Search result processing method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of big data, and particularly discloses a search result processing method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: in response to an input operation for a search box, obtaining a search text input by the input operation; recalling at least one search result matched with the search text; and for any search result, determining at least one target search word matched with the search text in the expected search words associated with the search result, and according to the association probability information between each target search word and the search result, adjusting the association degree information between the search text and the search result. Obtaining adjusted association degree information between the search text and the search result; and according to the adjusted association degree information between the search text and each search result, reordering each search result. By adopting the method, the beneficial effect of improving the sorting efficiency of the search results can be achieved.

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a search result processing method, apparatus, computer equipment, storage medium and computer program product. Background Art

[0002] With the development of Internet technology, people can find the information resources they need by entering search terms in various search engines.

[0003] At present, in business scenarios, business personnel can obtain the search flow data of user accounts through the search system, and use manual annotation methods to train a search result ranking model. The trained search result ranking model can output personalized search result ranking information for user accounts. However, this method of training the search result ranking model depends on the amount of search flow data and manual annotation, which requires high labor and operating costs, and requires a certain amount of model training time, and its efficiency in sorting search results is not high.

[0004] Therefore, there is a problem in the traditional technology that the efficiency of sorting search results is low. Summary of the invention

[0005] Based on this, it is necessary to provide a search result processing method, apparatus, computer device, computer-readable storage medium and computer program product that can improve the efficiency of sorting search results in response to the above technical problems.

[0006] A search result processing method, the method comprising:

[0007] In response to an input operation on the search box, obtaining a search text input by the input operation;

[0008] Recall at least one search result that matches the search text; the search result is associated with at least one expected search term, and association probability information between the at least one expected search term and the search result; the association probability information represents the probability of a user interacting with the search result after entering the expected search term;

[0009] For any search result, at least one target search word matching the search text is determined from each expected search word associated with the search result, and the correlation degree information between the search text and the search result is adjusted according to the correlation probability information between each target search word and the search result to obtain the adjusted correlation degree information between the search text and the search result;

[0010] The search results are reordered according to the adjusted correlation degree information between the search text and the search results.

[0011] In an exemplary embodiment, according to the association probability information between each target search word and the search result, the association degree information between the search text and the search result is adjusted to obtain the adjusted association degree information between the search text and the search result, including:

[0012] Summarize the association probability information between each target search term and the search result to obtain summary probability information;

[0013] According to the summary probability information, the correlation degree information between the search text and the search results is adjusted to obtain the adjusted correlation degree information between the search text and the search results.

[0014] In an exemplary embodiment, the method further comprises:

[0015] Obtaining historical trigger operation data for each search result; the historical trigger operation data includes trigger operation data for any search result triggered by a user inputting any desired search term within a preset historical time range;

[0016] According to the historical trigger operation data, the association probability information between any search result and any expected search term is determined; the association probability information between any search result and any expected search term represents the probability information of the user interacting with the search result after inputting the expected search term.

[0017] In an exemplary embodiment, determining association probability information between any search result and any desired search term based on historical trigger operation data includes:

[0018] For any search result, determine the number of times the search result is triggered by each expected search term based on historical trigger operation data;

[0019] According to the information of the number of times the search results are triggered by each expected search word, the association probability information between the search results and each expected search word is determined.

[0020] In an exemplary embodiment, the number of times information includes the daily cumulative number of times the search results are triggered by each expected search term in each historical date; the historical date is a date within a preset historical time range; according to the number of times the search results are triggered by each expected search term, determining the association probability information between the search results and each expected search term includes:

[0021] Get the time decay factor for each historical date;

[0022] According to the time decay factor of each historical date, the daily cumulative number of times the search results are triggered by any expected search term in each historical date is adjusted to determine the daily number of times the search results are triggered by any expected search term in each historical date;

[0023] Accumulate the number of times the search result is triggered by any expected search term on each historical date to obtain the total number of times the search result is triggered by any expected search term;

[0024] According to the total number of times that the search results are triggered by each expected search term, the association probability information between the search results and each expected search term is determined.

[0025] In an exemplary embodiment, determining association probability information between the search results and each expected search term according to the total number of times the search results are triggered by each expected search term includes:

[0026] Obtain the total number of searches for each desired search term within a preset historical time range;

[0027] According to the total number of times that the search results are triggered by each expected search term and the total number of times that each expected search term is searched, the association probability information between the search results and each expected search term is determined.

[0028] A search result processing device, comprising:

[0029] A response module, used for obtaining the search text input by the input operation in response to the input operation on the search box;

[0030] A recall module, used to recall at least one search result that matches the search text; the search result is associated with at least one expected search term, and association probability information between the at least one expected search term and the search result; the association probability information represents the probability of a user interacting with the search result after inputting the expected search term;

[0031] An adjustment module is used to determine, for any search result, at least one target search word that matches the search text among the desired search words associated with the search result, and adjust the correlation degree information between the search text and the search result according to the correlation probability information between each target search word and the search result, so as to obtain the adjusted correlation degree information between the search text and the search result;

[0032] The reordering module is used to reorder the search results according to the adjusted correlation degree information between the search text and the search results.

[0033] In an exemplary embodiment, the adjustment module is specifically used to: summarize the association probability information between each target search term and the search result to obtain summarized probability information; adjust the association degree information between the search text and the search result based on the summarized probability information to obtain the adjusted association degree information between the search text and the search result.

[0034] In an exemplary embodiment, the device also includes: an acquisition module for acquiring historical trigger operation data for each search result; the historical trigger operation data includes trigger operation data for triggering any search result after a user inputs any expected search term within a preset historical time range; based on the historical trigger operation data, determining the association probability information between any search result and any expected search term; the association probability information between any search result and any expected search term represents the probability information of the user interacting with the search result after inputting the expected search term.

[0035] In an exemplary embodiment, the acquisition module is specifically used to: for any search result, determine the number of times the search result is triggered by each expected search term based on historical trigger operation data; and determine the association probability information between the search result and each expected search term based on the number of times the search result is triggered by each expected search term.

[0036] In an exemplary embodiment, the number information includes the daily cumulative number of times the search results are triggered by each expected search term in each historical date; the historical date is a date within a preset historical time range; an acquisition module is specifically used to: obtain a time decay factor for each historical date; according to the time decay factor of each historical date, adjust the daily cumulative number of times the search results are triggered by any expected search term in each historical date, and determine the daily number of times the search results are triggered by any expected search term in each historical date; add up the daily number of times the search results are triggered by any expected search term in each historical date, and obtain the total number of times the search results are triggered by any expected search term; according to the total number of times the search results are triggered by each expected search term, determine the association probability information between the search results and each expected search term.

[0037] In an exemplary embodiment, the acquisition module is specifically used to: obtain the total number of searches corresponding to each expected search term within a preset historical time range; determine the association probability information between the search results and each expected search term based on the total number of times the search results are triggered by each expected search term and the total number of searches corresponding to each expected search term.

[0038] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0039] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0040] A computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0041] The above-mentioned search result processing method, apparatus, computer device, storage medium and computer program product obtain the search text input by the input operation in response to the input operation on the search box; recall at least one search result matching the search text; the search result is associated with at least one expected search word, and the association probability information between the at least one expected search word and the search result; the association probability information represents the probability of the user interacting with the search result after entering the expected search word; for any search result, at least one target search word matching the search text is determined from each expected search word associated with the search result, and the association degree information between the search text and the search result is analyzed based on the association probability information between each target search word and the search result. Adjustments are made to obtain adjusted correlation information between the search text and the search results; and the search results are reordered according to the adjusted correlation information between the search text and the search results. In this way, the correlation information between the search text and the search results can be adjusted based on the relationship between the search text input by the user and the desired search terms associated with the search results, and then the search results can be reordered according to the adjusted correlation information. The recalled search results can be flexibly reordered based on the user's search intent, and the recalled search results can be quickly and accurately reordered without training a search result ranking model, thereby achieving the beneficial effect of improving the efficiency of sorting the search results. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] FIG1 is an application environment diagram of a search result processing method in one embodiment;

[0044] Figure 2 A flowchart of a search result processing method in one embodiment;

[0045] Figure 3 A flowchart of an adaptive search result sorting method in one embodiment;

[0046] Figure 4 A schematic flow chart of a step of adjusting the correlation degree information between the search text and the search results in one embodiment;

[0047] Figure 5 A schematic diagram of user behavior flow information in one embodiment;

[0048] Figure 6 is a flowchart of a search result processing method in another embodiment;

[0049] Figure 7 is a structural block diagram of a search result processing device in one embodiment;

[0050] FIG. 8 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. It should be noted that in the embodiments of the present application, certain software, groups, models and other existing solutions in the industry may be mentioned, which should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0052] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations. The acquisition, storage, use, and processing of data in the technical solution of this application are in compliance with relevant provisions of national laws and regulations.

[0053] The search result processing method provided in the embodiment of the present application can be applied to the application environment shown in FIG. 1. The terminal 102 communicates with the server 104 through a network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on a cloud or other network server. The server 104 obtains the search text input by the input operation in response to the input operation on the search box; the server 104 recalls at least one search result that matches the search text; the search result is associated with at least one expected search term, and the association probability information between at least one expected search term and the search result; the association probability information represents the probability of the user interacting with the search result after entering the expected search term; for any search result, the server 104 determines at least one target search term that matches the search text in each expected search term associated with the search result, and adjusts the association degree information between the search text and the search result according to the association probability information between each target search term and the search result, and obtains the adjusted association degree information between the search text and the search result; the server 104 reorders each search result according to the adjusted association degree information between the search text and each search result. The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car devices, projection devices, etc. Portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices may be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 may be an independent physical server, or a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.

[0054] In an exemplary embodiment, Figure 2 As shown, a search result processing method is provided, which is described by taking the method applied to the server 104 in FIG. 1 as an example, and includes the following steps 202 to 206. Among them:

[0055] Step 202: In response to an input operation on the search box, obtain the search text input by the input operation.

[0056] The search box may refer to an area displayed in the search page for inputting search text.

[0057] The input operation may refer to an operation corresponding to when a user inputs a search text in a search box.

[0058] The search text may refer to words, sentences, etc. input by the user in the search box.

[0059] Optionally, the server obtains the search text input by the user in response to the user's input operation on the search box.

[0060] Step 204, recall at least one search result that matches the search text; the search result is associated with at least one expected search term, and association probability information between the at least one expected search term and the search result; the association probability information represents the probability of the user interacting with the search result after entering the expected search term.

[0061] The search result may refer to a specific search content in a search result list presented on a search result page, and each search result in the search result list matches the search text input by the user.

[0062] Among them, the association probability information between the expected search term and the search result may refer to the association probability value between the expected search term and the search result. In actual applications, the association probability value between the expected search term and the search result may refer to the probability that the search result is triggered by the expected search term. For example, when search result A may be triggered by expected search terms 1, 2, and 3, the weight of search result A being triggered by expected search terms 1, 2, and 3 can be used as the probability that search result A is triggered by expected search terms 1, 2, and 3.

[0063] Optionally, the server recalls each search result that matches the search text input by the user based on the search result index.

[0064] Step 206, for any search result, determine at least one target search term that matches the search text among the expected search terms associated with the search results, and adjust the correlation degree information between the search text and the search results based on the correlation probability information between each target search term and the search result to obtain the adjusted correlation degree information between the search text and the search result.

[0065] The expected search term may refer to a search term pre-set according to the user's historical input; the target search term may refer to a search term among the expected search terms that matches the search text input by the user.

[0066] The association probability information between the target search term and the search result may refer to the probability of triggering the search result through the target search term.

[0067] The information on the degree of association between the search text and the search results may be a recall score corresponding to the search results when the search results matching the search text are recalled.

[0068] The adjusted correlation degree information between the search text and the search results may refer to the result obtained after adjusting the recall score of the search results.

[0069] Optionally, for any search result, the server determines each target search term that matches the search text entered by the user among each expected search term associated with the search result, and the server adjusts the degree of association between the search text entered by the user and the search result based on the association probability between each target search term and the search result, to obtain an adjusted degree of association between the search text entered by the user and the search result.

[0070] Step 208: reorder the search results according to the adjusted correlation information between the search text and the search results.

[0071] Optionally, the server reorders the search results according to the adjusted correlation degree values ​​between the search text input by the user and the search results.

[0072] In the above search result processing method, by responding to an input operation on a search box, a search text input by an input operation is obtained; at least one search result matching the search text is recalled; the search result is associated with at least one expected search term, and association probability information between at least one expected search term and the search result; the association probability information represents the probability of a user interacting with the search result after inputting the expected search term; for any search result, at least one target search term matching the search text is determined from each expected search term associated with the search result, and according to the association probability information between each target search term and the search result, the association degree information between the search text and the search result is adjusted to obtain the search text. The invention relates to an embodiment of the present invention which is to adjust the correlation degree information between the search text and the search results; and reorder the search results according to the adjusted correlation degree information between the search text and the search results; in this way, the correlation degree information between the search text and the search results can be adjusted based on the relationship between the search text input by the user and the expected search terms associated with the search results, and then the search results can be reordered according to the adjusted correlation degree information; the recalled search results can be flexibly reordered based on the search intention of the user; and the recalled search results can be quickly and accurately reordered without training a search result ranking model, thereby achieving the beneficial effect of improving the efficiency of sorting the search results.

[0073] In order to facilitate the understanding of those skilled in the art, Figure 3 An adaptive search result ranking process is exemplarily provided, and its main framework mainly includes the main process in the dotted part and other data processes.

[0074] The main process includes: Step 1: The user enters the search text in the search box to trigger the search; Step 2: Based on the search text according to the search index, the first N search results are retrieved; Step 3: For each search result, the expected search term score (i.e., label score) for each search result is calculated; Step 4: Based on the expected search term score for each search result, the re-ordered search results are output.

[0075] Among them, other data processes include: Step 1: Collecting users' clicks and effective browsing behavior data on each search result; Step 2: Analyzing the clicks and effective browsing behavior data; Step 3: Calculating the mapping relationship between the search results and each expected search term, as well as the weight of the expected search term relative to the search result; Step 4: Updating the expected search term and weight.

[0076] The above-mentioned search index is a search inverted index structure constructed based on the business data of the business scenario. In addition to the fields of the business dimension in the business data, the expected search term field and the corresponding weight value for self-learning are also added. The search results will be reordered according to the expected search term field and the corresponding weight value. The search index fields include ID, content1-contentN and queryWord, where ID is the unique identifier of the search result, generally the unique identifier of the business scenario data, content1-contentN is the business scenario data field, queryWord is the expected search term and weight, which are stored in the form of a list, and each element is "expected search term-weight". Among them, in the list storing "expected search term-weight", each row represents each expected search term associated with a search result and the weight corresponding to each expected search term under the search result. The weight represents the probability of association between the search result and the expected search term, that is, the probability of interaction with the search result by searching the expected search term, for example, the probability of clicking or effective browsing.

[0077] In an exemplary embodiment, Figure 4 As shown, step 206 includes steps 402 to 404. Among them:

[0078] Step 402: Summarize the association probability information between each target search term and the search result to obtain summary probability information.

[0079] The summary probability information may refer to the probability obtained by adding the association probabilities between each target search term and the search result.

[0080] Optionally, the server adds the association probabilities between each target search term and the search result to obtain a probability sum.

[0081] Step 404: adjusting the correlation degree information between the search text and the search results according to the summary probability information to obtain adjusted correlation degree information between the search text and the search results.

[0082] Optionally, the server adjusts the correlation degree value between the search text and the search result according to the sum of the probabilities as an adjustment coefficient to obtain an adjusted correlation degree value between the search text and the search result.

[0083] In this embodiment, the association probability information between each target search word and the search results is summarized to obtain summarized probability information, and then the association degree information between the search text and the search results is adjusted based on the summarized probability information to obtain the adjusted association degree information between the search text and the search results. The association degree information between the search text and the search results can be adjusted based on the summarized probability information, and the adjusted association degree information used as the basis for sorting the search results can be obtained taking into account the user's search intent, thereby achieving the beneficial effect of improving the accuracy of sorting the search results.

[0084] In an exemplary embodiment, the method also includes: obtaining historical trigger operation data for each search result; the historical trigger operation data includes trigger operation data for triggering any search result after a user inputs any expected search term within a preset historical time range; determining the association probability information between any search result and any expected search term based on the historical trigger operation data; the association probability information between any search result and any expected search term represents the probability information of the user interacting with the search result after inputting the expected search term.

[0085] The preset historical time range may refer to a preset historical time period.

[0086] Among them, the trigger operation data of any search result triggered by the user entering any desired search term

[0087] Among them, the association probability information between any search result and any expected search term may refer to the association probability between any search result and any expected search term. For the association probability between search result A and expected search term 1, the association probability represents the probability of the user effectively clicking or effectively browsing search result A after entering the expected search term 1 in the search box.

[0088] Optionally, the server obtains historical trigger operation data for each search result, that is, obtains the trigger operation data of triggering any search result after each user enters any expected search term within a preset historical time range. The server then determines the association probability value between any search result and any expected search term based on the historical trigger operation data, that is, determines the probability information of the user interacting with any search result (for example, effective clicks, effective browsing, etc.) after entering any expected search term.

[0089] In order to facilitate the understanding of those skilled in the art, Figure 5 An exemplary schematic diagram of user behavior flow information is provided. By analyzing the user behavior flow information, the user's historical trigger operation data for each search result can be obtained. Figure 5 The main information of the behavior flow shown includes: tracking number + search word + data ID + behavior score; tracking number refers to the unique identifier of the search behavior, which can be used to track subsequent search behaviors; search word refers to the search word entered by the user, such as credit card, activation, etc.; data ID refers to the search result ID; behavior score refers to the effective behavior score of the user in a specific search result. Figure 5 As shown in the figure, when a user searches for "credit card", 10 search results are returned. The user clicks on the first and third search results, which is the user's click behavior; the user clicks on the first search result and browses for 5 seconds, and clicks on the third search result and browses for 30 seconds, which is the user's browsing behavior. In actual applications, effective browsing and invalid browsing are determined by setting a threshold for browsing time. For example, browsing time greater than 10 seconds is considered effective browsing, and browsing time less than 10 seconds is considered invalid browsing. The score calculated for "click behavior" is 1, and the score calculated for "effective browsing" is 1 (plus the score of click is 2).

[0090] In this embodiment, historical trigger operation data for each search result is obtained; the historical trigger operation data includes trigger operation data that triggers any search result after the user inputs any expected search term within a preset historical time range; based on the historical trigger operation data, the association probability information between any search result and any expected search term is determined; the association probability information between any search result and any expected search term represents the probability information of the user interacting with the search result after inputting the expected search term; in this way, processing based on the historical trigger operation data can achieve the beneficial effect of accurately determining the association probability between the search result and the expected search term.

[0091] In an exemplary embodiment, the association probability information between any search result and any expected search term is determined based on historical trigger operation data, including: for any search result, determining the number of times the search result is triggered by each expected search term based on historical trigger operation data; determining the association probability information between the search result and each expected search term based on the number of times the search result is triggered by each expected search term.

[0092] Among them, for the information on the number of times the search result is triggered by each expected search term, for example, for search result A, search result A is associated with expected search terms 1, 2 and 3, then the information on the number of times search result A is triggered by each expected search term may refer to the number of times the user effectively clicks or effectively browses search result A after entering expected search term 1 in the search box, the number of times the user effectively clicks or effectively browses search result A after entering expected search term 2 in the search box, and the number of times the user effectively clicks or effectively browses search result A after entering expected search term 3 in the search box.

[0093] Optionally, for any search result, the server determines the number of times the search result is effectively triggered by each expected search term based on historical trigger operation data, and then determines the association probability between the search result and each expected search term based on the number of times the search result is effectively triggered by each expected search term.

[0094] In this embodiment, for any search result, the number of times the search result is triggered by each expected search term is determined based on the historical trigger operation data; the association probability information between the search result and each expected search term is determined based on the number of times the search result is triggered by each expected search term; in this way, the association probability between the search result and each expected search term can be determined based on the number of times the search result is triggered by each expected search term, thereby achieving the beneficial effect of accurately obtaining the association probability between the search result and each expected search term.

[0095] In an exemplary embodiment, the number information includes the daily cumulative number of times the search results are triggered by each expected search term in each historical date; the historical date is a date within a preset historical time range; based on the number of times the search results are triggered by each expected search term, determining the association probability information between the search results and each expected search term, including: obtaining a time decay factor for each historical date; adjusting the daily cumulative number of times the search results are triggered by any expected search term in each historical date according to the time decay factor of each historical date, and determining the daily number of times the search results are triggered by any expected search term in each historical date; accumulating the daily number of times the search results are triggered by any expected search term in each historical date to obtain the total number of times the search results are triggered by any expected search term; determining the association probability information between the search results and each expected search term according to the total number of times the search results are triggered by each expected search term.

[0096] Among them, for "the daily cumulative number of times the search result is triggered by any expected search term on each historical date", for example, when the search result is A, the expected search terms associated with the search result A include expected search terms 1, 2 and 3, and the historical dates are dates 1, 2, and 3, the daily cumulative number of times the search result A is triggered by any expected search term on each historical date may include: the cumulative number of effective clicks or effective views on the search result A after entering the expected search term 1 on date 1, the cumulative number of effective clicks or effective views on the search result A after entering the expected search term 1 on date 2, and the cumulative number of effective clicks or effective views on the search result A after entering the expected search term 1 on date 3. The cumulative number of effective clicks or effective browsings on search result A after entering the expected search term 2 on date 1, the cumulative number of effective clicks or effective browsings on search result A after entering the expected search term 2 on date 2, the cumulative number of effective clicks or effective browsings on search result A after entering the expected search term 2 on date 3, the cumulative number of effective clicks or effective browsings on search result A after entering the expected search term 2 on date 1, the cumulative number of effective clicks or effective browsings on search result A after entering the expected search term 3 on date 2, and the cumulative number of effective clicks or effective browsings on search result A after entering the expected search term 3 on date 3.

[0097] The value of the time attenuation factor can be set according to the update frequency of the frequency information and the characteristics of the scene. The farther the behavior occurs from the current time, the smaller the impact.

[0098] Among them, the daily cumulative number is the total number of times a certain search result is effectively triggered by a certain expected search term in a certain day; the daily number is the number obtained by multiplying the daily cumulative number by the time decay factor of that day. The daily number can be regarded as the daily cumulative effective number, which represents the effective number of times a certain search result is effectively triggered by a certain expected search term in a certain day; the total number is obtained by adding up the daily cumulative effective number of times a certain search result is effectively triggered by a certain expected search term in all historical dates.

[0099] Optionally, the server obtains a time decay factor for each historical date, and then uses the time decay factor for each historical date to adjust the daily cumulative number of times a search result is triggered by any expected search term in each historical date, and determines the daily cumulative effective number of times the search result is triggered by any expected search term in each historical date. The server then adds up the daily number of times the search result is triggered by a certain expected search term in each historical date, and obtains the total number of times the search result is triggered by a certain expected search term. The server determines the association probability between the search result and each expected search term based on the total number of times the search result is triggered by each expected search term.

[0100] In this embodiment, a time decay factor for each historical date is obtained; the daily cumulative number of times the search results are triggered by any expected search term in each historical date is adjusted according to the time decay factor of each historical date to determine the daily number of times the search results are triggered by any expected search term in each historical date; the daily number of times the search results are triggered by any expected search term in each historical date is accumulated to obtain the total number of times the search results are triggered by any expected search term; based on the total number of times the search results are triggered by each expected search term, the association probability information between the search results and each expected search term is determined; in this way, based on the distance between the time when the behavior occurs and the current time, the daily cumulative number of times the search results are triggered by any expected search term in each historical date can be decayed according to a certain gradient, thereby accurately determining the effective total number of times the search results are triggered by each expected search term, thereby achieving the beneficial effect of improving the acquisition of association probability information between the search results and each expected search term.

[0101] In an exemplary embodiment, the association probability information between the search results and each expected search word is determined based on the total number of times the search results are triggered by each expected search word, including: obtaining the total number of searches corresponding to each expected search word within a preset historical time range; determining the association probability information between the search results and each expected search word based on the total number of times the search results are triggered by each expected search word and the total number of searches corresponding to each expected search word.

[0102] The total number of searches for the expected search term may refer to the total number of times the user enters the expected search term in the search box.

[0103] Optionally, the server obtains the total number of searches for each expected search term within a preset historical time range, and then determines the association probability between the search result and each expected search term based on the total number of times a search result is triggered by each expected search term and the total number of searches for each expected search term.

[0104] In this embodiment, the total number of searches corresponding to each expected search term within a preset historical time range is obtained; the association probability information between the search results and each expected search term is determined based on the total number of times each expected search term is triggered by the search results and the total number of searches corresponding to each expected search term; in this way, the weight between the search results and each expected search term can be calculated based on the total number of times each expected search term is triggered by the search results and the total number of searches corresponding to each expected search term, thereby achieving the beneficial effect of accurately determining the association probability between the search results and each expected search term.

[0105] To facilitate understanding by those skilled in the art, a method for determining the probability of association between each search result and each desired search term based on historical trigger operation data is provided below. The method includes:

[0106] Step 1: By Figure 5 The user behavior flow information shown in the figure can obtain the statistical results of "search results - behavior scores of each expected search word associated with the search results" as shown in Table 1 below:

[0107] Table 1

[0108]

[0109] Among them, the mapping relationship between search results and expected search terms and the corresponding behavior scores within a preset time range according to the set update frequency, for example, if the update frequency is 1 day, the mapping relationship between the search results of that day and the scores of the expected search terms are statistically analyzed.

[0110] Step 2: Use the time decay factor K to adjust the data obtained in step 1. Considering that the recent user behavior can better reflect the correlation between the search results and the expected search terms, the time decay factor K (K<1) is added when updating the scores of the search results and the expected search terms. The farther the behavior occurs from the current time, the smaller the impact. The setting of K value can be set according to the update frequency and scene characteristics. For example, if the update frequency is 1 day, yesterday's data is shown in Table 2:

[0111] Table 2

[0112]

[0113] If the K value is set to 0.9, that is, the historical data decays at a gradient of 10%, then the results after adding the recent statistical data (as shown in Table 1) are shown in Table 3:

[0114] Table 3

[0115]

[0116] Thus, the current search results and the expected search term score data are obtained, as shown in Table 4.

[0117] Table 4

[0118]

[0119] Step 3: Update the expected search term weight corresponding to each search result. According to the behavior flow [tracking number + search term + search result ID + behavior score], the total number of searches for all search terms (search terms with clicks or effective browsing behaviors) can be calculated. If the total number of effective searches for credit cards is 400, the total number of effective searches for activation is 300, the total number of effective searches for reporting loss is 100, and the total number of effective searches for opening a card is 100, then the weight calculation is: expected search term score / (total number of effective searches*2), then the relationship between the index data and the expected search term weight is shown in Table 5:

[0120] Table 5

[0121]

[0122] After processing the data in Table 5, the final results shown in Table 6 are obtained:

[0123] Table 6

[0124]

[0125] Then the final result shown in Table 6 is updated to the search inverted index.

[0126] Through the above steps 1 to 3, the weight relationship between each search result and each expected search term in the business data is obtained, which represents the probability of selecting a search result based on historical user behavior when the user search term matches the expected search term. At the same time, due to the addition of the time decay factor, recent user behavior can be better reflected.

[0127] Furthermore, after the user inputs the search text, it is necessary to reorder the search results recalled based on the search text based on the search results-expected search word list and weights shown in Table 6, specifically:

[0128] Step 4: Recall search results based on the search text entered by the user, or set weights according to business logic. Taking the search text entered by the user as "report loss after credit card activation" as an example, the recalled search results are shown in Table 7.

[0129] Table 7

[0130]

[0131] Step 5: Based on the search results retrieved in step 4, obtain the data in the queryWord column and calculate the score using the following formula:

[0132] ;

[0133] Among them, S is the final search result ranking score, which corresponds to the adjusted correlation degree information between the search text of this application and the search results; S (relevance) is the original recall score corresponding to the search result, which corresponds to the original correlation degree information between the search text of this application and the search results; K is the influence weight of the expected search term on the overall ranking, and the greater the influence of the expected search term score on the ranking, the greater K will be; P word[i] The expected search words that are hit by the user's search text correspond to the target search words that match the search text in this application. When the expected search words are hit, the probability value corresponding to the expected search words is superimposed. For example, if the search text entered by the user is "report loss after credit card activation", the expected search words that are hit are "credit card", "activate card" and "report loss"; This corresponds to the summary probability information.

[0134] Finally, the search results are re-sorted by the S value and output to the user. The adaptive search sorting capability designed by this application is based on user behavior and can better fit the recent user's search intentions; at the same time, the adaptive search sorting process of this application has high scalability. In actual applications, the expected search terms and weights can be calculated according to user groups, and then more in line with the user's personal attributes to achieve personalized search result sorting; at the same time, this application can guide the search experience of other users by setting high-quality user groups.

[0135] In another embodiment, Figure 6 As shown, a search result processing method is provided, which is described by taking the method applied to the server 104 in FIG. 1 as an example, and includes the following steps:

[0136] Step S602, obtaining historical trigger operation data for each search result; the historical trigger operation data includes trigger operation data for triggering any search result after the user inputs any desired search term within a preset historical time range.

[0137] Step S604, determining the association probability information between any search result and any expected search term based on the historical trigger operation data; the association probability information between any search result and any expected search term represents the probability information of the user interacting with the search result after inputting the expected search term.

[0138] Step S606: in response to an input operation on the search box, obtaining a search text input by the input operation.

[0139] Step S608, recall at least one search result that matches the search text; the search result is associated with at least one expected search term, and association probability information between at least one expected search term and the search result; the association probability information represents the probability of the user interacting with the search result after entering the expected search term.

[0140] Step S610, for any search result, determine at least one target search term that matches the search text among the expected search terms associated with the search results, and adjust the correlation degree information between the search text and the search results based on the correlation probability information between each target search term and the search result to obtain the adjusted correlation degree information between the search text and the search result.

[0141] Step S612: reorder the search results according to the adjusted correlation information between the search text and the search results.

[0142] It should be noted that the specific limitations of the above steps can refer to the specific limitations of a search result processing method mentioned above.

[0143] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0144] Based on the same inventive concept, the embodiment of the present application also provides a search result processing device for implementing the search result processing method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more search result processing device embodiments provided below can refer to the limitations of the search result processing method above, and will not be repeated here.

[0145] In an exemplary embodiment, Figure 7 As shown, a search result processing device is provided, including: a response module 702, a recall module 704, an adjustment module 706 and a rearrangement module 708, wherein:

[0146] A response module 702 is used to obtain a search text input by the input operation in response to the input operation on the search box;

[0147] The recall module 704 is used to recall at least one search result that matches the search text; the search result is associated with at least one expected search term, and association probability information between the at least one expected search term and the search result; the association probability information represents the probability of the user interacting with the search result after entering the expected search term;

[0148] An adjustment module 706 is used to determine, for any search result, at least one target search term that matches the search text among the desired search terms associated with the search result, and adjust the association degree information between the search text and the search result according to the association probability information between each target search term and the search result, so as to obtain the adjusted association degree information between the search text and the search result;

[0149] The reordering module 708 is used to reorder the search results according to the adjusted correlation degree information between the search text and the search results.

[0150] In an exemplary embodiment, the adjustment module 706 is specifically used to: summarize the association probability information between each target search term and the search result to obtain summarized probability information; adjust the association degree information between the search text and the search result based on the summarized probability information to obtain the adjusted association degree information between the search text and the search result.

[0151] In an exemplary embodiment, the device also includes: an acquisition module for acquiring historical trigger operation data for each search result; the historical trigger operation data includes trigger operation data for triggering any search result after a user inputs any expected search term within a preset historical time range; based on the historical trigger operation data, determining the association probability information between any search result and any expected search term; the association probability information between any search result and any expected search term represents the probability information of the user interacting with the search result after inputting the expected search term.

[0152] In an exemplary embodiment, the acquisition module is specifically used to: for any search result, determine the number of times the search result is triggered by each expected search term based on historical trigger operation data; and determine the association probability information between the search result and each expected search term based on the number of times the search result is triggered by each expected search term.

[0153] In an exemplary embodiment, the number information includes the daily cumulative number of times the search results are triggered by each expected search term in each historical date; the historical date is a date within a preset historical time range; an acquisition module is specifically used to: obtain a time decay factor for each historical date; according to the time decay factor of each historical date, adjust the daily cumulative number of times the search results are triggered by any expected search term in each historical date, and determine the daily number of times the search results are triggered by any expected search term in each historical date; add up the daily number of times the search results are triggered by any expected search term in each historical date, and obtain the total number of times the search results are triggered by any expected search term; according to the total number of times the search results are triggered by each expected search term, determine the association probability information between the search results and each expected search term.

[0154] In an exemplary embodiment, the acquisition module is specifically used to: obtain the total number of searches corresponding to each expected search term within a preset historical time range; determine the association probability information between the search results and each expected search term based on the total number of times the search results are triggered by each expected search term and the total number of searches corresponding to each expected search term.

[0155] Each module in the above search result processing device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0156] In an exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be shown in FIG8. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store search result processing data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a search result processing method is implemented.

[0157] Those skilled in the art will understand that the structure shown in FIG. 8 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0158] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above-mentioned XXX method. The steps of the search result processing method here may be the steps of the search result processing method in each of the above-mentioned embodiments.

[0159] In one embodiment, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the processor executes the steps of the above-mentioned search result processing method. The steps of the search result processing method here can be the steps of the search result processing method in each of the above-mentioned embodiments.

[0160] In one embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, causes the processor to execute the steps of the above-mentioned search result processing method. The steps of the search result processing method here may be the steps of the search result processing method in each of the above-mentioned embodiments.

[0161] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0162] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0163] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0164] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A search result processing method, characterized in that: The method comprises: In response to an input operation on the search box, obtaining a search text input by the input operation; Recalling at least one search result that matches the search text; the search result is associated with at least one expected search term, and association probability information between the at least one expected search term and the search result; the association probability information represents the probability of a user interacting with the search result after inputting the expected search term; For any of the search results, at least one target search term that matches the search text is determined from among the expected search terms associated with the search results, and according to association probability information between each of the target search terms and the search results, the association degree information between the search text and the search results is adjusted to obtain adjusted association degree information between the search text and the search results; The search results are reordered according to the adjusted correlation degree information between the search text and the search results.

2. The method according to claim 1, characterized in that The step of adjusting the association degree information between the search text and the search results according to the association probability information between each of the target search terms and the search results to obtain the adjusted association degree information between the search text and the search results includes: Summarizing the association probability information between each of the target search terms and the search results to obtain summary probability information; The association degree information between the search text and the search results is adjusted according to the summary probability information to obtain the adjusted association degree information between the search text and the search results.

3. The method according to claim 1, characterized in that: The method further comprises: Acquire historical trigger operation data for each of the search results; the historical trigger operation data includes trigger operation data for triggering any of the search results after the user inputs any of the desired search terms within a preset historical time range; According to the historical trigger operation data, the association probability information between any of the search results and any of the expected search terms is determined; the association probability information between any of the search results and any of the expected search terms represents the probability information of the user interacting with the search result after entering the expected search term.

4. The method according to claim 3, characterized in that The determining, based on the historical trigger operation data, the association probability information between any of the search results and any of the expected search terms includes: For any of the search results, determining information on the number of times the search result is triggered by each of the expected search terms according to the historical trigger operation data; According to information about the number of times the search results are triggered by each of the expected search terms, association probability information between the search results and each of the expected search terms is determined.

5. The method according to claim 4, characterized in that The number information includes the daily cumulative number of times the search result is triggered by each of the expected search terms on each historical date; the historical date is a date within the preset historical time range; the association probability information between the search result and each of the expected search terms is determined based on the number of times the search result is triggered by each of the expected search terms, including: Obtaining a time decay factor for each of the historical dates; According to the time decay factor of each of the historical dates, the daily cumulative number of times the search result is triggered by any of the expected search terms in each of the historical dates is adjusted to determine the daily number of times the search result is triggered by any of the expected search terms in each of the historical dates; Accumulating the number of times the search result is triggered by any of the expected search terms on each of the historical dates to obtain the total number of times the search result is triggered by any of the expected search terms; According to the total number of times that the search results are triggered by the expected search terms, the association probability information between the search results and the expected search terms is determined.

6. The method according to claim 5, characterized in that The determining, according to the total number of times that the search results are triggered by the expected search terms, the association probability information between the search results and each of the expected search terms comprises: Obtain the total number of searches for each of the desired search terms within the preset historical time range; According to the total number of times the search results are triggered by each of the expected search terms and the total number of times each of the expected search terms is searched, the association probability information between the search results and each of the expected search terms is determined.

7. A search result processing device, characterized in that: The device comprises: A response module, configured to obtain a search text inputted by the input operation in response to an input operation on the search box; A recall module, used to recall at least one search result matching the search text; the search result is associated with at least one expected search term, and association probability information between the at least one expected search term and the search result; the association probability information represents the probability of a user interacting with the search result after inputting the expected search term; An adjustment module, configured to determine, for any of the search results, at least one target search term that matches the search text among the expected search terms associated with the search results, and adjust the degree of association information between the search text and the search results based on the probability information of association between each of the target search terms and the search results, so as to obtain the adjusted degree of association information between the search text and the search results; The reordering module is used to reorder the search results according to the adjusted correlation degree information between the search text and the search results.

8. The device according to claim 7, characterized in that The adjustment module is specifically used to: summarize the association probability information between each target search word and the search result to obtain summarized probability information; and adjust the association degree information between the search text and the search result based on the summarized probability information to obtain the adjusted association degree information between the search text and the search result.

9. The device according to claim 7, characterized in that The device also includes: an acquisition module, used to acquire historical trigger operation data for each of the search results; the historical trigger operation data includes trigger operation data that triggers any of the search results after the user inputs any of the expected search terms within a preset historical time range; based on the historical trigger operation data, determine the association probability information between any of the search results and any of the expected search terms; the association probability information between any of the search results and any of the expected search terms represents the probability information of the user interacting with the search result after inputting the expected search term.

10. The device according to claim 9, characterized in that The acquisition module is specifically used to: for any of the search results, determine the number of times the search result is triggered by each of the expected search terms based on the historical trigger operation data; and determine the association probability information between the search result and each of the expected search terms based on the number of times the search result is triggered by each of the expected search terms.

11. The device according to claim 10, characterized in that The number information includes the daily cumulative number of times the search results are triggered by each of the expected search terms on each historical date; the historical date is a date within the preset historical time range; the acquisition module is specifically used to: obtain a time decay factor for each of the historical dates; according to the time decay factor of each of the historical dates, adjust the daily cumulative number of times the search results are triggered by any of the expected search terms on each of the historical dates, and determine the daily number of times the search results are triggered by any of the expected search terms on each of the historical dates; The number of days on which the search results are triggered by any of the expected search terms in each of the historical dates is accumulated to obtain the total number of times the search results are triggered by any of the expected search terms; based on the total number of times the search results are triggered by each of the expected search terms, the association probability information between the search results and each of the expected search terms is determined.

12. The device according to claim 7, characterized in that The acquisition module is specifically used to: obtain the total number of searches corresponding to each of the expected search terms within the preset historical time range; determine the association probability information between the search results and each of the expected search terms based on the total number of times the search results are triggered by each of the expected search terms and the total number of searches corresponding to each of the expected search terms.

13. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

14. 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.

15. A computer program product comprising a computer program, 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