A method and device for displaying historical search information

Abstract

The specification discloses a method and device for displaying historical search information. For each historical search information of a target user, input data is determined and input into a trained prediction model according to the historical search information and behavior data of the target user on the search result of the historical search information, and a predicted click rate of the historical search information is determined. Each historical search information is sorted according to the corresponding predicted click rate, and at least part of the historical search information is displayed in a historical search module of a search interface according to the sorting result. The predicted click rate can be output by the prediction model, and the historical search information can be sorted and displayed based on the predicted click rate. Since the predicted click rate is obtained by inputting the input data determined based on the historical search information and the behavior data of the target user into the prediction model, the historical search information is more reasonably sorted and displayed based on the search time of the historical search information.

Description

Technical Field

[0001] This specification relates to the field of search technology, and in particular to a method and apparatus for displaying historical search information. Background Technology

[0002] Historical search information is important data that reflects user preferences and recent search intentions. For clients with search functions, the search interface corresponding to the search entry usually has a historical search module to display the historical search information that the user recently entered through the search entry.

[0003] Displaying historical search information can reduce user input and improve search efficiency, while also stimulating new user intent. Since the information that can be displayed in the historical search module is limited, the current method is to sort and display historical search results based on the search time.

[0004] However, sorting and displaying historical search information based on time is too one-sided and unreasonable, making it difficult for the displayed historical search information to match the user's current search intent or stimulate the user's new search intent. Summary of the Invention

[0005] This specification provides a method and apparatus for displaying historical search information, in order to partially solve the aforementioned problems existing in the prior art.

[0006] The following technical solution is adopted in this specification:

[0007] This manual provides a method for displaying historical search information, including:

[0008] Obtain historical search information for the target user;

[0009] For each historical search record, input data is determined based on that historical search record and the target user's behavioral data regarding the search results of that historical search record;

[0010] The input data is fed into the trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model; wherein, the prediction model is trained based on: training samples determined by the historical search information input by the test user and the test user's behavior data on the search results of the historical search information, and the labels of the training samples determined by the test user's operation on the historical search information displayed by the historical search module;

[0011] Based on the predicted click-through rate corresponding to each historical search result, sort the historical search results.

[0012] Based on the sorting results, at least some of the historical search information is displayed in the historical search module of the search interface.

[0013] Optionally, obtain the target user's historical search information, specifically including:

[0014] In response to a target user's click on the global search entry point of the client, the system obtains the target user's historical search information entered at the global search entry point of the client and the target user's historical search information entered at the business search entry points corresponding to each business dimension of the client.

[0015] Optionally, for each historical search record, input data is determined based on the historical search record and the target user's behavioral data regarding the search results of that historical search record, specifically including:

[0016] The system determines that historical search information containing sensitive words in each of the target user's historical search information is invalid; and / or, based on each of the target user's historical search information, it determines the search results corresponding to each historical search information, and determines that historical search information without corresponding search results is invalid.

[0017] After deleting the invalid information from each of the historical search records, for each historical search record, input data is determined based on the historical search record and the target user's behavior data regarding the search results of that historical search record.

[0018] Optionally, for each historical search record, input data is determined based on the historical search record and the target user's behavioral data regarding the search results of that historical search record, specifically including:

[0019] Determine the user information of the target user;

[0020] For each historical search record, behavioral characteristics of a specified type of behavioral data are determined based on the target user's behavioral data of the search results corresponding to that historical search record.

[0021] Input data is determined based on the user information, the historical search information, and the behavioral characteristics.

[0022] Optionally, the method further includes:

[0023] For each historical search result, the target user's level of attention to that historical search result is determined based on the target user's behavioral data regarding the search results for that historical search result.

[0024] Based on the user profile of the target user, determine the business dimension that matches the target user;

[0025] The weight of each historical search result is determined based on the business dimension corresponding to each historical search result and the business dimension matched with the target user.

[0026] The weighted attention level of each historical search result is determined based on its popularity and weight.

[0027] Based on the weighted attention of each historical search result, the historical search results are sorted, and at least some of the historical search results are displayed in the historical search module of the search interface according to the sorting results.

[0028] Optionally, based on the sorting results, at least some historical search information can be displayed in the historical search module of the search interface, specifically including:

[0029] Determine the business dimensions corresponding to each historical search result;

[0030] Based on the sorting results, at least some historical search information and its business dimensions are displayed in the historical search module of the search interface.

[0031] Optionally, the method further includes:

[0032] In response to the target user's click operation on the historical search information in the historical search module, the historical search information clicked by the target user is determined as the target information;

[0033] When the historical search module displays a business dimension corresponding to the target information, it determines the resource associated with the business dimension corresponding to the target information, and performs a search operation based on the target information and the resource to determine the search results from the resource.

[0034] When the business dimension corresponding to the target information is not displayed in the historical search module, the resources associated with each business dimension are determined, and a search operation is performed based on the target information and the resources associated with each business dimension to determine the search results from the resources associated with each business dimension.

[0035] This manual provides a device for displaying historical search information, including:

[0036] The acquisition module is used to acquire historical search information corresponding to the target user;

[0037] The input determination module is used to determine the input data for each historical search information based on the historical search information and the target user's behavior data regarding the search results of that historical search information.

[0038] An input module is used to input the input data into a trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model; wherein, the prediction model is trained based on: training samples determined by the historical search information input by the test user and the test user's behavior data on the search results of the historical search information, and the labels of the training samples determined by the test user's operation on the historical search information displayed by the historical search module;

[0039] The sorting module is used to sort historical search information based on the predicted click-through rate corresponding to each historical search information.

[0040] The display module is used to display at least a portion of the historical search information in the historical search module of the search interface, based on the sorting results.

[0041] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for displaying historical search information.

[0042] This specification provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-mentioned method for displaying historical search information.

[0043] The above-mentioned technical solutions adopted in this specification can achieve the following beneficial effects:

[0044] In the method for displaying historical search information provided in this specification, the following steps are taken: First, obtain the historical search information corresponding to the target user. For each historical search information, determine the input data based on that historical search information and the target user's behavioral data regarding the search results for that historical search information. Then, input the input data into a trained prediction model to determine the predicted click-through rate (CTR) of that historical search information. Based on the predicted CTR of each historical search information, sort the historical search information. Finally, based on the sorting results, display at least a portion of the historical search information in the historical search module of the search interface.

[0045] As can be seen from the above method, this method can output the predicted click-through rate (CTR) of the target user's historical search information based on the prediction model, and then sort and display the historical search information based on the predicted CTR. Since the predicted CTR is determined by the input data based on historical search information and the target user's behavioral data on the search results of that historical search information, and is obtained through the prediction model, it is more reasonable than simply sorting and displaying historical search information based on the time when the user searched for historical search information. Attached Figure Description

[0046] The accompanying drawings, which are included to provide a further understanding of this specification and form part of this specification, illustrate exemplary embodiments and are used to explain this specification, but do not constitute an undue limitation thereof. In the drawings:

[0047] Figure 1 This is a flowchart illustrating a method for displaying historical search information as described in this specification.

[0048] Figure 2 A schematic diagram of a search interface provided in this specification;

[0049] Figure 3 A schematic diagram of a device for displaying historical search information provided in this specification;

[0050] Figure 4 This is a schematic diagram of the structure of an electronic device provided in this specification. Detailed Implementation

[0051] Currently, most business platform clients provide search entry points for users to search. For example, in shopping platform clients, users can enter keywords through the search entry point, and the client can then perform a search to identify products corresponding to those keywords. In video platform clients, users can enter keywords through the search entry point, and the client can then perform a search to identify videos corresponding to those keywords. In food delivery platform clients, users can enter keywords through the search entry point, and the client can then perform a search to identify meals, fruits, shops, etc., corresponding to those keywords.

[0052] The aforementioned products, videos, food items, shops, fruits, etc., constitute the search results. Keywords entered by the user during the search will then be displayed as historical search information on the search interface—that is, previously entered search information. Of course, keywords are just examples; the search information entered by the user is not limited to text; for example, it can also include images.

[0053] Currently, client-side search interfaces typically include a "History Search" section that displays a user's recent search data. Users can initiate a search by clicking on historical search information in this section, eliminating the need for manual input and reducing the amount of typing required.

[0054] However, users typically have a large amount of historical search information, and it may not be possible to display all of the user's historical search information in the historical search module. Therefore, it is necessary to determine which part of the historical search information to display.

[0055] However, existing methods, when displaying historical search information, only sort and display the historical search information in reverse chronological order, placing the user's most recent historical search information at the top.

[0056] However, sorting historical search information solely based on time is too one-sided and unreasonable. A user's most recent search history is very likely driven by temporary interests or even accidental clicks, and cannot represent the user's recent interests, current search intent, or long-term stable search intent. As a result, the historical search information displayed based on the sorting results has low or no relevance to the user's current search intent, and is unlikely to stimulate new search intents from the user.

[0057] To at least partially address the aforementioned problems, this specification provides a method for displaying historical search information. Based on this method, the predicted click-through rate (CTR) of each historical search result can be determined using a prediction model based on user behavior data regarding search results for each historical search result. The historical search results are then sorted and displayed based on the predicted CTR. For each historical search result, the predicted CTR is defined as the probability that a target user would click on that historical search result if it were displayed in the historical search module.

[0058] Since user behavior data on search results of various historical search information can comprehensively reflect users' preferences and interest in different historical search information, and the predicted click-through rate is determined based on users' historical behavior data, it is more reasonable to sort historical search information based on predicted click-through rate than to sort based on time. The historical search information displayed based on the sorting results is more in line with the user's search intent or can better stimulate the user's search intent.

[0059] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0060] The technical solutions provided in the various embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0061] Figure 1 This is a flowchart illustrating a method for displaying historical search information as described in this specification, specifically including the following steps:

[0062] S100: Obtain historical search information corresponding to the target user.

[0063] In this specification, the method for displaying the historical search information can be executed by the client.

[0064] In one or more embodiments of this specification, a user can open the search interface corresponding to the search entry point by operating the client. The client can then display historical search information in the historical search module of the search interface based on the sorting results of the user's historical search information. This operation specifically refers to the user's click operation on the search entry point. A user performs at least two operations on the search entry point for each search: the first is a click operation, where the user opens the search interface by clicking the search entry point. Subsequently, the user can trigger a search operation by performing an input operation on the search interface, and the client can then respond to the user's search operation by performing search services.

[0065] This input operation can include the user manually entering search information at the search entry point, as well as the user performing the input operation by clicking on historical search information in the historical search module.

[0066] It should be noted that, in this specification, after the user first interacts with the search entry point of the client, that is, after the user clicks on the search entry point, the client responds to the user's click on the search entry point by executing steps S100 to S106, determining the predicted click-through rate of each historical search information and determining the sorting result based on the predicted click-through rate, and then executing step S108, which determines and displays the historical search information to the user in the historical search module of the search interface of the search entry point clicked by the user, based on the sorting result.

[0067] Alternatively, the client can also execute steps S100-S106 after the user performs an input operation at the search entry point to pre-determine and sort the predicted click-through rates of each of the user's historical search information. Then, when the user clicks the client's search entry point again to perform a search, in response to the user's click operation, based on the sorting results obtained from steps S100-S106, step S108 is executed to determine and display the historical search information to the user in the historical search module of the search interface at the user's clicked search entry point.

[0068] In this specification, for ease of description, the users whose historical search information is to be determined and whose historical search information is to be displayed are referred to as target users.

[0069] First, the client can obtain the target user's historical search information.

[0070] S102: For each historical search information, determine the input data based on the historical search information and the target user's behavior data regarding the search results of that historical search information.

[0071] Once the target user's historical search information is determined, the client can identify the search results corresponding to each historical search, as well as the behavioral data of the target user corresponding to the search results of each historical search.

[0072] Then, for each historical search result, the client can determine the input data for the prediction model based on that historical search result and the target user's behavioral data regarding the search results of that historical search result.

[0073] For each search result, the behavioral data corresponding to the target user and that search result is included; that is, the data related to the target user's behavior towards the search result. The behavioral data may include at least one of the following: whether the target user interacted with the search result, the duration of the target user's browsing of the search result's details page, whether the app crashed after the target user clicked on the search result, and the specific actions taken by the target user towards the search result.

[0074] Whether the target user interacts with the search result can specifically include at least one of the following: whether the target user clicks on the search result, saves the search result, comments on the search result, or places an order for the search result. The specific content of the target user's interaction with the search result can be a comment on the search result.

[0075] Whether the app crashes can be determined based on whether the interval between the time it takes for the target user to enter the search result details page and the time it takes to leave the search result details page is less than a preset time difference. If this interval is less than the preset time difference, the client can determine that the app crashed after the target user clicked the search result; otherwise, it can determine that the app did not crash after the target user clicked the search result.

[0076] S104: Input the input data into the trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model; wherein, the prediction model is trained based on: training samples determined by the historical search information input by the test user and the test user's behavior data on the search results of the historical search information, and the labels of the training samples determined by the test user's operation on the historical search information displayed by the historical search module.

[0077] After determining the input data, the client can input the determined input data into the trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model.

[0078] The prediction model is trained based on training samples determined by historical search information input by test users and behavioral data of test users on the search results of historical search information, as well as the labels of training samples determined by test users' operations on historical search information displayed by the historical search module.

[0079] For each historical search result displayed in the test user's historical search module, the test user's action on that historical search result includes either clicking or not clicking. The training process for the prediction model will not be described here.

[0080] S106: Sort the historical search information according to the predicted click-through rate corresponding to each historical search information.

[0081] Once the predicted click-through rate (CTR) for each historical search result of the target user is determined, the historical search results can be sorted based on the determined CTR.

[0082] S108: Based on the sorting results, display at least some of the historical search information in the historical search module of the search interface.

[0083] After sorting the target user's historical search information and obtaining the sorting results, the client can display at least a portion of the historical search information in the historical search module of the search interface based on the sorting results.

[0084] based on Figure 1 The method for displaying historical search information, as shown, involves acquiring each historical search result for the target user. For each historical search result, based on that historical search result and the target user's behavioral data regarding the search results for that historical search result, input data is determined. This input data is then fed into a trained prediction model to determine the predicted click-through rate (CTR) of that historical search result. The historical search results are then sorted according to their predicted CTRs. Finally, based on the sorting results, at least a portion of the historical search results are displayed in the historical search module of the search interface.

[0085] As can be seen from the above method, this method can output the predicted click-through rate (CTR) of the target user's historical search information based on the prediction model, and then sort and display the historical search information based on the predicted CTR. Since the predicted CTR is determined by the input data based on historical search information and the target user's behavioral data on the search results of that historical search information, and is obtained through the prediction model, it is more reasonable than simply sorting and displaying historical search information based on the time when the user searched for historical search information.

[0086] Additionally, during historical searches, users may have included sensitive words in their search results due to input method issues or typos. For sensitive words, the client cannot determine any search results. Similarly, searches containing non-sensitive words may also fail to yield corresponding results. Displaying these two types of historical search information to the user is meaningless; therefore, the client can delete these two types of historical search information and determine their predicted click-through rate.

[0087] Therefore, in step S102, when determining the input data for each historical search information based on the historical search information and the target user's behavioral data of the search results for that historical search information, specifically, the client can determine that historical search information containing sensitive words in each of the target user's historical search information is invalid information, and / or, based on each of the target user's historical search information, determine the search results corresponding to each historical search information, and determine that historical search information without corresponding search results is invalid information.

[0088] Afterwards, the identified invalid information can be deleted from each historical search record. Then, for each historical search record, the input data can be determined based on the historical search record and the target user's behavior data regarding the search results of that historical search record.

[0089] Of course, invalid information can be left undeleted, and this can be set as needed; this instruction manual does not impose any restrictions on this.

[0090] Furthermore, since user information reflects a user's stable preferences, and search intent resulting from stable preferences is usually stable, historical search information corresponding to a user's stable preferences is more likely to be clicked than historical search information corresponding to unstable preferences. Therefore, when determining input data in step S102, specifically, the client can also determine input data for each historical search information based on the target user's user information, the historical search information itself, and the behavioral data of the search results for that historical search information.

[0091] User information refers to a user's basic information, such as their place of origin, occupation, gender, and age. Users with different basic information usually have different preferences. Conversely, most users with similar basic information tend to have similar preferences.

[0092] In addition, user behavior regarding search results from historical search information typically includes both behaviors that clearly demonstrate the user's level of interest in historical search information and behaviors that do not demonstrate the user's level of interest in historical search information.

[0093] For example, when the client corresponds to a shopping platform or a food delivery platform, a long browsing time of the target user on the details page of a search result could indicate that the user is more interested in the search result and therefore browses the details page in detail. However, it could also mean that the user is not interested in the search result and is hesitant to place an order, thus trying to see if there is any information that attracts them, resulting in a longer browsing time. Therefore, the behavioral data corresponding to browsing behavior (the browsing time on the details page of a search result) may not necessarily clearly reflect the user's level of interest in the historical search information corresponding to the behavioral data.

[0094] Alternatively, the purpose of determining the predicted click-through rate (CTR) of historical search information is to identify and display historical search results that users are more interested in and more likely to click on. User behavior data includes data corresponding to positive and negative behaviors. Positive behaviors are actions taken by users when they are interested in historical search results, such as saving or placing an order. Negative behaviors are actions taken by users when they are not interested in historical search results, such as clicking on a search result and the app crashing.

[0095] Therefore, this client can also filter behavioral data from search results corresponding to each historical search of the target user. It filters out behavioral data corresponding to positive behaviors to determine the input data.

[0096] Therefore, in one or more embodiments of this specification, in step S102, the client may further determine the user information of the target user. And for each historical search record, behavioral data for a specified behavior is determined from the behavioral data of the search results corresponding to that historical search record. This behavioral data for the specified behavior may be behavioral data from historical search records that do not reflect the user's level of interest in the historical search records, and / or behavioral data corresponding to positive behaviors.

[0097] After determining the behavioral data for the specified behavior, the client can determine the input data based on the target user's user information, the historical search information, and the behavioral data for the specified behavior.

[0098] Alternatively, when determining the input data in step S102, the client can also determine the target user's user information to determine the behavioral characteristics of a specified type of behavioral data for each historical search information, based on the target user's behavioral data of the search results corresponding to that historical search information. The input data is then determined based on the user information, the historical search information, and the behavioral characteristics.

[0099] The specified type can be set as needed. For example, the specified type of behavioral data can be behavioral data corresponding to a user's comment behavior. This behavioral feature can be obtained by quantifying the comment behavior data (i.e., the comment content).

[0100] Alternatively, the specified type of behavioral data can be behavioral data reflecting the user's level of interest in historical search information. In this case, the behavioral characteristics of the determined specified type of behavioral data can be features obtained by quantifying whether the user is interested in the historical search information corresponding to the behavioral data.

[0101] Alternatively, in one or more embodiments of this specification, the client may not output the predicted click-through rate of historical search information based on the prediction model, nor may it rank the historical search information based on the predicted click-through rate. The client may also determine the target user's level of attention to each historical search item based on richer user information, i.e., user profiles, and determine the ranking result of each historical search item based on this level of attention.

[0102] User profiles of target users can reveal which historical search results they are interested in and which they are not. This specification defines the purpose of displaying historical search results in the historical search module not only to show those that are relevant to the user's search intent and of interest, but also to exclude those that are irrelevant or of little interest. Therefore, determining the target user's level of interest in each historical search result based on user profiles, and then ranking those results accordingly, is more comprehensive and reasonable.

[0103] Among them, attention level is positively correlated with the degree of interest users have in historical search information, and is also positively correlated with the probability that target users click on historical search information.

[0104] Therefore, in one or more embodiments of this specification, after the client obtains the historical search information corresponding to the target user in step S100, the client can determine the target user's attention to the historical search information based on the target user's behavioral data of the search results of the historical search information for each historical search information.

[0105] Then, the client can determine the business dimensions matching the target user based on the target user's user profile. It can then determine the weight of each historical search result based on the business dimensions corresponding to each historical search result and the business dimensions matching the target user.

[0106] Among them, the business dimensions for target user matching can be business dimensions that the target user shows interest in, determined based on the user profile of the target user, and / or business dimensions that the target user shows no interest in.

[0107] After determining the weight of each historical search result, the client can determine the weighted attention of each historical search result based on its attention level and weight.

[0108] Finally, the client can sort the historical search information according to the weighted attention of each historical search information, and display at least some of the historical search information in the historical search module of the search interface according to the sorting results.

[0109] In one or more embodiments of this specification, when determining the target user's attention to the historical search information based on the target user's behavioral data of the search results of the historical search information, the client can sum the behavioral data and / or behavioral features of each search result of the historical search information to obtain the attention level.

[0110] For example, behavioral data can take the following form: (whether it was clicked, whether it was saved, whether it was commented on, and the duration of browsing), which includes four behavioral dimensions. Assuming 1 represents yes and 0 represents no, the behavioral data (1, 2, 0, 3) of one of the search results in the historical search information indicates that the user clicked and saved the search result, browsed the search result for 3 seconds after clicking, but did not comment on the search result.

[0111] The client can sum the values ​​of the four behavioral dimensions to obtain the attention score of the search result, which would be 1 + 2 + 0 + 3 = 6. Since the value of the browsing time behavioral dimension is usually greater than the values ​​of the other behavioral dimensions, the values ​​of each behavioral dimension can also be weighted and summed to obtain the score of the search result.

[0112] The total attention score of a historical search result can be obtained by summing the attention scores corresponding to each search result. Of course, other methods can also be used to determine the attention score of a search result. For example, the attention score can be determined based on only some behavioral dimensions. For instance, if the behavioral dimension includes whether an order was placed, then the value of this behavioral dimension can be used as the attention score of the corresponding search result.

[0113] In addition, the client can also identify popular search terms based on each user's historical search information. Among a user's historical search data, target users may pay more attention to historical searches that are considered popular compared to those that are less popular.

[0114] Therefore, in one or more embodiments of this specification, the client may also determine the rating of each historical search information based on the target user's behavioral data of the search results of the historical search information, when determining the target user's attention to the historical search information based on the behavioral data of the target user's behavioral data of the search results of the historical search information.

[0115] Next, from the historical search information, the historical search information belonging to the popular search information is identified as popular information. Then, based on the preset weights corresponding to popular information, the scores of each popular search information are weighted to obtain the attention level. For non-popular information, its score can be used as the attention level.

[0116] Of course, in the embodiment of sorting historical search information based on the attention of historical search information, each piece of historical search information used to determine the attention can also be the historical search information after the determined invalid information has been deleted from each piece of historical search information.

[0117] In one or more embodiments provided in this specification, the client can provide services across multiple business dimensions and has multiple search entry points specifically configured. For example, it may include a global search entry point for searching resources across all business dimensions corresponding to the client, and a business search entry point for searching resources only corresponding to a specific business dimension. The search results are determined based on the resources.

[0118] For example, the business dimensions of a business platform may include: food delivery, grocery shopping, medicine purchase, movies, hotels, etc. Each business dimension can then correspond to a search entry point, i.e., a business search entry point.

[0119] Taking the search term "tomato" as an example, when a user enters "tomato" through the food delivery search portal, the client will only search for tomato-related results (such as tomato beef brisket) based on food delivery resources. When a user enters "tomato" through the grocery search portal, the client will only search for tomato-related results (such as raw tomatoes) based on grocery search resources.

[0120] Taking the search term "orange" as an example, when a user enters "orange" through the search entry point for food delivery, the client will only search for results related to oranges (such as "fruit orange") based on resources in the food delivery category. When a user enters "orange" through the search entry point for hotels, the client will only search for results related to oranges (such as "hotels named orange") based on resources in the hotel category.

[0121] When a user enters search information through the client's global search entry, the client can determine the search results from all business resources across all business dimensions of the corresponding business platform. Therefore, when a user enters "tomato," they can get search results for multiple business dimensions, such as "tomato beef brisket" and "raw tomato."

[0122] Currently, historical search information across different search entry points on various business platforms is not interconnected. When a user clicks on the global search entry point, the client only displays historical search information entered by the user at that global search entry point, while historical search information entered by the user at specific business search entry points is not displayed. This makes it impossible for users to find more comprehensive historical search results covering multiple business dimensions on the search interface corresponding to the global search entry point, making it difficult to determine historical search information that matches the current search intent from the historical search information displayed by the client.

[0123] To address this issue, in this specification, when a target user clicks on the search entry, the client responds to the target user's click on the client's search entry by executing steps S100 to S106 to determine the predicted click-through rate of each historical search result and determine the ranking result based on the predicted click-through rate. Then, when executing step S108, specifically, when obtaining the historical search information corresponding to the target user in step S100, the client can respond to the target user's click on the client's global search entry to obtain the historical search information entered by the target user in the client's global search entry, as well as the historical search information entered by the target user in the business search entries corresponding to each business dimension of the client.

[0124] Of course, since when a target user clicks on the business search entry, the target user usually wants to get search results under a specific business dimension, the client can respond to the target user's click on the client's business search entry and obtain the target user's historical search information entered in the business search entry.

[0125] When the client executes steps S100-S106 after the target user performs an input operation through the search entry, it pre-determines and sorts the predicted click-through rates of the user's historical search information. When the target user clicks the client's search entry again, in response to the target user's click operation, based on the sorting results obtained in steps S100-S106, in step S108, the client can, after the target user performs an input operation through the search entry, execute step S100 to obtain the target user's historical search information entered through the client's global search entry, as well as the historical search information entered through the business search entries corresponding to each business dimension of the client. The obtained historical search information is used as the first historical search information. The sorting results of each first historical search information are then determined through steps S102-S106.

[0126] Furthermore, after the target user performs an input operation through the search entry, the client can execute step S100 to obtain the target user's historical search information entered through the business search entry corresponding to each business dimension, which serves as the second historical search information for that business dimension. The ranking results of each second historical search information are then determined through steps S102 to S106.

[0127] When a target user clicks on a search entry in the client application, the client application can determine the search entry that the target user clicked.

[0128] When the search entry is a global search entry, the client can respond to the target user's click operation on the client's search entry and display at least a portion of the first historical search information in the historical search module of the search interface according to the sorting results of the first search information.

[0129] When the search entry is a business search entry, the client can respond to the target user's click operation on the client's business search entry and display at least a portion of the second historical search information in the historical search module of the search interface according to the sorting results of the second search information of the business search entry clicked by the user.

[0130] Furthermore, in step S102, the client can also determine the input data for each historical search information based on the historical search information, the corresponding business dimension, and the target user's behavioral data regarding the search results of that historical search information. Correspondingly, when training the prediction model, training samples can be determined for each historical search information based on the historical search information, the corresponding business dimension, and the target user's behavioral data regarding the search results of that historical search information. The tags are still determined by the test user's actions on the historical search information displayed by the historical search module.

[0131] The predictive model outputs the predicted click-through rate (CTR) of historical search information, which represents the probability that a user will click on that historical search result. This predicted CTR is then used to rank the historical search results to determine whether or not to display them.

[0132] In other words, the output of the prediction model affects whether historical search information is displayed in the user's historical search module.

[0133] For each training sample used to train the prediction model, the label for that training sample is obtained through testing. In this specification, the distinction between test users and target users is made solely for ease of description and differentiation. However, test users can also be target users, and they can also be considered target users during the label determination process.

[0134] Since the purpose of training the prediction model is to enable the prediction model to output the predicted click-through rate of users for historical search information, when determining the training samples for training the prediction model, for each test user, after determining the historical search information displayed by the historical search module corresponding to the test user, a training sample can be determined for each historical search information displayed by the historical search module corresponding to the test user, based on the behavioral data of the historical search information and the search results corresponding to the historical search information.

[0135] Alternatively, for each historical search result displayed by the historical search module corresponding to the test user, a training sample can be determined based on the historical search result, the corresponding business dimension, and the target user's behavioral data regarding the search results of that historical search result. Or, for each historical search result displayed by the historical search module corresponding to the test user, a training sample can be determined based on the test user's user information, the historical search result, and the behavioral characteristics of the specified type of behavioral data corresponding to that historical search result.

[0136] For each training sample, the test user may or may not click on the historical search information within that training sample. Based on whether the test user clicked on the historical search information of that training sample, we can assess whether it is reasonable to display the historical search information of that training sample in the user's search history module. Since the output of the prediction model directly affects whether historical search information is displayed in the user's search history module, whether the test user clicked on the historical search information of that training sample can be used as a label for that training sample.

[0137] Based on the identified training samples and their labels, the prediction model is trained with the goal of producing a higher predicted click-through rate (CTR) for historical search information displayed in the user's historical information module and clicked by the user, and a lower predicted CTR for historical search information displayed in the user's historical information module but not clicked by the user. This allows the prediction model to accurately output the predicted CTR based on the input historical search information and the behavioral data corresponding to the search results.

[0138] Because the search history displayed in a user's search history module changes, a single search result may appear multiple times, potentially leading to multiple clicks by the user. Therefore, the number of clicks on different search results displayed in the user's search history module will vary. Even with the same number of clicks, the predicted click-through rate (CTR) output by the prediction model should be higher for search results with a higher CTR than for search results with a lower CTR.

[0139] Therefore, when determining the labels for training samples, different labels can be assigned to training samples corresponding to historical search information that the user did not click, and to training samples corresponding to historical search information that the user clicked. Similarly, different labels can also be assigned to training samples corresponding to historical search information with varying numbers of clicks.

[0140] For example, a preset range of clicks can be used, with different labels corresponding to different ranges. For instance, if the number of times a user clicks on a historical search result is in the first range, the label of the training sample corresponding to that historical search result is determined to be 0.5. If the number of times a user clicks on a historical search result is in the second range, the label of the training sample corresponding to that historical search result is determined to be 1. The maximum value in the first range is less than the minimum value in the second range.

[0141] The label for the training samples corresponding to historical search information that the user did not click on can be 0.

[0142] After identifying the training samples and their labels, each training sample can be input into the prediction model to be trained to obtain the predicted click-through rate (CTR) of historical search information within that training sample. Based on the predicted CTR of historical search information in that training sample and its label, the loss for that training sample can be determined. The prediction model can then be trained with the goal of minimizing this loss.

[0143] There are no restrictions on the loss function that should be used to determine the loss; for example, the cross-entropy loss function can be applied.

[0144] It should be noted that the training process of this prediction model can be performed by the server. After the prediction model is trained, the client can configure the trained prediction model to output the predicted click-through rate of historical search information based on the prediction model.

[0145] Furthermore, in step S108 of this specification, when displaying at least some historical search information in the historical search module of the search interface according to the sorting results, specifically, the client can determine the size of each historical search information. And according to the sorting results, based on the remaining space in the historical search module of the search interface and the size of the historical search information, it sequentially determines whether each historical search information can be displayed in the historical search module, until it is determined that there is no remaining space in the historical search module, or that there is no undetermined historical search information.

[0146] When the historical search information is a string, its size can be the string length and font size. When the historical search information is an image, its size is the image width and height.

[0147] In one or more embodiments of this specification, since a user's input at the global search entry point may correspond to multiple business dimensions, the client can also determine the business dimensions corresponding to each historical search information in step S108. Based on the sorting results, at least a portion of the historical search information and its business dimensions are displayed in the historical search module of the search interface. Therefore, for the input "tomato," the historical search module can display tomatoes with the "takeout" business dimension and tomatoes with the "grocery shopping" dimension. Figure 2 As shown.

[0148] Figure 2 This diagram illustrates a search interface provided in this manual. As shown, the interface displays a search box and a search history module from top to bottom. The search box is the search entry point. Users can enter search information in the search box and click "Search" on the right side of the search box to initiate a search.

[0149] Figure 2As can be seen, the "Historical Search" module displays historical search information corresponding to different business dimensions. Circles represent the food delivery dimension, diamonds represent the grocery delivery dimension, and triangles represent the hotel dimension. Therefore, for the historical search information "tomato," since the target user searched for "tomato" in both the food delivery and grocery delivery dimensions, the current display shows historical search information for "tomato" in both the food delivery and grocery delivery dimensions.

[0150] In the food delivery category, the tomato icon is linked to the corresponding business interface. When a user clicks on the tomato icon in the food delivery category, the client can search for tomato-related results in the food delivery resources and display the search results on the corresponding business interface, such as "tomato fish" or "tomato beef brisket." Similarly, in the grocery shopping category, the tomato icon is linked to the corresponding business interface. When a user clicks on the tomato icon in the grocery shopping category, the client can search for tomato-related results in the grocery shopping resources and display the search results on the corresponding business interface, such as "cherry tomatoes" or "tomatoes."

[0151] Figure 2 In the hotel dimension, two historical search results were displayed: "homestay" and "hotel". Besides tomatoes, the food delivery dimension also displayed two historical search results: "fish-flavored shredded pork" and "barbecue". In addition to tomatoes, the grocery delivery dimension also displayed a historical search result for "potatoes".

[0152] In one or more embodiments of this specification, when a user clicks on the historical search information in the historical search module, the client can also respond to the target user's click operation on the historical search information in the historical search module and determine the historical search information clicked by the target user as the target information.

[0153] When the historical search module displays the business dimension corresponding to the target information, a search can be performed based on the target information and its business dimension to obtain search results under that business dimension. In other words, when the historical search module displays the business dimension corresponding to the target information, the client can determine the resources associated with that business dimension, and then perform a search based on the target information and those resources to obtain search results from those resources.

[0154] When the historical search module does not display the business dimension corresponding to the target information, the client can perform a search based on the target information to obtain search results under each business dimension. That is, when the historical search module does not display the business dimension corresponding to the target information, the resources associated with each business dimension are determined, and a search is performed based on the target information and the resources associated with each business dimension to determine the search results from the resources associated with each business dimension.

[0155] Additionally, it should be noted that as time passes, the target user's historical search information accumulates and gradually becomes enormous, potentially several times the amount of historical search information that the historical search module can display. To reduce computational load, when acquiring historical search information in step S100, a predetermined number of historical search information entries can be retrieved.

[0156] Specifically, a specified number of historical search results can be retrieved in reverse chronological order based on the search time of all historical search results of the target user.

[0157] Alternatively, all historical search information can be sorted based on the number of search results for all historical search information of the target user, so as to obtain a specified number of historical search information based on the sorting results.

[0158] Alternatively, based on the predicted click-through rate of historical search information or the statistically obtained click-through rate, all historical search information can be sorted to obtain a specified number of historical search information based on the sorting results.

[0159] The above describes one or more embodiments of a method for displaying historical search information provided in this specification. Based on the same idea, this specification also provides a corresponding device for displaying historical search information, such as... Figure 3 As shown.

[0160] Figure 3 This is a schematic diagram of a device for displaying historical search information provided in this specification. The device includes:

[0161] The acquisition module 200 is used to acquire historical search information corresponding to the target user.

[0162] The input determination module 201 is used to determine input data for each historical search information based on the historical search information and the target user's behavior data of the search results of the historical search information.

[0163] Input module 202 is used to input the input data into the trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model; wherein, the prediction model is trained based on: training samples determined by the historical search information input by the test user and the test user's behavior data on the search results of the historical search information, and the labels of the training samples determined by the test user's operation on the historical search information displayed by the historical search module;

[0164] The sorting module 203 is used to sort the historical search information according to the predicted click-through rate corresponding to each historical search information.

[0165] The display module 204 is used to display at least a portion of the historical search information in the historical search module of the search interface based on the sorting results.

[0166] Optionally, the acquisition module 200 is further configured to, in response to the target user's click operation on the global search entry of the client, acquire the target user's historical search information entered in the global search entry of the client and the target user's historical search information entered in the business search entry corresponding to each business dimension of the client.

[0167] Optionally, the input determination module 201 is further configured to determine that among the target user's historical search information, there are historical search information containing sensitive words, which are considered invalid information; and / or, based on the target user's historical search information, determine the search results corresponding to each historical search information, determine that historical search information without corresponding search results is considered invalid information, delete the invalid information from each historical search information, and then, for each historical search information, determine the input data based on the historical search information and the target user's behavioral data regarding the search results of that historical search information.

[0168] Optionally, the input determination module 201 is further configured to determine the user information of the target user, and for each historical search information, determine the behavioral characteristics of a specified type of behavioral data based on the behavioral data of the target user of the search results corresponding to the historical search information, and determine the input data based on the user information, the historical search information and the behavioral characteristics.

[0169] Optionally, the device further includes:

[0170] The attention determination module 205 is used to determine the target user's attention to each historical search information based on the target user's behavioral data of the search results of that historical search information; determine the business dimension matched by the target user based on the user profile of the target user; determine the weight of each historical search information based on the business dimension corresponding to each historical search information and the business dimension matched by the target user; determine the weighted attention of each historical search information based on its attention and weight; sort the historical search information based on its weighted attention; and display at least a portion of the historical search information in the historical search module of the search interface based on the sorting results.

[0171] Optionally, the display module 204 is used to determine the business dimension corresponding to each historical search information, and display at least some of the historical search information and its business dimension in the historical search module of the search interface according to the sorting result.

[0172] Optionally, the device further includes:

[0173] Search module 206 is configured to respond to a target user's click operation on historical search information in the historical search module, determine the historical search information clicked by the target user as target information, and when the historical search module displays a business dimension corresponding to the target information, determine the resources associated with the business dimension corresponding to the target information, and perform search operations based on the target information and the resources to determine search results from the resources. When the historical search module does not display a business dimension corresponding to the target information, determine the resources associated with each business dimension, and perform search operations based on the target information and the resources associated with each business dimension to determine search results from the resources associated with each business dimension.

[0174] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 The method for displaying historical search information.

[0175] This instruction manual also provides Figure 4 The diagram shows the structure of the electronic device. Figure 4 As shown, at the hardware level, this electronic device includes a processor, an internal bus, memory, and non-volatile storage, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile storage into memory and then executes it to achieve the above-mentioned functions. Figure 1 The method for displaying historical search information.

[0176] It should be noted that all actions involving the acquisition of signals, information, or data in this manual are performed in accordance with the relevant data protection laws and regulations of the country where the device is located, and with the authorization of the owner of the relevant device.

[0177] Of course, in addition to software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software. In other words, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0178] In the 1990s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many methodological improvements today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that a methodological improvement cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program and "integrate" a digital system onto a PLD themselves, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0179] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0180] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0181] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware components.

[0182] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0183] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0184] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0185] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0186] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0187] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0188] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0189] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0190] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0191] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0192] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0193] The above description is merely an embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for displaying historical search information, characterized in that, include: Obtain historical search information for the target user; For each historical search record, input data is determined based on that historical search record and the target user's behavioral data regarding the search results of that historical search record; The input data is fed into the trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model; wherein, the prediction model is trained based on: training samples determined by the historical search information input by the test user and the test user's behavior data on the search results of the historical search information, and the labels of the training samples determined by the test user's operation on the historical search information displayed by the historical search module; Based on the predicted click-through rate corresponding to each historical search result, sort the historical search results. Based on the sorting results, at least some of the historical search information is displayed in the historical search module of the search interface; The method further includes: For each historical search result, the target user's level of attention to that historical search result is determined based on the target user's behavioral data regarding the search results for that historical search result. Based on the user profile of the target user, determine the business dimension that matches the target user; The weight of each historical search result is determined based on the business dimension corresponding to each historical search result and the business dimension matched with the target user. The weighted attention level of each historical search result is determined based on its popularity and weight. Based on the weighted attention of each historical search result, the historical search results are sorted, and at least some of the historical search results are displayed in the historical search module of the search interface according to the sorting results.

2. The method as described in claim 1, characterized in that, Obtain the target user's historical search information, specifically including: In response to a target user's click on the global search entry point of the client, the system obtains the target user's historical search information entered at the global search entry point of the client and the target user's historical search information entered at the business search entry points corresponding to each business dimension of the client.

3. The method as described in claim 1 or 2, characterized in that, For each historical search record, based on that historical search record and the target user's behavioral data regarding the search results for that historical search record, the input data is determined, specifically including: The historical search information of the target user containing sensitive words is identified as invalid information. And / or, based on the target user's historical search information, determine the search results corresponding to each historical search information, and determine that historical search information for which no corresponding search results exist is invalid information; After deleting the invalid information from each of the historical search records, for each historical search record, input data is determined based on the historical search record and the target user's behavior data regarding the search results of that historical search record.

4. The method as described in claim 1, characterized in that, For each historical search record, based on that historical search record and the target user's behavioral data regarding the search results for that historical search record, the input data is determined, specifically including: Determine the user information of the target user; For each historical search record, behavioral characteristics of a specified type of behavioral data are determined based on the target user's behavioral data of the search results corresponding to that historical search record. Input data is determined based on the user information, the historical search information, and the behavioral characteristics.

5. The method as described in claim 1, characterized in that, Based on the sorting results, at least a portion of the historical search information is displayed in the historical search module of the search interface, specifically including: Determine the business dimensions corresponding to each historical search result; Based on the sorting results, at least some historical search information and its business dimensions are displayed in the historical search module of the search interface.

6. The method as described in claim 5, characterized in that, The method further includes: In response to the target user's click operation on the historical search information in the historical search module, the historical search information clicked by the target user is determined as the target information; When the historical search module displays a business dimension corresponding to the target information, it determines the resource associated with the business dimension corresponding to the target information, and performs a search operation based on the target information and the resource to determine the search results from the resource. When the business dimension corresponding to the target information is not displayed in the historical search module, the resources associated with each business dimension are determined, and a search operation is performed based on the target information and the resources associated with each business dimension to determine the search results from the resources associated with each business dimension.

7. A device for displaying historical search information, characterized in that, include: The acquisition module is used to acquire historical search information corresponding to the target user; The input determination module is used to determine the input data for each historical search information based on the historical search information and the target user's behavior data regarding the search results of that historical search information. An input module is used to input the input data into a trained prediction model to determine the predicted click-through rate of the historical search information output by the prediction model; wherein, the prediction model is trained based on: training samples determined by the historical search information input by the test user and the test user's behavior data on the search results of the historical search information, and the labels of the training samples determined by the test user's operation on the historical search information displayed by the historical search module; The sorting module is used to sort historical search information based on the predicted click-through rate corresponding to each historical search information. The display module is used to display at least a portion of the historical search information in the historical search module of the search interface based on the sorting results; Also includes: For each historical search result, the target user's level of attention to that historical search result is determined based on the target user's behavioral data regarding the search results for that historical search result. Based on the user profile of the target user, determine the business dimension that matches the target user; The weight of each historical search result is determined based on the business dimension corresponding to each historical search result and the business dimension matched with the target user. The weighted attention level of each historical search result is determined based on its popularity and weight. Based on the weighted attention of each historical search result, the historical search results are sorted, and at least some of the historical search results are displayed in the historical search module of the search interface according to the sorting results.

8. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 6.

Images