Method, device, server and terminal device for determining recommended boarding point

Abstract

This specification provides a method, apparatus, server, and terminal device for determining a recommended boarding point. Based on this method, the server can first obtain the text data of the location names of multiple candidate boarding points, as well as the corresponding associated data; and determine the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points; and then comprehensively utilize the location name word vectors of the above-mentioned candidate boarding points and the corresponding associated data to determine the recommended boarding points that meet the requirements from the multiple candidate boarding points and recommend them to the user. By introducing and utilizing the text data of the location names of the candidate boarding points combined with the associated data, it is possible to integrate feature information of more dimensions, accurately predict the boarding points that are suitable for the user and recommend them to the user, thereby improving the success rate of recommending boarding points and allowing users to obtain a better user experience.

Description

Technical Field

[0001] This specification belongs to the field of Internet technology, and in particular, relates to a method, device, server and terminal device for determining a recommended boarding point. Background Art

[0002] Usually, when a user uses a taxi-hailing app to hail a taxi, the platform server will search for several locations around the user and push them to the user as recommended boarding points for the user to choose.

[0003] However, the boarding point pushed to the user based on the above method is often not suitable for the user, and the recommendation success rate is relatively low. The user often needs to re-enter the boarding point that meets his needs, resulting in a relatively poor user experience.

[0004] Currently, no effective solution has been proposed to the above problems. Summary of the invention

[0005] This specification provides a method, device, server and terminal device for determining a recommended boarding point, which can integrate feature information of more dimensions, accurately predict a boarding point suitable for the user and recommend it to the user, thereby improving the success rate of the recommendation and allowing the user to obtain a better user experience.

[0006] An embodiment of the present specification provides a method for determining a recommended boarding point, which is applied to a server, and includes: obtaining text data of location names of multiple candidate boarding points, and associated data; determining location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of the multiple candidate boarding points; and determining a recommended boarding point that meets the requirements from the multiple candidate boarding points using the location name word vectors of the candidate boarding points and the corresponding associated data.

[0007] An embodiment of the present specification also provides a method for determining a recommended boarding point, which is applied to a terminal device, and includes: sending a boarding request to a server; wherein the boarding request carries at least a user's location parameter; receiving a recommended boarding point fed back by the server; wherein the recommended boarding point is determined by the server responding to the boarding request, obtaining and utilizing text data of the location name of the candidate boarding point and corresponding associated data; and displaying the recommended boarding point to the user.

[0008] An embodiment of the present specification also provides a device for determining a recommended boarding point, including: an acquisition module, used to obtain text data of location names of multiple candidate boarding points, and associated data; a first determination module, used to determine location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of the multiple candidate boarding points; and a second determination module, used to use the location name word vectors of the candidate boarding points and the corresponding associated data to determine a recommended boarding point that meets the requirements from the multiple candidate boarding points.

[0009] The embodiment of the present specification also provides a server, including a processor and a memory for storing instructions executable by the processor, and the processor implements the relevant steps of the method for determining the recommended boarding point when executing the instructions.

[0010] The embodiment of the present specification also provides a terminal device, including a processor and a memory for storing instructions executable by the processor, and the processor implements the relevant steps of the method for determining the recommended boarding point when executing the instructions.

[0011] The embodiments of the present specification also provide a computer-readable storage medium having computer instructions stored thereon, and when the instructions are executed by a processor, the relevant steps of the method for determining a recommended boarding point are implemented.

[0012] Based on the method, device, server and terminal device for determining the recommended boarding point provided in this specification, the server can first obtain the text data of the location names of multiple candidate boarding points, and the corresponding associated data; and determine the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points; and then comprehensively utilize the location name word vectors of the above-mentioned candidate boarding points and the corresponding associated data to determine the recommended boarding points that meet the requirements from the multiple candidate boarding points and recommend them to the user. By introducing and utilizing the text data of the location names of the candidate boarding points, combined with the associated data, it is possible to integrate feature information of more dimensions, accurately predict the boarding points suitable for the user and recommend them to the user, thereby improving the success rate of recommending boarding points and allowing users to obtain a better user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of this specification, the drawings required for use in the embodiments will be briefly introduced below. The drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 It is a schematic diagram of an embodiment of the structural composition of a system using the method for determining a recommended boarding point provided in an embodiment of this specification;

[0015] Figure 2 is a schematic diagram of an embodiment of a method for determining a recommended boarding point provided by an embodiment of this specification, in a scenario example;

[0016] Figure 3 is a flowchart of a method for determining a recommended boarding point provided by an embodiment of this specification;

[0017] Figure 4is a flowchart of a method for determining a recommended boarding point provided by an embodiment of this specification;

[0018] Figure 5 It is a schematic diagram of the structure of a server provided by an embodiment of this specification;

[0019] Figure 6 It is a schematic diagram of the structure of a device for determining a recommended boarding point provided by an embodiment of the present specification;

[0020] Figure 7 It is a schematic diagram of the structural composition of a device for determining a recommended boarding point provided by an embodiment of the present specification. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0022] This specification embodiment provides a method for determining a recommended boarding point. Figure 1 As shown, the method can be specifically applied to a system including a server and a terminal device.

[0023] In this embodiment, the server may specifically include a background server applied to one side of a network platform (e.g., XX ride service platform) that can realize functions such as data transmission and data processing. Specifically, the server may be, for example, an electronic device with data calculation, storage and network interaction functions. Alternatively, the server may also be a software program running in the electronic device to provide support for data processing, storage and network interaction. In this embodiment, the number of the servers is not specifically limited. The server may specifically be one server, or several servers, or a server cluster formed by several servers.

[0024] In this embodiment, the terminal device may specifically include a client that is applied to the user side and can realize functions such as data collection and data transmission. Specifically, the terminal device may be, for example, an electronic device such as a desktop computer, a tablet computer, a laptop computer, a smart phone, etc. Alternatively, the terminal device may also be a software application that can be run in the above electronic devices. For example, it may be an XX ride APP running on a smart phone.

[0025] In this embodiment, when the user needs to take a ride, the user can perform corresponding operations in the taxi-hailing interface of the XX ride APP of the terminal device.

[0026] The terminal device can respond to the above operation of the user and generate a corresponding ride request (e.g., ride order, etc.). While generating the ride request, the terminal device can also collect the user's current location parameters (e.g., the latitude and longitude data of the user's current location, etc.); and then send the ride request carrying at least the user's location parameters to the server.

[0027] First, after receiving a ride request, the server can respond to the ride request and first determine the user's current location based on the user's location parameters. Then, the server can query the map data to find multiple locations that are close to the user's current location (for example, less than or equal to 200 meters) and allow rides as candidate boarding points. The server can determine the text data of the location names of the above multiple candidate boarding points (for example, POI address names, etc.) based on the map data.

[0028] At the same time, the server can determine and utilize the distance between each candidate boarding point and the user's current location, the reception time of the ride request or the ride time indicated by the user, current weather parameters, etc., and combine the user's current location as associated data.

[0029] Next, the server may determine the location name word vectors of the candidate boarding points corresponding to the text data of the location names of the plurality of candidate boarding points according to a preset processing rule.

[0030] Specifically, the server may first perform word segmentation processing on the text data of the location names of multiple candidate boarding points, respectively, to obtain multiple phrase combinations; wherein one phrase combination corresponds to the text data of the location name of a candidate boarding point, and one phrase combination may include one or more phrases. Then, by using a preset word vector database, word vectors of the location names of multiple candidate boarding points are obtained by determining and using the word vectors of the phrases; wherein the preset word vector database stores word vectors corresponding to multiple preset phrases; the preset word vector database is constructed by processing corpus text data using the first model based on the first model rule. The first model may specifically be a skip-gram model; the first model rule may specifically include one or more of the following: preset word frequency parameter rules, preset context parameter rules, and preset word vector dimension parameter rules.

[0031] Furthermore, the server can concatenate the location name word vectors of each candidate boarding point with the corresponding associated data to obtain multiple data groups; wherein each data group corresponds to a candidate boarding point. Then, the pre-trained processing model is called to process the multiple data groups, and the user's acceptance probability for each candidate boarding point is predicted as the processing result. The second model can specifically be a model based on the LightGBM algorithm.

[0032] Furthermore, based on the processing results, the server can arrange the multiple candidate boarding points in descending order of acceptance probability values ​​to obtain multiple sorted candidate boarding points; and filter out a specified number (for example, 3) of the top ranked candidate boarding points as recommended boarding points; and send the above recommended boarding points to the terminal device.

[0033] Accordingly, see Figure 2 As shown, the terminal device can display the above-mentioned recommended boarding points through the ride order setting interface of the XX ride APP for users to choose.

[0034] The user can select the most suitable recommended boarding point from the displayed recommended boarding points as the actual boarding point by clicking on the above ride order setting interface according to the specific situation. At the same time, the user can also enter the address of the destination in the destination input box.

[0035] Correspondingly, the terminal device can obtain and use the location parameters of the actual boarding point and the location parameters of the target address to generate a corresponding ride order; and send the ride order to the server.

[0036] The server then distributes the ride order to drivers near the user's current location, waiting for the driver to accept the order.

[0037] The driver who accepts the ride order can go to the corresponding pick-up point according to the location parameters of the pick-up point in the ride order to wait for the user to get on the car.

[0038] Through the above-mentioned embodiments, the influence of the location names of the candidate boarding points on the user's selection of the boarding points is fully considered. By introducing and utilizing the text data of the location names of the candidate boarding points and combining them with other related data, it is possible to integrate feature information of more dimensions and more accurately predict the boarding points suitable for the user and recommend them to the user, thereby improving the success rate of the recommendation and allowing the user to have a better user experience.

[0039] See also Figure 3 As shown, the embodiment of this specification provides a method for determining a recommended boarding point. The method is specifically applied to the server side. When implemented specifically, the method may include the following contents.

[0040] S301: Obtain text data of location names of multiple candidate boarding points, and associated data.

[0041] S302: Determine the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of the multiple candidate boarding points.

[0042] S303: Determine a recommended boarding point that meets the requirements from multiple candidate boarding points by using the location name word vectors of the candidate boarding points and the corresponding associated data.

[0043] In some embodiments, the candidate boarding point may be specifically understood as a location point that is within a proximity area of ​​the user's current location and that allows the user to board the vehicle.

[0044] In some embodiments, the text data of the location name of the candidate boarding point may specifically include location name text described in Chinese or other non-alphabetic languages, such as "Xinzhuang Center West Gate", "Oriental Hospital Minzu Street West Station", etc. Of course, during specific implementation, depending on the specific situation, the text data of the location name of the candidate boarding point may also include location name text described in other languages.

[0045] It should be noted that the existing methods for determining recommended boarding points often do not take into account the impact of the text data of the location name of the candidate boarding point on the user's choice of the boarding point. In fact, the text data of the location name of the candidate boarding point is one of the first information related to the candidate boarding point that the user receives when selecting the boarding point; and, different from other information, the text data of the location name of the candidate boarding point is also a relatively intuitive information, which will have a greater impact on the user's final choice. For example, based on certain preferences, users may tend to choose "such and such bus stop" or "such and such intersection" as the actual boarding point. Since the existing methods ignore the impact of the text data of the location name of the candidate boarding point on the user's choice of the boarding point, the dimensional features considered when predicting the user's acceptance probability of the candidate boarding point are not comprehensive and accurate, which leads to a low success rate in recommending the user's boarding point and affects the user's experience.

[0046] In some embodiments, the above-mentioned associated data can be specifically understood as one or more types of information other than the text data of the location name of the candidate boarding point that may affect the user's selection of the boarding point.

[0047] In some embodiments, the above-mentioned associated data may specifically include at least one of the following: the user's current location, the distance between the user's current location and the candidate boarding point, the ride time, etc.

[0048] Of course, it should be noted that the above-listed related data is only a schematic illustration. In specific implementation, according to specific circumstances and processing requirements, the above-listed related data may also include other types of information data, such as the time when the ride request initiated by the user is received; another example is the current weather parameters; another example is the user's age, gender, etc.

[0049] In some embodiments, the above-mentioned acquisition of text data of the location names of multiple candidate boarding points may include the following contents during specific implementation:

[0050] S1: Determine the user's current location;

[0051] S2: querying the map data and selecting multiple locations close to the user's current location as candidate boarding points;

[0052] S3: According to the map data, obtain text data of the location names of multiple candidate boarding points.

[0053] In some embodiments, when a user has a need for a ride, a corresponding ride request may be generated and sent to a server through a terminal device, wherein the ride request carries at least the user's location parameter.

[0054] Specifically, for example, a user can initiate a trigger operation by clicking the "Call a Ride" icon in the main interface of the XX ride app installed on the terminal device. Correspondingly, the terminal device can receive and respond to the above trigger operation and generate a corresponding ride request. At the same time, the terminal device can also obtain the latitude and longitude data of the user's current location as the user's location parameter; and then send the ride request carrying the user's location parameter to the server.

[0055] It should be noted that the actions of obtaining user-related data that appear in this manual are all implemented through compliant means with the user's knowledge and consent; and in the process of obtaining user-related data, corresponding encryption methods will also be used to properly protect the user's data privacy.

[0056] In some embodiments, the server may first extract the user's location parameters from the ride request; and then determine the user's current location based on the user's location parameters.

[0057] In some embodiments, during specific implementation, first, the server may query map data matching the user's current location from a map database based on the user's current location, such as an area map of the block where the user is currently located.

[0058] Next, the server can query the map data in the following manner: in the map data, use the current location as the center of the range and draw a corresponding adjacent range, for example, a circular range with the user's current location as the center and a radius of 200 meters; then search for location points with location names in the map data within the adjacent range, and filter out location points where boarding is allowed as candidate boarding points.

[0059] Furthermore, the server may query and obtain text data of the location names of the candidate boarding points based on the map data.

[0060] In some embodiments, the server may first convert the text data of the location names of the candidate boarding points into corresponding word vectors, and then effectively utilize the information contained in the text data of the location names of the candidate boarding points that may affect the user's choice of boarding points in the form of word vectors, so as to accurately predict the user's acceptance probability of the candidate boarding points to be recommended.

[0061] In some embodiments, the above-mentioned determination of the position name word vectors of the multiple candidate boarding points corresponding to the text data of the position names of the multiple candidate boarding points may include: determining the position name word vector of the current candidate boarding point corresponding to the text data of the position name of the current candidate boarding point in the following manner:

[0062] S1: Perform word segmentation on the text data of the location name of the current candidate boarding point to obtain multiple phrases;

[0063] S2: using a preset word vector database, determining word vectors of multiple word groups corresponding to the multiple word groups respectively; wherein the preset word vector database stores word vectors corresponding to the multiple preset word groups; the preset word vector database is constructed by processing corpus text data using the first model based on the first model rule;

[0064] S3: Determine the word vector of the location name of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point based on the word vectors of the multiple phrases.

[0065] In the above manner, according to the preset processing rules, the text data of the location names of other candidate boarding points are processed to obtain the location name word vectors of the other candidate boarding points corresponding to the text data of the location names of other candidate boarding points, thereby effectively and accurately converting the text data of the location names of multiple candidate boarding points into the corresponding location name word vectors of multiple candidate boarding points.

[0066] In some embodiments, during specific implementation, the server may utilize the jieba library to perform specific word segmentation processing on the text data of the location name of the current candidate boarding point to obtain a plurality of phrases corresponding to the location name of the current candidate boarding point.

[0067] The above-mentioned jieba library can be specifically understood as a Chinese natural language processing word segmentation library.

[0068] In specific implementation, the jieba library can be used to perform specific word segmentation processing on the text data of the location name of the current candidate boarding point. Based on the jieba library, multiple segmentation schemes for the text data of the location name of the previous candidate boarding point can be generated, and the score value of each segmentation scheme can be calculated; the segmentation results corresponding to the segmentation scheme with the highest score value are filtered out to obtain the corresponding multiple phrases.

[0069] Specifically, for example, the jieba library can be used to perform word segmentation on the text data of the location name of the current candidate boarding point, "Dongfang Hospital Minzu Street West Station", to obtain multiple phrases as shown below: Dongfang, Hospital, Minzu Street, West Station.

[0070] In some embodiments, the above-mentioned preset word vector database can specifically be a pre-constructed database related to the location name, which stores word vectors corresponding to multiple preset phrases.

[0071] Furthermore, the above-mentioned preset word vector database can also store the correspondence between preset phrases and word vectors, as well as the word vectors of default phrases.

[0072] In some embodiments, the above-mentioned preset word vector database can be specifically constructed in advance based on the first model rules by using the first model to process a large amount of corpus text data.

[0073] In some embodiments, the first model may specifically include a skip-gram model; correspondingly, the first model rule may specifically include at least one of the following: a preset word frequency parameter rule, a preset context parameter rule, and a preset word vector dimension parameter rule.

[0074] The skip-gram model can be specifically understood as a type of Word2Vec model, which is improved based on the feedforward neural network model. The skip-gram model can be understood as a simplified neural network, which can predict the current word by inputting two words in the context, that is, the context words are input into the model, and then processed by the hidden layer of the neural network, and finally output to the output layer, and then normalized to obtain the distribution probability of the word.

[0075] The first model rule can be specifically understood as a rule for controlling the first model to learn and generate word vectors. The word frequency parameter can be specifically recorded as: min_count; the context parameter can be specifically recorded as: window; the word vector dimension parameter can be specifically recorded as: vector_size.

[0076] In some embodiments, the corpus text data used to construct a preset word vector database may specifically include corpus text related to the location name.

[0077] In some embodiments, during specific implementation, a preset word vector database can be constructed based on the first model rules by using the first model to process a large amount of corpus text data to learn and train word vectors related to location names.

[0078] In a specific implementation, based on a preset word frequency parameter rule, the word frequency parameter can be set to 1, for example, min_count = 1. That is, the first model is controlled to learn and train the word vectors whose word frequency in the corpus text is greater than or equal to 1.

[0079] This is because the frequency of phrases in corpus texts that are usually related to location names or address names is often low. If the frequency parameter is set to 2 or a larger value as in the conventional method, it is easy to cause the first model to learn the relevant word vectors incompletely. Based on the above preset frequency parameter rules, the first model can be controlled to conduct a more comprehensive learning and training of all word vectors in the corpus text.

[0080] In specific implementation, based on the preset context parameter rules, the context parameter can be set to m, for example, window = m. That is, the first model is controlled to take m phrases before and after the context for learning and training, and the value of m is determined according to the actual number of phrases contained in the context where the phrase is located.

[0081] For example, for the phrase "West Station" in the text data of the location name "Xinzhuang Center West Station" shown below, after word segmentation processing, it can be determined that the actual number of phrases contained in the context where the phrase is located is 3 (that is, respectively: Xinzhuang, Center, West Station). At this time, the window can be automatically set to 3.

[0082] In this way, the specific values ​​of the context parameters can be automatically set based on the preset context parameter rules. Different from the conventional setting of fixed context parameters, the first model can be controlled more intelligently and accurately to learn and train word vectors.

[0083] In specific implementation, the preset word vector dimension parameter rule can set the word vector dimension parameter to the data value associated with the context parameter and the word frequency parameter, for example, vector_size=min_count*window*10. That is, the actual word vector dimension parameter is set according to the number of phrases in the context window and the word frequency min_count, so that the first model can be controlled to automatically determine the word vector dimension parameter suitable for the word vector.

[0084] In some embodiments, the above-mentioned use of a preset word vector database to determine the word vectors of multiple phrases corresponding to the multiple phrases respectively may include: determining the word vector of the current phrase among the multiple phrases in the following manner: querying the preset word vector database to determine a preset phrase that matches the current phrase as a matching phrase; and determining the word vector of the matching phrase as the word vector of the current phrase.

[0085] Specifically, according to the current phrase, a preset word vector database can be queried to see whether there is a preset word vector database with a difference value less than or equal to a preset difference threshold value with the current phrase, as a matching word vector that matches the current phrase; when the above matching phrase is queried, the word vector of the matching phrase can be obtained and determined as the word vector of the current phrase according to the correspondence between the preset phrase and the word vector. For example, in the above manner, by querying the preset word vector database, it can be determined that the word vector of the phrase "West Station" in "Xinzhuang Center West Station" is: [0.7655,-0.574,1.9122,0.1493].

[0086] In some embodiments, after querying the preset word vector database, the method may also include, when implemented specifically, if no preset phrase matching the current phrase is found in the preset word vector database, using the word vector of the default phrase in the preset word vector database as the word vector of the current phrase.

[0087] Specifically, if a preset phrase whose difference value with the current phrase is less than or equal to a preset difference threshold is not found in the preset word vector database, the word vector of the default phrase stored in the preset word vector data can be directly used as the word vector of the current phrase. Among them, the above-mentioned default phrase can specifically be a phrase without clear semantics that is statistically obtained based on a large amount of corpus text related to the location name. Of course, the word vector of the preset phrase with the smallest difference value with the current phrase can also be found from the preset word vector database as the word vector of the current phrase.

[0088] In some embodiments, during specific implementation, multiple word vectors corresponding to the location name of the current candidate boarding point may be combined to obtain the location name word vector of the current candidate boarding point. For example, the word vectors of multiple phrases are added together to obtain the sum as the location name word vector of the current candidate boarding point. For another example, the word vectors of multiple phrases are added together to obtain the average, and the obtained average value is used as the location name word vector of the candidate boarding point.

[0089] In some embodiments, during specific implementation, the above-mentioned determination of the word vector of the location name of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point based on the word vectors of the multiple word groups may include the following contents: arranging the word vectors of the multiple word groups according to the order of the word groups in the text data of the location name of the current candidate boarding point to obtain the word vectors of the multiple word groups after ordering; and selecting the word vectors of the preset number of word groups with a lower order from the word vectors of the multiple word groups after ordering to combine them to obtain the word vector of the location name of the current candidate boarding point. For example, the word vectors of the preset number of word groups with a lower order may be added to obtain the vector sum as the word vector of the location name of the current candidate boarding point.

[0090] The value of the above preset number may be 2. Of course, in specific implementation, other values ​​may be set as the preset number according to specific circumstances and accuracy requirements.

[0091] In this embodiment, it is further considered that when selecting a boarding point, users often pay more attention to and care about the last few phrases in the text data of the location name of the candidate boarding point. Therefore, in order to reduce the amount of data processing and improve the overall processing efficiency, the word vectors of a preset number of phrases ranked at the back can be selected from the word vectors of the sorted multiple phrases in the above manner to obtain the word vector of the location name of the current candidate boarding point.

[0092] According to the above method of determining the word vector of the location name of the current candidate boarding point, the word vectors of the location names of other candidate boarding points can be determined in sequence, so as to obtain the word vectors of the location names of multiple candidate boarding points.

[0093] In some embodiments, the above method of using the location name word vectors of the candidate boarding points and the corresponding associated data to determine a recommended boarding point that meets the requirements from multiple candidate boarding points may include the following when implemented:

[0094] S1: Combining the location name word vectors of multiple candidate boarding points and the corresponding associated data respectively to obtain multiple data groups; wherein one data group corresponds to one candidate boarding point;

[0095] S2: Processing multiple data groups using the second model to obtain corresponding processing results;

[0096] S3: According to the processing result, a recommended boarding point that meets the requirements is screened out from a plurality of candidate boarding points.

[0097] Through the above embodiment, the second model can be used to simultaneously process the location name word vectors and corresponding associated data of the candidate boarding points, so as to effectively and comprehensively use multi-dimensional feature information and more comprehensively and accurately determine the boarding point that the user is likely to accept as the recommended boarding point that meets the requirements.

[0098] In some embodiments, the second model may specifically be a probability prediction model based on the LightGBM algorithm (or the LightGBM / LR fusion algorithm).

[0099] In some embodiments, when splicing is performed, the location name word vector of the candidate boarding point and the associated data corresponding to the candidate boarding point, such as the distance between the user's current location and the candidate boarding point, can be spliced ​​to obtain a data group for the candidate boarding point.

[0100] Furthermore, other related data such as the user's current location, the time of receiving the ride request, the ride time indicated by the user, weather parameters, etc. can be integrated into the data group to obtain a data group with relatively richer and more comprehensive feature information for the candidate boarding point.

[0101] In some embodiments, during specific implementation, the above data group can be used as model input and input into the second model; and the second model is run to process the data group to output the user's acceptance probability for the candidate boarding point as a processing result.

[0102] In some embodiments, the second model can be trained in the following manner:

[0103] S1: Obtain the operation record of the test user for the test boarding point, as well as the text data and associated data of the location name of the test boarding point;

[0104] S2: Determine the location name word vector of the test boarding point corresponding to the text data of the location name of the test boarding point;

[0105] S3: Combine the location name word vector of the test boarding point with the corresponding associated data to obtain sample data;

[0106] S4: labeling the sample data according to the operation record of the test user for the test boarding point to obtain the labeled sample data;

[0107] S5: Perform model training using the labeled sample data to obtain the second model.

[0108] In some embodiments, the operation record may specifically include a test user's click operation record on the recommended test boarding point. Based on the operation record, it is possible to accurately determine whether the test user has accepted the recommended test boarding point according to the test user's specific operation on the recommended test boarding point.

[0109] In some embodiments, during specific implementation, the text data of the location name of the test boarding point can be processed using preset word vector data according to preset processing rules to obtain the corresponding word vector of the location name of the test boarding point.

[0110] In some embodiments, the location name word vector of the test boarding point can be combined with the corresponding associated data to obtain a data group as sample data corresponding to the test boarding point.

[0111] In some embodiments, when performing specific marking, it can be determined based on the operation record whether the test user has accepted the recommended test boarding point. If it is determined that the user has accepted the recommended test boarding point, the sample data corresponding to the test boarding point can be marked as a positive sample; on the contrary, if it is determined that the user has not accepted the recommended test boarding point, the sample data corresponding to the test boarding point can be marked as a negative sample.

[0112] Through the above embodiments, the sample data can be labeled more precisely to obtain labeled sample data with higher accuracy and better effect.

[0113] In some embodiments, during specific implementation, an initial model based on the LightGBM algorithm can be constructed as the initial second model; then, the above-mentioned labeled sample data is used to continuously learn and train the initial second model to obtain a second model with accuracy that meets the requirements.

[0114] In some embodiments, during specific implementation, the user's acceptance probabilities for different candidate boarding points can be ranked in descending order according to the probability values ​​based on the processing results; then the candidate boarding points corresponding to the top specified number (e.g., 3) of probability values ​​are selected as recommended boarding points that meet the user's requirements. Alternatively, the candidate boarding points with probability values ​​greater than or equal to a preset probability threshold can be selected as recommended boarding points based on the processing results.

[0115] In some embodiments, after determining the recommended boarding point, the server may send the recommended boarding point to the terminal device. Accordingly, the terminal device receives and displays the recommended boarding point to the user for selection.

[0116] The user can find a recommended boarding point that meets his or her needs from the recommended boarding points displayed by the terminal device according to the specific situation and make a selection operation to determine the recommended boarding point as the actual boarding point.

[0117] If the user feels that none of the recommended boarding points displayed by the terminal device meets his / her needs, he / she can input or set the actual boarding point. At this time, the terminal device can construct a corresponding boarding point recommendation record based on the above operation records and the recommended boarding points, and mark the boarding point recommendation record as a failed recommendation. Then, the terminal device can feed back the failed boarding point recommendation record to the server.

[0118] Correspondingly, the server can make targeted adjustments and updates to the second model every preset time period (for example, 1 week, etc.) based on the failed boarding point recommendation records collected within the preset time period to improve the accuracy of the second model, so that the recommended boarding point suitable for the user can be determined more accurately in the future.

[0119] As can be seen from the above, based on the method for determining the recommended boarding point provided in the embodiments of this specification, the server can first obtain the text data of the location names of multiple candidate boarding points, and the corresponding associated data; and determine the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points; and then comprehensively utilize the location name word vectors of the above candidate boarding points and the corresponding associated data to determine the recommended boarding point that meets the requirements from the multiple candidate boarding points and recommend it to the user. By introducing and utilizing the text data of the location names of the candidate boarding points, combined with the associated data, it is possible to integrate feature information of more dimensions, accurately predict the boarding point suitable for the user and recommend it to the user, thereby improving the success rate of the recommendation and allowing the user to obtain a better user experience.

[0120] See also Figure 4 As shown, the embodiment of this specification also provides a method for determining a recommended boarding point. The method is specifically applied to the terminal device side. When implemented specifically, the method may include the following contents:

[0121] S401: Sending a ride request to a server; wherein the ride request carries at least a user's location parameter;

[0122] S402: receiving a recommended boarding point fed back by the server; wherein the recommended boarding point is determined by the server in response to the boarding request, obtaining and using text data of the location name of the candidate boarding point and corresponding associated data;

[0123] S403: Display the recommended boarding point to the user.

[0124] Through the above embodiment, when a user has a need to take a ride, he or she can interact with the server through the above terminal device to obtain a recommended boarding point that is highly targeted, has a high reference value, and is suitable for the user, so that the user does not need to determine and enter the boarding point separately, which simplifies the user operation and improves the user experience.

[0125] The embodiments of this specification also provide a server, including a processor and a memory for storing processor executable instructions. When the processor is implemented, it can perform the following steps according to the instructions: obtain text data of location names of multiple candidate boarding points, and associated data; determine location name word vectors of multiple candidate boarding points corresponding to the text data of location names of multiple candidate boarding points; use the location name word vectors of the candidate boarding points and the corresponding associated data to determine a recommended boarding point that meets the requirements from the multiple candidate boarding points.

[0126] In order to complete the above instructions more accurately, refer to Figure 5 As shown, the embodiment of this specification also provides another specific server, wherein the server includes a network communication port 501, a processor 502 and a memory 503, and the above structures are connected through internal cables so that each structure can perform specific data interaction.

[0127] The network communication port 501 can be specifically used to obtain text data of location names of multiple candidate boarding points, as well as associated data.

[0128] The processor 502 can be specifically used to determine the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points; and use the location name word vectors of the candidate boarding points and the corresponding associated data to determine a recommended boarding point that meets the requirements from the multiple candidate boarding points.

[0129] The memory 503 may be specifically used to store corresponding instruction programs.

[0130] In this embodiment, the network communication port 501 can be a virtual port that is bound to different communication protocols so that different data can be sent or received. For example, the network communication port can be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for email data communication. In addition, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it can also be a Bluetooth chip.

[0131] In this embodiment, the processor 502 may be implemented in any appropriate manner. For example, the processor may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) executable by the (micro)processor, a logic gate, a switch, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, etc. This specification does not limit this.

[0132] In this embodiment, the memory 503 may include multiple levels. In a digital system, anything that can store binary data can be a memory; in an integrated circuit, a circuit with a storage function but no physical form is also called a memory, such as RAM, FIFO, etc.; in a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0133] The embodiments of this specification also provide a terminal device, including a processor and a memory for storing processor executable instructions. When the processor is implemented, it can perform the following steps according to the instructions: send a boarding request to a server; wherein the boarding request carries at least the user's location parameters; receive a recommended boarding point fed back by the server; wherein the recommended boarding point is determined by the server responding to the boarding request, obtaining and using text data of the location name of the candidate boarding point and corresponding associated data; and displaying the recommended boarding point to the user.

[0134] The embodiments of this specification also provide a computer storage medium based on the above-mentioned method for determining the recommended boarding point, wherein the computer storage medium stores computer program instructions, which, when executed, implement the following: obtaining text data of location names of multiple candidate boarding points, and associated data; determining location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of the multiple candidate boarding points; and determining a recommended boarding point that meets the requirements from the multiple candidate boarding points using the location name word vectors of the candidate boarding points and the corresponding associated data.

[0135] The embodiments of this specification also provide a computer storage medium based on the above-mentioned method for determining the recommended boarding point, wherein the computer storage medium stores computer program instructions, which, when executed, implement: sending a boarding request to a server; wherein the boarding request carries at least the user's location parameters; receiving a recommended boarding point fed back by the server; wherein the recommended boarding point is determined by the server responding to the boarding request, obtaining and utilizing text data of the location name of the candidate boarding point and corresponding associated data; and displaying the recommended boarding point to the user.

[0136] In this embodiment, the storage medium includes, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a cache, a hard disk (HDD), or a memory card. The memory may be used to store computer program instructions. The network communication unit may be an interface for network connection communication set in accordance with the standard specified by the communication protocol.

[0137] In this embodiment, the functions and effects specifically implemented by the program instructions stored in the computer storage medium can be explained in comparison with other implementations and will not be described in detail here.

[0138] See also Figure 6 As shown, at the software level, the embodiment of this specification also provides a device for determining a recommended boarding point, which may specifically include the following structural modules:

[0139] The acquisition module 601 may be specifically used to acquire text data of location names of multiple candidate boarding points, and associated data;

[0140] The first determination module 602 may be specifically used to determine the location name word vectors of the plurality of candidate boarding points corresponding to the text data of the location names of the plurality of candidate boarding points;

[0141] The second determination module 603 may be specifically configured to determine a recommended boarding point that meets the requirements from a plurality of candidate boarding points by using the location name word vectors of the candidate boarding points and the corresponding associated data.

[0142] In some embodiments, when the acquisition module 601 is implemented, the text data of the location names of multiple candidate boarding points can be acquired in the following manner: determine the user's current location; query the map data to filter out multiple location points adjacent to the user's current location as candidate boarding points; and acquire the text data of the location names of multiple candidate boarding points based on the map data.

[0143] In some embodiments, when the first determination module 602 is implemented, the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points can be determined in the following manner: the location name word vector of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point is determined in the following manner: the text data of the location name of the current candidate boarding point is processed by word segmentation to obtain multiple phrases; the word vectors of the multiple phrases respectively corresponding to the multiple phrases are determined using a preset word vector database; wherein the preset word vector database stores word vectors corresponding to multiple preset phrases; the preset word vector database is constructed by processing corpus text data using the first model based on the first model rule; based on the word vectors of the multiple phrases, the location name word vector of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point is determined.

[0144] In some embodiments, when the first determination module 602 is implemented, the word vector of the current phrase among multiple phrases can be determined in the following manner: a preset word vector database is queried to determine a preset phrase that matches the current phrase as a matching phrase; and the word vector of the matching phrase is determined as the word vector of the current phrase.

[0145] In some embodiments, when the first determination module 602 is implemented, after querying the preset word vector database, if no preset phrase matching the current phrase is found in the preset word vector database, the word vector of the default phrase in the preset word vector database can be used as the word vector of the current phrase.

[0146] In some embodiments, when the first determination module 602 is implemented, the word vector of the location name of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point can be determined according to the word vectors of the multiple word groups in the following manner: according to the order of the word groups in the text data of the location name of the current candidate boarding point, the word vectors of the multiple word groups are arranged to obtain the word vectors of the sorted multiple word groups; from the word vectors of the sorted multiple word groups, the word vectors of a preset number of word groups with low order are selected and combined to obtain the word vector of the location name of the current candidate boarding point.

[0147] In some embodiments, the first model may specifically include a skip-gram model; correspondingly, the first model rule may specifically include at least one of the following: a preset word frequency parameter rule, a preset context parameter rule, a preset word vector dimension parameter rule, etc.

[0148] In some embodiments, the associated data may specifically include at least one of the following: the user's current location, the distance between the user's current location and the candidate boarding point, the ride time, etc.

[0149] In some embodiments, when the second determination module 603 is implemented, the location name word vectors and corresponding associated data of the candidate boarding points can be used in the following manner to determine a recommended boarding point that meets the requirements from multiple candidate boarding points: the location name word vectors and corresponding associated data of multiple candidate boarding points are respectively combined to obtain multiple data groups; wherein one data group corresponds to one candidate boarding point; the multiple data groups are processed using the second model to obtain corresponding processing results; and based on the processing results, a recommended boarding point that meets the requirements is screened out from the multiple candidate boarding points.

[0150] In some embodiments, the device may further include a modeling module, which can be specifically used to obtain the test user's operation records for the test boarding point, as well as the text data and associated data of the location name of the test boarding point; determine the location name word vector of the test boarding point corresponding to the text data of the location name of the test boarding point; combine the location name word vector of the test boarding point with the corresponding associated data to obtain sample data; label the sample data according to the test user's operation records for the test boarding point to obtain labeled sample data; and use the labeled sample data to perform model training to obtain the second model.

[0151] See also Figure 7 As shown, the embodiment of this specification also provides another device for determining a recommended boarding point, which may specifically include the following structural modules:

[0152] The sending module 701 may be specifically used to send a ride request to the server; wherein the ride request at least carries a user's location parameter;

[0153] The receiving module 702 may be specifically used to receive a recommended boarding point fed back by the server; wherein the recommended boarding point is determined by the server responding to the ride request, obtaining and using text data of the location name of the candidate boarding point and corresponding associated data;

[0154] The display module 703 may be specifically used to display the recommended boarding point to the user.

[0155] It should be noted that the units, devices or modules described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. For the convenience of description, the above devices are described separately by functions divided into various modules. Of course, when implementing this specification, the functions of each module can be implemented in the same or more software and / or hardware, or the modules that implement the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0156] Based on the above-mentioned device for determining the recommended boarding point, the text data of the location names of multiple candidate boarding points and the corresponding associated data are first obtained; and the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points are determined; and then the location name word vectors of the candidate boarding points and the corresponding associated data are comprehensively used to determine the recommended boarding points that meet the requirements from the multiple candidate boarding points and recommend them to the user. By introducing and utilizing the text data of the location names of the candidate boarding points, combined with the associated data, it is possible to integrate feature information of more dimensions, accurately predict the boarding points suitable for the user and recommend them to the user, thereby improving the success rate of the recommendation and allowing the user to obtain a better user experience.

[0157] Although the present specification provides method operation steps as described in the embodiments or flow charts, more or less operation steps may be included based on conventional or non-creative means. The order of steps listed in the embodiments is only one way of executing the order of many steps, and does not represent a unique execution order. When the device or client product in practice is executed, it can be executed in sequence or in parallel according to the method shown in the embodiments or the drawings (for example, a parallel processor or a multi-threaded processing environment, or even a distributed data processing environment). The term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, product or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such a process, method, product or device. In the absence of more restrictions, it is not excluded that there are other identical or equivalent elements in the process, method, product or device including the elements. The first, second, etc. words are used to represent the name, and do not represent any particular order.

[0158] Those skilled in the art also know that, in addition to implementing the controller in a purely computer-readable program code, the controller can be made to implement the same function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered as a hardware component, and the devices for implementing various functions included therein can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules for implementing the method and structures within the hardware component.

[0159] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media including storage devices.

[0160] Through the description of the above embodiments, it can be known that those skilled in the art can clearly understand that the present specification can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present specification can essentially be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a mobile terminal, a server, or a network device, etc.) to execute the methods described in each embodiment of the present specification or some parts of the embodiments.

[0161] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. This specification can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.

[0162] Although the present specification is described through embodiments, those skilled in the art will appreciate that there are many modifications and changes to the present specification without departing from the spirit of the present specification, and it is intended that the appended claims include these modifications and changes without departing from the spirit of the present specification.

Claims

1. A method for determining a recommended boarding point, characterized in that: Applicable to servers, including: Acquire text data of location names of multiple candidate boarding points, and associated data; wherein the text data of location names of the candidate boarding points includes location name texts described in Chinese or other non-alphabetic languages; Determine multiple candidate boarding point location name word vectors corresponding to the text data of the location names of the multiple candidate boarding points; wherein the candidate boarding point location name vectors are obtained by filtering out the word vectors of a preset number of phrases ranked lower from the word vectors of the multiple phrases ranked after the location names of the candidate boarding points; Using the location name word vectors of the candidate boarding points and the corresponding associated data, determine a recommended boarding point that meets the requirements from multiple candidate boarding points; The location name vector of the candidate boarding point is determined according to a preset word vector database; the preset word vector database is constructed based on a first model rule by using a first model to process corpus text data to learn and train word vectors related to the location name; the first model rule includes: a preset word frequency parameter rule, a preset context parameter rule, and a preset word vector dimension parameter rule; Specifically, based on the preset word frequency parameter rule, by setting the word frequency parameter to 1, the first model is controlled to learn and train the word vectors whose word frequency is greater than or equal to 1 in the corpus text; based on the preset context parameter rule, by setting the context parameter to m, the first model is controlled to take m phrases before and after the context for learning and training, and the value of m is determined according to the actual number of phrases contained in the context in which the phrase is located; based on the preset word vector dimension parameter rule, by setting the word vector dimension parameter to a data value associated with the context parameter and the word frequency parameter, the first model is controlled to automatically determine the appropriate word vector dimension parameter for the word vector.

2. The method according to claim 1, characterized in that: Get the text data of the location names of multiple candidate boarding points, including: Determine the user's current location; Query the map data and select multiple locations close to the user's current location as candidate boarding points; According to the map data, text data of location names of multiple candidate boarding points are obtained.

3. The method according to claim 1, characterized in that Determining the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points includes: The location name word vector of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point is determined in the following manner: Perform word segmentation processing on the text data of the location name of the current candidate boarding point to obtain multiple phrases; Using a preset word vector database, word vectors of multiple word groups corresponding to the multiple word groups are determined; wherein the preset word vector database stores word vectors corresponding to the multiple preset word groups; the preset word vector database is constructed by processing corpus text data using the first model based on the first model rule; According to the word vectors of the multiple word groups, a word vector of the location name of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point is determined.

4. The method according to claim 3, characterized in that Using a preset word vector database, word vectors of multiple word groups corresponding to the multiple word groups are determined, including: The word vector of the current word group among the multiple word groups is determined in the following manner: A preset word vector database is queried to determine a preset word group that matches the current word group as a matching word group; and the word vector of the matching word group is determined as the word vector of the current word group.

5. The method according to claim 4, characterized in that After querying the preset word vector database, the method further includes: When no preset phrase matching the current phrase is found in the preset word vector database, the word vector of the default phrase in the preset word vector database is used as the word vector of the current phrase.

6. The method according to claim 3, characterized in that Determining the word vector of the location name of the current candidate boarding point corresponding to the text data of the location name of the current candidate boarding point according to the word vectors of the multiple phrases includes: Arrange the word vectors of the multiple word groups according to the order of the word groups in the text data of the location name of the current candidate boarding point, and obtain the word vectors of the multiple word groups after the ordering; The word vectors of a preset number of word vectors with a lower sorting are selected from the word vectors of the sorted multiple word groups, and the word vectors are combined to obtain the location name word vector of the current candidate boarding point.

7. The method according to claim 3, characterized in that The first model includes a skip-gram model.

8. The method according to claim 1, characterized in that The associated data includes at least one of the following: the user's current location, the distance between the user's current location and the candidate boarding point, and the boarding time.

9. The method according to claim 8, characterized in that Using the location name word vectors of the candidate boarding points and the corresponding associated data, a recommended boarding point that meets the requirements is determined from multiple candidate boarding points, including: Combining the location name word vectors of multiple candidate boarding points and the corresponding associated data respectively to obtain multiple data groups; wherein one data group corresponds to one candidate boarding point; Processing the plurality of data groups using the second model to obtain corresponding processing results; According to the processing result, a recommended boarding point that meets the requirements is screened out from a plurality of candidate boarding points.

10. The method according to claim 9, characterized in that The second model is trained in the following way: Obtain the test user's operation record for the test boarding point, as well as the text data and associated data of the location name of the test boarding point; Determine the location name word vector of the test boarding point corresponding to the text data of the location name of the test boarding point; Combine the location name word vector of the test boarding point with the corresponding associated data to obtain sample data; According to the operation records of the test user for the test boarding point, the sample data is labeled to obtain the labeled sample data; The labeled sample data is used to perform model training to obtain the second model.

11. A method for determining a recommended boarding point, characterized in that: Applied to terminal equipment, including: Sending a ride request to a server; wherein the ride request carries at least a user's location parameter; Receive the recommended boarding point fed back by the server; wherein the recommended boarding point is determined by the server responding to the boarding request, obtaining and using the text data of the location name of the candidate boarding point and the corresponding associated data; wherein the text data of the location name of the candidate boarding point includes the location name text described in Chinese or other non-alphabetic languages; the recommended boarding point is specifically determined based on the location name word vector of the candidate boarding point and the corresponding associated data; the location name vector of the candidate boarding point is obtained by filtering out the word vector combination of a preset number of phrases ranked lower from the word vectors of multiple phrases after the location names of the candidate boarding points are ranked; Displaying the recommended boarding point to the user; The location name vector of the candidate boarding point is determined according to a preset word vector database; the preset word vector database is constructed based on a first model rule by using a first model to process corpus text data to learn and train word vectors related to the location name; the first model rule includes: a preset word frequency parameter rule, a preset context parameter rule, and a preset word vector dimension parameter rule; Specifically, based on the preset word frequency parameter rule, by setting the word frequency parameter to 1, the first model is controlled to learn and train the word vectors whose word frequency is greater than or equal to 1 in the corpus text; based on the preset context parameter rule, by setting the context parameter to m, the first model is controlled to take m phrases before and after the context for learning and training, and the value of m is determined according to the actual number of phrases contained in the context in which the phrase is located; based on the preset word vector dimension parameter rule, by setting the word vector dimension parameter to a data value associated with the context parameter and the word frequency parameter, the first model is controlled to automatically determine the appropriate word vector dimension parameter for the word vector.

12. A device for determining a recommended boarding point, characterized in that: include: An acquisition module, used to acquire text data of location names of multiple candidate boarding points, and associated data; wherein the text data of the location names of the candidate boarding points includes location name texts described in Chinese or other non-alphabetic languages; A first determination module is used to determine the location name word vectors of multiple candidate boarding points corresponding to the text data of the location names of multiple candidate boarding points; wherein the location name vectors of the candidate boarding points are obtained by selecting the word vectors of a preset number of phrases ranked lower from the word vectors of the multiple phrases after the location names of the candidate boarding points are ranked; A second determination module is used to determine a recommended boarding point that meets the requirements from multiple candidate boarding points by using the position name word vectors of the candidate boarding points and the corresponding associated data; The location name vector of the candidate boarding point is determined according to a preset word vector database; the preset word vector database is constructed based on a first model rule by using a first model to process corpus text data to learn and train word vectors related to the location name; the first model rule includes: a preset word frequency parameter rule, a preset context parameter rule, and a preset word vector dimension parameter rule; Specifically, based on the preset word frequency parameter rule, by setting the word frequency parameter to 1, the first model is controlled to learn and train the word vectors whose word frequency is greater than or equal to 1 in the corpus text; based on the preset context parameter rule, by setting the context parameter to m, the first model is controlled to take m phrases before and after the context for learning and training, and the value of m is determined according to the actual number of phrases contained in the context in which the phrase is located; based on the preset word vector dimension parameter rule, by setting the word vector dimension parameter to a data value associated with the context parameter and the word frequency parameter, the first model is controlled to automatically determine the appropriate word vector dimension parameter for the word vector.

13. A server, characterized in that: The invention comprises a processor and a memory for storing instructions executable by the processor, wherein the steps of the method according to any one of claims 1 to 10 are implemented when the processor executes the instructions.

14. A terminal device, characterized in that: The invention comprises a processor and a memory for storing processor-executable instructions, wherein the processor implements the steps of the method according to claim 11 when executing the instructions.

15. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the instructions are executed by a processor, the steps of the method described in any one of claims 1 to 10 or 11 are implemented.

Images