above method embodiments, this application provides a route generation device and an electronic device, which will be described below with reference to the accompanying drawings.

[0084] See Figure 5 This figure is a structural diagram of a route generation device provided in an embodiment of this application, as shown below. Figure 5 As shown, the device 500 includes: a first acquisition unit 501, a second acquisition unit 502, and a generation unit 503.

[0085] Specifically, the first acquisition unit 501 is used to acquire the route tag of the target user, which is used to indicate the special features of the route to be generated;

[0086] The second acquisition unit 502 is used to input route labels into the generation model and acquire at least one set of POI sequences output by the generation model. Each set of POI sequences includes the transition relationship between different POIs. The generation model is generated by pre-training based on training samples, which include route labels and POI sequences corresponding to the route labels.

[0087] The generation unit 503 is used to generate a route corresponding to any set of POI sequences based on the transfer relationship between different POIs indicated by the set of POI sequences.

[0088] In one possible implementation, the first acquisition unit 501 is specifically used to acquire the target location information of the target user; determine at least one initial route label based on the target location information and a preset correspondence, the correspondence recording the correspondence between the location information and the initial route label; display at least one initial route label; and determine the initial route label selected by the target user as the target user's route label.

[0089] The target location information can be the target user's current location information; or the location information of interest corresponding to the target user's drag or zoom operation on the preset page.

[0090] In one possible implementation, the first acquisition unit 501 is specifically used to acquire the input information of the target user; take the input information and the prompt words as input to the semantic alignment module, and obtain the route label output by the semantic alignment module, wherein the type of the route label belongs to the type indicated by the prompt words.

[0091] In one possible implementation, the generation unit 503 is specifically used to determine the starting position of the target user; determine the position of the last POI in the group of POI sequences as the ending position; and generate a route containing multiple POIs in the group of POI sequences and conforming to route labels based on the starting position, the ending position and the transfer relationship between different POIs indicated by the group of POI sequences.

[0092] In one possible implementation, the training process of the generative model includes: obtaining multiple sets of POI sequences and route labels corresponding to each set of POI sequences, wherein each set of POI sequences includes multiple POIs and the order between the multiple POIs; training the initial generative model with route labels as input and the POI sequences corresponding to the route labels as ground truth, until the preset conditions are met.

[0093] In one possible implementation, obtaining the route label corresponding to each set of POI sequences includes: taking a set of POI sequences and prompt words as input to a semantic alignment module, obtaining the route label output by the semantic alignment module, wherein the type of the route label belongs to the type indicated by the prompt words.

[0094] It should be noted that the specific implementation of each unit in this embodiment can be found in the relevant descriptions in the above method embodiments. The division of units in this application embodiment is illustrative and only represents a logical functional division; in actual implementation, there may be other division methods. The functional units in this application embodiment can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. For example, in the above embodiments, the processing unit and the sending unit can be the same unit or different units. The integrated unit can be implemented in hardware or as a software functional unit.

[0095] Based on the route generation method provided in the above-described method embodiments, this application also provides an electronic device, including: one or more processors; a storage device storing one or more programs thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the route generation method described in any of the above embodiments.

[0096] The following is for reference. Figure 6This document illustrates a structural schematic diagram of an electronic device 600 suitable for implementing embodiments of this application. The terminal devices in these embodiments may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Android Devices), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs (televisions), desktop computers, etc. Figure 6 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0097] like Figure 6 As shown, electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from storage device 608 into random access memory (RAM) 603. RAM 603 also stores various programs and data required for the operation of electronic device 600. Processing device 601, ROM 602, and RAM 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.

[0098] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 6 An electronic device 600 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0099] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, it performs the functions defined in the methods of the embodiments of this application.

[0100] The electronic device provided in this application embodiment and the route generation method provided in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0101] Based on the route generation method provided in the above embodiments, this application provides a computer-readable medium storing a computer program thereon, wherein the program, when executed by a processor, implements the route generation method as described in any of the above embodiments.

[0102] It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium, a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0103] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0104] The units described in the embodiments of this application can be implemented in software or in hardware. The name of the unit / module does not necessarily limit the unit itself; for example, a voice data acquisition module can also be described as a "data acquisition module".

[0105] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

[0106] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0107] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0108] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A route generation method, characterized in that, The method includes: Obtain the route tags of the target user, which are used to indicate the special features of the route to be generated; The route labels are input into the generation model to obtain at least one set of Points of Interest (POI) sequences output by the generation model; each set of POI sequences includes the transition relationships between different POIs. For any set of POI sequences, generate the corresponding route based on the transition relationships between different POIs indicated by the set of POI sequences.

2. The method according to claim 1, characterized in that, The process of obtaining the target user's route tags includes: Obtain the target user's target location information; Based on the target location information and a preset correspondence, at least one initial route label is determined, wherein the correspondence records the correspondence between the location information and the initial route label; Display at least one initial route label; The initial route label selected by the target user is determined as the target user's route label.

3. The method according to claim 2, characterized in that, The target location information includes: The target user's current location information; or, The target user's drag or zoom operation on the preset page corresponds to the location of interest information.

4. The method according to claim 1, characterized in that, The process of obtaining the target user's route tags includes: Obtain the target user's input information; The input information and prompt words are used as input to the semantic alignment module to obtain the route label output by the semantic alignment module. The type of the route label belongs to the type indicated by the prompt words.

5. The method according to claim 1, characterized in that, The step of generating the route corresponding to the set of POI sequences based on the transfer relationships between different POIs indicated by the set of POI sequences includes: Determine the starting position of the target user; The position of the last POI in this POI sequence is determined as the endpoint position; Based on the starting point location, the ending point location, and the transfer relationships between different POIs indicated by the POI sequence, a route containing multiple POIs from the POI sequence and conforming to the route label is generated.

6. The method according to claim 1, characterized in that, The training process of the generative model includes: Obtain multiple sets of POI sequences and the route label corresponding to each set of POI sequences, wherein each set of POI sequences includes multiple POIs and the order between the multiple POIs; The initial generative model is trained using route labels as input and the POI sequences corresponding to the route labels as ground truth until preset conditions are met.

7. The method according to claim 6, characterized in that, Obtain the route label corresponding to each POI sequence, including: A set of POI sequences and prompt words are used as input to the semantic alignment module to obtain route labels output by the semantic alignment module. The type of the route labels belongs to the type indicated by the prompt words.

8. A route generation device, characterized in that, The device includes: The first acquisition unit is used to acquire the route tags of the target user, wherein the route tags are used to indicate the special features of the route to be generated; The second acquisition unit is used to input the route label into the generation model and acquire at least one set of Points of Interest (POI) sequences output by the generation model, wherein each set of POI sequences includes the transition relationship between different POIs. The generation unit is used to generate a route corresponding to any set of POI sequences based on the transition relationships between different POIs indicated by the set of POI sequences.

9. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the route generation method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on a computer, the computer enables the route generation method as described in any one of claims 1-7.

Images