Recalling method, device, equipment, computer medium and computer program product

Abstract

The application discloses a recall method, device, equipment, computer medium and computer program product, which can be applied to various scenes such as video search, information search and intelligent question answering. The method comprises the following steps: in response to receiving a search request for target search content, recalling a recall result of the target search content based on multiple recall strategies to obtain a first recall result; when a first data amount of the first recall result is not less than a preset threshold, selecting a target recall strategy from the multiple recall strategies based on the first recall result, so that when a next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result, recall resources are saved, and recall efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to a recall method, apparatus, device, computer medium, and computer program product. Background Technology

[0002] In video search scenarios, after a user enters their search query, the system sends requests to data sources or channels to retrieve the necessary data. After the initial data retrieval, the system merges the data from different data sources. Data merging is a crucial step, involving integrating data of different formats and sources into a unified dataset. The merged dataset will be used for subsequent analysis and processing.

[0003] After data merging, the system performs data volume verification to ensure that the merged dataset contains a sufficient amount of data to meet the needs of subsequent analysis or processing. Data volume verification typically involves assessing the quantity, quality, and completeness of the data. If the verification finds that the data volume meets the requirements, the recalled data is processed directly. If the verification finds that the data volume does not meet the requirements, a second concurrent data recall is performed to supplement missing data or improve data quality to meet the needs of subsequent processing. In the existing recall model, each recall is based on a full recall strategy, which consumes a lot of time, has high recall costs, and low recall efficiency. Summary of the Invention

[0004] This application provides a recall method, apparatus, device, computer medium, and computer program product to solve the technical problems of high recall time, high recall cost, and low recall efficiency in related technologies.

[0005] On the one hand, a recall method is provided, the method comprising:

[0006] In response to receiving a search request for the target search content, the system recalls the recall results of the target search content based on multiple recall strategies to obtain a first recall result;

[0007] When the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the plurality of recall strategies based on the first recall result, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

[0008] On the other hand, a recall device is provided, the method comprising:

[0009] The recall unit is used to respond to receiving a search request for the target search content, and to recall the recall results of the target search content based on multiple recall strategies to obtain a first recall result;

[0010] The selection unit is configured to select a target recall strategy from the plurality of recall strategies based on the first recall result when the first data volume of the first recall result is not less than a preset threshold, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

[0011] On the other hand, a computer-readable storage medium is provided that stores a computer program adapted for loading by a processor to perform the steps in the method described in any of the above embodiments.

[0012] On the other hand, a computer device is provided, the computer device including a processor and a memory, the memory storing a computer program, the processor executing the steps of the method as described in any of the above embodiments by calling the computer program stored in the memory.

[0013] On the other hand, a computer program product is provided, including computer instructions that, when executed by a processor, implement the steps in the method described in any of the above embodiments.

[0014] This application proposes a scheme: in response to receiving a search request for a target search content, the recall results of the target search content are recalled based on multiple recall strategies to obtain a first recall result; when the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the multiple recall strategies based on the first recall result, so that when a next search request for the target search content is received, the recall results of the target search content are recalled based on the target recall strategy to obtain a second recall result. Through the recall method of this application, when the recall results of the initial recall meet the data volume requirements, a recall strategy with strong recall capabilities can be determined and stored for use in the next recall. Subsequent recalls only need to be based on the stored recall strategy, eliminating the need to use a large number of recall strategies to re-recall data, thus saving recall resources and improving recall efficiency. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating the recall method provided in an embodiment of this application.

[0017] Figure 2 This is a schematic diagram of a scenario for the recall method provided in an embodiment of this application.

[0018] Figure 3 This is a schematic diagram of the recall device provided in an embodiment of this application.

[0019] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] The embodiments of this application can be applied to various scenarios such as video search, information search, and intelligent question answering.

[0022] This application provides a recall method, apparatus, device, computer medium, and computer program product. Specifically, the recall method of this application can be executed by a computer device, which can be a terminal or a server. The terminal can be a smartphone, tablet, laptop, smart voice interaction device, smart home appliance, wearable smart device, aircraft, smart vehicle terminal, etc. The terminal can also include a client, which can be a video client, browser client, or instant messaging client, etc. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0023] First, some of the nouns or terms that appear in the description of the embodiments of this application are explained as follows:

[0024] Data recall is a crucial process in data mining, primarily applied in recommendation systems and search engines. It involves quickly identifying candidate data relevant to user needs from massive datasets. In recommendation systems, the recall phase aims to trigger as many correct results as possible from the entire information set and return these results to the subsequent ranking phase. The main purpose of data recall is to improve the accuracy and relevance of recommendations or search results, thereby meeting user needs. Through recall, the system can filter content matching user interests or needs from large amounts of data, providing a foundation for subsequent ranking and presentation.

[0025] Second check: If the number of people recalled in the first data recall does not meet the required number of people to be recalled, a second supplementary recall will be carried out.

[0026] Data caching: Caching data is an effective way to improve data access speed and performance. Data caching refers to temporarily storing data in memory or other high-speed storage media for fast access and retrieval.

[0027] Response time: The total time from when a user initiates a search to when the data is returned.

[0028] Database: In this application, it refers to the recall database, specifically the collection of datasets obtained after offline data has been classified.

[0029] Recall Strategy: In a recall scenario, a recall strategy refers to a set of methods and rules that trigger as many correct results as possible from the entire set of information. These strategies aim to improve the accuracy and efficiency of recall to meet the needs of users or businesses.

[0030] Fan-out: Request fan-out refers to distributing a request across multiple processing nodes or devices to process it in parallel. This technique is often used to improve system processing capacity and response speed, especially in scenarios requiring the processing of large amounts of data or requests. In this article, fan-out refers to a request being redistributed multiple times.

[0031] Recall loop: A single recall will concurrently distribute multiple recall strategies, and each recall strategy is a recall loop.

[0032] In a recall scenario, "long-tail database" is a concept with a specific meaning. It mainly refers to a collection of relatively niche data that is accessed less frequently but has a wide variety of types. In a recall system, the long-tail database refers to a large number of data sets that are accessed less frequently compared to popular, frequently accessed data sets. This data may be accessed less often due to niche user interests, low timeliness, or other reasons.

[0033] In a user recall scenario, "doc" typically refers to a document, a broad concept encompassing various forms of information carriers. In applications such as search engines, recommendation systems, or advertising systems, a document specifically refers to a data unit indexed and stored by the system, containing information that users may be interested in. In a user recall scenario, a document refers to the basic information unit processed, indexed, and stored by the system. It is the foundational data used by the system to match user queries or needs. Documents come in various forms, including web pages, product information, news articles, video descriptions, and image tags. These different forms of documents all contain rich information to meet diverse user needs.

[0034] In video search scenarios, after a user enters their search query, the system sends requests to data sources or channels to retrieve the necessary data. After the initial data retrieval, the system merges the data from different data sources. Data merging is a crucial step, involving integrating data of different formats and sources into a unified dataset. The merged dataset will be used for subsequent analysis and processing.

[0035] After data merging, the system performs data volume verification to ensure that the merged dataset contains a sufficient amount of data to meet the needs of subsequent analysis or processing. Data volume verification typically involves assessing the quantity, quality, and completeness of the data. If the verification finds that the data volume meets the requirements, the recalled data is processed directly. If the verification finds that the data volume does not meet the requirements, a second concurrent data recall is performed. This step is to supplement missing data or improve data quality to meet the needs of subsequent processing. The second recall supplements some other relevant and long-tail data to ensure the overall number of documents returned to the user's request. In the existing recall model, each recall is based on a full recall strategy, which consumes a lot of time, has high recall costs, and low recall efficiency.

[0036] This application proposes a novel recall method that can identify and store a strong recall strategy when the recall results of the first recall meet the data volume requirements, for use in the next recall. Subsequent recalls only need to be based on the stored recall strategy, eliminating the need to use a large number of recall strategies to recall data again. This reduces response time and resource consumption of low-optimal databases, improves user experience, saves recall resources, and increases recall efficiency.

[0037] This application provides a recall system, which includes a terminal and a server; the terminal and the server are connected via a network, such as a wired or wireless network.

[0038] The terminal can be used to display a graphical user interface. Specifically, the terminal interacts with the user through the graphical user interface, such as downloading and installing a client application, running a program, or displaying a website's graphical user interface.

[0039] In this embodiment, the terminal can be a terminal device that provides a user with an entry point for inputting search content. The user can input target search content in a webpage or client interface displayed on the terminal and initiate a search. This causes the terminal to send a search request for the target search content to a server. The server, in response to receiving the search request, recalls the results of the target search content based on multiple recall strategies to obtain a first recall result. When the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the multiple recall strategies based on the first recall result. This allows the target recall strategy to be used to recall the results of the target search content when a subsequent search request for the target search content is received, resulting in a second recall result.

[0040] Optionally, when the first data volume of the first recall result is less than a preset threshold, the recall results of the target search content can be recalled again based on multiple recall strategies or multiple new recall strategies to obtain supplementary recall results. The aforementioned multiple new recall strategies may include multiple existing recall strategies and new recall strategies. The supplementary recall results and the first recall results can be used together as analytical material to determine the search results for the target search content.

[0041] The solution proposed in this application includes processing methods for different data volumes of the first recall result, which can cover the corresponding processing methods for more data volume situations, thereby improving the comprehensiveness and completeness of the recall mode.

[0042] Optionally, the aforementioned recall strategies may include at least two of the following: full-word matching, full-core-word matching, and arbitrary-core-word matching. Wherein:

[0043] Whole-word matching refers to matching an entire search term or query as a single unit. This matching method requires that the matched content completely contain the search term or query, and that the order and format remain consistent. In recall strategies, whole-word matching is typically used for precise searches, such as when a user wants to find products, articles, or information that exactly match their search term.

[0044] Full keyword matching means that all keywords in the search terms or query must be matched. Keywords are usually key information in the search terms or query, and are crucial for understanding the search intent. In recall strategies, full keyword matching requires that the matched content must contain all keywords, but the order and form of the keywords can vary.

[0045] Arbitrary keyword matching refers to matching any one of the core keywords in the search term or query. This matching method has relatively low requirements; as long as the matched content contains any one of the core keywords in the search term or query, it can be considered relevant. In recall strategies, arbitrary keyword matching is often used to improve the recall rate.

[0046] When you need to find specific information precisely, you can choose whole-word matching.

[0047] When you want to cover more relevant content but don't want it to be too broad, you can choose to match all core keywords.

[0048] When pursuing a high recall rate but accepting a certain degree of precision loss, one can choose to match any core keyword.

[0049] Furthermore, the server can further analyze the second recall results, such as further sorting and filtering the second recall results, to determine the search results, and then send the search results to the terminal for the terminal to display the search results.

[0050] Optionally, in the server-side search architecture, the online part mainly consists of two parts: the retrieval access layer and the recall cluster. The retrieval access layer will use different strategies to obtain and concurrently request different recall strategies and recall clusters according to different scenarios. When the data volume is large, the offline recall cluster will divide the data into databases according to the importance of the data, such as the latest database, the general latest database, P0, P1, P2 and other databases.

[0051] The server can respond to a search request for a target search content based on its search architecture and recall the results of the target search content based on multiple recall strategies.

[0052] It should be noted that the order of description of the following embodiments is not intended to limit the priority of the embodiments.

[0053] This application provides a recall method in various embodiments, which can be executed by a terminal or a server, or by both a terminal and a server. This application uses the execution of the method by a server as an example for illustration.

[0054] Figure 1This is a flowchart illustrating the recall method provided in an embodiment of this application. The method includes the following steps S101-S103:

[0055] S101. In response to receiving a search request for the target search content, recall the recall results of the target search content based on multiple recall strategies to obtain a first recall result;

[0056] The search request for the target search content can be sent by the search terminal device, which can be the aforementioned terminal.

[0057] Optionally, the target search content can be in the form of text information and / or image information. The target search content can be entered through the search box displayed on the search terminal device. When the search terminal device detects the command to start the search, it sends a search request for the target search content to the server.

[0058] Optionally, when the target search content is in the form of image information, the user can open the local photo album by clicking the upload image button inside the search box. The user can then select the corresponding image from the album and confirm the image, thereby inputting the image information.

[0059] Optionally, when the target search content is in the form of image information, the user can open the camera by clicking the camera button inside the search box, take a picture with the camera, and confirm the picture to input the image information.

[0060] The aforementioned upload image button and camera button can be the same button.

[0061] In some optional embodiments of this application, the server may recall the results of the target search content based on multiple databases.

[0062] Optionally, the aforementioned recall strategies can be recall strategies within the same database.

[0063] Optionally, the aforementioned recall strategies can be recall strategies from multiple databases.

[0064] Optionally, the multiple databases may include at least two of the following databases: Time-to-Update Database, General Time-to-Update Database, P0 Database, P1 Database, and P2 Database. Wherein:

[0065] A "timely update database" primarily refers to a collection of data that is the latest, most frequently accessed, or highly relevant to current events. It's used to quickly respond to user queries, especially when the query is closely related to the most up-to-date information. The database is updated frequently to ensure its timeliness and accuracy.

[0066] Compared to the latest data library, the generalized timeliness library encompasses a broader range of data and has relatively lower timeliness requirements. It includes data from a specific period, not just the most recent data. In recall scenarios, the generalized timeliness library provides more comprehensive data support, offering a richer candidate set when user queries extend beyond the latest information. The generalized timeliness library can include a combination of high-quality and timely data. High-quality data refers to data with high accuracy, completeness, consistency, and timeliness. This data excels in multiple aspects, meeting the needs of various analyses, decision-making, and applications.

[0067] High accuracy: High-quality data ensures its content is accurate and avoids data errors or misleading information. This is crucial for scenarios that rely on data for decision-making and analysis, as erroneous data can lead to incorrect conclusions and decisions.

[0068] High completeness: High-quality data encompasses all necessary information, with no omissions or gaps. Completeness ensures the comprehensiveness and availability of data, enabling analysts to access all the necessary information to make accurate judgments.

[0069] High consistency: High-quality data remains consistent across different sources, times, and formats. Consistency helps eliminate inconsistencies and conflicts in the data, improving its reliability and credibility.

[0070] High timeliness: High-quality data is updated in real time or promptly, reflecting the latest situations and trends. Timeliness is especially important for scenarios requiring rapid response and decision-making, as outdated data may not provide valuable information.

[0071] The P0 library can be understood as a collection of the most important and highest-priority data. This data may be what users care about most, what they query most frequently, or what the system considers most valuable to users. Data updates and maintenance in the P0 library are typically given high priority to ensure its accuracy and availability.

[0072] The data in the P1 database may be second only to the P0 database in importance, but it still has a high priority. This data may be frequently queried by users, or data that the system deems valuable to users. Updating and maintaining the data in the P1 database is also relatively important, but its update frequency and priority may be slightly lower than that of the P0 database.

[0073] The data in the P2 database may not be the most relevant or frequently queried information for users, but it still has some value. This data may be used to provide more comprehensive background or supplementary information. The P2 database may be updated and maintained less frequently than the P0 and P1 databases, but it still needs to maintain a certain level of data quality and availability.

[0074] In some optional embodiments of this application, in S101, the recall results of the target search content are recalled based on multiple recall strategies to obtain a first recall result, including the following steps S1011-S1022:

[0075] S1011. Traverse each database in multiple databases, and recall the recall results of the target search content according to each recall strategy in each database to obtain the third recall result corresponding to each database, and then obtain multiple third recall results corresponding to the multiple databases.

[0076] S1012. The plurality of third recall results are used as the first recall result.

[0077] Each database may include one or more recall strategies.

[0078] It should be noted that in this application, the database includes one or more recall strategies, meaning that the database performs data recall based on the one or more recall strategies.

[0079] For a database including one or more recall strategies, recalling the recall results of the target search content for each recall strategy in the database to obtain a third recall result corresponding to the database may include: recalling the recall results of the target search content based on each recall strategy in the database to obtain a fourth recall result corresponding to the recall strategy, and then obtaining one or more fourth recall results corresponding to one or more recall strategies; and using the one or more fourth recall results as the third recall result corresponding to the database.

[0080] In some optional embodiments of this application, when traversing each database in multiple databases, they can be traversed in a certain order. Accordingly, the method further includes: obtaining priority sorting information of the multiple databases; traversing each database in multiple databases includes: traversing each database in multiple databases according to the priority sorting information.

[0081] Optionally, the priority ordering information for multiple databases can be set by relevant personnel, with higher priority databases being traversed first.

[0082] Optionally, the more important the database, the higher its priority.

[0083] Optionally, when multiple databases include: the Time-update database, the General Time-update database, the P0 database, the P1 database, and the P2 database, the priority of these multiple databases from high to low can be: the Time-update database, the General Time-update database, the P0 database, the P1 database, and the P2 database.

[0084] S102. When the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the plurality of recall strategies based on the first recall result, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

[0085] Optionally, in the aforementioned S102, before recalling the results of the target search content based on the target recall strategy, the method further includes:

[0086] Store the identifier of the target recall strategy;

[0087] The target recall strategy is determined based on the identifier of the target recall strategy.

[0088] The storage of the identifier of the target recall strategy enables the rapid determination of the target recall strategy upon receiving the next search request for the target search content, improving data processing speed and thus recall speed and efficiency. Using the recall method of this application, when the recall results of the initial recall meet the data volume requirements, a recall strategy with strong recall capabilities can be determined and stored for use in the next recall. Subsequent recalls only need to be based on the stored recall strategy, eliminating the need to use numerous recall strategies to re-recall data, thus saving recall resources and improving recall efficiency.

[0089] Optionally, the preset threshold can be set by relevant personnel.

[0090] Optionally, setting the first data volume of the first recall result to a preset threshold can be used to ensure that the first recall result meets the minimum requirements for data analysis.

[0091] By setting preset thresholds, we can ensure that the recall results meet the minimum requirements for data analysis, reduce search response time and resource consumption of low-quality databases, and improve the user experience.

[0092] Among them, the target recall strategy is a partial recall strategy among multiple recall strategies.

[0093] Optionally, if the first data volume of the first recall result is less than a preset threshold, no processing is required.

[0094] Optionally, when the first data volume of the first recall result is less than a preset threshold, the recall results of the target search content can be recalled again based on multiple recall strategies or multiple new recall strategies to obtain supplementary recall results. The aforementioned multiple new recall strategies may include multiple existing recall strategies and new recall strategies. The supplementary recall results and the first recall results can be used together as analytical material to determine the search results for the target search content.

[0095] The solution proposed in this application includes processing methods for different data volumes of the first recall result, which can cover the corresponding processing methods for more data volume situations, thereby improving the comprehensiveness and completeness of the recall mode.

[0096] In some optional embodiments of this application, in the aforementioned S102, the step of selecting a target recall strategy from the plurality of recall strategies based on the first recall result includes the following steps S1021-S1022:

[0097] S1021. Analyze the first recall result to obtain at least one recall strategy to be analyzed among the multiple recall strategies involved in the first recall result.

[0098] Optionally, at least one recall strategy to be analyzed in the first recall result refers to the recall strategy used to obtain the first recall result among multiple recall strategies.

[0099] S1022. Determine the target recall strategy based on the at least one recall strategy to be analyzed.

[0100] In some optional embodiments of this application, in the aforementioned S1022, determining the target recall strategy based on the at least one recall strategy to be analyzed includes: using the at least one recall strategy to be analyzed as the target recall strategy.

[0101] In some other optional embodiments of this application, in the aforementioned S1022, determining the target recall strategy according to the at least one recall strategy to be analyzed includes: randomly selecting a preset number of recall strategies to be analyzed from the at least one recall strategy to be analyzed as the target recall strategy.

[0102] In some optional embodiments of this application, the method further includes: sorting the plurality of recall strategies according to their importance to obtain a sorting result;

[0103] Specifically, the sorting rules can be formulated by relevant personnel.

[0104] In some other optional embodiments of this application, in the aforementioned S1022, determining the target recall strategy based on the at least one recall strategy to be analyzed includes: selecting the recall strategy with the lowest importance from the at least one recall strategy to be analyzed as the target recall strategy.

[0105] Optionally, in the aforementioned S1022, determining the target recall strategy based on the at least one recall strategy to be analyzed includes the following steps S01-S02:

[0106] S01. Obtain the recall strategy with the lowest importance among the at least one recall strategy to be analyzed;

[0107] S02. Select at least one recall strategy from the ranking results that has an importance level not lower than the recall strategy with the lowest importance level as the target recall strategy.

[0108] In some other optional embodiments of this application, in S102, selecting a target recall strategy from the plurality of recall strategies based on the first recall result includes the following steps S11-S13:

[0109] S11. Obtain the recall results to be analyzed from the first recall results, where the amount of data is equal to the preset threshold.

[0110] Optionally, the recall result in which the amount of data recalled first in the first recall result is equal to the preset threshold can be used as the recall result to be analyzed.

[0111] Optionally, the recall result with the best quality data volume in the first recall result that is equal to the preset threshold can be used as the recall result to be analyzed.

[0112] Optionally, any recall result in the first recall result whose data volume is equal to the preset threshold can be used as the recall result to be analyzed.

[0113] S12. Analyze the recall results to be analyzed to obtain at least one recall strategy involved in the recall results to be analyzed among the multiple recall strategies.

[0114] Optionally, at least one recall strategy involved in the recall results to be analyzed refers to the recall strategy used to obtain the recall results among multiple recall strategies.

[0115] S13. Determine the target recall strategy based on the at least one recall strategy to be analyzed.

[0116] In some optional embodiments of this application, in the aforementioned S13, determining the target recall strategy based on the at least one recall strategy to be analyzed includes: using the at least one recall strategy to be analyzed as the target recall strategy.

[0117] In some other optional embodiments of this application, in the aforementioned S13, determining the target recall strategy according to the at least one recall strategy to be analyzed includes: randomly selecting a preset number of recall strategies to be analyzed from the at least one recall strategy to be analyzed as the target recall strategy.

[0118] In some other optional embodiments of this application, in the aforementioned S13, determining the target recall strategy based on the at least one recall strategy to be analyzed includes: selecting the recall strategy with the lowest importance from the at least one recall strategy to be analyzed as the target recall strategy.

[0119] At this point, the target recall strategy is only one of many recall strategies.

[0120] Optionally, in the aforementioned S13, determining the target recall strategy based on the at least one recall strategy to be analyzed includes the following steps S1-S1:

[0121] S1. Obtain the recall strategy with the lowest importance among the at least one recall strategy to be analyzed;

[0122] S2. Select at least one recall strategy from the ranking results that has an importance level not lower than the recall strategy with the lowest importance level as the target recall strategy.

[0123] Optionally, the identifier of the target recall strategy can be cached.

[0124] Optionally, the method further includes: when the identifier of the target recall strategy is stored locally, recalling the recall results of the target search content based on the target recall strategy to obtain a second recall result; when the identifier of the target recall strategy is not stored locally, recalling the recall results of the target search content based on the multiple recall strategies to obtain a target recall result.

[0125] Different processing methods are applied depending on whether the identifier of the target recall strategy is stored locally, which further improves the reliability of the recall process.

[0126] In some optional embodiments of this application, the method further includes: when the second data volume of the second recall result is less than the preset threshold, upon receiving a next search request for the target search content, recalling the recall result of the target search content based on the multiple recall strategies. Here, the next search request can be considered as a new search request for the target search content.

[0127] When the second data volume of the second recall result is less than the preset threshold, it indicates that the recall capability of the stored target recall strategy is poor. At this time, the identifier of the target recall strategy can be cleared.

[0128] It should be noted that when storing the identifier of the target recall strategy, the target search content and the identifier of the target recall strategy should be stored together, and there should be a corresponding relationship between the target search content and the identifier of the target recall strategy.

[0129] Furthermore, the server can store multiple search terms and the identifiers of the corresponding recall strategies for each search term. Each search term and its corresponding recall strategy identifier constitute a data pair.

[0130] In some optional embodiments of this application, the ordering rule for the aforementioned recall strategies is that the higher the importance, the higher the ranking. The identifier for the recall strategy can be letters or Arabic numerals.

[0131] Furthermore, when the target recall strategy is determined by selecting the recall strategy with the lowest importance from at least one recall strategy to be analyzed as the target recall strategy, the aforementioned recall of the target search content based on the target recall strategy to obtain a second recall result includes: obtaining the recall strategy ranked before the target recall strategy from the plurality of recall strategies; and recalling the target search content based on the recall strategy ranked before the target recall strategy and the target recall strategy to obtain a second recall result.

[0132] Furthermore, since this application involves the sorting of recall strategies, the sorting of recall data can be kept consistent with the recall strategy during the process of recalling data based on the target recall strategy, thereby improving the data processing efficiency of recall data.

[0133] Furthermore, this application provides a specific embodiment in which the determination method of the target recall strategy is as described in S11-S13 above, and the specific determination method of the target recall strategy is as follows: when the at least one recall strategy to be analyzed is taken as the target recall strategy, when the identifier of the recall strategy is an Arabic numeral, and when the sorting rule of multiple recall strategies is that the smaller the identifier, the greater the importance, if the number of multiple recall strategies is 22, and the search term is TV series ABC, for the first search request for TV series ABC, it is necessary to recall the recall results of TV series ABC based on multiple recall strategies to obtain recall data. After recall, all recall data are merged. It is found that the returned document only uses the data recalled by the first 7 recall strategies, and the amount of recall data meets the preset threshold. Then, [TV series ABC, 1234567] is cached locally. After receiving the second request for TV series ABC, the recall strategy identifier in the cache is first obtained, and then the first 7 recall strategies corresponding to the recall strategy identifier are used for recall to reduce invalid requests downstream. If the retrieved data based on the seven recall strategies does not meet the preset threshold, the cache will be actively cleared and the first retrieval process will be repeated on the next request.

[0134] Furthermore, this application provides another specific embodiment. When the determination method of the target recall strategy is as described in S11-S13 above, and the specific determination method of the target recall strategy is as follows: when the recall strategy with the lowest importance is selected from the at least one recall strategy to be analyzed as the target recall strategy, when the identifier of the recall strategy is an Arabic numeral, and when the sorting rule of multiple recall strategies is that the smaller the identifier, the greater the importance, if the number of multiple recall strategies is 22, and the search term is TV series ABC, for the first search request for TV series ABC, the recall results of TV series ABC need to be recalled based on multiple recall strategies to obtain recall data. After recall, all recall data are merged. It is found that the returned document only uses the data recalled by the first 7 recall strategies, and the amount of recall data meets the preset threshold. Then, [TV series ABC, 7] is cached locally. After the second request for TV series ABC is received, the recall strategy identifier in the cache is obtained first, and then the recall is performed based on the first 7 recall strategies to reduce invalid requests downstream.

[0135] To describe the scheme of this application in more detail, the following is in conjunction with the accompanying drawings. Figure 2 The scheme of this application will be further explained.

[0136] After a user inputs their target search terms through a terminal, the terminal sends a search request for those terms to the server. The server parses the search request through the retrieval access layer to obtain the request parsing results. It then checks its local cache for a recall strategy identifier. If not, it retrieves the target search terms based on multiple recall strategies to obtain a first recall result. This can be done through parallel distribution, retrieving the target search terms based on multiple recall strategies. If a recall strategy is found, it retrieves the target search terms based on the recall strategy corresponding to the identifier.

[0137] The aforementioned recall strategies can be contained in multiple databases, and each database can include multiple data slices. Figure 2 In this context, Set-0, Set-N, and Set-M refer to data slices.

[0138] This application proposes a scheme: in response to receiving a search request for a target search content, the recall results of the target search content are recalled based on multiple recall strategies to obtain a first recall result; when the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the multiple recall strategies based on the first recall result, so that when a next search request for the target search content is received, the recall results of the target search content are recalled based on the target recall strategy to obtain a second recall result. Through the recall method of this application, when the recall results of the initial recall meet the data volume requirements, a recall strategy with strong recall capabilities can be determined and stored for use in the next recall. Subsequent recalls only need to be based on the stored recall strategy, eliminating the need to use a large number of recall strategies to re-recall data, thus saving recall resources and improving recall efficiency.

[0139] All of the above technical solutions can be combined in any way to form optional embodiments of this application, and will not be described in detail here.

[0140] To facilitate better implementation of the recall method in this application, this application also provides a recall device. Please refer to [link to relevant documentation]. Figure 3 , Figure 3 This is a schematic diagram of the recall device provided in an embodiment of this application.

[0141] The recall device includes:

[0142] The recall unit 31 is used to respond to receiving a search request for the target search content, and to recall the recall results of the target search content based on multiple recall strategies to obtain a first recall result;

[0143] Selection unit 32 is configured to select a target recall strategy from the plurality of recall strategies based on the first recall result when the first data volume of the first recall result is not less than a preset threshold, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

[0144] In some optional embodiments of this application, before recalling the results of the target search content based on the target recall strategy, the device is further configured to:

[0145] Store the identifier of the target recall strategy;

[0146] The target recall strategy is determined based on the identifier of the target recall strategy.

[0147] In some optional embodiments of this application, the apparatus is further used for:

[0148] When the identifier of the target recall strategy is stored locally, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result;

[0149] When the identifier of the target recall strategy is not stored locally, the recall results of the target search content are retrieved based on the multiple recall strategies to obtain the target recall results.

[0150] In some optional embodiments of this application, when the device is used to select a target recall strategy from the plurality of recall strategies based on the first recall result, it is specifically used for:

[0151] Analyze the first recall result to obtain at least one recall strategy to be analyzed among the multiple recall strategies involved in the first recall result;

[0152] The target recall strategy is determined based on the at least one recall strategy to be analyzed.

[0153] In some optional embodiments of this application, when the device is used to select a target recall strategy from the plurality of recall strategies based on the first recall result, it is specifically used for:

[0154] From the first recall result, obtain the recall result to be analyzed with a data volume equal to the preset threshold.

[0155] The recall results to be analyzed are analyzed to obtain at least one recall strategy involved in the recall results to be analyzed among the multiple recall strategies.

[0156] The target recall strategy is determined based on the at least one recall strategy to be analyzed.

[0157] In some optional embodiments of this application, the apparatus is further configured to: sort the plurality of recall strategies according to their importance to obtain a sorting result;

[0158] When the device is used to determine a target recall strategy based on the at least one recall strategy to be analyzed, it is specifically used for:

[0159] Obtain the recall strategy with the lowest importance among the at least one recall strategy to be analyzed;

[0160] At least one of the recall strategies in the ranking results that has an importance level no lower than the recall strategy with the lowest importance level shall be selected as the target recall strategy.

[0161] In some optional embodiments of this application, when the device is used to recall the recall results of the target search content based on multiple recall strategies to obtain a first recall result, it is specifically used for:

[0162] The system iterates through each database in multiple databases, and retrieves the retrieval results of the target search content for each retrieval strategy in each database, thereby obtaining the third retrieval results corresponding to each database, and then obtaining multiple third retrieval results corresponding to the multiple databases.

[0163] The multiple third recall results are used as the first recall result.

[0164] In some optional embodiments of this application, the apparatus is further used for:

[0165] Obtain the priority sorting information of the multiple databases;

[0166] Traversing each database in multiple databases includes: traversing each database in multiple databases according to the priority sorting information.

[0167] In some optional embodiments of this application, the apparatus is further used for:

[0168] When the second data volume of the second recall result is less than the preset threshold, upon receiving the next search request for the target search content, the recall result of the target search content is recalled based on the multiple recall strategies.

[0169] Each unit of the aforementioned recall device can be implemented entirely or partially through software, hardware, or a combination thereof. These units can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can invoke and execute the corresponding operations of each unit.

[0170] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0171] The aforementioned recall device can be integrated into a terminal or server that has storage and a processor and thus computing power, or the recall device can be the terminal or server itself.

[0172] Optionally, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0173] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. The computer device may be... Figure 1 The terminal or server shown. For example... Figure 4 As shown, the computer device 400 may include: a communication interface 401, a memory 402, a processor 403, and a communication bus 404. The communication interface 401, memory 402, and processor 403 communicate with each other via the communication bus 404. The communication interface 401 is used for data communication between the computer device 400 and external devices. The memory 402 can be used to store software programs and modules, and the processor 403 runs the software programs and modules stored in the memory 402, such as the software programs for the corresponding operations in the foregoing method embodiments.

[0174] Optionally, the processor 403 can invoke software programs and modules stored in the memory 402 to perform the following operations:

[0175] In response to receiving a search request for the target search content, the system recalls the recall results of the target search content based on multiple recall strategies to obtain a first recall result;

[0176] When the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the plurality of recall strategies based on the first recall result, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

[0177] This application also provides a computer-readable storage medium for storing a computer program. This computer-readable storage medium can be applied to a computer device, and the computer program causes the computer device to execute the corresponding processes in the methods described in the embodiments of this application; for brevity, these will not be elaborated further here.

[0178] This application also provides a computer program product including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the corresponding processes in the methods described in the embodiments of this application. For brevity, these details will not be elaborated further here.

[0179] This application also provides a computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the corresponding processes in the methods described in the embodiments of this application. For the sake of brevity, these details will not be elaborated further here.

[0180] It should be understood that the processor in the embodiments of this application may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method embodiments can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor described above can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0181] It is understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DR RAM). It should be noted that the memory used in the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0182] It should be understood that the above-described memory is exemplary and not a limiting description. For example, the memory in the embodiments of this application may also be static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DR RAM), etc. That is to say, the memory in the embodiments of this application is intended to include, but is not limited to, these and any other suitable types of memory.

[0183] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0184] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0185] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0186] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0187] In addition, the functional units in the embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0188] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer or a server) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0189] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A recall method, characterized in that, The method includes: In response to receiving a search request for the target search content, the system recalls the recall results of the target search content based on multiple recall strategies to obtain a first recall result; When the first data volume of the first recall result is not less than a preset threshold, a target recall strategy is selected from the plurality of recall strategies based on the first recall result, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

2. The method according to claim 1, characterized in that, Before recalling the results of the target search content based on the target recall strategy, the method further includes: Store the identifier of the target recall strategy; The target recall strategy is determined based on the identifier of the target recall strategy.

3. The method according to claim 1, characterized in that, The method further includes: When the identifier of the target recall strategy is stored locally, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result; When the identifier of the target recall strategy is not stored locally, the recall results of the target search content are retrieved based on the multiple recall strategies to obtain the target recall results.

4. The method according to claim 1, characterized in that, The step of selecting a target recall strategy from the plurality of recall strategies based on the first recall result includes: Analyze the first recall result to obtain at least one recall strategy to be analyzed among the multiple recall strategies involved in the first recall result; The target recall strategy is determined based on the at least one recall strategy to be analyzed.

5. The method according to claim 1, characterized in that, The step of selecting a target recall strategy from the plurality of recall strategies based on the first recall result includes: From the first recall result, obtain the recall result to be analyzed with a data volume equal to the preset threshold. The recall results to be analyzed are analyzed to obtain at least one recall strategy involved in the recall results to be analyzed among the multiple recall strategies. The target recall strategy is determined based on the at least one recall strategy to be analyzed.

6. The method according to claim 4 or 5, characterized in that, The method further includes: ranking the multiple recall strategies according to their importance to obtain a ranking result; The step of determining the target recall strategy based on the at least one recall strategy to be analyzed includes: Obtain the recall strategy with the lowest importance among the at least one recall strategy to be analyzed; At least one of the recall strategies in the ranking results that has an importance level no lower than the recall strategy with the lowest importance level shall be selected as the target recall strategy.

7. The method according to claim 1, characterized in that, The first recall result is obtained by recalling the target search content based on multiple recall strategies, including: The system iterates through each database in multiple databases, and retrieves the retrieval results of the target search content for each retrieval strategy in each database, thereby obtaining the third retrieval results corresponding to each database, and then obtaining multiple third retrieval results corresponding to the multiple databases. The multiple third recall results are used as the first recall result.

8. The method according to claim 7, characterized in that, The method further includes: Obtain the priority sorting information of the multiple databases; Traversing each database in multiple databases includes: traversing each database in multiple databases according to the priority sorting information.

9. The method according to claim 3, characterized in that, The method further includes: When the second data volume of the second recall result is less than the preset threshold, upon receiving the next search request for the target search content, the recall result of the target search content is recalled based on the multiple recall strategies.

10. A recall device, characterized in that, The device includes: The recall unit is used to respond to receiving a search request for the target search content, and to recall the recall results of the target search content based on multiple recall strategies to obtain a first recall result; The selection unit is configured to select a target recall strategy from the plurality of recall strategies based on the first recall result when the first data volume of the first recall result is not less than a preset threshold, so that when the next search request for the target search content is received, the recall result of the target search content is recalled based on the target recall strategy to obtain a second recall result.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted for loading by a processor to perform the steps of the recall method as described in any one of claims 1-9.

12. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing a computer program, and the processor executing the steps of the recall method according to any one of claims 1-9 by calling the computer program stored in the memory.

13. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the recall method according to any one of claims 1-9.

Images