Search processing method and device, equipment, storage medium and program product

Abstract

The invention provides a search processing method and device, equipment, a storage medium and a program product, and belongs to the technical field of Internet. By adopting the method, the selection reference information corresponding to each candidate sorting mode corresponding to the keyword can be continuously updated on the basis of operations of different users on the search result of the keyword, and in each search process, one sorting mode is selected from the candidate sorting modes on the basis of the selection reference information and is used for generating the search result. Therefore, the sequencing mode does not need to be determined through a sequencing model, so that model training is not needed, a large number of samples do not need to be marked, and the method does not need a long preparation period.

Description

Technical Field

[0001] The present disclosure relates to the field of Internet technology, and in particular to a search processing method, device, equipment, storage medium and program product. Background Art

[0002] The backend database of an application provides various types of content, which can generally be collectively referred to as heterogeneous resources. For example, the backend database of an audio playback application provides various types of content such as playlists, songs, singers, and albums, while the backend database of a life service application provides various types of resources such as stores, takeaways, and commodities.

[0003] The search processing method in the prior art is to search for resource information of multiple resource types according to the keywords in the search request, input the keywords and the multiple resource types into a trained sorting model, and obtain a sorting mode, which is used to indicate the sorting between the multiple resource types. Then, the search results are determined according to the resource information of the multiple resources and the sorting mode, and fed back to the user.

[0004] However, using the above method requires labeling a large number of samples and using the samples to train the model before the model can be put into use, which will result in a longer preparation cycle for the above method. Summary of the invention

[0005] In order to solve the related technical problems, the present disclosure provides a search processing method, device, equipment, storage medium and program product. The technical solution is as follows:

[0006] In a first aspect, a search processing method is provided, the method comprising:

[0007] receiving a first search request carrying a keyword;

[0008] When resource information of multiple resource types corresponding to the keyword is obtained, multiple candidate sorting modes corresponding to the keyword are obtained, wherein the candidate sorting modes are used to indicate a display order between the multiple resource types;

[0009] Determine the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode;

[0010] Selecting a target sorting mode from the plurality of candidate sorting modes according to the mode selection probability of each candidate sorting mode;

[0011] Sending the resource information of the multiple resource types and the target ranking mode to a sending end of the first search request, so that the sending end displays the resource information of the multiple resource types according to the target ranking mode;

[0012] When operation information for the displayed resource information of the multiple resource types returned by the sending end is obtained, the selection reference information corresponding to the target sorting mode is updated based on the operation information.

[0013] In a possible implementation, the acquiring multiple candidate ranking modes corresponding to the keyword includes:

[0014] If multiple candidate sorting modes corresponding to the keyword have been stored, obtaining the multiple candidate sorting modes corresponding to the keyword;

[0015] If multiple candidate sorting modes corresponding to the keyword are not stored, various possible sortings of the multiple resource types are determined to obtain multiple candidate sorting modes corresponding to the keyword, and the multiple candidate sorting modes corresponding to the keyword are stored.

[0016] In a possible implementation, the selection reference information corresponding to the candidate sorting mode includes a first parameter and a second parameter, an initial value of the first parameter is a first specified value, and an initial value of the second parameter is a second specified value;

[0017] The updating of the selection reference information corresponding to the target sorting mode based on the operation information includes:

[0018] When the operation information satisfies a specified condition, the first parameter in the selection reference information corresponding to the target sorting mode is updated; when the operation information does not satisfy the specified condition, the second parameter in the selection reference information corresponding to the target sorting mode is updated;

[0019] Among them, the specified conditions include: performing a specified operation on the resource information of the target resource type in the search results, the target resource type is the resource type ranked first in the search results, and the specified operation includes at least one of the completion rate reaching a threshold, collection, sharing, and downloading.

[0020] In a possible implementation, determining the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode includes:

[0021] For each candidate sorting mode, determine a probability density function of the candidate sorting mode based on selection reference information corresponding to the candidate sorting mode, and randomly determine a mode selection probability of the candidate sorting mode based on the probability density function;

[0022] The selecting a target sorting mode from the plurality of candidate sorting modes according to the respective mode selection probabilities of each candidate sorting mode includes:

[0023] Among the multiple candidate sorting modes, the candidate sorting mode corresponding to the largest mode selection probability is selected as the target sorting mode.

[0024] In a possible implementation manner, determining the probability density function of the candidate sorting mode based on the selection reference information corresponding to the candidate sorting mode includes:

[0025] Determine a beta distribution function of the selection reference information corresponding to the candidate sorting mode;

[0026] A probability density function of the candidate sorting pattern is determined based on the Beta distribution function.

[0027] In a possible implementation, the first search request is a search request sent by an account in a low-traffic account set, and the method further includes:

[0028] When the judgment period is reached, determining a to-be-standardized sorting mode that satisfies a specified convergence condition from a plurality of candidate sorting modes corresponding to the keyword;

[0029] Determine the selection probability of the to-be-standardized mode for each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode;

[0030] Determine the to-be-standardized sorting pattern corresponding to the largest to-be-standardized pattern selection probability as the standardized sorting pattern corresponding to the keyword;

[0031] When a second search request carrying the keyword is received and resource information of multiple resource types corresponding to the keyword is obtained, the resource information of the multiple resource types and the standardized sorting mode are sent to the sender of the second search request, so that the sender of the second search request displays the resource information of the multiple resource types according to the standardized sorting mode; the second search request is a search request sent by an account in a large-traffic account set.

[0032] In a possible implementation, determining the selection probability of the to-be-standardized mode for each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode includes:

[0033] Determine the probability density function of each of the to-be-standardized sorting modes based on the selection reference information corresponding to each of the to-be-standardized sorting modes;

[0034] Based on the probability density function corresponding to each to-be-standardized sorting mode, the to-be-standardized mode selection probability of each to-be-standardized sorting mode is determined.

[0035] In a possible implementation, the specified convergence condition is that the number of times the candidate sorting mode is selected reaches a number threshold.

[0036] In a second aspect, a search processing device is provided, the device comprising:

[0037] A receiving module, configured to receive a first search request carrying a keyword;

[0038] An acquisition module, configured to acquire, when resource information of multiple resource types corresponding to the keyword is acquired, multiple candidate sorting modes corresponding to the keyword, wherein the candidate sorting modes are used to indicate a display order between the multiple resource types;

[0039] A determination module, used to determine the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode;

[0040] A selection module, configured to select a simulated sorting mode from among the plurality of candidate sorting modes according to the respective mode selection probability of each candidate sorting mode;

[0041] A sending module, used for sending the resource information of the multiple resource types and the target sorting mode to a sending end of the first search request, so that the sending end displays the resource information of the multiple resource types according to the target sorting mode;

[0042] The updating module is used to update the selection reference information corresponding to the target sorting mode based on the operation information returned by the sending end for the resource information of the multiple resource types displayed.

[0043] In a possible implementation, the acquisition module is used to:

[0044] If multiple candidate sorting modes corresponding to the keyword have been stored, obtaining the multiple candidate sorting modes corresponding to the keyword;

[0045] If multiple candidate sorting modes corresponding to the keyword are not stored, various possible sortings of the multiple resource types are determined to obtain multiple candidate sorting modes corresponding to the keyword, and the multiple candidate sorting modes corresponding to the keyword are stored.

[0046] In a possible implementation, the selection reference information corresponding to the candidate sorting mode includes a first parameter and a second parameter, an initial value of the first parameter is a first specified value, and an initial value of the second parameter is a second specified value;

[0047] The update module is used to:

[0048] When the operation information satisfies a specified condition, the first parameter in the selection reference information corresponding to the target sorting mode is updated; when the operation information does not satisfy the specified condition, the second parameter in the selection reference information corresponding to the target sorting mode is updated;

[0049] Among them, the specified conditions include: performing a specified operation on the resource information of the target resource type in the search results, the target resource type is the resource type ranked first in the search results, and the specified operation includes at least one of the completion rate reaching a threshold, collection, sharing, and downloading.

[0050] In a possible implementation manner, the determining module is used to:

[0051] For each candidate sorting mode, determine a probability density function of the candidate sorting mode based on selection reference information corresponding to the candidate sorting mode, and randomly determine a mode selection probability of the candidate sorting mode based on the probability density function;

[0052] The selection module is used to:

[0053] Among the multiple candidate sorting modes, the candidate sorting mode corresponding to the largest mode selection probability is selected as the target sorting mode.

[0054] In a possible implementation manner, the determining module is used to:

[0055] Determine a beta distribution function of the selection reference information corresponding to the candidate sorting mode;

[0056] A probability density function of the candidate sorting pattern is determined based on the Beta distribution function.

[0057] In a possible implementation, the first search request is a search request sent by an account in a low-traffic account set, and the method further includes a second search module, which is used to:

[0058] When the judgment period is reached, determining a to-be-standardized sorting mode that satisfies a specified convergence condition from a plurality of candidate sorting modes corresponding to the keyword;

[0059] Determine the selection probability of the to-be-standardized mode for each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode;

[0060] Determine the to-be-standardized sorting pattern corresponding to the largest to-be-standardized pattern selection probability as the standardized sorting pattern corresponding to the keyword;

[0061] When a second search request carrying the keyword is received and resource information of multiple resource types corresponding to the keyword is obtained, the resource information of the multiple resource types and the standardized sorting mode are sent to the sender of the second search request, so that the sender of the second search request displays the resource information of the multiple resource types according to the standardized sorting mode; the second search request is a search request sent by an account in a large-traffic account set.

[0062] In a possible implementation manner, the determining module is used to:

[0063] Determine the probability density function of each of the to-be-standardized sorting modes based on the selection reference information corresponding to each of the to-be-standardized sorting modes;

[0064] Based on the probability density function corresponding to each to-be-standardized sorting mode, the to-be-standardized mode selection probability of each to-be-standardized sorting mode is determined.

[0065] In a possible implementation, the specified convergence condition is that the number of times the candidate sorting mode is selected reaches a number threshold.

[0066] In a third aspect, a computer device is provided, comprising a memory and a processor, wherein the memory is used to store computer instructions, and the processor executes the computer instructions stored in the memory so that the computer device executes the method provided by the first aspect and its possible implementation methods.

[0067] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program code. When the computer program code is executed by a computer device, the computer device executes the method provided by the first aspect and its possible implementation methods.

[0068] In a fifth aspect, a computer program product is provided. The computer program product includes a computer program code. When the computer program code is executed by a computer device, the computer device executes the method provided by the first aspect and its possible implementation methods.

[0069] The technical solution provided by the embodiments of the present disclosure has the following beneficial effects:

[0070] By adopting the above method, the selection reference information corresponding to each candidate sorting mode corresponding to the keyword can be continuously updated based on the operations of different users on the search results of the keyword. In each search process, a sorting mode is selected from the candidate sorting modes based on the selection reference information to generate the search results. In this way, there is no need to determine the sorting mode through the sorting model, and thus, there is no need to train the model or label a large number of samples. It can be seen that this method does not require a long preparation period. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 is a structural diagram of a server provided by an embodiment of the present disclosure;

[0072] Figure 2 It is a schematic diagram of a processing flow of a search processing method provided by an embodiment of the present disclosure;

[0073] Figure 3 is a page schematic diagram of an application provided by an embodiment of the present disclosure;

[0074] Figure 4 is a schematic diagram showing a search result provided by an embodiment of the present disclosure;

[0075] Figure 5 It is a schematic diagram of a processing flow of selecting a target sorting mode provided by an embodiment of the present disclosure;

[0076] Figure 6 is a schematic diagram of a processing flow for processing a second search request provided by an embodiment of the present disclosure;

[0077] Figure 7 is a schematic diagram of two probability density functions provided by an embodiment of the present disclosure;

[0078] Figure 8 It is a structural diagram of a search processing device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0079] In order to make the objectives, technical solutions and advantages of the present disclosure more clear, the embodiments of the present disclosure will be further described in detail below with reference to the accompanying drawings.

[0080] The disclosed embodiment provides a method for search processing. The execution subject of the method may be a server, which may be a single server or a server group. If it is a single server, the server may be responsible for all processing in the following scheme. If it is a server group, different servers in the server group may be responsible for different processing in the following scheme. The specific processing allocation may be arbitrarily set by the technician according to actual needs, and will not be described in detail here.

[0081] The server may be a background server of an application, and the application may be an audio player application, a video player application, a life service application, a shopping application, etc. The disclosed embodiment takes the background server search process of the music player software as an example to explain the detailed description of the scheme, and other situations are similar to them, which will not be described in detail in this embodiment.

[0082] Figure 1is a schematic diagram of a server structure provided by an embodiment of the present disclosure. From the perspective of hardware composition, the server structure can be as follows: Figure 1 As shown, it includes a processor 110 , a memory 120 and a communication component 130 .

[0083] The processor 110 may be a central processing unit (CPU) or a system on chip (SoC), etc. The processor 110 may be used to process various operation instructions, for example, searching based on a first search request or a second search request carrying keywords.

[0084] The memory 120 may include various volatile memories or non-volatile memories, such as a solid state disk (SSD), a dynamic random access memory (DRAM), etc. The memory 120 may be used to store initial data, intermediate data, and result data used in the relevant processing, for example, the correspondence between the standardized sorting mode and the keywords, the selection reference information corresponding to the multiple candidate sorting modes, etc.

[0085] The communication component 130 may be a wired network connector, ultra wide band (UWB), wireless fidelity (WiFi) module, Bluetooth module, cellular network communication module, etc. The communication component 130 may be used to transmit data with other devices, and the other devices may be other servers or terminals, etc. For example, a first search request carrying a keyword is received.

[0086] People often use audio playback applications. The servers behind audio playback applications store a large number of heterogeneous resources, such as singers, playlists, songs, etc. Users often search in such applications. For example, users enter keywords to search, and the server feeds back search results to users based on the keywords.

[0087] The present disclosure provides a method for search processing. The following is a detailed description of the search processing process. Figure 2 As shown, the following steps are included:

[0088] 201. Receive a first search request carrying a keyword.

[0089] There is usually a search box in the application, and users often need to search in the search box and select resource information that meets their intentions in the search results. For example Figure 3The main page of the music player software is shown. There is a search box in the main page. After the user enters a keyword in the search box, the terminal is triggered to send a first search request. The server receives the first search request, which carries the keyword.

[0090] 202 , when resource information of multiple resource types corresponding to the keyword is obtained, multiple candidate sorting modes corresponding to the keyword are obtained.

[0091] The candidate sorting mode is used to indicate the display order among multiple resource types.

[0092] A heterogeneous resource database may include various types of content, such as songs, singers, playlists, albums, etc. Resource information may include content identifiers and related information. For example, the resource information of a song may include information such as the song identifier and song name, and the resource information of a playlist may include the song identifier, playlist name, part or all of the playlist content, etc.

[0093] After receiving the keyword, various types of content in the heterogeneous resource database are searched. If content matching the keyword can be found in a certain type of content, the resource type is recorded, and the corresponding resource information is recorded, and finally multiple resource types and resource information corresponding to each resource type are obtained. For example, if the keyword entered by the user is "Zhang San", the single ID and single name of Zhang San, the album ID and album name of Zhang San, the singer ID and singer name related to Zhang San, the playlist ID and playlist name related to Zhang San, etc. may be searched.

[0094] Different keywords may have different numbers and types of resource types, so different keywords correspond to different candidate sorting modes. For example, the keyword is "Zhang San", which corresponds to five resource types: "singer", "playlist", "single", "video", and "album". The keyword is "Li Si", which corresponds to three resource types: "single", "playlist", and "album". Therefore, the number and types of candidate sorting modes corresponding to these two keywords are also different.

[0095] If multiple candidate sorting modes corresponding to the keyword have been stored, then the multiple candidate sorting modes corresponding to the stored keyword are obtained. If multiple candidate sorting modes corresponding to the keyword have not been stored, then all possible sortings of multiple resource types are determined to obtain multiple candidate sorting modes corresponding to the keyword, and multiple candidate sorting modes corresponding to the keyword are stored. For example, a keyword is input to obtain resource information of multiple resource types, and the multiple resource types are "singer", "playlist", and "album". The possible sortings are singer, playlist, album, singer, album, playlist, playlist, singer, album, playlist, album, singer, album, singer, songlist, album, songlist, singer, so the keyword corresponds to the above six candidate sorting modes.

[0096] In addition, the technician can also preset to exclude certain sorting, for example, preset to not allow the "video" resource type to be ranked first.

[0097] 203 : Determine the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode.

[0098] The selection reference information is used to directly or indirectly indicate the mode selection probability of the corresponding candidate sorting mode.

[0099] 204 , selecting a target sorting mode from a plurality of candidate sorting modes according to the mode selection probability of each candidate sorting mode.

[0100] The candidate sorting mode with a high probability of pattern selection is used as the target sorting mode. The detailed selection process is described in detail in the following process.

[0101] 205 , sending resource information of multiple resource types and the target ranking mode to the sending end of the first search request, so that the sending end displays the resource information of multiple resource types according to the target ranking mode.

[0102] The resource information of multiple resource types and the target sorting mode are combined into search results, and the search results are sent to the sending end of the first search request. The sending end sorts the resource information of multiple resource types based on the target sorting mode and displays it to the user.

[0103] 206 , when the operation information for the resource information of the multiple resource types displayed is obtained from the sending end, the selection reference information corresponding to the target sorting mode is updated based on the operation information.

[0104] When the search results are displayed to the user, the user can perform a series of operations on the search results. For example, Figure 4As shown, when the keyword input by the user is "Zhang San", the page displayed to the user includes multiple resource types such as "single", "playlist", and "album". "Single" includes multiple songs of Zhang San, "playlist" includes multiple playlists related to Zhang San, and "album" includes multiple albums of Zhang San. The user can click on any resource information in this page.

[0105] The selection reference information corresponding to the candidate sorting mode includes a first parameter and a second parameter, wherein the initial value of the first parameter is a first specified value, and the initial value of the second parameter is a second specified value, and the initial values ​​of the first parameter and the second parameter can be set by a technician. For example, the initial values ​​of the first parameter and the second parameter are both 1.

[0106] The first parameter is used to characterize the relevant information of the "successful sorting" of the candidate sorting mode, which can be specifically the number of successes, and the second parameter is used to characterize the relevant information of the "failed sorting" of the candidate sorting mode, which can be specifically the number of failures. For example, when the sorting is successful, the resource type that the user is most interested in is ranked first. When the first parameter in the selection reference information is updated (the operation information meets the specified conditions), it means "successful sorting", and when the second parameter in the selection reference information is updated (the operation information does not meet the specified conditions), it means "failed sorting".

[0107] The method of updating the selection reference information corresponding to the target sorting mode may be: if the operation information satisfies the specified condition, the first parameter in the selection reference information corresponding to the target sorting mode is added with the third specified value, and if the operation information does not satisfy the specified condition, the second parameter in the selection reference information corresponding to the target sorting mode is added with the third specified value. For example, the third specified value and the fourth specified value are both 1, and the first parameter and the second parameter are both 1. When the operation information satisfies the specified condition, the first parameter is updated to 2, and the second parameter remains unchanged.

[0108] The specified condition may include: performing a specified operation on resource information of a target resource type in the search results, wherein the target resource type is the resource type ranked first in the search results, and the specified operation includes at least one of a completion rate reaching a threshold, collection, sharing, and downloading. For example, the specified condition is that in the search results, the user selects to download a song in the target resource type.

[0109] For example, when the first account collects resource information of the target resource type in the search results, and the first parameter in the selection reference information corresponding to the target sorting mode is 1, when the first account collects resource information of the target resource type in the search results, the first parameter in the selection reference information corresponding to the target sorting mode can be updated to 2.

[0110] In a possible implementation, the process of selecting the target sorting mode can be as follows: Figure 5 As shown, the following steps are included:

[0111] 501. For each candidate sorting mode, determine the probability density function of the candidate sorting mode based on the selection reference information corresponding to the candidate sorting mode, and randomly determine the mode selection probability of the candidate sorting mode based on the probability density function.

[0112] The beta distribution function of the selection reference information corresponding to the candidate sorting pattern may be determined first, and then, based on the beta distribution function, the probability density function of the candidate sorting pattern may be determined.

[0113] The Beta distribution function is α is the first parameter for selecting reference information, β is the second parameter for selecting reference information, and Γ(z) is the gamma function. The probability density function is x is the mode selection probability of the candidate sorting mode, f(x) is the probability density function value of the mode selection probability taking different values, the horizontal axis of the probability density function is the mode selection probability, and the vertical axis is the probability density.

[0114] According to the probability density function of the mode selection probability of the candidate sorting mode, a mode selection probability (i.e., the target mode selection probability) is randomly determined. The larger the probability density corresponding to the mode selection probability, the higher the probability of being randomly selected.

[0115] 502. Select, from among multiple candidate sorting patterns, a candidate sorting pattern corresponding to a maximum target pattern selection probability as a target sorting pattern.

[0116] For example, if the first parameter and the second parameter of the selection reference information corresponding to a candidate sorting mode M are 2 and 4 respectively, then the Beta distribution function is B(2,4), and the probability density function of the mode selection probability of M is The first parameter and the second parameter of the selection reference information corresponding to another candidate sorting mode N are 5 and 1 respectively, then the Beta distribution function is B(5,1), and the probability density function of the mode selection probability of N is If the target mode selection probability randomly determined according to B(2,4) is 0.3, and the target mode selection probability randomly determined according to B(5,1) is 0.75, 0.75 is greater than 0.3, so the candidate sorting mode N corresponding to the target mode selection probability of 0.75 is used as the target sorting mode.

[0117] In a possible implementation, the first search request is a search request sent by an account in a small flow account set, and the second search request is a search request sent by an account in a large flow account set. The processing flow for processing the second search request may be as follows: Figure 6 As shown, the following steps are included:

[0118] 601 , when the judgment period is reached, a to-be-standardized sorting pattern that satisfies a specified convergence condition is determined from a plurality of candidate sorting patterns corresponding to the keywords.

[0119] The judgment period can be set by the technician based on experience, and can be 1 hour, 2 hours, etc. The specified convergence condition can be pre-set by the technician, and the specified convergence condition can be that the number of times the candidate sorting mode is selected as the target sorting mode reaches the number threshold, and the number threshold can be set by the technician, such as 100 times, etc. The specified convergence condition can also be that the number of completed broadcasts, the number of shares, the number of favorites, and the number of downloads reach the number threshold, and the number threshold can be pre-set by the technician, such as 50 times, etc.

[0120] 602 : Determine a selection probability of a to-be-standardized mode for each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode.

[0121] For each to-be-standardized sorting mode that meets the specified convergence condition, the probability density function of each to-be-standardized sorting mode is determined based on the selection reference information corresponding to each to-be-standardized sorting mode, and then the to-be-standardized mode selection probability of each to-be-standardized sorting mode is determined based on the probability density function corresponding to each to-be-standardized sorting mode. The maximum value of the probability density function can be taken as the to-be-standardized mode selection probability, or a random value can be taken as the to-be-standardized mode selection probability.

[0122] For example, there are two to-be-standardized sorting patterns M and N that meet the specified convergence conditions. The first parameter and the second parameter of the selection reference information of M are 100 and 50 respectively, the Beta distribution function is B(100,50), and the probability density function of M is The probability of selection to be standardized is 0.77, the first parameter and the second parameter of the reference information for selecting N are 30 and 100 respectively, the Beta distribution function is B(30, 100), and the probability density function of N is The probability of selection to be standardized is 0.24, and the graphs of the probability density functions of M and N can be shown as Figure 7 shown.

[0123] 603 , determining the to-be-standardized sorting mode corresponding to the maximum to-be-standardized selection probability as the standardized sorting mode corresponding to the keyword.

[0124] Compare the selection probabilities to be standardized of multiple probability density functions, determine the largest selection probability to be standardized, and determine the standardized sorting mode corresponding to the largest selection probability to be standardized as the standardized sorting mode corresponding to the keyword. For example, the two standardized sorting modes M and N that meet the specified convergence conditions have standardized selection probabilities of 0.77 and 0.24, respectively, so M is determined as the standardized sorting mode corresponding to the keyword.

[0125] If there is no to-be-standardized sorting pattern that meets the specified convergence condition when the judgment cycle is reached, the judgment result this time is that there is no to-be-standardized sorting pattern that can be selected as the standardized sorting pattern corresponding to the keyword.

[0126] A corresponding relationship table of standardized sorting modes corresponding to keywords can be established in advance. When there is a sorting mode to be standardized that meets the specified convergence condition, in the next cycle, the first parameter and the second parameter in the selection reference information corresponding to the other multiple sorting modes to be standardized are respectively set to the first specified value and the second specified value. When there is no sorting mode to be standardized that meets the specified convergence condition, in the next cycle, the first parameter and the second parameter in the selection reference information corresponding to the multiple sorting modes to be standardized are updated on the basis of the previous cycle, and do not need to be set to the initial value.

[0127] 604. When a second search request carrying a keyword is received and resource information of multiple resource types corresponding to the keyword is obtained, the resource information of the multiple resource types and the standardized sorting mode are sent to the sending end of the second search request, so that the sending end of the second search request displays the resource information of the multiple resource types according to the standardized sorting mode.

[0128] The second search request is a search request sent by an account in the high-traffic account set, and correspondingly, the first search request is a search request sent by an account in the low-traffic account set. A portion of accounts can be randomly selected from the total account set as the low-traffic account set, and the remaining accounts as the high-traffic account set. The accounts in the high-traffic account set and the accounts in the low-traffic account set constitute all the accounts of the application.

[0129] The resource information of multiple resource types and the standardized sorting mode are sent as search results to the sender of the second search request, and the sender of the second search request displays the resource information of multiple resource types according to the standardized sorting mode.

[0130] In the disclosed embodiment, based on the operations of different users on the search results of the keyword, the selection reference information corresponding to each candidate sorting mode corresponding to the keyword can be continuously updated. In each search process, based on the selection reference information, a sorting mode is selected from the candidate sorting modes to generate the search results. In this way, there is no need to determine the sorting mode through a sorting model, and thus, there is no need to train the model or label a large number of samples. It can be seen that this method does not require a long preparation period.

[0131] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0132] Based on the same technical concept, the embodiment of the present disclosure also provides a search processing device, such as Figure 8 As shown, the device comprises:

[0133] The receiving module 810 is used to receive a first search request carrying a keyword, and can specifically implement the processing function of the above step 201 and other implicit steps.

[0134] The acquisition module 820 is used to acquire multiple candidate sorting modes corresponding to the keyword when resource information of multiple resource types corresponding to the keyword is acquired, wherein the candidate sorting mode is used to indicate the display order between the multiple resource types. Specifically, the processing function of the above step 202 and other implicit steps can be realized.

[0135] The determination module 830 is used to determine the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode. Specifically, it can realize the processing function of the above step 203 and other implicit steps.

[0136] The selection module 840 is used to select a simulated sorting mode from a plurality of candidate sorting modes according to the mode selection probability of each candidate sorting mode, and can specifically implement the processing function of the above step 204 and other implicit steps.

[0137] The sending module 850 is used to send the resource information of multiple resource types and the target sorting mode to the sender of the first search request, so that the sender can display the resource information of multiple resource types according to the target sorting mode. Specifically, the processing function of the above step 205 and other implicit steps can be realized.

[0138] The updating module 860 is used to update the selection reference information corresponding to the target sorting mode based on the operation information returned by the sending end for the resource information of the displayed multiple resource types, and can specifically implement the processing function of the above step 206 and other implicit steps.

[0139] In a possible implementation, the acquisition module 820 is used to: if multiple candidate sorting modes corresponding to the keyword are stored, then the multiple candidate sorting modes corresponding to the stored keyword are acquired; if multiple candidate sorting modes corresponding to the keyword are not stored, then various possible sortings of multiple resource types are determined to acquire multiple candidate sorting modes corresponding to the keyword, and the multiple candidate sorting modes corresponding to the keyword are stored. Specifically, the processing function of the above step 202 and other implicit steps can be realized.

[0140] In a possible implementation, the selection reference information corresponding to the candidate sorting mode includes a first parameter and a second parameter, the initial value of the first parameter is a first specified value, and the initial value of the second parameter is a second specified value; the update module 860 is used to: when the operation information meets the specified condition, update the first parameter in the selection reference information corresponding to the target sorting mode; when the operation information does not meet the specified condition, update the second parameter in the selection reference information corresponding to the target sorting mode; wherein the specified condition includes: performing a specified operation on the resource information of the target resource type in the search results, the target resource type is the resource type ranked first in the search results, and the specified operation includes at least one of the completion rate reaching a threshold, collection, sharing, and downloading. Specifically, the processing function of the above step 206 and other implicit steps can be implemented.

[0141] In a possible implementation, the determination module 830 is used to: for each candidate sorting mode, determine the probability density function of the candidate sorting mode based on the selection reference information corresponding to the candidate sorting mode, and randomly determine the mode selection probability of the candidate sorting mode based on the probability density function; the selection module 840 is used to: select the candidate sorting mode corresponding to the largest mode selection probability among multiple candidate sorting modes as the target sorting mode. Specifically, the processing functions of the above steps 501 and 502 and other implicit steps can be implemented.

[0142] In a possible implementation, the determination module 830 is used to: determine the Beta distribution function of the selection reference information corresponding to the candidate sorting mode; and determine the probability density function of the candidate sorting mode based on the Beta distribution function. Specifically, the processing function of the above step 501 and other implicit steps can be implemented.

[0143] In a possible implementation, the first search request is a search request sent by an account in a small flow account set, and the method further includes a second search module 870, which is used to: when the judgment cycle is reached, determine the to-be-standardized sorting mode that meets the specified convergence condition from the multiple candidate sorting modes corresponding to the keyword; determine the to-be-standardized mode selection probability of each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode; determine the to-be-standardized sorting mode corresponding to the largest to-be-standardized mode selection probability as the standardized sorting mode corresponding to the keyword; when receiving a second search request carrying a keyword and obtaining resource information of multiple resource types corresponding to the keyword, send the resource information of multiple resource types and the standardized sorting mode to the sender of the second search request, so that the sender of the second search request displays the resource information of multiple resource types according to the standardized sorting mode; the second search request is a search request sent by an account in a large flow account set. Specifically, the processing functions of the above steps 601, 602, 603, and 604, as well as other implicit steps, can be implemented.

[0144] In a possible implementation, the determination module 830 is used to: determine the probability density function of each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode; determine the to-be-standardized mode selection probability of each to-be-standardized sorting mode based on the probability density function corresponding to each to-be-standardized sorting mode. Specifically, the processing function of the above step 602 and other implicit steps can be implemented.

[0145] In a possible implementation, the specified convergence condition is that the number of times the candidate sorting mode is selected reaches a number threshold. Specifically, the processing function of the above step 601 and other implicit steps can be implemented.

[0146] The above-mentioned receiving module 810, acquiring module 820, determining module 830, selecting module 840, sending module 850, updating module 860, and second searching module 870 may be implemented by a processor, or by a processor in conjunction with a memory and a display.

[0147] In the disclosed embodiment, based on the operations of different users on the search results of the keyword, the selection reference information corresponding to each candidate sorting mode corresponding to the keyword can be continuously updated. In each search process, based on the selection reference information, a sorting mode is selected from the candidate sorting modes to generate the search results. In this way, there is no need to determine the sorting mode through a sorting model, and thus, there is no need to train the model or label a large number of samples. It can be seen that this method does not require a long preparation period.

[0148] When the search processing device provided in the above embodiment is running, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the electric device is divided into different functional modules to complete all or part of the functions described above. In addition, the search processing device provided in the above embodiment and the method embodiment of the search processing belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0149] The embodiment of the present disclosure also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that can be stored by a computer device or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk). The computer-readable storage medium includes instructions that instruct the computing device to perform a search processing method.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for search processing, characterized in that: The method comprises: receiving a first search request carrying a keyword; When resource information of multiple resource types corresponding to the keyword is obtained, multiple candidate sorting modes corresponding to the keyword are obtained, wherein the candidate sorting modes are used to indicate a display order between the multiple resource types; Determine the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode; Selecting a target sorting mode from the plurality of candidate sorting modes according to the mode selection probability of each candidate sorting mode; Sending the resource information of the multiple resource types and the target ranking mode to a sending end of the first search request, so that the sending end displays the resource information of the multiple resource types according to the target ranking mode; When operation information for the displayed resource information of the multiple resource types returned by the sending end is obtained, the selection reference information corresponding to the target sorting mode is updated based on the operation information.

2. The method according to claim 1, characterized in that The obtaining of multiple candidate ranking modes corresponding to the keyword includes: If multiple candidate sorting modes corresponding to the keyword have been stored, obtaining the multiple candidate sorting modes corresponding to the keyword; If multiple candidate sorting modes corresponding to the keyword are not stored, various possible sortings of the multiple resource types are determined to obtain multiple candidate sorting modes corresponding to the keyword, and the multiple candidate sorting modes corresponding to the keyword are stored.

3. The method according to claim 1, characterized in that The selection reference information corresponding to the candidate sorting mode includes a first parameter and a second parameter, the initial value of the first parameter is a first specified value, and the initial value of the second parameter is a second specified value; The updating of the selection reference information corresponding to the target sorting mode based on the operation information includes: When the operation information satisfies a specified condition, the first parameter in the selection reference information corresponding to the target sorting mode is updated; when the operation information does not satisfy the specified condition, the second parameter in the selection reference information corresponding to the target sorting mode is updated; Among them, the specified conditions include: performing a specified operation on the resource information of the target resource type in the search results, the target resource type is the resource type ranked first in the search results, and the specified operation includes at least one of the completion rate reaching a threshold, collection, sharing, and downloading.

4. The method according to claim 1, characterized in that The determining of the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode includes: For each candidate sorting mode, determine a probability density function of the candidate sorting mode based on selection reference information corresponding to the candidate sorting mode, and randomly determine a mode selection probability of the candidate sorting mode based on the probability density function; The selecting a target sorting mode from the plurality of candidate sorting modes according to the respective mode selection probabilities of each candidate sorting mode includes: Among the multiple candidate sorting modes, the candidate sorting mode corresponding to the largest mode selection probability is selected as the target sorting mode.

5. The method according to any one of claim 4, characterized in that: The determining, based on the selection reference information corresponding to the candidate sorting mode, the probability density function of the candidate sorting mode comprises: Determine a beta distribution function of the selection reference information corresponding to the candidate sorting mode; A probability density function of the candidate sorting pattern is determined based on the Beta distribution function.

6. The method according to claim 1, characterized in that The first search request is a search request sent by an account in a low-traffic account set, and the method further includes: When the judgment period is reached, determining a to-be-standardized sorting mode that satisfies a specified convergence condition from a plurality of candidate sorting modes corresponding to the keyword; Determine the selection probability of the to-be-standardized mode for each to-be-standardized sorting mode based on the selection reference information corresponding to each to-be-standardized sorting mode; Determine the to-be-standardized sorting pattern corresponding to the largest to-be-standardized pattern selection probability as the standardized sorting pattern corresponding to the keyword; When a second search request carrying the keyword is received and resource information of multiple resource types corresponding to the keyword is obtained, the resource information of the multiple resource types and the standardized sorting mode are sent to the sender of the second search request, so that the sender of the second search request displays the resource information of the multiple resource types according to the standardized sorting mode; the second search request is a search request sent by an account in a large-traffic account set.

7. The method according to claim 6, characterized in that The determining, based on the selection reference information corresponding to each of the to-be-standardized sorting modes, the to-be-standardized mode selection probability of each to-be-standardized sorting mode includes: Determine the probability density function of each of the to-be-standardized sorting modes based on the selection reference information corresponding to each of the to-be-standardized sorting modes; Based on the probability density function corresponding to each to-be-standardized sorting mode, the to-be-standardized mode selection probability of each to-be-standardized sorting mode is determined.

8. The method according to claim 6, characterized in that The specified convergence condition is that the number of times the candidate sorting mode is selected reaches a number threshold.

9. A search processing device, characterized in that: The device comprises: A receiving module, configured to receive a first search request carrying a keyword; An acquisition module, configured to acquire, when resource information of multiple resource types corresponding to the keyword is acquired, multiple candidate sorting modes corresponding to the keyword, wherein the candidate sorting modes are used to indicate a display order between the multiple resource types; A determination module, used to determine the mode selection probability of each candidate sorting mode based on the selection reference information corresponding to each candidate sorting mode; A selection module, configured to select a simulated sorting mode from among the plurality of candidate sorting modes according to the mode selection probability of each candidate sorting mode; A sending module, used for sending the resource information of the multiple resource types and the target sorting mode to a sending end of the first search request, so that the sending end displays the resource information of the multiple resource types according to the target sorting mode; The updating module is used to update the selection reference information corresponding to the target sorting mode based on the operation information returned by the sending end for the resource information of the multiple resource types displayed.

10. A computer device, characterized in that: The computer device comprises a memory and a processor, wherein the memory is used to store computer instructions; The processor executes the computer instructions stored in the memory to enable the computer device to perform the method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program code. When the computer program code is executed by a computer device, the computer device executes the method according to any one of claims 1 to 8.

12. A computer program product, characterized in that The computer program product comprises a computer program code. When the computer program code is executed by a computer device, the computer device executes the method according to any one of claims 1 to 8.

Images