Search algorithm optimization method, device, equipment and storage medium

Abstract

The present disclosure relates to a search algorithm optimization method, apparatus, device, and storage medium. The method includes: determining at least one candidate optimization solution corresponding to a current search algorithm based on a set of relevant search results for a search term, wherein the set of relevant search results includes multiple relevant search results; determining multiple candidate search algorithms from a set of search algorithms based on a target optimization solution, wherein the target optimization solution is part or all of the at least one candidate optimization solution; and determining a target optimization algorithm based on the multiple candidate search algorithms. This method can automatically optimize the search algorithm and improve the optimization efficiency of the search algorithm.

Description

Technical Field

[0001] The present disclosure relates to the field of Internet technology, and in particular to a search algorithm optimization method, apparatus, device, and storage medium. Background Art

[0002] A search engine system automatically collects information from the internet, organizes it, and then presents it to users for query. Users enter their needs through a search box, and the system uses a search algorithm to identify and display content that meets their needs. Continuous optimization of the search engine system after launch is necessary to ensure that the content displayed by the optimized system better meets the user's current needs.

[0003] In the existing technology, the optimization of search engine systems can be divided into the optimization of search algorithms and the optimization of search results. For the optimization of search algorithms, problems with the current search algorithms can be found through manual analysis of search results, and then optimization strategies can be formulated and technical personnel can optimize the search algorithms; for the optimization of search results, search results can be optimized and adjusted based on online user data.

[0004] However, the former search algorithm optimization method results in low work efficiency, while the latter method can only automatically optimize the search results but not the algorithm. Therefore, automatic optimization of the search algorithm cannot be achieved. Summary of the Invention

[0005] In order to solve the above technical problems, the present disclosure provides a search algorithm optimization method, device, equipment and storage medium, which can realize automatic optimization of the search algorithm and improve the optimization efficiency of the search algorithm.

[0006] In a first aspect, the present disclosure provides a method for optimizing a search algorithm, comprising:

[0007] Determining at least one candidate optimization solution corresponding to a current search algorithm based on a set of relevant search results for the search term, wherein the set of relevant search results includes a plurality of relevant search results;

[0008] Determining a plurality of candidate search algorithms from a set of search algorithms according to a target optimization scheme, wherein the target optimization scheme is part or all of the at least one candidate optimization scheme;

[0009] A target optimization algorithm is determined based on the multiple candidate search algorithms.

[0010] Optionally, before determining at least one candidate optimization solution corresponding to the current search algorithm based on the set of search results related to the search term, the method further includes:

[0011] Input the search term into the current search algorithm to obtain a search result set corresponding to the search term;

[0012] Determining the relevance of all search results in the search result set based on feedback data information of all search results;

[0013] The relevant search result set is determined from the search result set according to the relevance of all the search results and a preset relevance.

[0014] Optionally, determining at least one candidate optimization solution corresponding to the current search algorithm based on a set of search results related to the search term includes:

[0015] Determining optimization features corresponding to the search term classification based on a set of search results related to the search term;

[0016] Determine the at least one candidate optimization solution based on the optimization features corresponding to the search term classification.

[0017] Optionally, determining the optimization features corresponding to the search term classification based on the set of search results related to the search term includes:

[0018] Determining the optimized feature corresponding to the search term according to the score of each feature of the relevant search result set in the current search algorithm and the score of each feature of the search result set in the current search algorithm;

[0019] According to the optimization features corresponding to the search terms, the optimization features corresponding to the search term categories are determined.

[0020] Optionally, the step of determining multiple candidate search algorithms from a search algorithm set according to the target optimization scheme includes:

[0021] The plurality of candidate search algorithms are determined according to the optimization features corresponding to the search term classification in the target optimization solution and the relevance between all search algorithms in the search algorithm set and each feature.

[0022] Optionally, determining a target optimization algorithm based on the multiple candidate search algorithms includes:

[0023] Determining a weight according to the importance of each of the plurality of candidate search algorithms;

[0024] Fusion forms multiple candidate optimization algorithms based on the multiple candidate search algorithms, the current search algorithm, and the weights;

[0025] The target optimization algorithm is determined from the multiple candidate optimization algorithms.

[0026] Optionally, determining the target optimization algorithm from the multiple candidate optimization algorithms includes:

[0027] Push the multiple candidate optimization algorithms online and obtain feedback data information within a preset time period;

[0028] Determine the target optimization algorithm based on the feedback data information within the preset time period corresponding to the multiple candidate optimization algorithms

[0029] In a second aspect, the present disclosure provides an optimization device for a search algorithm, comprising:

[0030] A first determination module is configured to determine at least one candidate optimization solution corresponding to a current search algorithm based on a set of relevant search results for a search term, wherein the set of relevant search results includes a plurality of relevant search results;

[0031] A second determination module is configured to determine a plurality of candidate search algorithms from a set of search algorithms according to a target optimization scheme, wherein the target optimization scheme is part or all of the at least one candidate optimization scheme;

[0032] The third determination module is used to determine a target optimization algorithm based on the multiple candidate search algorithms.

[0033] In a third aspect, the present disclosure provides an electronic device, comprising: a processor, wherein the processor is configured to execute a computer program stored in a memory, wherein the computer program implements the steps of the method described in the first aspect when executed by the processor.

[0034] In a fourth aspect, the present disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in the first aspect.

[0035] In the technical solution provided by the present disclosure, at least one candidate optimization scheme corresponding to the current search algorithm is determined based on a set of relevant search results of the search term, and the relevant search result set includes multiple relevant search results; based on the target optimization scheme, multiple candidate search algorithms are determined from the search algorithm set, and the target optimization scheme is part or all of at least one candidate optimization scheme; based on multiple candidate search algorithms, the target optimization algorithm is determined, which can realize automatic optimization of the search algorithm and improve the optimization efficiency of the search algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0037] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0038] Figure 1 A schematic diagram of a flow chart of a search algorithm optimization method provided by the present disclosure;

[0039] Figure 2 A schematic diagram of a flow chart of another search algorithm optimization method provided by the present disclosure;

[0040] Figure 3 A schematic flow chart of another search algorithm optimization method provided by the present disclosure;

[0041] Figure 4 A schematic flow chart of another search algorithm optimization method provided by the present disclosure;

[0042] Figure 5 A schematic flow chart of another search algorithm optimization method provided by the present disclosure;

[0043] Figure 6 A schematic flow chart of another search algorithm optimization method provided by the present disclosure;

[0044] Figure 7 A schematic flow chart of another search algorithm optimization method provided by the present disclosure;

[0045] Figure 8 A schematic diagram of the structure of an optimization device for a search algorithm provided by the present disclosure;

[0046] Figure 9 A schematic structural diagram of an electronic device provided by the present disclosure. DETAILED DESCRIPTION

[0047] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0049] Figure 1 A flowchart of an optimization method for a search algorithm provided by the present disclosure is shown in FIG. Figure 1Shown, including:

[0050] S102: Determine at least one candidate optimization solution corresponding to the current search algorithm based on a set of search results related to the search term.

[0051] The relevant search result set includes multiple relevant search results.

[0052] Based on the current search algorithm, multiple search results corresponding to the search terms can be recalled, and at least some of the multiple search results correspond to feedback data information. The feedback data information can be user viewing and clicking behavior data, and / or manual relevance annotations of the search results. The feedback data information includes the search terms, multiple search results determined to be relevant or irrelevant under each search term, and the relevance score of each search result. The relevance score is obtained through calculation, for example, it can be obtained by weighted average of click-through rate, dwell time, conversions such as orders completed in the results, and manual relevance scores. Among all search results, if the relevance score is greater than the preset relevance, the search result is considered to be a relevant search result. If the relevance score is less than or equal to the preset relevance, the search result is considered to be an irrelevant search result.

[0053] For example, based on the current search algorithm, the search results corresponding to the search term "Qilixiang" are "Qilixiang plant", "Qilixiang song" and "Jay Chou". Based on the user's click-through rate, stay time, etc., the relevance scores of "Qilixiang song" and "Jay Chou" are both greater than the preset relevance, then "Qilixiang song" and "Jay Chou" are the relevant search results for the search term "Qilixiang".

[0054] Based on multiple relevant search results of the search term and all search results of the search term in the current search algorithm, the optimization direction of the current search algorithm can be determined, thereby determining at least one candidate optimization scheme. For example, based on the above embodiment, the search results obtained by the search term "Seven-year-old Fragrant Osmanthus fragrans" based on the current search algorithm are "plant Seven-year-old Fragrant Osmanthus fragrans", "song Seven-year-old Fragrant Osmanthus fragrans" and "Jay Chou", and the relevant search results of the search term "Seven-year-old Fragrant Osmanthus fragrans" are "song Seven-year-old Fragrant Osmanthus fragrans" and "Jay Chou". Obviously, it is necessary to increase the proportion of feature 1 corresponding to "song Seven-year-old Fragrant Osmanthus fragrans", increase the proportion of feature 2 corresponding to "Jay Chou", and reduce the proportion of feature 3 corresponding to "plant Seven-year-old Fragrant Osmanthus fragrans". In this way, the candidate optimization scheme corresponding to the current search algorithm can be to increase the proportion of feature 1 by 10%-30%, increase the proportion of feature 2 by 10%-40%, and reduce the proportion of feature 3 by 20%-70%.

[0055] S104: Determine multiple candidate search algorithms from the search algorithm set according to the target optimization solution.

[0056] The target optimization solution is part or all of the at least one candidate optimization solution.

[0057] A target optimization solution can be manually determined from at least one candidate optimization solution to determine whether the optimization direction is correct. In this case, the candidate optimization solution can be directly determined as the target optimization solution, or the candidate optimization solution can be fine-tuned to determine the target optimization solution. The target optimization solution can include optimized features and their optimization weights. For example, based on the above embodiment, the candidate optimization solution can be determined as the target optimization solution by increasing the weight of feature 1 by 10%-30%, the weight of feature 2 by 10%-40%, and the weight of feature 3 by 20%-70%. The candidate optimization solution can also be fine-tuned to increase the weight of feature 1 by 15%-20%, increase the weight of feature 2 by 10%-25%, and reduce the weight of feature 3 by 25%-45%.

[0058] The search algorithm set includes a variety of complete search algorithm models and formulas available for use, as well as algorithm models and formulas for partial features. All search algorithms in the search algorithm set are derived from end-to-end search algorithms defined by technicians, end-to-end algorithm models trained by deep learning algorithms, search algorithms manually formulated for specific features or modules, and models trained by deep learning algorithms for specific features or modules. Each search algorithm in the search algorithm set has its relevance to each feature defined in advance. Since the target optimization solution provides the optimized features and their optimized weights, multiple search algorithms that match the optimized features and their optimized weights in the target optimization solution can be found from the search algorithm set as candidate search algorithms. For example, based on the above embodiment, based on search algorithm 1, the weight of feature 1 can be increased by 15%-20%, the weight of feature 2 can be increased by 10%-25%, and based on search algorithm 2, the weight of feature 3 can be reduced by 25%-45%. In this way, search algorithm 1 and search algorithm 2 can be determined as candidate search algorithms.

[0059] S106: Determine a target optimization algorithm based on the multiple candidate search algorithms.

[0060] By integrating parts of multiple candidate search algorithms with the current search algorithm, we can obtain a candidate optimization algorithm. This way, we can generate multiple candidate optimization algorithms. These multiple candidate optimization algorithms are pushed online, and each candidate optimization algorithm is assigned a certain number of users. After a certain period of feedback, we obtain feedback data for that period.

[0061] Based on the user click-through rate in the feedback data, the candidate optimization algorithm with the highest user click-through rate can be determined as the target optimization algorithm. The candidate optimization algorithm with the highest user click-through rate can also be repeatedly executed to retain it, and multiple candidate optimization algorithms similar to the candidate optimization algorithm with the highest user click-through rate can be determined. Feedback data information is obtained within a certain period of time. After a long period of iteration, the final target optimization algorithm can be obtained. The target optimization algorithm is the optimized search algorithm in the search engine system. In this way, the search algorithm can be automatically optimized, improving the optimization efficiency of the search algorithm.

[0062] In this embodiment, at least one candidate optimization scheme corresponding to the current search algorithm is determined based on a set of relevant search results for the search term, and the relevant search result set includes multiple relevant search results; based on the target optimization scheme, multiple candidate search algorithms are determined from the search algorithm set, and the target optimization scheme is part or all of at least one candidate optimization scheme; based on multiple candidate search algorithms, the target optimization algorithm is determined, which can realize automatic optimization of the search algorithm and improve the optimization efficiency of the search algorithm.

[0063] Figure 2 A schematic diagram of a flow chart of another search algorithm optimization method provided by the present disclosure, Figure 2 for Figure 1 Based on the embodiment shown, before executing S102, the following steps are further included:

[0064] S1011: Input the search term into the current search algorithm to obtain a search result set corresponding to the search term.

[0065] The search term is input into the current search algorithm. After the current search algorithm processes the search term, the search results corresponding to the search term can be recalled. For the same search term, different search algorithms may recall different search results with different contents and / or rankings.

[0066] S1012: Determine the relevance of all search results in the search result set based on the feedback data information of all search results.

[0067] Feedback data can include user viewing and click behavior data and / or manual relevance annotations of search results. Feedback data includes the search terms, multiple search results determined to be relevant or irrelevant for each search term, and the relevance score for each search result. The relevance score is calculated, for example, by taking a weighted average of click-through rate, dwell time, conversions (such as orders) completed within the results, and manual relevance scores.

[0068] S1013: Determine the relevant search result set from the search result set according to the relevance of all search results and a preset relevance.

[0069] Among all search results, if the relevance score is greater than the preset relevance, the search result is considered to be a relevant search result. If the relevance score is less than or equal to the preset relevance, the search result is considered to be an irrelevant search result. In this way, all search results in the search result set with a relevance score greater than the preset relevance are relevant search results, and all relevant search results constitute the relevant search result set.

[0070] In this embodiment, by inputting the search term into the current search algorithm, a search result set corresponding to the search term is obtained; based on the feedback data information of all search results in the search result set, the relevance of all search results is determined; based on the relevance of all search results and the preset relevance, a relevant search result set is determined from the search result set. Obviously, the relevance of the search results in the relevant search result set is higher. Therefore, optimizing the search algorithm based on the relevant search result set can improve the accuracy of the search results.

[0071] Figure 3 A flowchart of another search algorithm optimization method is provided for the present disclosure. Figure 3 for Figure 1 Based on the illustrated embodiment, a possible implementation method for executing S102 is specifically described as follows:

[0072] S1021: Determine optimization features corresponding to the search term classification based on a set of search results related to the search term.

[0073] Each search term corresponds to a set of relevant search results. All search terms are categorized into different search term categories. A search term category can include multiple search terms of the same type. Based on the optimization features corresponding to all search terms within a search term category, the optimization features corresponding to that search term category can be determined. Different search term categories may correspond to different optimization features. For example, the price feature of shopping category search terms can be increased by 10%-30%, and the popularity feature of news category search terms can be increased by 5%-7%.

[0074] S1022: Determine the at least one candidate optimization solution based on the optimization features corresponding to the search term classification.

[0075] The candidate optimization scheme includes at least one optimization feature and the optimization proportion of each optimization feature. For example, based on the above embodiment, the candidate optimization scheme can be: increasing the product price feature of the shopping category search term by 10%-30%, and increasing the popularity feature of the news category search term by 5%-7%.

[0076] In this embodiment, the optimization features corresponding to the search word classification are determined based on the search word and its related search result set; at least one candidate optimization scheme is determined based on the optimization features corresponding to the search word classification. The optimization features can be determined based on the search word category, and there is no need to determine the optimization features for each search word. The number of optimization features can be reduced, thereby reducing the computational complexity of the target optimization algorithm and improving the optimization efficiency of the search algorithm.

[0077] Figure 4 A flowchart of another search algorithm optimization method provided by the present disclosure is shown below. Figure 4 for Figure 3 Based on the illustrated embodiment, a possible implementation method for executing S1021 is specifically described as follows:

[0078] S201 : Determine an optimized feature corresponding to the search term according to the score of each feature of the relevant search result set in the current search algorithm and the score of each feature of the search result set in the current search algorithm.

[0079] The search algorithm is associated with multidimensional features, which can be at least one of content features, applicable scenario features, device features, user features and search behavior features. Among them, content features can be at least one of text relevance, semantic relevance and popularity, applicable scenario features can be at least one of home scenes, travel scenes and shopping scenes, device features can be a large-screen car system, or a voice-screenless system, etc., and search behavior features can be at least one of voice search, short text search and question-and-answer search.

[0080] It should be noted that this embodiment only uses text relevance, semantic relevance and popularity as examples to illustrate content features. In actual applications, content features can also be other content such as search volume, and this embodiment does not impose specific restrictions on this.

[0081] This embodiment only uses home scenes, travel scenes and shopping scenes as examples to illustrate the applicable scene features. In actual applications, other scenes such as service scenes can also be used, and this embodiment does not impose specific restrictions on this.

[0082] This embodiment only uses voice search, short text search, and question-and-answer search as examples to illustrate the search behavior characteristics. In actual applications, other search behaviors such as image search may also be used, and this embodiment does not impose specific restrictions on this.

[0083] For each feature, the score of the relevant search result set and the score of the search result set are determined. Based on the scores of the relevant search result set and the scores of the search result set under the same feature, the optimized feature and the proportion of the optimized feature can be determined. For example, the relevant search result set for search term A is B, and the search result set for search term A is C. For feature 1, the score of relevant search result set B is lower than the score of search result set C, and for feature 2, the score of relevant search result set B is higher than the score of search result set C. In this way, the optimized features can be determined to be feature 1 and feature 2, and the proportion of feature 1 needs to be reduced, while the proportion of feature 2 needs to be increased. In addition, based on the specific scores, the optimization proportion of feature 1 and the optimization proportion of feature 2 can be determined.

[0084] S202: Determine the optimization features corresponding to the search term category according to the optimization features corresponding to the search term.

[0085] For each search term, its corresponding optimization feature can be determined. By categorizing all search terms, different search term categories can be obtained. For example, search term categories can be shopping categories, news categories, video categories, etc. The optimization features and optimization ratios corresponding to all search terms within a search term category are combined to calculate the optimization features and optimization ratios corresponding to the search term category. For example, it is calculated that search terms in the shopping category increase the price feature by 10%-30%, and search terms in the news category increase the popularity feature by 5%-7%.

[0086] Figure 5 A flowchart of another search algorithm optimization method provided by the present disclosure is shown below. Figure 5 for Figure 1 Based on the illustrated embodiment, a possible implementation method for executing S104 is specifically described as follows:

[0087] S104', determining the plurality of candidate search algorithms according to the optimization features corresponding to the search term classification in the target optimization solution and the relevance between all search algorithms in the search algorithm set and each feature.

[0088] The target optimization plan includes the optimization features corresponding to different search term categories, as well as the optimization ratio of each optimization feature. For each search algorithm in the search algorithm set, the relevance between the search algorithm and each feature can be determined. Based on the optimization features and their optimization ratios, multiple candidate search algorithms that match the target optimization plan can be identified from all search algorithms. For example, if the target optimization plan is to increase the product price feature of shopping category search terms by 5%-15% and the popularity feature of news category search terms by 5%-7%, and it is determined that search algorithm 1 and search algorithm 2 match the target optimization plan, then search algorithm 1 and search algorithm 2 are determined to be candidate search algorithms.

[0089] In this embodiment, multiple candidate search algorithms are determined based on the optimization features corresponding to the search term classification in the target optimization scheme and the relevance of all search algorithms in the search algorithm set to each feature, so that the optimized search algorithm can search for more accurate search results.

[0090] Figure 6 A flowchart of another search algorithm optimization method provided by the present disclosure is shown below. Figure 6 for Figure 1 Based on the illustrated embodiment, a possible implementation method for executing S106 is specifically described as follows:

[0091] S1061: Determine a weight according to the importance of each of the multiple candidate search algorithms.

[0092] Based on the importance of multiple candidate search algorithms, a weight can be determined for each candidate search algorithm. Different importances result in different corresponding weights. For example, based on the above embodiment, the importance of candidate search algorithm 1 is 0.6, and the importance of candidate search algorithm 2 is 0.2. It can be determined that the weight of candidate search algorithm 1 is 0.6, and the weight of candidate search algorithm 2 is 0.2.

[0093] S1062: Fusion-form multiple candidate optimization algorithms based on the multiple candidate search algorithms, the current search algorithm, and the weights.

[0094] Multiple candidate search algorithms are multiplied by their corresponding weights and fused with the current search algorithm to obtain multiple candidate optimization algorithms. For example, candidate search algorithms 1, 2, and the current candidate search algorithm are fused to obtain candidate optimization algorithm M1, and candidate search algorithms 3, 4, and the current candidate search algorithm are fused to obtain candidate optimization algorithm M2.

[0095] S1063: Determine the target optimization algorithm from the multiple candidate optimization algorithms.

[0096] As a specific implementation of a possible implementation of S1063, Figure 7 As shown:

[0097] S301: Push the multiple candidate optimization algorithms online to obtain feedback data information within a preset time period.

[0098] For example, five candidate optimization algorithms are pushed online, and each candidate optimization algorithm is assigned 20% of users. After three days, feedback data information of all search results is obtained, and the feedback data information is used as verification of the search accuracy of the candidate optimization algorithm.

[0099] S302: Determine the target optimization algorithm according to the feedback data information within the preset time period corresponding to the multiple candidate optimization algorithms.

[0100] Exemplarily, the feedback data information includes user click-through rates. Based on the user click-through rates of the search results of the five candidate optimization algorithms within three days, the candidate optimization algorithm corresponding to the search result with the highest user click-through rate is determined to be the target optimization algorithm. In other embodiments, the candidate optimization algorithm corresponding to the search result with the highest user click-through rate can be retained based on the user click-through rates of the search results of the five candidate optimization algorithms within three days. Then, five sets of candidate optimization algorithms similar to the retained candidate optimization algorithms are calculated, and the above operation is repeated. Finally, after two weeks of iteration, the candidate optimization algorithm corresponding to the search result with the highest user click-through rate is determined to be the target optimization algorithm. The target optimization algorithm is the new search algorithm in the final browser.

[0101] The present disclosure also provides an optimization device for a search algorithm, Figure 8 A schematic diagram of the structure of an optimization device for a search algorithm provided by the present disclosure, such as Figure 8 As shown, the optimization device includes:

[0102] The first determination module 110 is configured to determine at least one candidate optimization solution corresponding to a current search algorithm based on a set of relevant search results of a search term, wherein the set of relevant search results includes a plurality of relevant search results.

[0103] The second determination module 120 is configured to determine a plurality of candidate search algorithms from the search algorithm set according to a target optimization scheme, wherein the target optimization scheme is part or all of the at least one candidate optimization scheme.

[0104] The third determination module 130 is configured to determine a target optimization algorithm based on the multiple candidate search algorithms.

[0105] Optionally, the first determination module 110 is also used to input the search term into the current search algorithm to obtain a search result set corresponding to the search term; determine the relevance of all search results in the search result set based on the feedback data information of all search results in the search result set; and determine the relevant search result set from the search result set based on the relevance of all search results and a preset relevance.

[0106] Optionally, the first determination module 110 is further configured to determine an optimization feature corresponding to a search term classification based on a set of relevant search results for the search term; and determine the at least one candidate optimization solution based on the optimization feature corresponding to the search term classification.

[0107] Optionally, the first determination module 110 is further used to determine the optimization features corresponding to the search term based on the scores of the relevant search result set for each feature in the current search algorithm, and the scores of the search result set for each feature in the current search algorithm; and determine the optimization features corresponding to the search term classification based on the optimization features corresponding to the search term.

[0108] Optionally, the second determination module 120 is further configured to determine the multiple candidate search algorithms based on the optimization features corresponding to the search term classification in the target optimization solution and the relevance of all search algorithms in the search algorithm set to each feature.

[0109] Optionally, the third determination module 130 is further used to determine weights based on the importance of each of the multiple candidate search algorithms; to fuse multiple candidate optimization algorithms based on the multiple candidate search algorithms, the current search algorithm and the weights; and to determine the target optimization algorithm from the multiple candidate optimization algorithms.

[0110] Optionally, the third determination module 130 is further used to push the multiple candidate optimization algorithms online and obtain feedback data information within a preset time period; and determine the target optimization algorithm based on the feedback data information within the preset time period corresponding to the multiple candidate optimization algorithms.

[0111] The present disclosure also provides an electronic device, Figure 8 This is a schematic diagram of the structure of an electronic device provided by the present disclosure. Figure 8 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present disclosure is shown. Figure 8 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0112] like Figure 8 As shown, the electronic device 12 is implemented as a general-purpose computing device. Components of the electronic device 12 may include, but are not limited to, one or more processors 16, a system memory 28, and a bus 18 connecting various system components (including the system memory 28 and the processor 16).

[0113] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0114] The electronic device 12 typically includes a variety of computer system readable media. These media can be any media that can be accessed by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.

[0115] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be configured to read and write non-removable, non-volatile magnetic media ( Figure 8 Not shown, often called a "hard drive"). Although Figure 8 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data media interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of various embodiments of the present invention.

[0116] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods described in the embodiments of the present invention.

[0117] The processor 16 executes at least one of the multiple programs stored in the system memory 28 to perform various functional applications and data processing, such as implementing the steps of the method provided in the embodiment of the present invention.

[0118] The present disclosure also provides a computer-readable storage medium having a computer program stored thereon, which implements any one of the methods provided in the embodiments of the present invention when the computer program is executed by a processor.

[0119] Any combination of one or more computer-readable media can be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductors, or any combination thereof. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by an instruction execution system, device or device or used in combination with it.

[0120] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical cable, RF, etc., or any suitable combination thereof.

[0121] The computer program code for performing the operations of the present invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet). The device of this embodiment can be used to perform the steps of the above-mentioned method embodiment, and its implementation principles and technical effects are similar and will not be repeated here.

[0122] The present disclosure also provides a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the steps of the above method embodiments.

[0123] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0124] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for optimizing a search algorithm, characterized in that: include: Determining at least one candidate optimization solution corresponding to a current search algorithm based on a set of relevant search results for the search term, wherein the set of relevant search results includes a plurality of relevant search results; Determining, from a search algorithm set, a plurality of candidate search algorithms according to a target optimization scheme, the target optimization scheme being part or all of the at least one candidate optimization scheme, wherein the target optimization scheme includes optimization features and optimization weights corresponding to the optimization features, and the plurality of candidate search algorithms being a plurality of search algorithms found from the search algorithm set that match the optimization features and the optimization weights corresponding to the optimization features in the target optimization scheme; Determining a target optimization algorithm based on the multiple candidate search algorithms; The step of determining at least one candidate optimization solution corresponding to the current search algorithm based on a set of search results related to the search term includes: Determining optimization features corresponding to the search term classification based on a set of search results related to the search term; Determine the at least one candidate optimization solution based on the optimization features corresponding to the search term classification.

2. The method according to claim 1, characterized in that Before determining at least one candidate optimization solution corresponding to the current search algorithm based on the set of search results related to the search term, the method further includes: Input the search term into the current search algorithm to obtain a search result set corresponding to the search term; Determining the relevance of all search results in the search result set based on feedback data information of all search results; The relevant search result set is determined from the search result set according to the relevance of all the search results and a preset relevance.

3. The method according to claim 1, characterized in that Determining the optimization features corresponding to the search term classification based on the set of search results related to the search term includes: Determining the optimized feature corresponding to the search term according to the score of each feature of the relevant search result set in the current search algorithm and the score of each feature of the search result set in the current search algorithm; According to the optimization features corresponding to the search terms, the optimization features corresponding to the search term categories are determined.

4. The method according to claim 1 or 2, characterized in that The method further comprises determining a plurality of candidate search algorithms from a search algorithm set according to the target optimization scheme, including: The plurality of candidate search algorithms are determined according to the optimization features corresponding to the search term classification in the target optimization solution and the relevance between all search algorithms in the search algorithm set and each feature.

5. The method according to claim 1 or 2, characterized in that Determining a target optimization algorithm based on the multiple candidate search algorithms includes: Determining a weight according to the importance of each of the plurality of candidate search algorithms; Fusion forms multiple candidate optimization algorithms based on the multiple candidate search algorithms, the current search algorithm, and the weights; The target optimization algorithm is determined from the multiple candidate optimization algorithms.

6. The method according to claim 5, characterized in that Determining the target optimization algorithm from the multiple candidate optimization algorithms includes: Push the multiple candidate optimization algorithms online and obtain feedback data information within a preset time period; The target optimization algorithm is determined according to the feedback data information within the preset time period corresponding to the multiple candidate optimization algorithms.

7. A search algorithm optimization device, characterized in that: include: A first determination module is configured to determine at least one candidate optimization solution corresponding to a current search algorithm based on a set of relevant search results for a search term, wherein the set of relevant search results includes a plurality of relevant search results; a second determination module, configured to determine, from a search algorithm set, a plurality of candidate search algorithms according to a target optimization scheme, wherein the target optimization scheme is part or all of the at least one candidate optimization scheme, wherein the target optimization scheme includes optimization features and optimization weights corresponding to the optimization features, and the plurality of candidate search algorithms are a plurality of search algorithms found from the search algorithm set that match the optimization features and the optimization weights corresponding to the optimization features in the target optimization scheme; A third determination module is used to determine a target optimization algorithm based on the multiple candidate search algorithms; The first determination module is configured to determine an optimization feature corresponding to a search term classification based on a set of search results related to the search term; and determine the at least one candidate optimization solution based on the optimization feature corresponding to the search term classification.

8. An electronic device, characterized in that: include: A processor, wherein the processor is configured to execute a computer program stored in a memory, wherein the computer program, when executed by the processor, implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images