Ticket recommendation method and system, travel server and program

Abstract

The embodiment of the invention provides a ticket business recommendation method and system, a travel server and a program. The method comprises the following steps: acquiring a ticket business search request of a user; determining search results of recommended traffic tickets and replacing multiple candidate replacement search information of the traffic tickets; aiming at any candidate alternative search information, determining a search preference score of the user on the candidate alternative search information; according to the candidate alternative search information and any candidate alternative search information, on the basis of a plurality of ticket features related to a user ticket decision, scores of feature differences, corresponding to the recommended traffic tickets, of the candidate alternative search information in all ticket features are determined, and feature difference scores of the candidate alternative search information are obtained; according to the search preference score and the feature difference score of each item of candidate alternative search information, determining the score of each item of candidate alternative search information; and determining recommended alternative search information according to the scores of the candidate alternative search information. According to the embodiment of the invention, the recommendation accuracy of alternative traffic ticket recommendation can be improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer Internet technology, and specifically to a ticket recommendation method, system, travel server and program. Background Art

[0002] Travel service platforms are online platforms that provide travel services to users, such as OTP (Online Travel Platform), OTA (Online Travel Agency) platforms, etc. With the continuous development of intelligent technology, travel service platforms not only provide ticketing services such as query, reservation, change and refund of air tickets, train tickets and other transportation tickets, but also provide diversified travel solutions such as hotel reservations, travel packages, and car rentals.

[0003] Ticket recommendation is an important part of the ticket service of the travel service platform. It is mainly used to recommend search results of recommended transportation tickets such as air tickets and train tickets that meet the ticket search conditions based on the ticket search conditions submitted by the user, such as the departure place, destination and departure date. Furthermore, in order to meet the diverse travel needs and travel preferences of users, in addition to recommending search results of recommended transportation tickets that meet the ticket search conditions, ticket recommendation will also recommend alternative transportation tickets.

[0004] Alternative transportation ticket recommendation refers to recommending alternative transportation tickets that are not completely consistent with the ticket search conditions but are related to the user, so as to provide users with more choices for travel, thereby helping users find suitable air tickets, train tickets and other transportation tickets. The accuracy of alternative transportation ticket recommendations is crucial to improving the user's travel experience and increasing the ticket click-through rate and conversion rate of travel service platforms. Therefore, how to improve the accuracy of alternative transportation ticket recommendations has become a technical problem that technicians in this field need to solve urgently. Summary of the invention

[0005] In view of this, embodiments of the present application provide a ticket recommendation method, system, travel server and program to improve the accuracy of recommendation of alternative transportation tickets.

[0006] To achieve the above objectives, the embodiments of the present application provide the following technical solutions.

[0007] In a first aspect, an embodiment of the present application provides a ticket recommendation method, comprising: Obtaining a ticket search request from a user, wherein the ticket search request includes a ticket search condition; Determining search results of recommended transportation tickets that match the ticket search criteria; Determine a plurality of candidate replacement search information of alternative transportation tickets related to the ticket search condition, each candidate replacement search information includes a ticket search condition for searching for alternative transportation tickets and is associated with a plurality of ticket features related to the user's ticket decision; For any candidate alternative search information, determine the user's search preference score for the candidate alternative search information based on the user's user characteristic information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets; and, for any item of candidate alternative search information, determining a score of a feature difference between the candidate alternative search information and the recommended transportation ticket corresponding to each ticket feature based on the plurality of ticket features, and obtaining a feature difference score of the candidate alternative search information based on the score of the feature difference between the candidate alternative search information and each ticket feature; Determining the score of each candidate alternative search information according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information; According to the scores of each candidate alternative search information, recommended alternative search information is determined from the plurality of candidate alternative search information.

[0008] In a second aspect, an embodiment of the present application provides a ticket recommendation system, including: A request acquisition module, used to acquire a user's ticket search request, wherein the ticket search request includes a ticket search condition; A ticket search engine, used to determine search results of recommended transportation tickets that meet the ticket search criteria; Alternative transportation ticketing recommendation subsystem for: Determine multiple candidate alternative search information of alternative transportation tickets related to the ticket search condition, each candidate alternative search information includes a ticket search condition for searching for alternative transportation tickets and is associated with multiple ticket features related to the user's ticketing decision; for any candidate alternative search information, determine the user's search preference score for the candidate alternative search information based on the user's user feature information, historical behavior information, the ticket search condition of the candidate alternative search information, and the search results of the recommended transportation tickets; and, for any candidate alternative search information, determine the score of the feature difference corresponding to each ticket feature of the candidate alternative search information compared with the recommended transportation ticket based on the multiple ticket features, and obtain the feature difference score of the candidate alternative search information based on the feature difference score corresponding to each ticket feature of the candidate alternative search information; determine the score of each candidate alternative search information based on the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information; and determine the recommended alternative search information from the multiple candidate alternative search information based on the score of each candidate alternative search information.

[0009] In a third aspect, an embodiment of the present application provides a travel server, comprising a memory and a processor, wherein the memory stores computer-executable instructions, and the processor calls the computer-executable instructions to execute the ticket recommendation method as described in the first aspect above.

[0010] In a fourth aspect, an embodiment of the present application provides a computer program product, including computer execution instructions, which, when executed, implement the ticket recommendation method as described in the first aspect above.

[0011] The embodiment of the present application combines the user's search preference score for candidate alternative search information and the feature difference score of the candidate alternative search information, and can comprehensively consider the user's search preference and the actual situation of the alternative transportation ticketing corresponding to the candidate alternative search information, and generate accurate scores for the candidate alternative search information. Furthermore, this dual evaluation system of the search preference score and the feature difference score for the candidate alternative search information can reduce the error caused by one-sided consideration of a single factor, more accurately recommend alternative search information of alternative transportation ticketing to users, and improve the accuracy of the recommendation of alternative transportation ticketing; thereby providing users with more choice possibilities for travel, improving users' travel experience, and thereby improving the ticket click-through rate and conversion rate of the travel service platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0013] Figure 1 This is an example page of a travel service platform.

[0014] Figure 2 A flowchart of a ticket recommendation method provided in an embodiment of the present application.

[0015] Figure 3 A flowchart for determining a search preference score provided in an embodiment of the present application.

[0016] Figure 4 A block diagram of an alternative transportation ticket recommendation subsystem provided in an embodiment of the present application.

[0017] Figure 5 A block diagram of a main network module provided in an embodiment of the present application.

[0018] Figure 6 A flow chart for determining a feature difference score provided in an embodiment of the present application.

[0019] Figure 7 A block diagram of a monotonic balanced network module provided in an embodiment of the present application.

[0020] Figure 8 A block diagram of a ticket recommendation system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0022] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0023] To facilitate understanding of alternative transportation ticket recommendations, Figure 1 An example of a page example of a travel service platform is shown in FIG. Figure 1 As shown, the search page is a page provided by the travel service platform for users to search for air tickets, train tickets and other transportation tickets; when the user selects a mode of transportation (such as airplanes, trains and other modes of transportation), the search page includes a search box for users to enter ticket search conditions such as ODD, where ODD is the abbreviation of Origin, Destination and Departure Date; specifically, the departure place is the starting city or place of the user's trip, such as the departure place corresponds to the search field of depcity, the destination is the end city or place of the user's trip, such as the destination corresponds to the search field of arrcity, and the departure date is the date of the user's trip, such as the departure date corresponds to the search field of deptime.

[0024] For ease of explanation, the departure place when the user searches for tickets is called the target departure place, the destination when the user searches for tickets is called the target destination, and the departure date when the user searches for tickets is called the target departure date.

[0025] Users can submit ticket search requests to the travel service platform through the search page. The ticket search request can indicate the user's ticket search conditions such as the target departure point, target destination and target departure date when the mode of transportation is selected; further, the travel service platform can search and recommend search results of recommended transportation tickets that meet the ticket search conditions based on the user's ticket search conditions (for example, when the mode of transportation is selected, search and recommend search results of recommended transportation tickets that meet the target departure point, target destination and target departure date); at the same time, in order to meet users' diverse travel needs and travel preferences, the travel service platform also recommends alternative transportation tickets to users through alternative transportation ticket recommendations, which are not completely consistent with the ticket search conditions but are related and meet user needs (such as price requirements, distance requirements, etc.), thereby searching and recommending search results of alternative transportation tickets through the alternative search information of alternative transportation tickets. For example, if the search results for recommended transportation tickets that meet the ticket search conditions show that there are no remaining tickets or are not preferred by the user, the alternative search information of the alternative transportation tickets recommended by the alternative transportation ticket recommendation can provide a further transportation ticket search entry for the user's travel, thereby providing the user with more options for travel, helping the user find suitable transportation tickets, thereby improving the user's travel experience and increasing the ticket click-through rate and conversion rate of the travel service platform.

[0026] Combination Figure 1 As shown, the search result page (Listing Page) is a page provided by the travel service platform for displaying search results to users after users search for tickets on the travel service platform; the search result page can display search results of recommended transportation tickets that meet the user's ticket search conditions, as well as alternative search information of alternative transportation tickets recommended by alternative transportation ticket recommendations. Specifically, the search results of recommended transportation tickets can be one or more items, such as one or more search results of recommended transportation tickets that meet the target departure point, target destination, and target departure date; wherein a search result of a recommended transportation ticket includes the departure time (deptime), arrival time (arrtime), price (price), discount (discount), other attribute information (attribute information), and other information of a transportation schedule (direct transportation schedule or transfer transportation schedule) that meets the target departure point, target destination, and target departure date.

[0027] It should be noted that transportation schedules, such as airplane flights, train schedules (such as high-speed train schedules, etc.), bus schedules, etc., depend on the specific ticketing scenario and the selected mode of transportation when the user searches for tickets, and the embodiments of the present application do not limit this; other attribute information is additional attribute information related to transportation schedules (such as flights). Taking flights as an example, other attribute information includes flight class, airline, whether it is a direct flight, etc.

[0028] The recommendation of alternative transportation tickets can be a multi-mode recommendation, which is mainly used to recommend alternative search information of alternative transportation tickets corresponding to multiple alternative recommendation modes, wherein different alternative recommendation modes indicate different alternative transportation ticket situations; for the convenience of distinction, the search information used to search for alternative transportation tickets is called alternative search information, and the alternative search information of alternative transportation tickets may at least include ticket search conditions for alternative transportation tickets; specifically, the alternative search information of alternative transportation tickets may include ticket search conditions such as the departure place, destination, departure date and selected transportation mode in the corresponding alternative recommendation mode, so that the alternative search information of alternative transportation tickets can be used to search for search results of alternative transportation tickets that are consistent with the departure place, destination, departure date and transportation mode in the corresponding alternative recommendation mode.

[0029] Example, combined Figure 1 As shown, the various alternative recommendation modes for alternative transportation ticket recommendations may include but are not limited to nearby travel modes, intermodal modes (such as an airplane-train intermodal mode that combines intermodal services of airplanes and trains), other direct transportation modes (such as a high-speed rail direct mode in an air ticket search scenario), etc.; among them, alternative transportation ticket recommendations for nearby travel modes may also be referred to as nearby ticket recommendations, which recommend alternative search information for alternative transportation tickets based on the proximity of the departure place, destination or departure date, thereby searching for corresponding search results for transportation tickets for nearby departure places, nearby destinations or nearby departure dates; further, nearby travel modes may be specifically divided into nearby OD (departure place / destination) modes and nearby date modes.

[0030] The proximity OD mode may also be referred to as the proximity location mode, which mainly recommends alternative search information of alternative transportation tickets based on the proximity of the departure place and / or the destination, and belongs to the proximity ticket recommendation in the alternative transportation ticket recommendation; specifically, the proximity OD mode corresponds to one or more alternative search information of alternative transportation tickets, and is used to recommend alternative search information of nearby departure places close to the target departure place, and / or nearby destinations close to the target destination, when the target departure date remains unchanged, so as to search for search results of alternative transportation tickets that match the target departure date, nearby departure place and / or nearby destination under the selected transportation mode.

[0031] In the nearby OD mode, the alternative search information for alternative transportation tickets may include information such as the departure place (target departure place or nearby departure place), destination (target destination or nearby destination), departure date (target departure date), and price, and the transportation mode is the transportation mode selected by the user when searching for tickets on the search page; wherein, the departure place in the alternative search information for alternative transportation tickets in the nearby OD mode is a nearby departure place, and / or the destination is a nearby destination, and the price is the price of the transportation ticket that matches the departure place, destination, and departure date in the alternative search information, such as the minimum price.

[0032] It should be noted that, based on the specific mode of transportation, the nearby departure place can be regarded as a city or place with a transportation station for airplanes or trains within the preset geographical range of the target departure place, and the number of nearby departure places may be one or more; similarly, the nearby destination can be regarded as a city or place with a transportation station for airplanes or trains within the preset geographical range of the target destination, and the number of nearby destinations may be one or more. For example, taking the air ticket scenario as an example, the city or place with an airport within a preset radius distance with the target departure place as the center is the nearby departure place, and the city or place with an airport within a preset radius distance with the target destination as the center is the nearby destination.

[0033] The approaching date mode mainly recommends alternative search information of alternative transportation tickets based on the proximity of the departure date, and belongs to the recommendation of the approaching ticket in the recommendation of alternative transportation tickets; specifically, the approaching date mode corresponds to one or more alternative search information of alternative transportation tickets, and is used to recommend alternative search information of alternative transportation tickets of the approaching departure date close to the target departure date when the target departure place and the target destination remain unchanged, so as to search for the search results of alternative transportation tickets that match the target departure place, the target destination, and the approaching departure date under the selected transportation mode. For example, the approaching departure date close to the target departure date is such as a date within a preset number of days from the target departure date, and the approaching departure date can be one or more, depending on the specific number of days of the preset number of days. In the approaching date mode, the alternative search information of the alternative transportation ticket may include information such as the departure place (target departure place), the destination (target destination), the departure date (approaching departure date), and the price, and the transportation mode is the transportation mode when the user searches for tickets on the search page; wherein, the price is the price of the transportation ticket that matches the departure place, the destination, and the departure date in the alternative search information, such as the minimum price.

[0034] The intermodal mode (such as the air-rail intermodal mode) corresponds to one or more alternative search information of alternative transportation tickets, and is used to recommend alternative search information of alternative transportation tickets that are a combination of multiple means of transportation when the target departure place, target destination, and target departure date remain unchanged, so as to search for search results of a combination of multiple means of transportation that are consistent with the target departure place, target destination, and target departure date under the transportation travel mode that is a combination of multiple means of transportation; that is, the intermodal mode is an alternative recommendation mode that combines different types of transportation (such as airplanes, trains, buses, etc.) to recommend to users a combination of tickets for multiple means of transportation that meet the travel needs of the target departure place, target destination, and target departure date. In the intermodal mode, the alternative search information of alternative transportation tickets may include the departure place (target departure place), destination (target destination), departure date (target departure date), intermodal transportation vehicle identification (intermodal transportation vehicle identification is used to identify multiple transportation vehicles that are combined, such as the identification of airplanes and high-speed trains in the air-fire intermodal mode), and price and other information, wherein the price is the price of a transportation ticket combination that combines multiple transportation vehicles that match the departure place, destination, and departure date in the alternative search information, such as the minimum price.

[0035] Other direct transportation modes (such as the high-speed rail direct mode in the air ticket search scenario) correspond to one or more alternative search information of alternative transportation tickets, which are used to recommend alternative search information of alternative transportation tickets of other direct transportation when the target departure place, target destination, and target departure date remain unchanged, so as to search for search results of alternative transportation tickets that match the target departure place, target destination, and target departure date under other direct transportation modes; for example, in the air ticket search scenario, if there is a direct high-speed rail between the target departure place and the target destination, the high-speed rail direct mode can be used to recommend the corresponding alternative search information of alternative transportation tickets to search for search results of direct high-speed rail that match the target departure place, target destination, and target departure date. In other direct transportation modes, the alternative search information of alternative transportation tickets may include the departure place (target departure place), destination (target destination), departure date (target departure date), direct transportation vehicle identification (such as the identification of the high-speed rail transportation vehicle in the high-speed rail direct mode), and price, etc., wherein the price is the price of other direct transportation tickets that match the departure place, destination, and departure date in the alternative search information, such as the minimum price.

[0036] The alternative search information of the alternative transportation ticket may include ticket search conditions of the alternative transportation ticket, which is used to search for search results of transportation tickets that match the departure place, destination, departure date, and corresponding transportation mode of the alternative search information. For example, after the user selects the alternative search information of a certain alternative transportation ticket on the search results page, the departure place, destination, departure date, and transportation mode of the alternative search information of the alternative transportation ticket can be used as further ticket search conditions, so that the travel service platform can enter the further search results page, and this further search results page can display the search results of transportation tickets that match the departure place, destination, departure date, and transportation mode of the alternative search information selected by the user, so as to provide more ticket selection possibilities for the user's travel. The above-mentioned further search results page will not only display the search results of transportation tickets that match the ticket search conditions, but also can further display the alternative search information of alternative transportation tickets corresponding to multiple alternative recommendation modes through the alternative transportation ticket recommendation again, so as to continuously provide users with further search opportunities for alternative transportation tickets on the search results page. That is, the recommendation of alternative transportation tickets is not limited to being triggered by the user performing a ticket search through a search page, or by performing a ticket search through alternative search information of alternative transportation tickets displayed on a search result page.

[0037] Furthermore, the search results of recommended transportation tickets that meet the user's ticket search criteria may be blank information. For example, the target departure point and / or target destination does not have a transportation station corresponding to the transportation mode selected by the user, so that the search result page returned to the user by the travel service platform does not contain search results that meet the user's ticket search criteria; at this time, alternative transportation ticket recommendations (especially nearby ticket recommendations in alternative transportation ticket recommendations) can recommend alternative search information of alternative transportation tickets near the departure point and / or near the destination to the user, thereby providing the user with a further search entry for transportation tickets near the departure point and / or near the destination to meet the user's travel needs.

[0038] It can be seen that the alternative search information of alternative transportation tickets is displayed as a search entry for searching alternative transportation tickets on the search results page as a recommendation of alternative transportation tickets. When the search results of recommended transportation tickets that meet the ticket search conditions displayed on the search results page show that there are no remaining tickets, or are not liked by users, or are blank information, it can provide users with further search opportunities for alternative transportation tickets, thereby helping users find suitable transportation tickets to meet their travel needs, thereby improving users' travel experience and increasing the click-through rate and conversion rate of tickets on the travel service platform. Therefore, the recommendation of alternative transportation tickets is an important part of the travel service platform's ticket recommendation, and improving the accuracy of the recommendation of alternative transportation tickets has important technical significance.

[0039] Based on this, the embodiment of the present application provides a ticket recommendation solution applied to a travel service platform to improve the accuracy of recommendations for alternative transportation tickets, thereby improving the user's travel experience and increasing the ticket click-through rate and conversion rate of the travel service platform.

[0040] As an optional implementation, Figure 2 An optional flow chart of the ticket recommendation method provided in the embodiment of the present application is shown as an example, and the ticket recommendation method can be applied to travel service platforms, such as OTP, OTA platform, etc. In an optional implementation, the travel service platform (such as OTP, OTA platform) may include a server on the service side, called a travel server (such as OTP server, OTA server), so that the ticket recommendation method provided in the embodiment of the present application can be specifically applied to the travel server; in an optional implementation, the travel service platform (such as OTP, OTA platform) can be regarded as an online service platform formed by a server cluster formed by the travel server by setting a variety of software and hardware functional architectures.

[0041] Reference Figure 2 The ticket recommendation method provided in the embodiment of the present application may include the following steps.

[0042] Step S210: Acquire a user's ticket search request, where the ticket search request includes a ticket search condition.

[0043] In an optional implementation, a ticket search request can be triggered through a search page. For example, when the user selects a mode of transportation, he or she enters and submits ticket search conditions such as the target departure point, target destination, and target departure date through the search page. The user's user-end device can then transmit the ticket search request corresponding to the ticket search conditions to the travel service platform so that the travel service platform can perform a ticket search.

[0044] In an optional implementation, the ticket search request may also be triggered by alternative search information of alternative transportation tickets. In an embodiment of the present application, the alternative search information of alternative transportation tickets includes ticket search conditions of alternative transportation tickets, which can be used to search for alternative transportation tickets. Thus, the embodiment of the present application may also support the user's ticket search request to be triggered by the alternative search information of alternative transportation tickets; for example, if the user clicks on the alternative search information of an alternative transportation ticket displayed on the search results page, the user's user terminal device may transmit the ticket search request corresponding to the alternative search information to the travel service platform.

[0045] As an optional implementation, the ticket search conditions include but are not limited to the target departure place, target destination, target departure date, etc.; further, the ticket search conditions may also indicate the mode of transportation, etc.

[0046] For example, if a ticket search request is triggered through a search page, the mode of transportation can be determined in advance. For example, before the user's user-end device enters the search page, the user has selected search topics such as air ticket search and train ticket search in advance, so that the mode of transportation such as airplane or train when the user-end device enters the search page can be determined in advance; further, the departure place entered by the user on the search page is the target departure place in the ticket search conditions, the destination entered by the user on the search page is the target destination in the ticket search conditions, and the departure date entered by the user on the search page is the target departure date in the ticket search conditions.

[0047] For example, if the ticket search request is triggered by the alternative search information of the alternative transportation ticket, the departure place contained in the alternative search information is the target departure place in the ticket search condition, the destination contained in the alternative search information is the target destination in the ticket search condition, and the departure date contained in the alternative search information is the target departure date in the ticket search condition. Furthermore, if the alternative search information of the alternative transportation ticket also contains the identification of the transportation mode, such as the identification of the intermodal transportation tool in the alternative search information corresponding to the intermodal mode, the identification of the direct transportation tool in the alternative search information corresponding to other direct transportation modes, etc., then the transportation mode corresponding to the identification of the transportation mode in the alternative search information is the transportation mode of the ticket search condition; if the identification of the transportation mode is not included in the alternative search information, then the transportation mode of the ticket search condition can be the transportation mode when the user searches for tickets on the search page, such as the transportation mode of the alternative search information corresponding to the near OD mode and the near date mode is the transportation mode when the user searches for tickets on the search page.

[0048] Step S220: Determine search results of recommended transportation tickets that meet the ticket search conditions.

[0049] After obtaining the ticket search request, the travel service platform can perform a ticket search to search and recommend the search results of recommended transportation tickets that meet the ticket search conditions. For ease of explanation, the transportation tickets that meet the ticket search conditions and are searched and recommended by the travel service platform are called recommended transportation tickets. Step S220 can be regarded as a normal ticket recommendation of the travel service platform. In an optional implementation example, the travel service platform can set a ticket search engine to search and recommend the search results of recommended transportation tickets that meet the ticket search conditions based on the ticket search conditions.

[0050] Step S230: Determine a plurality of candidate replacement search information of alternative transportation tickets related to the ticket search condition, each candidate replacement search information includes the ticket search condition for searching for alternative transportation tickets and is associated with a plurality of ticket features related to the user's ticket decision.

[0051] In an embodiment of the present application, in addition to normal ticket recommendations, the travel service platform also recommends alternative transportation tickets, so as to recommend to users alternative search information of alternative transportation tickets that are not completely consistent with the ticket search conditions but are related, so as to search for alternative transportation tickets. In an optional implementation, the recommendation of alternative transportation tickets can be divided into a recall stage and a scoring recommendation stage; in the recall stage, the embodiment of the present application can first preliminarily recall multiple candidate alternative search information of alternative transportation tickets that are not completely consistent with the ticket search conditions but are related to the ticket search conditions. For ease of explanation, the alternative search information of alternative transportation ticket candidates is referred to as candidate alternative search information; in the scoring recommendation stage, the embodiment of the present application can score each candidate alternative search information, and then determine the final recommended alternative search information based on the score of each candidate alternative search information.

[0052] Step S230 may be regarded as a recall phase of the recommendation of alternative transportation tickets, which is used to determine a plurality of candidate alternative search information of alternative transportation tickets that are not completely consistent with the ticket search condition but are related to the ticket search condition.

[0053] In an optional implementation, the recommendation of alternative transportation tickets may have a variety of alternative recommendation modes, including but not limited to a nearby location mode (i.e., a nearby OD mode), a nearby date mode, an intermodal mode (such as an airplane-train intermodal mode), and other direct transportation modes (such as a high-speed rail direct mode), etc. The embodiment of the present application may determine the candidate alternative search information of the alternative transportation tickets corresponding to various alternative recommendation modes, thereby obtaining multiple candidate alternative search information of the alternative transportation tickets, wherein each alternative recommendation mode corresponds to at least one candidate alternative search information, depending on the specific circumstances and definition of the alternative recommendation mode.

[0054] That is to say, the candidate alternative search information of alternative transportation tickets corresponding to each alternative recommendation mode is determined based on different logics and rules. Specifically, by combining the ticket search conditions with each alternative recommendation mode, the candidate alternative search information of alternative transportation tickets related to the ticket search conditions under each alternative recommendation mode is determined to recall multiple candidate alternative search information of alternative transportation tickets.

[0055] For example, taking the nearby location mode as an example, the embodiments of the present application can set a preset geographical range for the target departure place and the target destination, so as to screen out nearby departure places and nearby destinations where there are transportation stations (transportation stations corresponding to the transportation mode of the ticket search condition) through the set preset geographical range; thereby, based on the unchanged target departure date, when the departure place is switched to a nearby departure place and / or the destination is switched to a nearby destination, at least one candidate alternative search information of an alternative transportation ticket corresponding to the nearby location mode is determined.

[0056] For example, taking the approaching date mode as an example, the embodiment of the present application can set a date range approaching the departure date, such as a date within a preset number of days from the target departure date as the approaching departure date, so that the transportation mode, target departure place and target destination based on the ticket search conditions remain unchanged, and when the departure date is switched to each approaching departure date, at least one candidate alternative search information of the alternative transportation ticket corresponding to the approaching date mode is determined.

[0057] For example, taking the intermodal mode as an example, the embodiments of the present application can set multiple means of transportation that need to be used in combination, so as to determine at least one candidate alternative search information of alternative transportation tickets corresponding to the intermodal mode based on the target departure place, target destination and target departure date remaining unchanged, when the transportation mode is adjusted to a combination of multiple means of transportation in the intermodal mode.

[0058] For example, taking other direct transportation modes as an example, the embodiments of the present application can set other direct transportation modes, so as to determine at least one candidate alternative search information of alternative transportation tickets corresponding to other direct transportation modes based on the unchanged target departure place, target destination and target departure date, when the transportation mode is adjusted to other direct transportation modes.

[0059] After determining multiple candidate alternative search information of alternative transportation tickets related to the ticket search conditions, each candidate alternative search information may include ticket search conditions for searching for alternative transportation tickets, such as departure place, destination, departure date and other ticket search conditions for searching for alternative transportation tickets; further, the ticket search conditions of the candidate alternative search information may also include transportation modes corresponding to the alternative transportation tickets, such as the identification of the transportation vehicle, etc.

[0060] In the embodiment of the present application, the candidate alternative search information of the alternative transportation ticket includes not only the ticket search conditions of the alternative transportation ticket, which is used to search for the alternative transportation ticket, but also multiple ticket features related to the user's ticket decision. Among them, the ticket features related to the user's ticket decision can be understood as the ticket feature information that affects the user's decision when the user selects the transportation ticket, which is used to further score each candidate alternative search information in the scoring recommendation stage after recalling multiple candidate alternative search information.

[0061] In an optional implementation, the multiple ticket features associated with the candidate alternative search information of the alternative transportation ticket and relevant to the user's ticketing decision may include but are not limited to: The price feature indicates the price of the alternative transportation ticket corresponding to the candidate alternative search information, so as to be used for the subsequent analysis of the price difference between the alternative transportation ticket and the recommended transportation ticket; price is one of the decision factors for users when selecting transportation tickets, and the price and the user's tendency to select transportation tickets may be negatively correlated, for example, the lower the price of the transportation ticket, the higher the user's tendency to select the transportation ticket; specifically, compared with the recommended transportation ticket, the alternative search information is used as a search entry for alternative transportation tickets. If the price of the alternative transportation ticket is lower than the recommended transportation ticket, the user's tendency to select the alternative transportation ticket is higher, so that each alternative search information may be associated with a price feature to recommend alternative transportation tickets with a lower price than the recommended transportation ticket as much as possible; in an optional implementation, the price feature associated with the candidate alternative search information may be determined based on the ticket search conditions contained in the candidate alternative search information, such as based on the ticket search conditions such as the departure place, destination, departure date and transportation mode of the candidate alternative search information, the price of the matching transportation ticket (such as the minimum price) is determined as the price feature associated with the candidate alternative search information; The distance feature represents the distance between the departure place and the target departure place of the alternative transportation ticket corresponding to the candidate alternative search information, and the distance between the destination and the target destination of the alternative transportation ticket corresponding to the candidate alternative search information, such as the distance between the departure place and the target departure place of the alternative transportation ticket, and the comprehensive distance between the destination and the target destination of the alternative transportation ticket; the distance feature associated with the candidate alternative search information is used for subsequent analysis of the distance difference with the distance feature of the recommended transportation ticket, wherein the distance feature of the recommended transportation ticket is represented by the distance between the departure place and the target departure place of the recommended transportation ticket, and the distance between the destination and the target destination, such as the comprehensive distance; the distance feature is one of the decision factors for users when selecting transportation tickets. Specifically, when the departure place of the alternative transportation ticket is close to the departure place and / or the destination is close to the destination, the distance feature can be used to calculate the distance difference with the distance feature of the recommended transportation ticket, thereby helping to determine the alternative. When the departure place of the alternative transportation ticket is a nearby departure place and / or the destination is a nearby destination, whether it can provide the user with convenient travel options, for example, the alternative transportation ticket with a smaller distance difference can provide the user with convenient travel options, and thus has a higher tendency to be selected by the user; for example, in the nearby location mode, the user will expect the recommended nearby departure place to be closer to the departure place of the recommended transportation ticket (such as the target departure place), and the recommended nearby destination to be closer to the destination of the recommended transportation ticket (such as the target destination); in an optional implementation, the distance feature associated with the candidate alternative search information can be determined based on the ticket search conditions contained in the candidate alternative search information, for example, based on the ticket search conditions such as the departure place and destination of the candidate alternative search information, the distance between the departure place of the candidate alternative search information and the target departure place, and the distance between the destination of the candidate alternative search information and the target destination are determined as the distance feature associated with the candidate alternative search information; The time feature represents the departure time of the alternative transportation ticket corresponding to the candidate alternative search information, such as the departure date and departure time of the alternative transportation ticket, so as to be used for subsequent analysis of the time difference between the departure time of the alternative transportation ticket and the departure time of the recommended transportation ticket. The departure time of the recommended transportation ticket is such as the departure date and departure time of the recommended transportation ticket. The time feature is one of the decision-making factors for users when selecting transportation tickets. Specifically, in the approaching date mode, the departure time of the alternative transportation ticket can be used to calculate the time difference with the departure time of the recommended transportation ticket, thereby helping the travel service platform to give priority to recommending alternative transportation tickets with a smaller time difference. In an optional implementation, the time feature associated with the candidate alternative search information can be determined based on the ticket search conditions contained in the candidate alternative search information, such as based on the ticket search conditions such as the departure place, destination and departure date of the candidate alternative search information, the departure time (such as the earliest departure time) of the matching transportation ticket is determined as the time feature associated with the candidate alternative search information.

[0062] It should be noted that the ticket characteristics associated with the candidate alternative search information may be related to but not exactly the same as the ticket search conditions contained in the candidate alternative search information; specifically, the ticket search conditions contained in the candidate alternative search information are mainly used to search for alternative transportation tickets, and the ticket characteristics associated with the candidate alternative search information are optimization and adjustment factors used by the embodiment of the present application to determine the final recommended alternative search information, which affect the priority of the final recommendation; that is, whether a certain candidate alternative search information will be finally recommended needs to consider ticket characteristics such as price characteristics (used to determine price difference), distance characteristics (used to determine distance difference) and time characteristics (used to determine time difference).

[0063] Step S240: for any candidate alternative search information, determine the user's search preference score for the candidate alternative search information based on the user's user characteristic information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets.

[0064] After completing step S230, the embodiment of the present application can recall multiple candidate alternative search information, thereby entering the scoring recommendation stage from step S240 to step S270, and finally recommending alternative search information by determining the score of each candidate alternative search information to improve the recommendation accuracy of alternative transportation ticket recommendations.

[0065] In an embodiment of the present application, the score of the candidate alternative search information is mainly composed of two parts, namely, the user's search preference score for the candidate alternative search information, and the feature difference score of the candidate alternative search information; wherein, the user's search preference score for the candidate alternative search information can be regarded as the user's preference score for selecting the candidate alternative search information for ticket search, such as the user's preference score for selecting the departure place, destination and departure date of the candidate alternative search information for ticket search; at the level of multiple ticket features related to the user's ticketing decision, the feature difference score of the candidate alternative search information can be regarded as a combination of the feature difference scores of the candidate alternative search information and the recommended transportation ticket in each ticket feature; thus, the feature difference score of the candidate alternative search information is integrated with the user's search preference score for the candidate alternative search information to obtain the score of the candidate alternative search information, which is used to determine the recommended alternative search information from multiple candidate alternative search information, wherein the recommended alternative search information is displayed on the search results page together with the search results of the recommended transportation ticket.

[0066] Specifically, the user's search preference score for candidate alternative search information can reflect the user's preference for candidate alternative search information when searching for tickets, that is, the user's preference for the ticket search conditions (such as departure place, destination, and departure date) contained in the candidate alternative search information, thereby quantifying the search preference for candidate alternative search information. The travel service platform can give priority to recommending candidate alternative search information with higher user search preference; at the same time, the characteristic difference score of the candidate alternative search information takes into account the differences in ticket characteristics between the candidate alternative search information and the recommended transportation tickets, such as differences in price, distance, and time. The above differences can effectively measure the degree of deviation between the alternative transportation tickets corresponding to the candidate alternative search information and the recommended transportation tickets in multiple dimensions of ticket characteristics, thereby helping the travel service platform to screen and recommend alternative transportation tickets that are closest to user needs, thereby improving the accuracy of alternative transportation ticket recommendations. That is to say, the embodiment of the present application combines the user's search preference score for candidate alternative search information and the feature difference score of the candidate alternative search information, and can comprehensively consider the user's search preference and the actual situation of the alternative transportation ticketing corresponding to the candidate alternative search information, and generate accurate scores for the candidate alternative search information. Furthermore, this dual evaluation system of the search preference score and the feature difference score for the candidate alternative search information can reduce the errors caused by one-sided consideration of a single factor, and more accurately match user preferences and recommend alternative search information that is close to user needs.

[0067] In the embodiment of the present application, step S240 is used to determine the user's search preference score for the candidate alternative search information in the scoring recommendation stage of the alternative transportation ticket recommendation; specifically, for any candidate alternative search information, the embodiment of the present application can determine the user's search preference score for the candidate alternative search information based on the user's user feature information, historical behavior information, ticket search conditions included in the candidate alternative search information, and search results of the recommended transportation ticket; wherein the input information on which the determination of the user's search preference score for the candidate alternative search information depends can include: User characteristic information of the user, which may include multiple items of basic user information of the user, and may further include multiple items of personalized information of the user; for example, multiple items of basic user information of the user may include the user's age, gender, occupation, location, etc., and multiple items of personalized information of the user may include the user's travel frequency, budget preference, etc.; the user characteristic information of the user may help the travel service platform understand the user's basic needs and habits, and then make personalized adjustments to the recommended results of alternative transportation tickets; The user's historical behavior information may be multiple pieces of historical behavior information, for example, multiple pieces of historical behavior information represent multiple behavior records of the user on the travel service platform, including but not limited to multiple behavior records related to transportation tickets, such as historical searches for transportation tickets, historical purchases of transportation tickets, historical browsing of transportation tickets, and historical changes and refunds of transportation tickets on the travel service platform; each piece of historical behavior information (such as each behavior record) may include multiple historical behavior features that record the historical behavior, such as the time when the historical behavior occurred, the type of behavior, and the details of the transportation tickets involved; the user's historical behavior information can help the travel service platform analyze the user's preferences and trends for transportation tickets; The ticket search conditions of the candidate alternative search information may be multiple, such as the departure place, destination, departure date, etc. used to search for alternative transportation tickets in the candidate alternative search information; The search results of recommended transportation tickets are the search results of recommended transportation tickets that meet the ticket search conditions and have been recommended by the travel platform service based on the user's ticket search request; recommended transportation tickets are transportation tickets selected by the travel service platform according to the recommendation strategy to meet the user's ticket search needs. The recommended transportation tickets will not only consider the user's current ticket search conditions, but also further optimize the recommendation results based on the user's historical behavior and preferences; in this embodiment of the application, the search results of recommended transportation tickets and the recommended alternative search information can be displayed together on the search results page, so that the search results of recommended transportation tickets can be used as context information of the alternative search information to be recommended, so as to determine the user's search preference score for the candidate alternative search information; as the context information of the alternative search information to be recommended, the search results of recommended transportation tickets can be one or more, and each search result includes information such as the departure place, destination, and other attribute information.

[0068] It should be noted that the information used to determine the search preference score may not include ticket features related to the user's ticketing decision, such as price features, distance features, time features, and other ticket features; for example, the user feature information used to determine the search preference score does not include price features, distance features, time features, and other ticket features; the user's historical behavior information used to determine the search preference score does not include price features, distance features, time features, and other ticket features; the ticket search conditions of the candidate alternative search information used to determine the search preference score do not include price features, distance features, time features, and other ticket features; the search results of the recommended transportation tickets used to determine the search preference score do not include price features, distance features, time features, and other ticket features. It should be further noted that the embodiment of the present application uses price features, distance features, time features, and other ticket features related to the user's ticketing decision when determining the feature difference score of the candidate alternative search information, and uses user feature information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets other than price features, distance features, time features, and other ticket features when determining the search preference score.

[0069] As an example of an optional implementation, Figure 3 An optional flow chart for determining a search preference score provided in an embodiment of the present application is shown as an example. Figure 3 The optional process of determining the search preference score in the embodiment of the present application may include the following steps.

[0070] Step S310: embed each item of user feature information of the user into corresponding each item of user feature vector, and concatenate each item of user feature vector into a user feature vector combination.

[0071] The process of embedding is to map discrete, high-dimensional features into a low-dimensional vector space; specifically, the user feature information of a user is multi-dimensional, including multiple discrete information, such as the user's gender, age, personalized preferences and other user feature information, wherein each user feature information can be converted into a low-dimensional vector through embedding processing to obtain each user feature vector corresponding to each user feature information; that is, each dimension of the user's user feature information (i.e., each user feature information of the user) will be converted into a low-dimensional vector to obtain the corresponding user feature vector. Furthermore, the embodiment of the present application can splice each user feature vector into a user feature vector combination, that is, the user feature vectors are spliced ​​and integrated into a comprehensive vector representation, which is convenient for subsequent processing and analysis.

[0072] In an optional implementation, the various user feature information of the user processed in step S310 may not include price features, distance features, time features, and other ticket features related to the user's ticketing decision.

[0073] In an optional implementation, Figure 4 The block diagram of the alternative transportation ticket recommendation subsystem provided by the embodiment of the present application is shown as an example. Figure 4 As shown, the alternative transportation ticket recommendation subsystem 400 is used as the alternative transportation ticket recommendation part of the ticket recommendation system of the travel service platform. The alternative transportation ticket recommendation subsystem 400 may include: A recall module 410, configured to determine a plurality of candidate alternative search information of alternative transportation tickets related to the ticket search condition of the ticket search request; The main network module 420 is used to implement step S240, that is, the main network module 420 can determine the user's search preference score for any candidate alternative search information based on the user's user feature information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets.

[0074] As an optional implementation, Figure 5 The optional block diagram of the main network module provided in the embodiment of the present application is shown in FIG. Figure 4 and Figure 5 As shown, the main network module 420 may include: The main network data input module 510 is used to obtain the user's user feature information, historical behavior information, ticket search conditions (such as departure place, destination, departure date) of various candidate alternative search information, and search results of recommended transportation tickets; in an optional implementation, the information obtained by the main network data input module 510 may not include price features, distance features, time features and other ticket features related to the user's ticketing decision; The basic network module 520 is used to determine the user's search preference score for each candidate alternative search information based on the user feature information, historical behavior information, ticket search conditions of each candidate alternative search information, and search results of recommended transportation tickets obtained by the main network data input module 510.

[0075] Specifically, the basic network module 520 may include a user feature processing module 521 for implementing step S310, that is, the user feature processing module 521 may be used to embed and process various user feature information of the user (excluding ticket features such as price features, distance features, time features, etc.) into corresponding user feature vectors, and splice the various user feature vectors into a user feature vector combination.

[0076] Example, combined Figure 5As shown, the user feature processing module 521 may include multiple embedding processing units and splicing units; in the user feature processing module 521, each embedding processing unit is used to embed and process each user feature information of the user into a corresponding user feature vector to obtain each user feature vector; in the user feature processing module 521, the splicing unit is used to splice each user feature vector into a user feature vector combination.

[0077] Step S320: For any piece of historical behavior information of the user, embed each historical behavior feature in the historical behavior information into corresponding historical behavior feature vectors, and concatenate each historical behavior feature vector into a historical behavior feature vector combination to obtain a historical behavior feature vector combination corresponding to the historical behavior information.

[0078] As an optional implementation, each piece of historical behavior information of the user can be processed by embedding and splicing to form a corresponding historical behavior feature vector combination, so that each piece of historical behavior information can correspond to a historical behavior feature vector combination. Specifically, for any piece of historical behavior information of the user, the embodiment of the present application can embed the various historical behavior features in the historical behavior information to obtain the corresponding historical behavior feature vectors, and then splice the various historical behavior feature vectors together to form a historical behavior feature vector combination corresponding to the historical behavior information. By forming a corresponding historical behavior feature vector combination for each piece of historical behavior information of the user, the user can be accurately profiled by analyzing each piece of historical behavior information.

[0079] In an optional implementation, combined with Figure 5 As shown, the basic network module 520 may include a historical behavior processing module 522 for implementing step S320, that is, the historical behavior processing module 522 can embed and process various historical behavior features in the historical behavior information into corresponding historical behavior feature vectors for any historical behavior information of the user, and splice the various historical behavior feature vectors into a historical behavior feature vector combination to obtain a historical behavior feature vector combination corresponding to the historical behavior information.

[0080] Example, combined Figure 5As shown, the historical behavior processing module 522 may include multiple embedding processing unit groups and multiple splicing units, wherein one embedding processing unit group may include multiple embedding processing units for embedding a piece of historical behavior information, and one splicing unit for splicing a historical behavior feature vector after embedding a piece of historical behavior information; specifically, in the historical behavior processing module 522, one embedding processing unit group is used to embed and process each historical behavior feature in a piece of historical behavior information of a user into corresponding historical behavior feature vectors; in the historical behavior processing module 522, one splicing unit is used to splice each historical behavior feature vector after embedding a piece of historical behavior information into a corresponding combination of historical behavior feature vectors.

[0081] Step S330: Based on the combination of historical behavior feature vectors corresponding to each piece of historical behavior information, the potential relationship of the historical behavior information is extracted through the self-attention mechanism and converted into the user's historical behavior representation.

[0082] As an optional implementation, after obtaining the historical behavior feature vector combination corresponding to each piece of historical behavior information, the embodiment of the present application can input the historical behavior feature vector combination corresponding to each piece of historical behavior information into the Transformer Layer for processing, thereby using the self-attention mechanism of the Transformer Layer to extract the potential relationship of the user's historical behavior information based on the historical behavior feature vector combination corresponding to each piece of historical behavior information, and convert the potential relationship of the historical behavior information into the user's historical behavior representation for subsequent recommendation or analysis.

[0083] It should be noted that Transformer Layer is a neural network layer based on the self-attention mechanism, which is used to process sequence data, especially in natural language processing and time series analysis. The core idea of ​​Transformer is to capture the relationship between each element in the sequence through the self-attention mechanism, thereby enhancing the representation ability of the model. In the optional implementation, Transformer Layer can include a multi-head attention layer and a feedforward neural network layer; specifically, after the historical behavior feature vector combination corresponding to each piece of historical behavior information is input into Transformer Layer, Transformer Layer calculates the relationship between the historical behavior feature vector combination corresponding to different historical behavior information through the self-attention mechanism to understand the historical behavior information that affects the user's preferences and decisions; and Transformer Layer can simultaneously focus on different parts of the historical behavior feature vector combination corresponding to multiple pieces of historical behavior information according to multi-head attention, thereby obtaining a rich context representation to identify the pattern and change of the user's historical behavior; then, Transformer Layer outputs a comprehensive vector that combines multiple historical behavior feature vector combinations, that is, the user's historical behavior representation.

[0084] In an optional implementation, combined with Figure 5 As shown, the basic network module 520 may include a transformer layer 523 for implementing step S330, that is, the transformer layer 523 may extract the potential relationship of the historical behavior information through a self-attention mechanism based on the combination of historical behavior feature vectors corresponding to each piece of historical behavior information, and convert it into a historical behavior representation of the user.

[0085] Step S340: embed each ticket search condition of the candidate replacement search information into corresponding search condition vectors, and concatenate each search condition vector into a corresponding search condition vector combination.

[0086] As an optional implementation, the ticket search conditions of the candidate alternative search information processed by step S340 do not include ticket features related to the user's ticketing decision, such as price features, distance features, and time features. For example, the ticket search conditions of the candidate alternative search information processed by step S340 may include: the departure place, destination and departure date of the candidate alternative search information, that is, the ODD of the candidate alternative search information; the various ticket search conditions of the above-mentioned candidate alternative search information, as discrete features, can be mapped to the vector space through embedding processing, so that each ticket search condition of the candidate alternative search information is converted into a corresponding search condition vector, that is, a vector representation of the ticket search condition. After each ticket search condition is converted into a corresponding search condition vector, each search condition vector can be spliced ​​into a long vector, that is, a search condition vector combination, representing the comprehensive feature vector of the ticket search conditions of the candidate alternative search information.

[0087] In an optional implementation, combined with Figure 5 As shown, the basic network module 520 may include a search condition processing module 524 for implementing step S340, that is, the search condition processing module 524 may embed and process each ticket search condition of the candidate replacement search information into corresponding search condition vectors, and splice each search condition vector into a corresponding search condition vector combination.

[0088] Example, combined Figure 5 As shown, the search condition processing module 524 may include multiple embedding processing units and splicing units; in the search condition processing module 524, each embedding processing unit is used to embed and process each ticket search condition of the candidate alternative search information into a corresponding search condition vector to obtain each search condition vector; in the search condition processing module 524, the splicing unit is used to splice each search condition vector into a corresponding search condition vector combination.

[0089] Step S350: embed each search result of the recommended transportation ticket into a corresponding search result vector, and concatenate each search result vector into a corresponding search result vector combination.

[0090] In the embodiment of the present application, the search results of the recommended transportation ticketing are used as the context information of the alternative search information to be recommended, and participate in the search preference score confirmation of the candidate alternative search information. As an optional implementation, the search results of the recommended transportation ticketing processed by step S350 do not include ticket features related to the user's ticketing decision, such as price features, distance features, and time features. In the embodiment of the present application, the search results of the recommended transportation ticketing, except for ticket features such as price features, distance features, and time features, can be embedded into corresponding search result vectors, and then the search result vectors of each item can be spliced ​​into a corresponding search result vector combination.

[0091] In an optional implementation, combined with Figure 5 As shown, the basic network module 520 may include a search result processing module 525 for implementing step S350, that is, the search result processing module 525 may embed and process various search results of recommended transportation tickets into corresponding search result vectors, and splice various search result vectors into corresponding search result vector combinations.

[0092] Example, combined Figure 5 As shown, the search result processing module 525 may include multiple embedding processing units and splicing units; in the search result processing module 525, each embedding processing unit is used to embed and process each search result of the recommended transportation ticketing into a corresponding search result vector; in the search result processing module 525, the splicing unit is used to splice each search result vector into a corresponding search result vector combination.

[0093] Step S360: Based on the user's historical behavior representation and the search condition vector combination, a comprehensive representation vector of the ticket search condition that integrates the user's historical behavior and the candidate alternative search information is determined through a self-attention mechanism.

[0094] As an optional implementation, after obtaining the user's historical behavior representation through step S330 and obtaining the search condition vector combination through step S340, the embodiment of the present application can input the user's historical behavior representation and the search condition vector combination into the Attention Layer for processing, thereby extracting the potential relationship between the user's historical behavior and the ticket search conditions of the candidate alternative search information through the self-attention mechanism, and obtaining a comprehensive representation vector of the ticket search conditions that integrates the user's historical behavior and the candidate alternative search information.

[0095] It should be noted that the Attention Layer calculates the correlation or dependency between different inputs through the self-attention mechanism, and dynamically assigns weights to each input to indicate the importance of the input in the current context. Specifically, the user's historical behavior representation can reflect the user's historical behavior and can capture the user's interests and preferences; the search condition vector combination can represent the conditional features related to the alternative transportation ticket search of the candidate alternative search information (such as departure place, destination, departure date, etc.), so that in the Attention Layer, the Attention Layer can analyze the correlation between the historical behavior representation and the search condition vector combination, determine the user behavior that has a strong influence on the ticket search conditions of the candidate alternative search information, and assign a higher weight; further, after the user's historical behavior representation and the search condition vector combination are processed by the Attention Layer, a new weighted representation will be obtained, which integrates the relationship between the user's historical behavior and the ticket search conditions of the candidate alternative search information, that is, the comprehensive representation vector.

[0096] In an optional implementation, combined with Figure 5 As shown, the basic network module 520 may include an attention layer 526 for implementing step S360, that is, the attention layer 526 may determine a comprehensive representation vector of the ticket search condition that integrates the user's historical behavior and the candidate alternative search information through a self-attention mechanism based on the user's historical behavior representation and the search condition vector combination.

[0097] Step S370: merge the user feature vector combination, the comprehensive representation vector, the search condition vector combination, and the search result vector combination into a merged vector; and obtain the user's search preference score for the candidate alternative search information by multi-layer feature mapping and non-linear processing of the merged vector.

[0098] After obtaining the user feature vector combination through step S310, obtaining the search condition vector combination through step S340, obtaining the search result vector combination through step S350, and obtaining the comprehensive representation vector of the ticket search conditions that integrates the user's historical behavior and the candidate alternative search information through step S360, the embodiment of the present application can merge the user feature vector combination, the comprehensive representation vector, the search condition vector combination, and the search result vector combination into a merged vector.

[0099] In an optional implementation, the embodiment of the present application may input the merged vector into multiple fully connected (FC) layers and non-linear activation functions (such as ReLU activation function or other non-linear activation functions) for processing, thereby transforming and enhancing the merged vector by performing multi-layer feature mapping and non-linear processing on the merged vector, and then outputting the user's search preference score for the candidate alternative search information.

[0100] In an optional implementation, combined with Figure 5 As shown, the basic network module 520 may include a merging module 527 for implementing step S370, and a combination 528 of multiple fully connected layers and non-linear activation functions; wherein the merging module 527 is used to merge the user feature vector combination, the comprehensive representation vector, the search condition vector combination, and the search result vector combination into a merged vector; and the combination 528 of multiple fully connected layers and non-linear activation functions is used to obtain the user's search preference score for the candidate alternative search information through multi-layer feature mapping and non-linear processing of the merged vector.

[0101] It should be noted that the fully connected layer can map the input to a new feature space through a linear transformation, thereby integrating different inputs and capturing the relationship and interaction between the inputs; the embodiment of the present application can set up multiple fully connected layers, and connect a nonlinear activation function after each fully connected layer. Nonlinear activation functions, such as ReLU (Rectified Linear Unit) activation function, introduce nonlinear characteristics through nonlinear activation functions, so that more complex patterns and relationships can be learned, rather than just performing linear transformations. For example, the ReLU activation function can change each negative value to zero, while the positive number remains unchanged, thereby helping to fit the relationship between complex inputs and outputs.

[0102] In an optional implementation, the embodiment of the present application adopts a combination of multiple fully connected layers and nonlinear activation functions, such as a stack of multiple fully connected layers and nonlinear activation functions, to extract higher-level feature representations layer by layer; specifically, after the merged vector is feature mapped by the first fully connected layer, it is nonlinearly processed by the nonlinear activation function after the first fully connected layer; thereafter, it is feature mapped by the second fully connected layer, and then nonlinearly processed by the nonlinear activation function after the second fully connected layer, and so on, until the combination of multiple fully connected layers and nonlinear activation functions is completed, and the user's search preference score for candidate alternative search information is output. In other words, a nonlinear activation function (such as a ReLU activation function) is connected after each fully connected layer to perform a nonlinear transformation on the output of the fully connected layer, and then each layer of the fully connected layer is connected to a stack of nonlinear activation functions, so that each layer can learn more abstract features based on the previous layer, and then through multiple layers of feature mapping and nonlinear processing, the representation ability of the merged vector is gradually improved, and finally the user's search preference score for candidate alternative search information is output.

[0103] Each candidate alternative search information is passed through Figure 3By processing the process shown, the search preference scores of each candidate alternative search information can be obtained, wherein the user's search preference score for the candidate alternative search information can be regarded as a score output by the model of the alternative transportation ticket recommendation subsystem of the travel service platform, expressed as Logit1; wherein Logit is a term in machine learning, especially used in logistic regression models or binary classification tasks, representing a predicted value. In the embodiment of the present application, the higher the Logit1 score of the candidate alternative search information, the higher the user's preference for the candidate alternative search information.

[0104] Back to Figure 2 As shown, in step S250, for any candidate alternative search information, based on multiple ticket features related to the user's ticketing decision, the score of the feature difference between the candidate alternative search information and the recommended transportation ticket corresponding to each ticket feature is determined, and based on the score of the feature difference between the candidate alternative search information and each ticket feature, the feature difference score of the candidate alternative search information is obtained.

[0105] In an embodiment of the present application, step S250 is used to determine the feature difference score of the candidate alternative search information in the scoring and recommendation stage of the alternative transportation ticket recommendation. Specifically, for any candidate alternative search information, the embodiment of the present application can determine the difference between the candidate alternative search information and the recommended transportation ticket in each ticket feature (such as price feature, distance feature, time feature), and thus based on the difference in each ticket feature, obtain the feature difference score of the candidate alternative search information compared to the recommended transportation ticket in each ticket feature, referred to as the feature difference score of the candidate alternative search information in each ticket feature, and then synthesize the feature difference scores of the candidate alternative search information in each ticket feature to obtain the feature difference score of the candidate alternative search information; the feature difference score of the candidate alternative search information can reflect the difference between the candidate alternative search information and the recommended transportation ticket in multiple ticket features, thereby improving the accuracy of recommending the alternative search information based on the feature difference score of the candidate alternative search information.

[0106] As an example of an optional implementation, Figure 6 An optional flow chart for determining a feature difference score provided in an embodiment of the present application is shown as an example, referring to Figure 6 , the optional process of determining the feature difference score in the embodiment of the present application may include the following steps.

[0107] Step S610: Pooling multiple ticket features associated with each search result of recommended transportation tickets to obtain multiple pooled ticket features of recommended transportation tickets.

[0108] The search results for recommended transportation tickets serve as context information for the alternative search information to be recommended. Each search result for recommended transportation tickets may be respectively associated with multiple ticketing features related to the user's ticketing decision. For example, each search result for recommended transportation tickets may be respectively associated with price features, distance features, time features, etc. The relevant concepts of price features, distance features, and time features can be referred to the previous description and will not be expanded here.

[0109] The embodiment of the present application can perform a pooling process on multiple ticket features respectively associated with each search result of the recommended transportation ticket, thereby determining the pooled ticket features that meet the pooling requirements in each feature dimension, and obtaining multiple pooled ticket features of the recommended transportation ticket, that is, the multiple pooled ticket features of the recommended transportation ticket can include the ticket features of the recommended transportation ticket that meet the pooling requirements in each feature dimension, which are used to compare with the ticket features of the candidate alternative search information in each feature dimension. It should be noted that the pooling process is a feature dimensionality reduction operation, which is used to aggregate and reduce the dimension of a group of features in any feature dimension. For example, the minimum pooling process can select the feature with the smallest feature value from a group of features in any feature dimension.

[0110] That is to say, the ticket features related to the user's ticketing decision have multiple feature dimensions, such as the price dimension of the price feature, the distance dimension of the distance feature, the time dimension of the time feature, etc., and the multiple ticket features associated with the search results of the recommended transportation tickets are pooled. Then, the pooled ticket features whose feature values ​​meet the pooling requirements can be selected from each feature dimension to obtain the pooled ticket features of the recommended transportation tickets in each feature dimension, such as the pooled price features of the recommended transportation tickets in the price dimension, the pooled distance features in the distance dimension, and the pooled time features in the time dimension, thereby forming multiple pooled ticket features for the recommended transportation tickets.

[0111] For example, taking the minimum pooling operation as an example, multiple ticket features associated with each search result of recommended transportation tickets are subjected to minimum pooling processing, and the pooled ticket features with the smallest feature value can be selected from each feature dimension, such as the pooled price feature with the smallest price, the pooled distance feature with the shortest distance, and the pooled time feature with the smallest time associated with each search result of recommended transportation tickets, to form multiple pooled ticket features for recommended transportation tickets.

[0112] In an optional implementation, combined with Figure 4As shown, the alternative transportation ticket recommendation subsystem 400 may include a monotonic balance network module 430, which is used to implement step S250, that is, the monotonic balance network module 430 can determine, for any candidate alternative search information, based on multiple ticket features related to the user's ticketing decision, a score of the feature difference between the candidate alternative search information and the recommended transportation ticket in each ticket feature, and obtain the feature difference score of the candidate alternative search information based on the score of the feature difference of the candidate alternative search information in each ticket feature.

[0113] As an optional implementation, Figure 7 The optional block diagram of the monotonic balanced network module provided in the embodiment of the present application is shown as an example. Figure 4 and Figure 7 As shown, the monotonic balancing network module 430 may include: A monotonic network data input module 710, for obtaining a plurality of ticket features respectively associated with each search result of the recommended transportation ticket, and a plurality of ticket features respectively associated with each candidate alternative search information; The data preprocessing module 720 is used to implement step S610, that is, the data preprocessing module 720 can pool multiple ticket features associated with each search result of the recommended transportation ticket to obtain multiple pooled ticket features of the recommended transportation ticket.

[0114] In an optional implementation, the data preprocessing module 720 can be a pooling module, which takes multiple ticket features respectively associated with each search result of recommended transportation tickets as input, and outputs pooled ticket features whose feature dimensions meet the pooling requirements, i.e., multiple pooled ticket features of recommended transportation tickets; for example, the pooling module can be a minimum pooling module, which takes multiple ticket features respectively associated with each search result of recommended transportation tickets as input, and outputs pooled ticket features whose feature values ​​in each feature dimension are the smallest, i.e., multiple pooled ticket features of recommended transportation tickets.

[0115] Step S620: Perform difference processing on each ticket feature of the candidate replacement search information and each pooled ticket feature of the recommended transportation ticket to obtain feature differences of each ticket feature of the candidate replacement search information.

[0116] After obtaining multiple pooled ticket features of recommended transportation tickets, the embodiment of the present application can perform difference processing on the ticket features of the candidate alternative search information and the pooled ticket features of the recommended transportation tickets in each feature dimension, thereby obtaining the feature difference of the ticket features of the candidate alternative search information in each feature dimension, that is, the feature difference of each ticket feature of the candidate alternative search information.

[0117] In an optional implementation, the embodiment of the present application can perform difference processing on the ticket features of the candidate alternative search information and the pooled ticket features of the recommended transportation tickets through a subtraction operation. For example, for any feature dimension, the pooled ticket features of the recommended transportation tickets are subtracted from the ticket features of the candidate alternative search information to obtain the feature difference of the ticket features of the candidate alternative search information in this feature dimension; further, the ticket features of the candidate alternative search information and the pooled ticket features of the recommended transportation tickets are difference processed in each feature dimension to obtain the feature difference of each ticket feature of the candidate alternative search information. For example, in the price dimension, the price features of the candidate alternative search information are subtracted from the pooled price features of the recommended transportation tickets to obtain the feature difference of the price features of the candidate alternative search information, referred to as the price feature difference; in the distance dimension, the distance features of the candidate alternative search information are subtracted from the pooled distance features of the recommended transportation tickets to obtain the feature difference of the distance features of the candidate alternative search information, referred to as the distance feature difference; in the time dimension, the time features of the candidate alternative search information are subtracted from the pooled time features of the recommended transportation tickets to obtain the feature difference of the time features of the candidate alternative search information, referred to as the time feature difference.

[0118] In an optional implementation, combined with Figure 7 As shown, the monotonic balancing network module 430 may include: a plurality of difference calculation units 730, Figure 7 Taking three difference calculation units as an example, multiple difference calculation units 730 are used to implement step S620, that is, multiple difference calculation units 730 can perform difference processing on each ticket feature of the candidate alternative search information and each pooled ticket feature of the recommended transportation ticket, respectively, to obtain the feature difference of each ticket feature of the candidate alternative search information. In an optional implementation, a difference calculation unit can perform difference processing on the ticket features of the candidate alternative search information and the pooled ticket features of the recommended transportation ticket in one feature dimension, to obtain the feature difference of the ticket features of the candidate alternative search information in one feature dimension; thus, multiple difference calculation units can perform difference processing on the ticket features of the candidate alternative search information and the pooled ticket features of the recommended transportation ticket in multiple feature dimensions, respectively, to obtain the feature difference of the ticket features of the candidate alternative search information in multiple feature dimensions, that is, the feature difference of each ticket feature of the candidate alternative search information.

[0119] For example, taking the multiple ticket features including price features, distance features, and time features as an example, the multiple difference calculation units may include: a difference calculation unit for performing difference processing on the price feature of the candidate replacement search information and the pooled price feature of the recommended transportation ticket, the difference calculation unit outputting the price feature difference of the candidate replacement search information; a difference calculation unit for performing difference processing on the distance feature of the candidate alternative search information and the pooled distance feature of the recommended transportation ticket, the difference calculation unit outputting the distance feature difference of the candidate alternative search information; a difference calculation unit for performing difference processing on the time feature of the candidate alternative search information and the pooled time feature of the recommended traffic ticket, the difference calculation unit outputting the time feature difference of the candidate alternative search information; For ease of explanation, for any candidate alternative search information, the embodiment of the present application may define the price feature difference of the candidate alternative search information as Δp, the distance feature difference of the candidate alternative search information as Δd, and the time feature difference of the candidate alternative search information as Δt.

[0120] Step S630: quantize the feature differences of each ticket feature of the candidate alternative search information into binary vectors to obtain binary vectors corresponding to the feature differences of each ticket feature of the candidate alternative search information.

[0121] After obtaining the characteristic differences of each ticket feature of the candidate alternative search information (such as price characteristic difference Δp, distance characteristic difference Δd, time characteristic difference Δt), the embodiment of the present application can quantize the characteristic differences of each ticket feature respectively, thereby quantizing the characteristic differences of each ticket feature into corresponding binary vectors, and obtaining binary vectors corresponding to the characteristic differences of each ticket feature of the alternative search information.

[0122] In an optional implementation, the characteristic difference of each ticket feature can be quantized through gear division and reverse operation to obtain the corresponding binary vector. Specifically, for any ticket feature, the embodiment of the present application can determine the characteristic difference gear in which the characteristic difference of the ticket feature of the candidate alternative search information is located from the multiple characteristic difference gears pre-divided by the ticket feature; and, based on the original binary vectors corresponding to each characteristic difference gear, determine the original binary vector corresponding to the characteristic difference gear in which it is located to obtain the original binary vector corresponding to the characteristic difference of the ticket feature of the candidate alternative search information; and then, perform a reverse operation on the original binary vector corresponding to the characteristic difference of the ticket feature of the candidate alternative search information to obtain the binary vector corresponding to the characteristic difference of the ticket feature of the candidate alternative search information.

[0123] For example, taking the quantization processing of the price characteristic difference Δp of the candidate replacement search information as an example, the embodiment of the present application can pre-divide multiple price characteristic difference gears. Taking the pre-division of P price characteristic difference gears as an example, the range of the price characteristic difference is divided into P intervals, such as the 1st price characteristic difference gear to the Pth price characteristic difference gear, corresponding to the P intervals into which the price characteristic difference is divided; for example, assuming that P is 3, 3 price characteristic difference gears can be pre-divided, that is, the range of the price characteristic difference is divided into 3 intervals, such as the 1st price characteristic difference gear is the interval of price difference between 0 and 20 yuan, the 2nd price characteristic difference gear is the interval of price difference between 21 yuan and 50 yuan, and the 3rd price characteristic difference gear is the interval of price difference between 51 yuan and 100 yuan; of course, the above-mentioned specific numerical values ​​and gear divisions are only examples, and the number of specific price characteristic difference gears and the price difference interval corresponding to each price characteristic difference gear can be set according to actual conditions, and the embodiment of the present application is not limited thereto; Further, for any price feature difference gear, the embodiment of the present application can set a price feature difference gear index corresponding to the price feature difference gear to identify each pre-divided price feature difference gear, for example, the price feature difference gear indexes from the 1st price feature difference gear to the Pth price feature difference gear can be identified as price feature difference gear indexes 1 to P in sequence, that is, P price feature difference gears correspond to P price feature difference gear indexes; The embodiment of the present application can further set the original binary vector corresponding to each price feature difference level, that is, each price feature difference level corresponds to an original binary vector; thus, the embodiment of the present application can determine the price feature difference level at which the price feature difference Δp of the candidate alternative search information is located from the multiple pre-divided price feature difference levels, and based on the original binary vectors corresponding to each price feature difference level, determine the original binary vector corresponding to the price feature difference level at which the price feature difference of the candidate alternative search information is located, so as to obtain the original binary vector corresponding to the price feature difference of the candidate alternative search information; for ease of explanation, the original binary vector corresponding to the price feature difference of the candidate alternative search information can be expressed as ; For example, assuming that the price feature difference Δp of the candidate alternative search information is in the first price feature difference level, and the corresponding price feature difference level index is 1, then the original binary vector corresponding to the first price feature difference level is the original binary vector corresponding to the price feature difference of the candidate alternative search information ; Assume that the original binary vector corresponding to the first price feature difference level is [0,0,...0,1,...,1], where The first term is 0, and the next term is 1, then is [0,0,...0,1,...,1]; The original binary vector corresponding to the price feature difference of the candidate alternative search information is obtained Afterwards, the embodiment of the present application can convert the original binary vector Perform a reversal operation, such as converting the original binary vector The 1 in is inverted to 0, and the 0 is inverted to 1, so as to obtain the binary vector corresponding to the price feature difference of the candidate alternative search information; for the convenience of explanation, the binary vector corresponding to the price feature difference of the candidate alternative search information can be expressed as ,Right now It is through Perform the inversion operation to obtain .

[0124] It is understandable that in After the reversal operation, if the binary vector corresponding to the price feature difference of a candidate alternative search information The more 1s there are in , the greater the price difference between the price feature of the candidate alternative search information and the pooled price feature of the recommended transportation ticket (such as the minimum price feature) compared to other candidate alternative search information, the cheaper the price feature of the candidate alternative search information is, and the more it meets the price requirements of the user. Therefore, by quantizing the price feature difference of the candidate alternative search information into a binary vector , the impact of price feature differences on the final recommended alternative search information can be more accurately considered when making alternative transportation ticket recommendations.

[0125] The quantization processing of the distance feature difference Δd and the time feature difference Δt of the candidate replacement search information is the same as above, and a corresponding simplified explanation is given below.

[0126] In an optional implementation, the embodiment of the present application can pre-divide multiple distance feature difference gears, for example, pre-divide D distance feature difference gears, and at the same time, each distance feature difference gear corresponds to a distance feature difference gear index, and each distance feature difference gear is set with a corresponding original binary vector; thus, the embodiment of the present application can determine the distance feature difference gear in which the distance feature difference Δd of the candidate alternative search information is located from the multiple pre-divided distance feature difference gears, and based on the original binary vectors corresponding to each distance feature difference gear, determine the original binary vector corresponding to the distance feature difference gear in which the distance feature difference of the candidate alternative search information is located, so as to obtain the original binary vector corresponding to the distance feature difference of the candidate alternative search information; for ease of explanation, the original binary vector corresponding to the distance feature difference of the candidate alternative search information can be expressed as ; Furthermore, the embodiment of the present application can Perform a reversal operation, such as The 1 in the image is inverted to 0, and the 0 is inverted to 1, and the binary vector corresponding to the distance feature difference of the candidate alternative search information is obtained; for the convenience of explanation, the binary vector corresponding to the distance feature difference of the candidate alternative search information can be expressed as ,Right now Through It should be noted that the binary vector corresponding to the distance feature difference of a candidate replacement search information is The more 0s there are in , the smaller the distance difference between the candidate alternative search information and the recommended transportation ticket is compared with other candidate alternative search information, that is, the closer the distance of the candidate alternative search information is.

[0127] In an optional implementation, the embodiment of the present application can pre-divide multiple time feature difference gears, such as pre-dividing T time feature difference gears, and at the same time, each time feature difference gear corresponds to a time feature difference gear index, and each time feature difference gear is set with a corresponding original binary vector; thus, the embodiment of the present application can determine the time feature difference gear in which the time feature difference Δt of the candidate alternative search information is located from the multiple pre-divided time feature difference gears, and based on the original binary vectors corresponding to each time feature difference gear, determine the original binary vector corresponding to the time feature difference gear in which the time feature difference of the candidate alternative search information is located, so as to obtain the original binary vector corresponding to the time feature difference of the candidate alternative search information; for ease of explanation, the original binary vector corresponding to the time feature difference of the candidate alternative search information can be expressed as ; Furthermore, the embodiment of the present application can Perform a reversal operation, such as The 1 in the image is inverted to 0, and the 0 in the image is inverted to 1, and the binary vector corresponding to the time feature difference of the candidate alternative search information is obtained; for the convenience of explanation, the binary vector corresponding to the time feature difference of the candidate alternative search information can be expressed as ,Right now Through It should be noted that the binary vector corresponding to the time feature difference of a candidate alternative search information is The more 0s there are in , the smaller the time difference between the candidate alternative search information and the recommended transportation ticket is compared with other candidate alternative search information, that is, the time of the candidate alternative search information is closer.

[0128] In an optional implementation, combined with Figure 7 As shown, the monotonic balanced network module 430 may include: a plurality of monotonic encoders 740, Figure 7Taking three monotone encoders as an example, multiple monotone encoders 740 are used to implement step S630, that is, multiple monotone encoders can quantize the feature differences of each ticket feature of the candidate alternative search information into binary vectors, respectively, to obtain binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature. Specifically, a monotone encoder can quantize the feature difference of a ticket feature of the candidate alternative search information into a corresponding binary vector, so that multiple monotone encoders can output binary vectors corresponding to the feature differences of the candidate alternative search information in multiple ticket features.

[0129] In an optional implementation example, taking the multiple ticket features including price features, distance features, and time features as an example, the multiple monotonic encoders may include: The price feature difference Δp of the candidate alternative search information is quantified into the binary vector corresponding to the price feature difference of the candidate alternative search information Monotonic encoder of ; The distance feature difference Δd of the candidate alternative search information is quantified into a binary vector corresponding to the distance feature difference of the candidate alternative search information: Monotonic encoder of ; The time feature difference Δt of the candidate alternative search information is quantified into the binary vector corresponding to the time feature difference of the candidate alternative search information The monotone encoder.

[0130] Step S640: Determine the scores of the feature differences of the candidate alternative search information corresponding to each ticket feature according to the binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature.

[0131] After obtaining the binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature, the embodiment of the present application can determine the scores of the feature differences of the candidate alternative search information in each ticket feature based on the binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature.

[0132] In an optional implementation, for any candidate replacement search information, the embodiment of the present application can be combined with Figure 3The search condition vector combination obtained in step S340 and the search result vector combination obtained in step S350 are used to determine the scores of the feature differences corresponding to the candidate alternative search information in each ticket feature; for example, the search condition vector combination and the search result vector combination are respectively combined with the binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature, thereby obtaining the scores of the feature differences corresponding to the candidate alternative search information in each ticket feature. For example, for any candidate alternative search information, the embodiment of the present application can perform a dot multiplication operation on the search condition vector combination and the search result vector combination with the binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature, respectively, to obtain the scores of the feature differences corresponding to the candidate alternative search information in each ticket feature.

[0133] As an implementation example, for any candidate replacement search information, let the search condition vector combination be , the search result vector combination is , then taking the scores of the price feature difference Δp, the distance feature difference Δd, and the time feature difference Δt of the candidate alternative search information as an example, when determining the score of the price feature difference Δp, we can , as well as (the binary vector corresponding to the price feature difference of the candidate alternative search information) is multiplied to obtain the score of the price feature difference Δp; let the score of the candidate alternative search information in the price feature difference Δp be Logit p , then Logit p It can be expressed as: ; When determining the score of the distance feature difference Δd, we can , as well as (binary vector corresponding to the distance feature difference of candidate alternative search information) is multiplied to obtain the score of the distance feature difference Δd; let the score of the candidate alternative search information in the distance feature difference Δd be Logit d , then Logit d It can be expressed as: ; When determining the score of the time characteristic difference Δt, we can , as well as (binary vector corresponding to the time feature difference of candidate alternative search information) is multiplied to obtain the score of the time feature difference Δt; let the score of the candidate alternative search information at the time feature difference Δt be Logit t , then Logit t It can be expressed as: .

[0134] In an optional implementation, combined with Figure 7 As shown, the monotonic balance network module 430 may include: a plurality of feature difference scoring units 750, Figure 7 Taking three feature difference scoring units as an example, multiple feature difference scoring units 750 are used to implement step S640. Specifically, one feature difference scoring unit can determine the score of the feature difference of the candidate alternative search information corresponding to a ticket feature. For example, one feature difference scoring unit can determine the score of the feature difference of the candidate alternative search information corresponding to the ticket feature based on the binary vector corresponding to the feature difference of the candidate alternative search information in the ticket feature. Thus, multiple feature difference scoring units can output the scores of the feature differences of the candidate alternative search information corresponding to multiple ticket features, that is, the scores of the feature differences of the candidate alternative search information corresponding to each ticket feature.

[0135] In an optional implementation, the feature difference scoring unit may be a dot multiplication unit, which is used to perform a dot multiplication operation on the combination of search condition vectors and the combination of search result vectors with the binary vector corresponding to the feature difference of the candidate alternative search information in the ticket feature. Thus, by arranging a plurality of dot multiplication units, the combination of search condition vectors and the combination of search result vectors may be dot multiplied with the binary vector corresponding to the feature difference of the candidate alternative search information in each ticket feature, respectively, to obtain the score of the feature difference of the candidate alternative search information in each ticket feature.

[0136] For example, taking the multiple ticket features including price features, distance features, and time features as an example, the multiple point multiplication units may include: Combine the search criteria vector , search result vector combination , the binary vector corresponding to the price feature difference of candidate alternative search information A dot product unit performs a dot product operation, which outputs the score Logit of the price feature difference of the candidate alternative search information p ; Combine the search criteria vector , search result vector combination , the binary vector corresponding to the distance feature difference of the candidate alternative search information A dot product unit that performs a dot product operation, which outputs the score Logit of the candidate alternative search information in the distance feature difference d ; Combine the search criteria vector , search result vector combination , the binary vector corresponding to the temporal feature difference of the candidate alternative search information A dot product unit that performs a dot product operation, which outputs the score Logit of the candidate alternative search information in the time feature differencet .

[0137] Step S650: The feature difference scores of the candidate alternative search information corresponding to each ticket feature are synthesized to obtain the feature difference scores of the candidate alternative search information.

[0138] After obtaining the scores of the feature differences of the candidate alternative search information corresponding to each ticket feature, for example, obtaining the score Logit of the price feature difference of the candidate alternative search information p , the score Logit of the distance feature difference d , the score Logit of the time characteristic difference t Afterwards, the embodiment of the present application may integrate the feature difference scores of the candidate alternative search information corresponding to each ticket feature, such as weighted summation, to obtain the feature difference score of the candidate alternative search information.

[0139] In an optional implementation, the embodiment of the present application can pre-set the weights of each ticket feature, thereby performing weighted summation processing based on the weights of each ticket feature and the scores of the feature differences of the candidate alternative search information corresponding to each ticket feature to obtain the feature difference scores of the candidate alternative search information.

[0140] For example, taking multiple ticket features including price features, distance features, and time features as an example, the embodiment of the present application can be combined with the pre-set weight of the price feature (set as w p ), the weight of the distance feature (set to w d ), the weight of the time feature (set to w t ), so the weight of the price feature w p Logit score of price feature difference with candidate alternative search information p Multiplication, weight w of distance feature d Logit score of the difference in distance feature with candidate alternative search information d Multiplication, weight w of time feature t Logit score of the difference in time characteristics with candidate alternative search information t Multiply, then add up the products, complete the weighted summation process, and obtain the feature difference score of the candidate alternative search information.

[0141] For example, the feature difference score of the candidate alternative search information can be regarded as a score output by the alternative transportation ticket recommendation subsystem of the travel service platform using the model, expressed as Logit2, and .

[0142] In an optional implementation, combined with Figure 7As shown, the monotonic balance network module 430 may include: a feature difference score synthesis unit 760, which is used to implement step S650, that is, the feature difference score synthesis unit 760 can synthesize the feature difference scores of the candidate alternative search information corresponding to each ticket feature to obtain the feature difference score of the candidate alternative search information.

[0143] In an optional implementation, the feature difference score synthesis unit 760 may be a weighted summation unit, which is used to perform weighted summation based on the weights of each ticket feature and the feature difference scores of the candidate alternative search information corresponding to each ticket feature, to obtain the feature difference scores of the candidate alternative search information. For example, the weighted summation unit may be based on the weight w of the price feature. p and the Logit score of the price feature difference between candidate alternative search information p , the weight of the distance feature w d and the candidate alternative search information in the distance feature difference score Logit d , and the weight of the time feature w t and the score Logit of the difference in time characteristics between candidate alternative search information t , perform weighted summation processing and output the feature difference score of the candidate alternative search information.

[0144] Each candidate alternative search information is passed through Figure 6 By processing the process shown, the characteristic difference score Logit2 of each candidate alternative search information can be obtained. The higher the Logit2 of the candidate alternative search information, the closer the candidate alternative search information is to the user's needs.

[0145] Back to Figure 2 As shown, in step S260, the score of each candidate alternative search information is determined according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information.

[0146] For any candidate alternative search information, after determining the user's search preference score Logit1 for the candidate alternative search information and the characteristic difference score Logit2 of the candidate alternative search information, the embodiment of the present application can at least comprehensively consider the search preference score Logit1 of the candidate alternative search information and the characteristic difference score Logit2 of the candidate alternative search information to obtain the score of the candidate alternative search information.

[0147] In an optional implementation, when determining the score of the candidate alternative search information, the embodiment of the present application can further combine the search preference weight and the set bias item parameter. Among them, the search preference weight is the weight corresponding to the search preference score, which can be expressed as w1. The bias item parameter is a bias item introduced to prevent the user from accidentally touching or operating the item, which can be expressed as Specifically, the bias parameter is an additional constant term used to adjust the model output. In the recommendation system, the bias parameter can be a constant to prevent misoperation or excessive bias towards certain features. It ensures the rationality of the output by adjusting the score to avoid unreasonable results introduced by user misoperation or excessive bias towards features.

[0148] In an optional implementation, for any candidate alternative search information, the embodiment of the present application can multiply the search preference weight w1 by the user's search preference score Logit1 for the candidate alternative search information, and add the product to the characteristic difference score Logit2 of the candidate alternative search information and the set bias item parameter to obtain the score of the candidate alternative search information.

[0149] For example, let the score of a candidate alternative search information be Logit, then Logit is expressed as: .

[0150] In a further optional implementation, the sum of the search preference weight and the weight of each ticket feature is 1; for example, the search preference weight w1, the price feature weight w p , the weight of the distance feature w d , the weight of the time feature w t The sum can be 1, that is, w1+w p +w d +w t =1, that is, the combined impact of the above four weights is equal to 1.

[0151] In a further optional implementation, the embodiment of the present application can adjust the values ​​of each weight according to the actual situation and data characteristics. For example, when the search preference is more important, the search preference weight w1 has a greater impact on the recommendation than other weights. Then the search preference weight w1 can be set to the maximum, such as w1=0.5, w p =0.2, w d =0.2, w t =0.1; for example, when the price feature is more important, such as for price-sensitive users, the weight of the price feature is w p If the influence on the recommendation is greater than other weights, the weight of the price feature can be set to w p Maximum, for example, w1=0.2, w p =0.5, w d =0.2, w t =0.1; In some scenarios, the travel service platform's recommendation of alternative transportation tickets may focus on distance and time, so the weight of the distance feature w d , the weight of the time feature w tThere may be larger weight values; of course, the above four weights can be regarded as equally important, so as to set balanced weights, for example, the above four weights are equal, all 0.25.

[0152] In an optional implementation, combined with Figure 4 As shown, the alternative transportation ticket recommendation subsystem 400 may include: a final score synthesis module 440, which can be used to implement step S260, that is, the final score synthesis module can determine the score of each candidate alternative search information according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information.

[0153] For example, the final score synthesis module can be in the form of a weighted summation unit, which can be used to multiply the search preference weight w1 by the user's search preference score Logit1 for the candidate alternative search information for any candidate alternative search information, and add the product to the characteristic difference score Logit2 of the candidate alternative search information and the set bias item parameter to obtain the score Logit of the candidate alternative search information.

[0154] Back to Figure 2 As shown, in step S270, recommended alternative search information is determined from multiple candidate alternative search information according to the scores of each candidate alternative search information.

[0155] After obtaining the Logit score of each candidate alternative search information, the embodiment of the present application can determine the alternative search information whose score meets the recommendation requirement from multiple candidate alternative search information based on the score of each candidate alternative search information as the recommended alternative search information.

[0156] In an optional implementation, the embodiment of the present application can sort multiple candidate alternative search information according to their scores, and the sorting can be in descending order (i.e., the candidate alternative search information with the highest score is placed in front) or in ascending order; thereby, based on the sorting result, a set recommended number of candidate alternative search information is selected from the candidate alternative search information with the highest score as the recommended alternative search information, and the number of recommended alternative search information (i.e., the set recommended number) can be set according to the specific situation, and the embodiment of the present application is not limited thereto.

[0157] In other optional implementations, the embodiments of the present application may also support using candidate alternative search information whose scores meet the scoring criteria as recommended alternative search information, such as candidate alternative search information whose scores are higher than the scoring criteria as recommended alternative search information.

[0158] Furthermore, the recommended alternative search information and the search results of recommended transportation tickets can be displayed together on the search results page. In a further optional implementation, the display method of the recommended alternative search information on the search results page includes but is not limited to any one of horizontal sliding display, list display, card display, etc.; wherein, horizontal sliding display means that the user can view the recommended alternative search information on the search results page by sliding horizontally on the screen by dragging, scrolling or clicking arrows; for example, when the recommended alternative search information is displayed by horizontal sliding display, when there are multiple recommended alternative search information, the multiple recommended alternative search information will not be displayed on the search results page at once, but will be presented in a horizontal sliding manner, that is, the user can view all the recommended alternative search information by sliding, and the horizontal sliding display can be applicable to the search results page of the mobile terminal or the web page to display the recommended alternative search information.

[0159] In an optional implementation, combined with Figure 4 As shown, the alternative transportation ticket recommendation subsystem 400 may include: a final recommendation module 450, which is used to implement step S270, that is, the final recommendation module can determine the recommended alternative search information from multiple candidate alternative search information according to the scores of each candidate alternative search information.

[0160] The ticket recommendation method provided in the embodiment of the present application, in addition to recommending search results of recommended transportation tickets that meet the ticket search conditions, also recommends alternative search information of alternative transportation tickets related to the ticket search conditions when the user searches for tickets, so as to facilitate the user to search for alternative transportation tickets; when recommending alternative search information, the embodiment of the present application can first determine a plurality of candidate alternative search information of alternative transportation tickets related to the ticket search conditions, wherein each candidate alternative search information includes the ticket search conditions for searching for alternative transportation tickets and is associated with a plurality of ticket features related to the user's ticketing decision; then, each candidate alternative search information is scored, and in the embodiment of the present application, the scores of the candidate alternative search information are scored. The score includes the user's search preference score for the candidate alternative search information and the feature difference score of the candidate alternative search information; the user's search preference score for the candidate alternative search information can reflect the user's preference for selecting the candidate alternative search information when searching for tickets, and the feature difference score of the candidate alternative search information takes into account the differences in ticket characteristics related to the user's ticketing decision between the candidate alternative search information and the recommended transportation ticket. The differences can effectively measure the degree of deviation between the alternative transportation ticket corresponding to the candidate alternative search information and the recommended transportation ticket in multiple ticket feature dimensions, thereby helping the travel service platform to screen and recommend the alternative transportation ticket that is closest to the user's needs, thereby improving the accuracy of the recommendation of alternative transportation tickets.

[0161] Thus, the embodiments of the present application can determine the user's search preference score for candidate alternative search information based on the user's user feature information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets; determine the feature difference scores of the candidate alternative search information compared to the recommended transportation ticket in each ticket feature based on multiple ticket features related to the user's ticketing decision, and obtain the feature difference scores of the candidate alternative search information based on the feature difference scores of the candidate alternative search information in each ticket feature; further, determine the scores of each candidate alternative search information based on the user's search preference scores for each candidate alternative search information and the feature difference scores of each candidate alternative search information; and finally determine the recommended alternative search information from multiple candidate alternative search information based on the scores of each candidate alternative search information.

[0162] It can be seen that the embodiment of the present application combines the user's search preference score for candidate alternative search information and the characteristic difference score of the candidate alternative search information, and can comprehensively consider the user's search preference and the actual situation of the alternative transportation ticketing corresponding to the candidate alternative search information, and generate accurate scores for the candidate alternative search information. Furthermore, this dual evaluation system of the search preference score and the characteristic difference score for the candidate alternative search information can reduce the errors caused by one-sided consideration of a single factor, more accurately recommend alternative search information of alternative transportation tickets to users, and improve the accuracy of the recommendation of alternative transportation ticketing; thereby providing users with more choice possibilities for travel, improving users' travel experience, and thereby improving the ticket click-through rate and conversion rate of the travel service platform.

[0163] Furthermore, in the construction of the alternative transportation ticket recommendation subsystem, the embodiment of the present application introduces a monotonic balanced network architecture, namely a monotonic balanced network module, so as to model the influence of multiple ticket characteristics related to the user's ticketing decision (such as price characteristics, distance characteristics, and time characteristics) on the recommendation of alternative search information, so that the model can be smoothly adjusted based on the differences in ticket characteristics, thereby enhancing the accuracy of the recommendation.

[0164] Furthermore, the score of the candidate alternative search information determined in the embodiment of the present application is the Logit fusion result of multiple models, specifically the fusion result of the Logit score of the main network module and the Logit score of the monotonic balanced network module. The score of the candidate alternative search information is obtained by fusing the Logit of multiple models, which can reduce the negative impact caused by overfitting of a model or excessive sensitivity to certain features, thereby enhancing stability; that is, the Logit result of a single model may be affected by outliers, or appear unstable on certain features, but through the Logit fusion of multiple models, the Logit results of multiple models can be combined through weighted averaging or voting mechanisms, thereby improving the accuracy and robustness of recommendations.

[0165] Taking the recommendation of nearby tickets as an example, compared with the use of traditional deep neural network (DNN) algorithm for nearby ticket recommendation, the use of the ticket recommendation scheme provided in the embodiment of the present application for nearby ticket recommendation can have obvious effects on the increase of AUC (Area Under Curve) and CTR (Click Through Rate); specifically, through offline experiments, compared with the use of DNN algorithm for nearby ticket recommendation, the ticket recommendation scheme provided in the embodiment of the present application has a 3.4% AUC increase in nearby ticket recommendation, which improves the ability of the travel service platform to distinguish different user needs and preferences, thereby improving the recommendation accuracy of alternative transportation ticket recommendations; through online experiments, compared with the use of DNN algorithm for nearby ticket recommendation, the ticket recommendation scheme provided in the embodiment of the present application has a 14.1% CTR increase in nearby ticket recommendation, which improves the user's click-through rate.

[0166] In an optional implementation, the embodiment of the present application further provides a ticket recommendation system, which is applied to a travel service platform. Figure 8 An optional block diagram of the ticket recommendation system provided by the embodiment of the present application is shown exemplarily. Figure 8 The ticket recommendation system provided in the embodiment of the present application may include: A request acquisition module 810, configured to acquire a user's ticket search request, wherein the ticket search request includes a ticket search condition; A ticket search engine 820, for determining search results of recommended transportation tickets that meet the ticket search conditions; The alternative transportation ticket recommendation subsystem 400 is used to determine multiple candidate alternative search information of alternative transportation tickets related to the ticket search condition, each candidate alternative search information includes a ticket search condition for searching for alternative transportation tickets and is associated with multiple ticket features related to the user's ticketing decision; for any candidate alternative search information, determine the user's search preference score for the candidate alternative search information based on the user's user feature information, historical behavior information, the ticket search condition of the candidate alternative search information, and the search result of the recommended transportation ticket; and, for any candidate alternative search information, determine the score of the feature difference corresponding to each ticket feature of the candidate alternative search information compared with the recommended transportation ticket based on the multiple ticket features, and obtain the feature difference score of the candidate alternative search information based on the feature difference score corresponding to each ticket feature of the candidate alternative search information; determine the score of each candidate alternative search information based on the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information; and determine the recommended alternative search information from the multiple candidate alternative search information based on the score of each candidate alternative search information.

[0167] In an optional implementation, the relevant components and functions of the alternative transportation ticket recommendation subsystem 400 can refer to the above description and will not be expanded here.

[0168] In a further optional implementation, the embodiment of the present application also provides a travel server, which is regarded as a server or server cluster set up in the travel service platform; in an optional implementation, the travel server may include a memory and a processor, the memory stores computer execution instructions, and the processor calls the computer execution instructions stored in the memory to execute the ticket recommendation method provided in the embodiment of the present application.

[0169] In a further optional implementation, the embodiment of the present application also provides a storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed (for example, when the computer-executable instructions are executed by a processor), the ticket recommendation method provided in the embodiment of the present application is implemented.

[0170] In a further optional implementation, the embodiment of the present application also provides a computer program product, including computer execution instructions, which, when executed (for example, when the computer execution instructions are executed by a processor), implement the ticket recommendation method provided in the embodiment of the present application.

[0171] The above describes multiple implementation schemes provided by the embodiments of the present application. The various optional methods introduced in each implementation scheme can be combined and cross-referenced with each other without conflict, thereby extending a variety of possible implementation schemes, which can all be considered as implementation schemes disclosed and open in the embodiments of the present application.

[0172] Although the embodiments of the present application are disclosed above, the present application is not limited thereto. Any person skilled in the art may make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be subject to the scope defined by the claims.

Claims

1. A ticket recommendation method, characterized in that: include: Obtaining a ticket search request from a user, wherein the ticket search request includes a ticket search condition; Determining search results of recommended transportation tickets that match the ticket search criteria; Determine a plurality of candidate replacement search information of alternative transportation tickets related to the ticket search condition, each candidate replacement search information includes a ticket search condition for searching for alternative transportation tickets and is associated with a plurality of ticket features related to the user's ticket decision; For any candidate alternative search information, determine the user's search preference score for the candidate alternative search information based on the user's user characteristic information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets; and, for any item of candidate alternative search information, determining a score of a feature difference between the candidate alternative search information and the recommended transportation ticket corresponding to each ticket feature based on the plurality of ticket features, and obtaining a feature difference score of the candidate alternative search information based on the score of the feature difference between the candidate alternative search information and each ticket feature; Determining the score of each candidate alternative search information according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information; According to the scores of each candidate alternative search information, recommended alternative search information is determined from the plurality of candidate alternative search information.

2. The method according to claim 1, characterized in that The step of determining, based on the plurality of ticket features, a score of a feature difference between the candidate alternative search information and the recommended transportation ticket corresponding to each ticket feature comprises: Pooling multiple ticket features associated with each search result of recommended transportation tickets to obtain multiple pooled ticket features of recommended transportation tickets; Performing difference processing on each ticket feature of the candidate replacement search information and each pooled ticket feature of the recommended transportation ticket, respectively, to obtain the feature difference of each ticket feature of the candidate replacement search information; quantizing the feature differences of each ticket feature of the candidate alternative search information into binary vectors respectively, so as to obtain the binary vectors corresponding to the feature differences of each ticket feature of the candidate alternative search information; According to the binary vectors corresponding to the feature differences of the candidate replacement search information in each ticket feature, the scores of the feature differences of the candidate replacement search information corresponding to each ticket feature are determined respectively.

3. The method according to claim 2, characterized in that The feature differences of each ticket feature of the candidate alternative search information are quantized into binary vectors respectively to obtain binary vectors corresponding to the feature differences of each ticket feature of the candidate alternative search information, including: For any ticket feature, determine the feature difference level at which the feature difference of the ticket feature of the candidate replacement search information is located from a plurality of feature difference levels pre-divided by the ticket feature; Based on the original binary vectors corresponding to the various feature difference levels, determining the original binary vector corresponding to the feature difference level in which the candidate alternative search information is located, so as to obtain the original binary vector corresponding to the feature difference of the ticket feature; Inverting the original binary vector corresponding to the feature difference of the ticket feature of the candidate replacement search information to obtain the binary vector corresponding to the feature difference of the ticket feature of the candidate replacement search information; The feature difference scores of the candidate alternative search information are obtained based on the feature difference scores corresponding to the various ticket features of the candidate alternative search information, including: Based on the weight of each ticket feature and the feature difference score of the candidate replacement search information corresponding to each ticket feature, a weighted summation process is performed to obtain the feature difference score of the candidate replacement search information.

4. The method according to claim 3, characterized in that: Determining the user's search preference score for the candidate alternative search information based on the user's user characteristic information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets includes: Embed each user feature information of the user into corresponding user feature vectors, and concatenate each user feature vector into a user feature vector combination; For any piece of historical behavior information of the user, embed each historical behavior feature in the historical behavior information into corresponding historical behavior feature vectors, and concatenate each historical behavior feature vector into a historical behavior feature vector combination to obtain a historical behavior feature vector combination corresponding to the historical behavior information; Based on the combination of historical behavior feature vectors corresponding to each piece of historical behavior information, the potential relationship of historical behavior information is extracted through the self-attention mechanism and converted into the user's historical behavior representation; Embed each ticket search condition of the candidate replacement search information into corresponding search condition vectors, and concatenate each search condition vector into a corresponding search condition vector combination; Embed each search result of the recommended transportation ticketing service into a corresponding search result vector, and concatenate each search result vector into a corresponding search result vector combination; Based on the user's historical behavior representation and the combination of search condition vectors, a comprehensive representation vector of ticket search conditions that integrates the user's historical behavior and candidate alternative search information is determined through a self-attention mechanism; The user feature vector combination, the comprehensive representation vector, the search condition vector combination, and the search result vector combination are combined into a combined vector; the combined vector is subjected to multi-layer feature mapping and nonlinear processing to obtain the user's search preference score for the candidate alternative search information.

5. The method according to claim 4, characterized in that The step of determining the scores of the feature differences of the candidate alternative search information corresponding to each ticket feature according to the binary vectors corresponding to the feature differences of the candidate alternative search information in each ticket feature comprises: Perform a dot multiplication operation on the combination of search condition vectors and the combination of search result vectors with the binary vectors corresponding to the feature differences of candidate alternative search information in each ticket feature, respectively, to obtain the scores of the feature differences of candidate alternative search information in each ticket feature; Determining the score of each candidate alternative search information according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information includes: For any candidate alternative search information, the search preference weight is multiplied by the user's search preference score for the candidate alternative search information, and the product is added to the feature difference score of the candidate alternative search information and the set bias parameter to obtain the score of the candidate alternative search information; The sum of the search preference weight and the weight of each ticket feature is 1.

6. The method according to any one of claims 1 to 5, characterized in that: The multiple ticket features associated with the candidate alternative search information and relevant to the user's ticket decision include: a price feature, indicating the price of the alternative transportation ticket corresponding to the candidate alternative search information; A distance feature, which indicates the distance between the departure place and the target departure place of the alternative transportation ticket corresponding to the candidate alternative search information, and the distance between the destination and the target destination of the alternative transportation ticket corresponding to the candidate alternative search information; wherein the departure place when the user searches for tickets is the target departure place, and the destination when the user searches for tickets is the target destination; The time feature indicates the departure time of the alternative transportation ticket corresponding to the candidate alternative search information.

7. A ticket recommendation system, characterized in that: include: A request acquisition module, used to acquire a user's ticket search request, wherein the ticket search request includes a ticket search condition; A ticket search engine, used to determine search results of recommended transportation tickets that meet the ticket search criteria; Alternative transportation ticketing recommendation subsystem for: Determine a plurality of candidate alternative search information of alternative transportation tickets related to the ticket search condition, each candidate alternative search information includes a ticket search condition for searching for alternative transportation tickets and is associated with a plurality of ticket features related to the user's ticketing decision; for any candidate alternative search information, determine a user's search preference score for the candidate alternative search information based on the user's user feature information, historical behavior information, the ticket search condition of the candidate alternative search information, and the search results of the recommended transportation tickets; and, for any item of candidate alternative search information, determining a score of a feature difference between the candidate alternative search information and the recommended transportation ticket corresponding to each ticket feature based on the plurality of ticket features, and obtaining a feature difference score of the candidate alternative search information based on the score of the feature difference between the candidate alternative search information and each ticket feature; Determining the score of each candidate alternative search information according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information; According to the scores of each candidate alternative search information, recommended alternative search information is determined from the plurality of candidate alternative search information.

8. The system according to claim 7, characterized in that The alternative transportation ticket recommendation subsystem includes: A recall module, configured to determine a plurality of candidate alternative search information of alternative transportation tickets related to the ticket search condition; The main network module is used to determine the user's search preference score for any candidate alternative search information based on the user's user characteristic information, historical behavior information, ticket search conditions of the candidate alternative search information, and search results of recommended transportation tickets; a monotone balance network module, for determining, for any candidate alternative search information, a score of a feature difference between the candidate alternative search information and the recommended transportation ticket corresponding to each ticket feature based on the plurality of ticket features, and obtaining a feature difference score of the candidate alternative search information based on the score of the feature difference between the candidate alternative search information and each ticket feature; A final score synthesis module, used to determine the score of each candidate alternative search information according to the user's search preference score for each candidate alternative search information and the feature difference score of each candidate alternative search information; The final recommendation module is used to determine the recommended alternative search information from multiple candidate alternative search information according to the scores of each candidate alternative search information.

9. A travel server, characterized in that: The system comprises a memory and a processor, wherein the memory stores computer-executable instructions, and the processor calls the computer-executable instructions to execute the ticket recommendation method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer-executable instructions, which, when executed, implement the ticket recommendation method according to any one of claims 1 to 6.

Images