WO2024174865A1 - Search method and apparatus, and device and storage medium - Google Patents

Abstract

According to the embodiments of the present disclosure, provided are a search method and apparatus, and a device and a storage medium. The method comprises: in response to receiving, at a server, the current query from a client device, determining a target region associated with the current query; at least on the basis of historical target region data corresponding to the target region, generating a prediction result for the current query; sending the prediction result to the client device; in response to a search engine generating a search result for the current query, comparing the prediction result with the search result; and on the basis of a comparison result, sending, to the client device, supplementary information related to subsequent processing of the prediction result. In this way, the speed and accuracy of a server providing search services can be improved.

Description

Search method, device, equipment and storage medium

This application claims priority to the Chinese invention patent application with application number 202310152761.8 filed on February 22, 2023 and titled “Search method, apparatus, device and storage medium”, the entire contents of which are incorporated by reference into this application.

Technical Field

Example embodiments of the present disclosure generally relate to the field of computers, and more particularly, to search methods, devices, apparatuses, and computer-readable storage media.

Background Art

With the development of information technology, diverse and massive data resources are constantly generated. In order to find the required information from the massive data, users usually search for the required information by querying. For example, a user can use a client device to send a query request to a server and receive the corresponding query results from the server. Usually, the server needs a certain amount of processing time to provide the corresponding search results after receiving the query request. However, when the server's transmission resources, computing resources, etc. are insufficient and/or the network infrastructure is imperfect, the search results may take a long time to be provided from the server to the client device, resulting in a poor user experience.

Summary of the invention

In a first aspect of the present disclosure, a search method is provided. The method includes: in response to receiving a current query from a client device at a server, determining a target area associated with the current query; generating a prediction result for the current query based at least on target area historical data corresponding to the target area; sending the prediction result to the client device; in response to a search engine generating search results for the current query, comparing the prediction result with the search results; and sending, based on the comparison result, a message to the client device related to subsequent processing of the prediction result. Additional information.

In a second aspect of the present disclosure, a search method is provided. The method includes: sending a current query to a server at a client device; receiving a prediction result for the current query from the server, wherein the prediction result is generated based on target area historical data corresponding to a target area, and the target area is associated with the current query; receiving supplementary information related to subsequent processing of the prediction result from the server, the supplementary information being generated based on a comparison of the prediction result and a search result generated by a search engine; and performing an operation on the prediction result based on the supplementary information.

In a third aspect of the present disclosure, a search device is provided. The device includes: a current region determination module configured to determine a target region associated with the current query in response to receiving a current query from a client device at a server; a prediction result generation module configured to generate a prediction result for the current query based on at least target region historical data corresponding to the target region; a prediction result sending module configured to send the prediction result to the client device; a comparison module configured to compare the prediction result with the search result in response to a search engine generating a search result for the current query; and a supplementary information sending module configured to send supplementary information related to subsequent processing of the prediction result to the client device based on the comparison result.

In a fourth aspect of the present disclosure, a search device is provided. The device includes: a query sending module configured to send a current query to a server at a client device; a prediction result receiving module configured to receive a prediction result for the current query from the server, wherein the prediction result is generated based on target area historical data corresponding to a target area, and the target area is associated with the current query; a supplementary information receiving module configured to receive supplementary information related to subsequent processing of the prediction result from the server, wherein the supplementary information is generated based on a comparison between the prediction result and a search result generated by a search engine; and an operation execution module configured to execute an operation on the prediction result based on the supplementary information.

In a fifth aspect of the present disclosure, an electronic device is provided. The device includes at least one processing unit; and at least one memory, the at least one memory is coupled to the at least one processing unit and stores instructions for execution by the at least one processing unit. When the instructions are executed by the at least one processing unit, the device executes the method of the first aspect or the second aspect.

In a sixth aspect of the present disclosure, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and the computer program can be executed by a processor to implement the method of the first aspect or the second aspect.

It should be understood that the contents described in the summary of the present invention are not intended to limit the key features or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

FIG1 shows a schematic diagram of an example environment in which embodiments of the present disclosure can be implemented;

FIG2 shows an exemplary interaction diagram of a search process according to some embodiments of the present disclosure;

FIG3 is a schematic diagram showing a search process performed on a server side according to some embodiments of the present disclosure;

FIG4 shows a schematic diagram of providing supplementary information according to some embodiments of the present disclosure;

FIG5 shows a flow chart of a search method performed on a server side according to some embodiments of the present disclosure;

FIG6 shows a flow chart of a search method performed on a client device according to some embodiments of the present disclosure;

FIG7 shows a block diagram of a search device according to some embodiments of the present disclosure;

FIG8 shows a block diagram of a search device according to some other embodiments of the present disclosure; and

FIG. 9 shows a block diagram of a device capable of implementing various embodiments of the present disclosure.

DETAILED DESCRIPTION

It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

For example, in response to receiving an active request from a user, a prompt message is sent to the user. The user is clearly prompted that the operation requested by the user will require the acquisition and use of the user's personal information. Thus, the user can independently choose whether to provide personal information to the electronic device, application, server or storage medium or other software or hardware that executes the operation of the technical solution of the present disclosure according to the prompt information.

As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet the relevant laws and regulations may also be applied to the implementation of the present disclosure.

It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

It should be noted that the titles of any sections/subsections provided herein are not restrictive. Various embodiments are described throughout this article, and any type of embodiment may be included under any section/subsection. In addition, the embodiments described in any section/subsection may be combined in any manner with any other embodiments described in the same section/subsection and/or different sections/subsections.

In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The term "some embodiments" should be understood as "at least some embodiments". Other explicit and implicit definitions may also be included below. The terms "first", "second" and "third" should be understood as "at least some embodiments". The following may also include other explicit and implicit definitions.

As briefly mentioned above, when the server's transmission resources, computing resources, etc. are insufficient and/or the network infrastructure is imperfect, search results may take a long time to be provided from the server to the client device. For example, when the server is located remotely or even overseas, the client device is usually far away from the server. In particular, in some areas, the network infrastructure is poor, resulting in large network transmission delays. In some cases, the number of equipment in the computer room where the server is located is insufficient and the service demand continues to grow, which will cause the transmission time to continue to deteriorate, for example, the time required for each query will increase by hundreds of milliseconds or even longer.

In order to solve the above and other potential problems, a search solution is proposed according to the technical solution of the present disclosure. In this solution, the predicted results for the query are first provided to the client device, and the predicted results are determined based on the historical data of the area associated with the query. Then, the client device is provided with supplementary information related to the subsequent processing of the predicted results, so as to improve the overall search results. Compared with the traditional solution of providing search results only with one data transmission (e.g., one data packet) for one query request, the technical solution of the present disclosure can provide search results through multiple data transmissions (e.g., multiple data packets). In this way, by generating the predicted data and issuing it in advance before the search engine provides the search results, the search engine time can be parallelized with the generation and issuance of the predicted results, thereby reducing service delays. Furthermore, the extracted and issued predicted results are determined based on the historical data related to the query. This is conducive to improving the accuracy of the predicted results, thereby reducing or even avoiding the issuance of subsequent supplementary information. In this way, it can further contribute to the reduction of service delays.

Example environment and interactive process

1 shows a schematic diagram of an example environment 100 in which embodiments of the present disclosure can be implemented. A server 110 for providing a search service is shown in the environment 100, and a client device 120 using the search service is also shown. The client device 120 sends a search request 101 to the server 120 according to user input (e.g., a keyword typed or spoken), and receives search results 102 returned by the server 110.

In some embodiments, an application may be installed in the client device 120. The user may The client device 120 and/or an attached device of the client device 120 interact with the application to initiate the search request 101. As an example, the application may be a content sharing application, which can at least provide users with services related to media content, such as browsing, commenting, forwarding, and creating media content. In this article, "media content" may include multiple forms, including but not limited to images, videos, audio, and the like.

The server 110 may be any type of server-side device with communication functions, which may include, for example, a computing system, such as a mainframe, an edge computing node, a computing device in a cloud environment, and the like. The server 110 may be a single server or a server cluster consisting of several servers. The server 110 may also be a cloud computing service center or a computing instance implemented by cloud computing resources. It should be understood that the above is merely an example of the server 110, not a limitation, and the server 110 may be a device of various forms or structures that can provide search services for the client device 120.

The client device 120 may be any type of terminal device with communication functions. The terminal device may be any type of mobile terminal, fixed terminal or portable terminal, including a mobile phone, a desktop computer, a laptop computer, a notebook computer, a netbook computer, a tablet computer, a media computer, a multimedia tablet, a personal communication system (PCS) device, a personal navigation device, a personal digital assistant (PDA), an audio/video player, a digital camera/camcorder, a positioning device, a television receiver, a radio broadcast receiver, an electronic book device, a gaming device or any combination of the foregoing, including accessories and peripherals of these devices or any combination thereof. The client device 120 may communicate with the server 110 via a wireless network or a wired network.

It should be understood that the above description of the structure and function of the environment 100 is only for exemplary purposes, and does not imply any limitation on the scope of the present disclosure. Some exemplary embodiments of the present disclosure will be described below with reference to FIGS.

FIG. 2 shows an exemplary interaction diagram 200 of a search process according to some embodiments of the present disclosure. The search process 200 of FIG. 2 shows a process in which a server provides a search service to a client device. For ease of discussion, the process 200 is described below with reference to the server 110 and the client device 120 in FIG. 1. It should be understood that this is merely exemplary and is not intended to limit the present disclosure in any way. The server and the client device in the process 200 may be implemented as other Any appropriate equipment.

As shown in FIG2 , the client device 120 sends ( 205 ) a query to the server 110 . For ease of discussion, the query is referred to as the “current query” below. After receiving ( 210 ) the current query from the client device 120 , the server 110 generates ( 215 ) a prediction result for the current query.

The prediction results may include one or more items predicted based on the current query. An item of the prediction results may be media content (e.g., video, audio, image), or other possible data predicted. For example, if the current query involves an animal, the prediction results may include videos, audio, images related to the animal, and may also include information about social accounts related to the animal (e.g., animal science accounts), such as avatars, introductions, published media content, etc. It should be understood that the information included in the prediction results complies with relevant regulations. For example, the user may have granted access rights to the information, or the information returned to the client is processed.

In some embodiments, the prediction results may include all items predicted based on the current query, and there may be no limit on the total number of items. As an alternative, in some embodiments, certain restrictions may be placed on the predicted items. For example, the prediction results may include only a predetermined number of items (e.g., 4 items) for display on the first screen of the client device.

In order to obtain the prediction results, a variety of data sources can be used. Examples of data sources include historical queries over a period of time in the past, candidate data created over a period of time in the past, and the profile of the user who initiated the current query. These data sources can be maintained by region. Accordingly, the server 110 determines the target area associated with the current query, and then considers the historical data of the target area to generate the prediction results. After describing the overall flow of process 200, an example embodiment of generating prediction results using these data sources will be described in detail below.

After generating the prediction result, the server 110 sends (220) the prediction result to the client device 120. In the description herein, the time taken to send the prediction result may include the time taken to package the data and the time taken to transmit the data. The client device 120 receives (225) the prediction result from the server 110. Additionally, in some embodiments, the client device 120 may perform presentation-related preprocessing on the prediction result in response to receiving (225) the prediction result. For example, the client device 110 may render each item in the prediction result but not display it. In this way, after receiving the supplementary information about the predicted structure, the client device 110 can quickly make adjustments and display them to the user. Therefore, it is helpful to further shorten the time spent on presenting the results to the user, thereby improving the user experience.

Continue to describe the server side. After receiving (210) the current query, the server 110 can generate search results for the current query through a local search engine or through a search engine that communicates with it. The search results can be generated using various known search engines, and the search engines will not be described in detail here. Similar to the content of the predicted results, the search results may include one or more items determined by the search engine based on the current query, each item may be media content (e.g., video, audio, image), or other possible data predicted, and so on. It should be understood that the information included in the search results also complies with relevant regulations. For example, the user may have granted access rights to the information, or the information returned to the client is processed.

The process of generating search results for the current query can be performed in parallel with the process of generating (215) and/or sending (220) the prediction results. In the embodiments of the present disclosure, the time sequence is not limited. In response to the search engine generating search results for the current query, the server 110 compares the prediction results with the search results (230). Then, based on the comparison result, the server 110 sends (235) supplementary information related to the subsequent processing of the prediction results to the client device 120. Accordingly, the client device 120 receives (240) the supplementary information from the server 110. Then, based on the supplementary information, the client device 120 performs (245) an operation on the prediction result.

Depending on the results of the comparison, the supplemental information may include various suitable types of data and/or indications.

In some embodiments, the supplemental information may include update information generated based on the results of the comparison. The update information is used to update the predicted result to the search result. Accordingly, the client device 120 updates the previously received predicted result to the search result based on the update information. If the client device 120 has previously rendered the predicted result, the rendering update may be performed and displayed. For example, the update result may indicate the data and operations required to adjust the search result to the predicted result. Alternatively, the update result may indicate the difference between the search result and the predicted result. In this case, the client device 120 may determine to update the predicted result based on the difference. The actions required to adjust the search results to the predicted results.

Depending on the specific implementation of the search results and the predicted results, the update information can indicate the required update operation to the client device 120 in a corresponding appropriate manner. In some embodiments, the predicted results may include a first group of entries, and the search results may include a second group of entries. In such an embodiment, the update result may indicate that the first group of entries arranged in a first order is updated to a second group of entries arranged in a second order. In this case, the update information may include, for example, but not limited to, the deletion of entries, the addition of entries, and/or the replacement of entries, etc. For example, the predicted results include entry 1, entry 2, entry 3, and entry 4, and the search results include entry 1, entry 5, entry 3, and entry 4. The update information may indicate that entry 2 is replaced with entry 5.

Since the prediction results may come from historical queries, the specific content of the items in the prediction results may not be the latest, so even for the same item, its specific content in the prediction results and the search results may be different. In this article, the same item may refer to items related to the same object or entity, for example, items related to the same video, items of the same user. In view of this, in some embodiments, the update information may further indicate an update to the object description in the item.

Specifically, if the entry in the predicted result (also referred to as the first entry) and the entry in the search result (also referred to as the second entry) involve the same object, the update information may indicate an object description element that the second entry has that is different from the first entry. The object description element may exist in the entry in the search results but not in the entry in the predicted result; or it may be different in the search results and the predicted result. Accordingly, the client device 120 may add an object description element to the previously received search results based on the supplementary information, or modify the object description element that originally existed in the search results. As an example, if the entries in the predicted result and the search results are both information about object A, the object description element may be the latest published content related to object A, such as videos, photos, text, etc. As another example, if the entries in the predicted result and the search structure both involve media content B, the object description element may be the number of likes, comments, and reposts of media content B, etc.

In some embodiments, the supplemental information may include instructions on how to process the search results. Specifically, the server 110 may determine, based on the result of the comparison (230), whether to process the predicted result. The number of operations required to update the prediction result to the search result. The number of operations may be, for example, the number of operations required to adjust the order of entries in the prediction result or update the content of the entry. If the number of operations is less than the first threshold number, a positive indication of the prediction result may be sent to the client device 120. Accordingly, the client device 120 may present the prediction result. If the client device 120 has previously rendered the prediction result, the client device 120 may directly display the rendered prediction result. It should be understood that the specific value of the first threshold number may depend on the requirements for aspects such as the accuracy of the query result. For example, if the requirements are more stringent, a positive indication may be sent only if the prediction result is consistent with the search result. For another example, if the requirements are more relaxed, a positive indication may be sent even if the prediction result is slightly different from the search result (for example, the entries included are the same but the order is different).

If the number of operations required to update the prediction result to the search result is greater than the second threshold number, the server 110 may send a negative indication of the prediction result to the client device 120. Accordingly, the client device 120 abandons presenting the prediction result. In this case, the server 110 may send the search result to the client device 120 for presentation. It should be understood that the second threshold number may depend on an estimate of the time required for the client device 120 to update the search result.

Traditionally, search results are provided with only one data transmission (e.g., one data packet) corresponding to one query request. In the example process 200, by generating and sending prediction data in advance before the search engine provides search results, the search engine time can be parallelized with the generation and sending of prediction results, thereby reducing service delays.

Generation of prediction results

As mentioned above with reference to Figure 2, a variety of data sources can be used to generate prediction results. One data source can be data of historical queries. For example, at least one historical query matching the current query can be determined from candidate historical queries, and a prediction result can be generated based on the matching historical queries.

In some embodiments, in order to obtain more candidate items for predicting results, the historical query that matches the current query does not necessarily have to be completely consistent with the current query. For example, if the semantic difference between the current query and the historical query is small, the two can be considered to be matched. If the important features in the semantics of the previous query and the historical query are highly consistent, for example, if the difference between the terms that contribute most to the current query and the historical query is less than a predetermined threshold, then the current query and the historical query can be considered to be matched. The following will also describe in detail how to determine the matching historical query with reference to FIG. 3.

In order to predict the result for the current query, the server 110 may pre-build an appropriate database to store and maintain information related to the candidate historical queries. In some embodiments, the server 110 may store the most recent search result for each query. Then the server 110 may find the most recent query time that matches the current query, and determine the predicted result based on the search results of the query. For example, the server 110 may determine the predicted result for the current query based on the last search result of the query for a certain animal.

Alternatively or additionally, in some embodiments, the server 110 may pre-store information related to queries that appeared within a period of historical time (also referred to as a "second historical time period"). Such information may include the time distribution of query appearances, frequency, search results of each time, etc. For example, the time when each query appeared within a period of time (e.g., the past 24 hours), the items that appeared the most times in the search results, etc. may be calculated every hour.

In this embodiment, after receiving the current query, the server 110 can determine multiple historical queries that appear during this historical period and match the current query, and can generate prediction results based on these historical queries. By introducing historical queries over a period of time in the past, more candidate data can be obtained. In this way, it can be helpful to improve the hit rate of the prediction results, thereby reducing the update operation of the client device on the prediction results.

In some embodiments, when using matched historical queries, the discreteness of the search results of these historical queries (which can also be regarded as multiple searches of the same query) can be further considered. Specifically, the discreteness of multiple historical search results for these multiple historical queries can be determined, and the prediction results are determined based on these historical search results only when the discreteness is low. A low discreteness indicates that the historical search results are good, so they can be used as prediction results. In some embodiments, if it is determined that the discreteness is less than a threshold discreteness, the prediction results are determined based on multiple historical search results.

Sometimes historical queries may not be real-time enough. For example, historical queries may include queries within 24 hours from the current time point. As an alternative to historical queries, in some embodiments, the data source of the prediction results may include data that is newly created or newly released within a period of historical time (i.e., the first historical time period, such as the last 6 hours) close to the current time point, such as videos, images, text, etc., which are referred to as "candidate search data" below. Then, the prediction results can be generated based on these candidate search data. In this way, streaming data can be used to supplement relatively real-time data, and streaming data and offline statistical data can be merged together to determine the prediction results.

For example, data with a creation time within 6 hours from the request initiation time can be obtained from the offline data stream as candidate search data. In some embodiments, the candidate search data can be stored in a database, for example, written to the same database as the historical query. For example, the timestamp of data creation can be used as a score, and the count within 6 hours can be recorded under another key value. The server 110 can obtain the candidate search data from the database and determine the prediction result accordingly.

In some embodiments, whether to use such more real-time candidate search data can be determined based on a predetermined strategy. For example, if the number of historical queries that match the current query is less than a threshold number, the data that matches the current query can be determined from the candidate search data as at least a part of the predicted result. For another example, if the historical queries that match the current query appear in a concentrated time close to the current time point, the candidate search data can be used when generating the predicted result. A small number of matching historical queries or a recent concentrated appearance of matching historical queries may mean that the current query has a stronger real-time nature. Therefore, the candidate search data can be used in this case.

Information related to candidate historical queries and/or candidate search data can be stored in a database as pre-stored data or historical data to be used as a data source for prediction. As briefly mentioned above, historical data can be stored and maintained separately according to different regions. Regions can be divided, for example, according to geographical location, language type, administrative divisions, etc. For example, different countries or regions located in the same continent can be divided into the same region, countries or regions with the same official language can be divided into the same region, and different regions can be distinguished according to the administrative divisions of countries and regions. It should be understood that the above description of regions is merely exemplary and not restrictive, and the embodiments of the present disclosure can be implemented through a variety of suitable The historical data may include the following data associated with a certain area, such as but not limited to, historical queries matching the current query, candidate search data created in the recent period, and the like.

By utilizing this historical data maintained by region, predictions by region can be performed. Specifically, the region associated with the current user (hereinafter also referred to as the "first region") can be determined based on the profile of the current user. For example, the first region may be the region indicated in the current user's registration information, or the region corresponding to the language used by the current user. As another example, the first region may be the region where the current user is located when initiating the current query. However, it should be understood that no matter what scheme is used to determine the first region, the user's authorization or permission should be obtained. Then, the prediction result can be determined based on the historical data corresponding to the first region, as described above. In this embodiment, the first region can be regarded as the target region.

In some embodiments, cross-region prediction can be performed for the current query. For example, if no result for the current query is found in the historical data corresponding to the first region, the region corresponding to the historical query (e.g., the most recent query) matching the current query (hereinafter also referred to as the "second region") can be determined as the target region. Then, the prediction result can be determined based on the historical data corresponding to the second region.

For example, multiple historical queries matching the current query can be searched in the historical data corresponding to the second region, and predictions can be made based on these historical queries. For another example, predictions can be made based on more real-time candidate search data corresponding to the second region. By making predictions across regions, the data sources for predictions can be further expanded, thereby improving the hit rate of prediction results.

Alternatively or additionally, in some embodiments, the data source of the prediction result may include the profile of the current user who initiated the current query. The user's profile may include basic information about the user, such as which accounts the user follows, like information, comment content, forwarding status, etc. Server 110 may determine whether the current query indicates another user associated with the current user based on the profile. If so, information related to the current user may be determined as at least part of the prediction result. For example, server 110 may first determine that the current query "Zhang XX" has the intent of "searching for users". Then, based on the current user's profile, it may determine that Zhang XX is another user followed by the current user. Accordingly, the information of "Zhang XX" may be included in the prediction result. For example, identification information, introduction information, a list of published media contents, etc. are used as at least a part of the prediction result.

It should be understood that the use of the user profile in the embodiments of the present disclosure has obtained the user's authorization in advance. In addition, the use of the user profile is intended to provide services to the user, but not intended to expose sensitive and private information.

According to the technical solution of the present disclosure, there are many data sources for determining the prediction results for the query, and the use strategies of these data sources can be flexibly adjusted/iterated, so that accurate prediction results can be obtained. The prediction results obtained in this way may only require a small number of update operations, or even no update operations. In addition, the technical solution of the present disclosure can quickly, flexibly and accurately provide search results for the user's search request, with few constraints, and is suitable for a wide range of applications.

Example search process at the server

FIG3 shows a schematic diagram of performing a search process 300 according to some embodiments of the present disclosure. The process 300 can be implemented on the server side that provides the search service. For ease of discussion, the process 300 of FIG3 is described below with reference to the environment 100 of FIG1 , in particular the server 110. It should be understood that this is merely exemplary and is not intended to limit the embodiments of the present disclosure in any way.

At block 310, the server 110 receives a query from the client device 120. In response to the query, the server 110 may find and load information of the user who initiated the query at the client device 120. The server 110 may then perform two branches in parallel, one branch being the determination of regular search results and the other branch being the generation and transmission of predicted results.

As shown in FIG3 , in the determination branch of the regular search results, the process 300 enters box 320, where the server 110 determines the regular search results for the query. Specifically, the server 110 performs engine recall sorting, and uses the search engine to generate search results for the query from the client device 120. Recall is the process of obtaining a set of candidate content related to the query based on the input query. Sorting can be to sort the content in the candidate content set. If the relevant candidate content set cannot be recalled, even if the subsequent sorting method is very optimized, it cannot provide search results for the query. Therefore, recall is very important for the search engine, which determines the quality of the search engine. Recall can be achieved through a variety of existing algorithms, such as Such as traditional word-based recall, vector-based semantic recall, etc., which will not be elaborated here.

The engine recall sorting process of frame 320 is usually time-consuming, for example, 600ms. In an embodiment according to the present disclosure, by executing the recall sorting of the search engine (frame 320) and the related processing of the prediction data (frames 330 to 350) in parallel, the processing time can be reduced and the processing efficiency can be improved.

In parallel with the engine recall ranking of block 320, in the prediction result generation and transmission branch, process 300 enters block 330. In block 330, server 110 generates prediction results. As mentioned above, the prediction results can have a variety of data sources. For example, at least one historical query matching the current query, candidate search data created within a certain period of historical time, the profile of the user who initiated the current query, and so on.

In some embodiments, offline statistical data may be used to determine the prediction result. The offline statistical data may be, for example, data related to historical queries over a period of time in the past.

Offline statistics can be classified by region, and the region can be divided, for example, by geographical location (e.g., by longitude, latitude, continent, etc.), language type, administrative division (e.g., by country, region, etc.), etc. The user's profile can include which region(s) the user is associated with. In this way, the region associated with the user making the query can be determined through the user's profile, so as to obtain offline statistics in the region, and then obtain one or more historical queries matching the current query.

If one or more historical queries matching the current query are not found in the offline statistics of the above region (also referred to as the first region), the offline statistics of another region (also referred to as the second region) can be searched. The second region can be determined, for example, in descending order of the number of queries with the same number as the first region. For example, the second region can be determined from the top N countries (N is a positive integer, such as 8) with the largest number of queries with the first region. For example, these N countries can be sorted in descending order according to the same number of queries, and if no historical queries matching the current query are found in the offline statistics of a certain region, the search can be continued in the region with a lower ranking.

In some embodiments, in order to improve the hit rate and maintainability, the area corresponding to each query can be recorded. If the same query is received later, the historical query matching the query can be directly determined from the offline statistical data in the recorded area. As an alternative, you can first search for historical queries matching the query in the offline statistics of the current area, and if no historical queries are found, then search in the recorded area.

The historical queries matching the query received from the client device 120 can be determined from the offline statistics in a variety of ways. In some embodiments, this can be done by query standardization or normalization. Alternatively, in some embodiments, the historical queries matching the query can be determined by term matching.

In an embodiment where query normalization is used to determine matching historical queries, it is possible to determine whether two queries match based on semantics. For example, if one query is "hello, world" and the other query is "hello world", then the two queries are considered to match because they have the same semantics and differ only in the punctuation used for assistance.

In some embodiments, considering the particularity of some languages, two-level processing can be performed. In the first level of processing, first, rules can be used for standardization. Such rules include, for example, but are not limited to, converting multiple spaces into one space, removing punctuation marks, eliminating accent marks of certain languages (such as European languages), and so on. Since such accent marks usually indicate emphasis without separate semantics, they are eliminated here for standardization. In the second level of processing, the query can be analyzed first, and the terms in the query can be standardized according to the analysis results. This standardization may have a certain impact on semantics, and may also remove information such as tense, singular and plural. For example, the term "apples" contained in the query may be adjusted to "apple". As an alternative, in an embodiment of a historical query that uses a term matching method to determine matching, each term after the query segmentation can be obtained after query analysis, and the contribution of each term to the entire query in terms of semantics, also known as term importance, can also be obtained. In addition, the intent of the query and/or other information can also be determined. Using this information, more matching historical queries can be obtained by removing a part of the semantics corresponding to the query, thereby improving the hit rate.

As a supplement to offline statistical data, in some alternative or additional embodiments, streaming data can be used to provide relatively real-time results, that is, streaming data is applied to the process of determining the prediction results. For example, candidate search data created within a certain period of time can be obtained from the offline data stream, such as the creation time is within 100 seconds of the request initiation time. Data within the historical time period (e.g., the 6 hours closest to the current time). In some embodiments, the candidate search data can be stored in a database. The server 110 can obtain the candidate search data from the database and decide which part of the data to package through a certain strategy.

As another alternative for determining the prediction result, in some cases, the server 110 can determine the user's search intention based on the query received from the client device 120, such as searching for content or searching for people. If it is found that the user query involves a person and there is less data related to the person, such as the user is searching for an ordinary friend, it is possible that there are few or no historical queries or related candidate search data for the query. In this case, in order to avoid being unable to obtain the prediction data or obtaining inaccurate prediction data, for example, the query from the client device 120 may sometimes include the user's unique identifier, such as uniq_id. Through the user's unique identifier and the current user's profile, it can be determined whether the current user wants to query his or her friends or the object of concern. If so, the accuracy and hit rate of the prediction result can be improved. In this way, for the case where the user's intention is to obtain the prediction data and the prediction data cannot be obtained, the prediction data of better quality can be obtained by querying the user profile.

After obtaining the prediction results, the server 110 packages the prediction results in block 340. Since packaging may also be a relatively time-consuming process, for example, 200 ms, this parallel processing method can effectively improve the processing efficiency of the server 110. Then, in block 350, the server 110 sends the packaged prediction data to the client device 120.

After receiving the prediction data, the client device 120 can pre-process the prediction data, such as rendering. Since rendering is a relatively time-consuming process, for example, 200ms, before the client device 120 displays the final result for the query, the final result may be the prediction data or the result after the prediction data is adjusted using the update information. The client device 120 can complete the relatively time-consuming rendering operation in advance during the long time that the server 110 uses the search engine to determine the search results, and only needs to complete the display operation after receiving the positive confirmation or update information from the server 110. In this way, the display efficiency of the client device 120 for the final result can be effectively improved.

After obtaining the prediction result and the search result, the server 110 compares the two. For example, in block 360, the server 110 determines whether the prediction result hits the search result provided by the search engine after recall sorting. Based on the comparison result, the server 110 sends supplementary information related to subsequent processing of the prediction result to the client device 120. The supplementary information may include, but is not limited to, a positive indication, a negative indication, update information, etc. for the prediction result.

Specifically, if it is determined in box 360 that the predicted result hits, the server 110 may send a positive indication of the predicted result to the client device 120 in box 370. After receiving the positive indication, the client device 120 may know that the predicted result is accurate and can be displayed without adjustment. If the predicted result does not hit, the server 110 may send a negative indication of the predicted result to the client device 120 in box 395. After receiving the negative indication, the client device 120 may know that the predicted result is inaccurate and may directly discard the predicted result. In the case where the predicted result is a partial hit, the server may generate update information for updating the predicted result to the search result based on the result of the comparison between the predicted result and the search result in box 380, and send the update information to the client device 120 in box 390. After receiving the update information, the client device 120 may update the previously received predicted result to the search result and present the search result.

From the example process described in FIG. 3 , from the perspective of time dimension, the time-consuming engine recall sorting is parallel to the prediction result generation, packaging, transmission, and optional client rendering. In contrast, in the traditional solution, the engine recall sorting, search result packaging, transmission, and client rendering are completely serial. Therefore, in the embodiments of the present disclosure, it is beneficial to improve the search service speed.

Example procedure for generating supplementary information

FIG. 4 shows a schematic diagram of a process 400 for providing supplemental information according to some embodiments of the present disclosure. The process 400 can be implemented on the server side. For ease of discussion, the process 400 of FIG. 4 is described below with reference to the environment 100 of FIG. 1 , particularly the server 110. It should be understood that this is merely exemplary and is not intended to limit the embodiments of the present disclosure in any way.

In block 410, the server 110 initializes the prediction results and search results. The prediction results may be obtained, for example, in block 215 or block 330, and may include ordered Multiple items (i.e., item sequence). The search results are, for example, a collection of multiple items for the current query obtained by a search engine, which can be implemented as an item sequence. The server 110 can sort the items in the prediction results and the search results respectively, thereby determining the order of the items in the prediction results and the order of the items in the search results.

In block 415, the server 110 reorders the prediction results. For example, the server 110 may reorder the items in the prediction results that are not in the same order according to the order of the items in the search results. In some embodiments, considering that the items displayed on the first screen need to be displayed to the user at an accelerated speed, only a portion of the prediction results may be reordered in block 415, for example, only the items in the prediction results that are used for first screen display (such as the first 6 items) are reordered.

In box 420, the minimum edit distance is calculated for the entries in the predicted results and the entries in the search results. In an embodiment of the present disclosure, the edit distance is a quantitative measure of the degree of difference between the entry sequences. The edit distance can be determined, for example, by determining how many times at least one sequence of entries is required to transform one sequence of entries into another sequence of entries. The edit distance can include, for example, the Levenshtein distance, which allows operations such as entry addition, entry deletion, and entry replacement, and can also include the Damerau-Levenshtein distance, etc. The Damerau-Levenshtein distance allows the exchange of two adjacent entries based on the Levenshtein distance. The calculation of the edit distance can be non-weighted or weighted. Different weights can be assigned to each operation involved in the edit distance. For example, the weights of the exchange, modification, addition, and deletion operations can be 0, 2, 1, and 3, respectively. It should be understood that the above examples of each operation and its weight are merely illustrative and not any limitation to the embodiments of the present disclosure. The embodiments of the present disclosure can set any appropriate weights for the operations in the process of determining the edit distance.

There may be multiple editing paths for editing the entry sequence of the prediction results into the entry sequence of the search results, and each editing path includes one or more editing operations. The editing distance of each editing path may be different. Therefore, the minimum editing distance for converting the entry sequence of the prediction results into the editing sequence of the search results can be calculated in box 420. In this way, the load of editing the prediction results of the client device 120 can be reduced, thereby improving the display speed of the search results.

At block 425, an edit operation is generated. For example, the minimum The edit distance derives one or more edit operations to be performed. Any suitable method may be used to derive the edit operation corresponding to the minimum edit distance. The scope of the present disclosure is not limited in this respect.

In block 430, an entry update package is generated. According to the editing operation obtained in block 425, update information for adjusting the position of the entry in the prediction result may be generated, and the update information may be packaged into an update package.

In box 435, the order of the entries is verified. Here, the order of the entries in the adjusted prediction results is compared with the order of the entries in the search results to determine whether the two are consistent. After the verification is completed, in box 440, an update package for the entry sub-elements is generated. The entry sub-elements can be the object description elements described above. For example, for entries involving media content, the entry sub-elements can include fields such as the number of likes, the number of comments, the number of forwardings, etc. For another example, for entries involving object information, the entry sub-elements can include a list of published videos related to the object, etc.

In block 445, a boundary condition check is performed. In this process, it can be checked whether the conditions for returning the update package are met. For example, the total number of replacement operations and addition operations cannot exceed 2.

If the boundary condition check succeeds, an additional update package for the entry is generated in block 450. The additional update package may include other information that needs to be notified to the client device 120, such as Easter eggs.

At block 455, the hit status is fed back based on the result of the boundary condition check. If the result of the check indicates that no editing operation is required, it means a complete hit. In this case, the server 110 may send a positive indication to the client device 120, for example, setting the ACK in the indication to a value of 0. It is understood that, in this case, the entry update package generated at block 430 may be empty, and the sub-element update package generated at block 440 may be empty or not empty.

If the condition for returning the update packet is not met, for example, the total number of replacement operations and addition operations exceeds 2, it means a miss. In this case, the server 110 may send a negative indication to the client device 120, for example, setting the ACK in the indication to a value of -1. It is understood that in this case, the entry update packet generated at block 430 and the entry update packet generated at block The sub-element update packages generated at 440 may all be empty.

If the condition for returning the update package is met, for example, the total number of replacement operations and addition operations does not exceed 2, it means that the prediction result can be updated to the search result. In this case, the server 110 can send update information to the client device 120, and the update information can be sent in the form of an update package, for example. The client device 120 can adjust the prediction result to the search result according to the entry editing operation indicated by the update information, so as to quickly obtain a more accurate search result in an efficient manner.

In some embodiments, the prediction results sent to the client device 120 include, for example, entries for first screen display, so the number of other entries to be provided to the client device 120 can be determined in block 460. For example, assuming that 6 entries can be displayed on the first screen at the client device 120, and the search results for home page display include a total of 10 entries, then in block 460 it can be determined that the number of entries that need to be provided to the client device 120 is 4 entries.

The first sending of the prediction results to the client device 120 can be implemented in a variety of ways. In some embodiments, if the first screen requires 6 items, M items (M is an integer greater than or equal to 6) can be actually sent from the server 110 to the client device 120, and the client device 120 selects 6 items from the M items to be displayed on the first screen.

In some alternative embodiments, when sending for the first time, only the entries that are most likely to appear on the first screen may be sent from the server 110 to the client device 120, without caring about the order of the entries.

In other alternative embodiments, different prediction result sending strategies can be determined for different sessions based on the historical recall of the current query. The historical recall situation may include a lot of information, such as but not limited to, the frequency of the query in the past 24 hours, the discreteness of the search results, when was the last search for the query, the intent of the query, the timeliness, the region where the user who made the query is located (the region can be used to approximate the network infrastructure situation), the performance score of the client device, and so on. The prediction result sending strategy can be made based on the above historical recall situation and/or other information related to the current query. For example, for data with a relatively high hit rate, the number of entries sent from the server 110 to the client device 120 is reduced. For another example, in the case of poor server performance or poor network infrastructure, the number of entries sent from the server 110 to the client device 120 is reduced. For another example, for data with a particularly low hit rate, the number of entries sent from the server 110 to the client device 120 is reduced. It should be understood that in such an embodiment, the acquisition of the user's area and/or the client device performance score is authorized by the user in advance.

Example methods and apparatus

FIG5 shows a flow chart of a search method 500 performed on a server side according to some embodiments of the present disclosure. The method 500 can be implemented on the server side. For ease of discussion, the method 500 of FIG5 is described below using the server 110 of FIG1 as an example. It should be understood that this is merely exemplary and is not intended to limit the embodiments of the present disclosure in any way.

At block 505, the server 110 determines a target area associated with the current query in response to receiving the current query from the client device 120. At block 510, the server 110 generates a prediction result for the current query based on at least target area historical data corresponding to the target area.

In some embodiments, the target area historical data includes at least one of the following: data of at least one historical query matching the current query and associated with the target area, candidate search data associated with the target area created within the first historical time period, and the like.

In some embodiments, at least one historical query includes multiple historical queries that occurred within a second historical time period, and generating a predicted result includes: determining a degree of discreteness of multiple historical search results for the multiple historical queries; and in response to determining that the degree of discreteness is less than a threshold degree of discreteness, determining a predicted result based on the multiple historical search results.

In some embodiments, to generate a prediction result, server 110 may determine data matching the current query from the candidate search data as at least a part of the prediction result in response to determining that the number of at least one historical query is less than a threshold number.

In some embodiments, in order to generate a predicted result, the server 110 may determine whether the current query indicates another user associated with the current user based on the profile of the current user who initiated the current query; and in response to determining that the current query indicates another user, determine information related to the current user as at least part of the predicted result.

In some embodiments, to determine the target area, the server 110 may initiate Based on the profile of the current user of the current query, determining a first area associated with the current user; searching for results for the current query in the first area historical data corresponding to the first area; and in response to not finding results for the current query in the first area historical data, determining a second area corresponding to the last query that matches the current query as a target area.

At block 520, the server 110 sends the predicted results to the client device 120. At block 530, the server 110 compares the predicted results with the search results in response to the search engine generating search results for the current query.

At block 540 , the server 110 transmits supplementary information related to subsequent processing of the prediction result to the client device 120 based on the result of the comparison.

In some embodiments, in order to send the supplementary information, the server 110 may generate update information for updating the prediction result as the search result based on the comparison result, and send the update information to the client device 120 .

In some embodiments, the prediction result includes a first group of entries, the search result includes a second group of entries, and the update information indicates that the first group of entries arranged in the first order are updated to the second group of entries arranged in the second order. The update information may include at least one of the following: deletion of an entry, addition of an entry, and replacement of an entry.

In some embodiments, the first entry in the predicted result and the second entry in the search result relate to the same object, and the update information indicates that the second entry has different object description elements from the first entry.

In some embodiments, in response to the comparison result indicating that the number of operations required to update the predicted result to the search result is less than the first threshold number, the server 110 may send a positive indication of the predicted result to the client device 120. Alternatively, in response to the comparison result indicating that the number of operations required to update the predicted result to the search result is greater than the second threshold number, the server 110 may send a negative indication of the predicted result to the client device 120.

FIG6 shows a flow chart of a search method 600 performed on a client device according to some embodiments of the present disclosure. The method 600 can be implemented on the client device side. For ease of discussion, the method 600 of FIG6 is described below using the client device 120 of FIG1 as an example. It should be understood that this is merely exemplary and is not intended to limit the embodiments of the present disclosure in any way.

At block 610 , the client device 120 sends a current query to the server.

At block 620, the client device 120 receives a prediction result for the current query from the server. The prediction result is generated based on target area historical data corresponding to the target area, and the target area is associated with the current query. In some embodiments, in response to receiving the prediction result, the client device 120 may perform pre-processing on the prediction result related to presentation.

At block 630 , the client device 120 receives from the server supplemental information related to subsequent processing of the prediction result, the supplemental information being generated based on a comparison of the prediction result and the search results generated by the search engine.

At block 640 , the client device 120 performs an operation on the prediction result based on the supplemental information.

In some embodiments, the client device 120 may update the predicted result into a search result based on the updated information included in the supplemental information, and present the search result.

In some embodiments, the prediction result includes a first group of entries, and the search result includes a second group of entries. The client device 120 may update the first group of entries arranged in the first order to a second group of entries arranged in the second order according to the entry editing operation indicated by the update information. The entry editing operation may include at least one of the following: deletion of an entry, addition of an entry, and replacement of an entry.

In some embodiments, the first item in the predicted result and the second item in the search result relate to the same object. In response to the update information indicating that the second result has an object description element different from the first item, the client device 120 may update the first item to have an object description element.

In some embodiments, client device 120 may present the predicted result in response to the supplemental information including a positive indication of the predicted result, or may forgo presenting the predicted result in response to the supplemental information including a negative indication of the predicted result.

7 shows a block diagram of a search apparatus 700 according to some embodiments of the present disclosure. The apparatus 700 may be implemented as or included in the server 110. Each module/component in the apparatus 700 may be implemented by hardware, software, firmware, or any combination thereof.

As shown in the figure, the apparatus 700 includes: a target area determination module 705, configured to determine a target area associated with the current query in response to receiving a current query from a client device at a server; a prediction result generation module 710, configured to generate a prediction result based on at least The target area historical data corresponding to the domain is used to generate a prediction result for the current query; the prediction result sending module 720 is configured to send the prediction result to the client device; the comparison module 730 is configured to compare the prediction result with the search result in response to the search engine generating the search result for the current query; and the supplementary information sending module 740 is configured to send supplementary information related to the subsequent processing of the prediction result to the client device based on the comparison result.

In some embodiments, the target area historical data includes at least one of the following: data of at least one historical query that matches the current query and is associated with the target area, or candidate search data associated with the target area created within a first historical time period.

In some embodiments, at least one historical query includes multiple historical queries that occurred within a second historical time period, and the prediction result generation module is also configured to: determine the degree of discreteness of multiple historical search results for the multiple historical queries; and in response to determining that the degree of discreteness is less than a threshold degree of discreteness, determine the prediction result based on the multiple historical search results.

In some embodiments, the prediction result generation module 710 is further configured to: in response to determining that the number of at least one historical query is less than a threshold number, determine data matching the current query from the candidate search data as at least a part of the prediction result.

In some embodiments, the prediction result generation module 710 is also configured to: determine whether the current query indicates another user associated with the current user based on the profile of the current user who initiated the current query; and in response to determining that the current query indicates another user, determine information related to the current user as part of the prediction result.

In some embodiments, the target area determination module 705 is also configured to: determine a first area associated with the current user based on the profile of the current user who initiated the current query; search for results for the current query in the first area historical data corresponding to the first area; and in response to not finding results for the current query in the first area historical data, determine a second area corresponding to a historical query that matches the current query as a target area.

In some embodiments, the supplementary information sending module 740 is further configured to: generate update information for updating the prediction result to the search result based on the comparison result; and send the update information to the client device.

In some embodiments, the predicted results include a first set of items and the search results include a second set of items. The update information indicates that the first group of entries arranged in the first order is updated to the second group of entries arranged in the second order. The update information may include at least one of the following: deletion of an entry, addition of an entry, and replacement of an entry.

In some embodiments, the first entry in the predicted result and the second entry in the search result relate to the same object, and the update information indicates that the second entry has different object description elements from the first entry.

In some embodiments, the supplemental information sending module 740 is also configured to: in response to the comparison result indicating that the number of operations required to update the predicted result to the search result is less than a first threshold number, send a positive indication of the predicted result to the client device; or, in response to the comparison result indicating that the number of operations required to update the predicted result to the search result is greater than a second threshold number, send a negative indication of the predicted result to the client device.

8 shows a block diagram of a search apparatus 800 according to some other embodiments of the present disclosure. The apparatus 800 may be implemented as or included in the client device 120. Each module/component in the apparatus 800 may be implemented by hardware, software, firmware, or any combination thereof.

As shown in the figure, the apparatus 800 includes a query sending module 810, which is configured to send a current query to a server at a client device; a prediction result receiving module 820, which is configured to receive a prediction result for the current query from the server, wherein the prediction result is generated based on target area historical data corresponding to the target area, and the target area is associated with the current query; a supplementary information receiving module 830, which is configured to receive supplementary information related to subsequent processing of the prediction result from the server, wherein the supplementary information is generated based on a comparison between the prediction result and a search result generated by a search engine; and an operation execution module 840, which is configured to perform an operation on the prediction result based on the supplementary information.

In some embodiments, the apparatus 800 may further include a preprocessing module configured to perform presentation-related preprocessing on the prediction result in response to receiving the prediction result.

In some embodiments, the operation execution module 840 is further configured to: update the predicted result to a search result based on the updated information included in the supplemental information; and present the search result.

In some embodiments, the prediction results include a first group of entries, the search results include a second group of entries, and the operation execution module 840 is further configured to: update the first group of entries arranged in the first order to the second group of entries arranged in the second order according to the entry editing operation indicated by the update information; A second group of entries in the column, wherein the entry editing operation includes at least one of the following: deletion of an entry, addition of an entry, and replacement of an entry.

In some embodiments, the first entry in the predicted result and the second entry in the search result relate to the same object, and the operation execution module 840 is also configured to: in response to the update information indicating that the second result has an object description element different from that of the first entry, update the first entry to have an object description element.

In some embodiments, the operation execution module 840 is further configured to present the prediction result in response to the supplemental information including a positive indication of the prediction result, or is configured to give up presenting the prediction result in response to the supplemental information including a negative indication of the prediction result.

FIG9 shows a block diagram of an electronic device 900 in which one or more embodiments of the present disclosure may be implemented. It should be understood that the electronic device 900 shown in FIG9 is merely exemplary and should not constitute any limitation on the functionality and scope of the embodiments described herein. The electronic device 900 shown in FIG9 may be used to implement the server 110 and/or the client device 120 of FIG1 .

As shown in FIG9 , the electronic device 900 is in the form of a general electronic device. The components of the electronic device 900 may include, but are not limited to, one or more processors or processing units 910, a memory 920, a storage device 930, one or more communication units 940, one or more input devices 950, and one or more output devices 960. The processing unit 910 may be an actual or virtual processor and is capable of performing various processes according to a program stored in the memory 920. In a multi-processor system, multiple processing units execute computer executable instructions in parallel to improve the parallel processing capability of the electronic device 900.

The electronic device 900 typically includes a plurality of computer storage media. Such media may be any retrievable media accessible to the electronic device 900, including but not limited to volatile and nonvolatile media, removable and non-removable media. The memory 920 may be a volatile memory (e.g., registers, cache, random access memory (RAM)), a nonvolatile memory (e.g., read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory), or some combination thereof. The storage device 930 may be a removable or non-removable medium and may include a machine-readable medium such as a flash drive, a disk, or any other medium that may be capable of being used to store information and/or data (e.g., using training data for training) and can be accessed within the electronic device 900.

The electronic device 900 may further include additional removable/non-removable, volatile/non-volatile storage media. Although not shown in FIG. 9 , a disk drive for reading or writing from a removable, non-volatile disk (e.g., a “floppy disk”) and an optical drive for reading or writing from a removable, non-volatile optical disk may be provided. In these cases, each drive may be connected to a bus (not shown) by one or more data media interfaces. The memory 920 may include a computer program product 925 having one or more program modules configured to perform various methods or actions of various embodiments of the present disclosure.

The communication unit 940 implements communication with other electronic devices through a communication medium. Additionally, the functions of the components of the electronic device 900 can be implemented with a single computing cluster or multiple computing machines that can communicate through a communication connection. Therefore, the electronic device 900 can operate in a networked environment using a logical connection with one or more other servers, a network personal computer (PC), or another network node.

The input device 950 may be one or more input devices, such as a mouse, a keyboard, a tracking ball, etc. The output device 960 may be one or more output devices, such as a display, a speaker, a printer, etc. The electronic device 900 may also communicate with one or more external devices (not shown) through the communication unit 940 as needed, such as a storage device, a display device, etc., communicate with one or more devices that allow a user to interact with the electronic device 900, or communicate with any device that allows the electronic device 900 to communicate with one or more other electronic devices (e.g., a network card, a modem, etc.). Such communication may be performed via an input/output (I/O) interface (not shown).

According to an exemplary implementation of the present disclosure, a computer-readable storage medium is provided, on which computer-executable instructions are stored, wherein the computer-executable instructions are executed by a processor to implement the method described above. According to an exemplary implementation of the present disclosure, a computer program product is also provided, which is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions, and the computer-executable instructions are executed by a processor to implement the method described above.

Reference is made here to methods, apparatuses, devices and computer program products implemented according to the present disclosure The flowcharts and/or block diagrams of the present disclosure are described in detail. It should be understood that each block of the flowcharts and/or block diagrams and the combination of each block in the flowcharts and/or block diagrams can be implemented by computer-readable program instructions.

These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processing unit of the computer or other programmable data processing device, a device that implements the functions/actions specified in one or more boxes in the flowchart and/or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and/or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions/actions specified in one or more boxes in the flowchart and/or block diagram.

Computer-readable program instructions can be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, so that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions/actions specified in one or more boxes in the flowchart and/or block diagram.

The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple implementations of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of a module, program segment or instruction includes one or more executable instructions for realizing the logical function of the specification. In some implementations as replacements, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and/or flow chart, and the combination of the square boxes in the block diagram and/or flow chart can be realized by a special hardware-based system that performs the function or action of the specification, or can be realized by a combination of special hardware and computer instructions.

The above descriptions are intended to be illustrative and not exhaustive. The present invention is not intended to be exhaustive and is not limited to the disclosed implementations. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described implementations. The terminology used herein is selected to best explain the principles of the implementations, practical applications, or improvements to the technology in the marketplace, or to enable other persons of ordinary skill in the art to understand the various implementations disclosed herein.

Claims (20)

A search method, comprising: In response to receiving a current query at the server from a client device, determining a target area associated with the current query; generating a prediction result for the current query based at least on target area historical data corresponding to the target area; Sending the prediction result to the client device; In response to a search engine generating search results for the current query, comparing the predicted results with the search results; and Based on the result of the comparison, supplementary information related to subsequent processing of the prediction result is sent to the client device. The method according to claim 1, wherein the target area historical data includes at least one of the following: data of at least one historical query that matches the current query and is associated with the target area, or Candidate search data associated with the target area created within a first historical time period. The method according to claim 2, wherein the at least one historical query includes a plurality of historical queries occurring within a second historical time period, and generating the prediction result includes: Determining a degree of dispersion of a plurality of historical search results for the plurality of historical queries; and In response to determining that the degree of dispersion is less than a threshold degree of dispersion, the predicted result is determined based on the plurality of historical search results. The method according to claim 2, wherein generating the prediction result comprises: In response to determining that the number of the at least one historical query is less than a threshold number, data matching the current query is determined from the candidate search data as at least a part of the predicted result. The method according to claim 1, further comprising: determining, based on a profile of a current user initiating the current query, whether the current query indicates another user associated with the current user; and In response to determining that the current query indicates the other user, information related to the current user is determined as part of the predicted result. The method of claim 1, wherein determining the target area comprises: determining a first region associated with a current user based on a profile of the current user initiating the current query; Searching for results for the current query in first region history data corresponding to the first region; and In response to not finding a result for the current query in the first region history data, a second region corresponding to a history query matching the current query is determined as the target region. The method according to claim 1, wherein sending the supplemental information comprises: Based on the comparison result, generating update information for updating the prediction result as the search result; and The update information is sent to the client device. The method of claim 7, wherein the predicted results include a first set of entries, the search results include a second set of entries, and the update information indicates: The first group of entries arranged in a first order is updated to the second group of entries arranged in a second order, and includes at least one of the following: Deletion of entries, Addition of entries, Replacement of entry. The method according to claim 7 or 8, wherein the first item in the predicted result and the second item in the search result relate to the same object, and the update information indicates: The second entry has different object description elements from those of the first entry. The method according to claim 1, wherein sending the supplemental information comprises: In response to the comparison result indicating that the number of operations required to update the predicted result to the search result is less than a first threshold number, sending a positive indication of the predicted result to the client device, or In response to a result of the comparison indicating that the number of operations required to update the predicted result to the search result is greater than a second threshold number, a negative indication of the predicted result is sent to the client device. A search method, comprising: sending, at the client device, a current query to the server; receiving a prediction result for the current query from the server, wherein the prediction result is generated based on target area historical data corresponding to a target area, and the target area is associated with the current query; receiving from the server supplemental information related to subsequent processing of the prediction result, the supplemental information being generated based on a comparison of the prediction result and a search result generated by a search engine; and Based on the supplemental information, an operation is performed on the prediction result. The method according to claim 11, further comprising: In response to receiving the prediction result, the prediction result is pre-processed related to presentation. The method according to claim 12, wherein performing an operation on the prediction result comprises: Based on the update information included in the supplementary information, updating the predicted result to the search result; and The search results are presented. The method of claim 13, wherein the predicted result includes a first set of items, the search result includes a second set of items, and updating the predicted result to the search result includes: According to the entry editing operation indicated by the update information, the first group of entries arranged in the first order is updated to the second group of entries arranged in the second order, The entry editing operation includes at least one of the following: Deletion of entries, Addition of entries, Replacement of entry. The method according to claim 13 or 14, wherein the first item in the predicted result and the second item in the search result relate to the same object, and updating the predicted result to the search result comprises: In response to the update information indicating that the second result has an object description element different from that of the first entry, the first entry is updated to have the object description element. The method according to claim 11, wherein performing an operation on the prediction result comprises: In response to the supplemental information including a positive indication of the predicted result, presenting the predicted result, or In response to the supplemental information including a negative indication of the predicted result, presenting the predicted result is abandoned. A search device, comprising: a target area determination module configured to, in response to receiving a current query at the server from a client device, determine a target area associated with the current query; A prediction result generating module, configured to generate a prediction result for the current query based at least on the target area historical data corresponding to the target area; A prediction result sending module, configured to send the prediction result to the client device; a comparison module configured to compare the predicted result with the search result in response to a search engine generating search results for the current query; and The supplementary information sending module is configured to send supplementary information related to subsequent processing of the prediction result to the client device based on the comparison result. A search device, comprising: a query sending module, configured to send a current query to a server at a client device; The prediction result receiving module is configured to receive the prediction result from the server. a prediction result of a query, wherein the prediction result is generated based on target area historical data corresponding to a target area, and the target area is associated with the current query; a supplementary information receiving module configured to receive supplementary information related to subsequent processing of the prediction result from the server, the supplementary information being generated based on a comparison between the prediction result and a search result generated by a search engine; and An operation execution module is configured to execute an operation on the prediction result based on the supplementary information. An electronic device, comprising: at least one processing unit; and At least one memory, the at least one memory being coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions, when executed by the at least one processing unit, causing the electronic device to perform the method according to any one of claims 1 to 10 or the method according to any one of claims 11 to 16. A computer-readable storage medium having a computer program stored thereon, wherein the computer program can be executed by a processor to implement the method according to any one of claims 1 to 10 or the method according to any one of claims 11 to 16.