Search result ranking method, apparatus, device, and storage medium

Abstract

This application provides a method, apparatus, device, and storage medium for ranking search results. The method includes: determining at least two recommended terms and at least two search results based on a search term entered by a user, wherein the recommended terms are historical search terms including the search term; determining a ranking score for each search result based on the search term, the at least two search results, and the at least two recommended terms; and ranking the at least two search results based on the ranking scores. This application can improve search efficiency by ranking the user's desired search content at the top of the search results.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus, device, and storage medium for sorting search results. Background Technology

[0002] Because of its instant search functionality, which allows users to initiate searches and receive results in real time as they type in their search terms, it avoids the need for users to click on search controls or press enter to initiate searches. This advantage has led to its increasing adoption in various applications (apps). For example, ... Figures 1a to 1c This is a screenshot showing the effect of instant search results.

[0003] However, when users search for content based on search terms in applications that support instant search, if the search term is short, a larger search range will be determined based on the short search term. This will result in a large number of search results, requiring users to spend a lot of time searching through multiple search results to find the content they want, resulting in low search efficiency. Summary of the Invention

[0004] This application provides a search result sorting method, apparatus, device, and storage medium that can sort the search content desired by the user to the top of the search results, thereby improving search efficiency.

[0005] In a first aspect, embodiments of this application provide a method for ranking search results, including:

[0006] Based on the search terms entered by the user, at least two recommended terms and at least two search results are determined, wherein the recommended terms are historical search terms that include the search terms.

[0007] A ranking score is determined for each search result based on the search term, the at least two search results, and the at least two recommended terms.

[0008] The at least two search results are sorted according to the sorting score.

[0009] Secondly, embodiments of this application provide a search result sorting apparatus, including:

[0010] The first determining module is used to determine at least two recommended terms and at least two search results based on the search terms entered by the user.

[0011] The second determining module is used to determine the ranking score of each search result based on the search term, the at least two search results, and the at least two recommended terms;

[0012] The result sorting module is used to sort the at least two search results according to the sorting score.

[0013] Thirdly, embodiments of this application provide an electronic device, including:

[0014] A processor and a memory, the memory being used to store a computer program, and the processor being used to invoke and run the computer program stored in the memory to perform the search result sorting method as described in the first aspect embodiment or its various implementations.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium for storing a computer program that causes a computer to perform the search result ranking method as described in the first aspect embodiment or its various implementations.

[0016] Fifthly, embodiments of this application provide a computer program product containing program instructions that, when executed on an electronic device, cause the electronic device to perform the search result sorting method as described in the first aspect embodiment or its various implementations.

[0017] The technical solutions disclosed in the embodiments of this application have at least the following beneficial effects:

[0018] By identifying multiple recommended terms and search results associated with the user's search term, and then determining a ranking score for each search result based on the search term, recommended terms, and search results, the search results are ranked according to their ranking scores. This allows the desired search content to be prioritized in the search results, thereby improving search efficiency. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figures 1a-1c This is a search effect diagram of an application software that supports instant search results.

[0021] Figure 2 This is a flowchart illustrating a search result sorting method provided in an embodiment of this application;

[0022] Figure 3a This is a diagram illustrating the latter method of sorting search results using traditional methods;

[0023] Figure 3bThis is a schematic diagram illustrating the sorting of search results based on user-inputted search terms, provided in an embodiment of this application.

[0024] Figure 4 This is a schematic diagram of a process for determining the ranking score of each search result, provided in an embodiment of this application.

[0025] Figure 5 This is a schematic diagram illustrating the determination of the ranking score for each search result based on a search term, at least two recommended terms, and at least two search results, provided in an embodiment of this application.

[0026] Figure 6 This is a flowchart illustrating another search result sorting method provided in an embodiment of this application;

[0027] Figure 7 This is a flowchart illustrating another search result sorting method provided in an embodiment of this application;

[0028] Figure 8 This is a schematic block diagram of a search result sorting device provided in an embodiment of this application;

[0029] Figure 9 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0032] In this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or solution described as "exemplary" or "for example" in this application should not be construed as being better or more advantageous than other embodiments or solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0033] In the description of the embodiments in this application, unless otherwise stated, "multiple" means two or more, that is, at least two. "At least one" means one or more.

[0034] Currently, when users search for content using search terms in applications that support instant search, short search terms lead to a large search scope and a large number of search results. This results in users spending a significant amount of time searching through multiple results, leading to poor search efficiency. This application provides a search result ranking scheme that prioritizes the user's desired search content in the search results, thereby improving search efficiency.

[0035] The technical solutions of this application will be described in detail below through some embodiments. The embodiments described below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0036] Figure 2 This is a flowchart illustrating a search result ranking method provided in an embodiment of this application. The search result ranking method provided in this application can be executed by a search result ranking device. This search result ranking device can consist of hardware and / or software and can be integrated into an electronic device. Optionally, the electronic device in this application can be various terminal devices, such as tablet computers, smartphones (e.g., Android phones, iOS phones, Windows Phones, etc.), laptops, personal digital assistants (PDAs), wearable devices, smart TVs, smart screens, high-definition TVs, 4K TVs, etc. This application does not limit the type of electronic device. The terminal device can also be user equipment (UE), a terminal, or a user device, etc., without any limitations.

[0037] like Figure 2 As shown, the method includes the following steps:

[0038] S101, based on the search term entered by the user, determine at least two recommended terms and at least two search results, wherein the recommended terms are historical search terms that include the search term.

[0039] S102, determine the ranking score for each search result based on the search term, at least two search results, and at least two recommended terms.

[0040] S103, Sort at least two search results based on the sorting score.

[0041] Optionally, when using any application that supports instant search, users can enter search terms in the search box provided by the application. The electronic device will then search the database in real time for multiple search results matching the user's input. Search terms can be, but are not limited to: Chinese characters, Pinyin, uppercase and lowercase English letters, and numbers.

[0042] The aforementioned application software can be software from any field, such as any music software or video playback software in the audio-visual field.

[0043] The search results may be videos, images, documents, or other forms of content, and this application makes no restrictions on this.

[0044] The database mentioned above can be determined based on the type of application software. For example, when the application software is music software, the database can be a music resource database. Similarly, when the application software is document software, the database can be a document resource database. This application does not impose any restrictions on the database.

[0045] For example, when a user is using a music software, they can enter the search term "r" in the search box provided by the software. The electronic device will then filter search results from its music library in real time, selecting those with the prefix "r" that match the user's input. Similarly, when a user enters the search term "ri" in the same search box, the electronic device can filter search results from its music library in real time, selecting those with the prefix "ri" that match the user's input.

[0046] In some embodiments, this application may also determine at least two related recommended terms based on the user's input search terms. This can supplement the user's input search terms based on the determined recommended terms, thereby laying the foundation for ranking the user's desired search content at the top of the search results.

[0047] Considering the differences in search content focus across different regions and / or different applications, for example, users in different regions may have different preferences for singers when searching for "ba" on music apps. A user in Western Europe searching for "ba" will get different results than a user in Central Asia searching for the same term. Furthermore, for cross-domain applications like music and news apps, music apps focus on recommending singers or songs, while news apps focus on recommending the latest news. Therefore, when a user enters the same search term "learn" into different types of applications, the search results will also differ.

[0048] Therefore, when determining at least two related recommended terms based on a user's input search term, this application can first determine the user's location, and then, based on the user's location, determine the target dictionary corresponding to that location within the currently used application software. Then, it can search the target dictionary for at least two recommended terms corresponding to the search term.

[0049] In this application, the dictionary corresponding to a region can be a key-value structure, and this dictionary can be updated periodically. The key is the search term, and the value includes two parts: the first part is multiple recommended terms associated with the search term, and the second part is the probability of each recommended term appearing. For example, assuming a user's region is Australia, then the user's region is represented by the abbreviation "AU" for Australia, and the target dictionary corresponding to "AU" can be shown in Table 1 below.

[0050] Table 1

[0051]

[0052] Therefore, when searching for at least two recommended terms corresponding to a search term in the target dictionary, the user-input search term can be used as the key to search for all recommended terms associated with that key in the target dictionary. In other words, the dictionary in the application software has a region-specific relationship; different regions have different dictionaries.

[0053] In this application, at least two related recommended terms are determined based on the user-input search term, which may be all historical search terms with prefix information of the aforementioned search term. For example, when the search term is "ta", recommended terms with the prefix "ta" can be determined in the target dictionary based on "ta", such as "tankxxx", "taylor aaa", and "tame bbb".

[0054] The probability of each recommended term appearing can be obtained by statistically analyzing historical query data from the application. For example, if historical query data shows that 53.8% of users selected the recommended term "baby xxx," then the probability of "baby xxx" appearing is 0.538. As another example, if historical query data shows that 23.1% of users ended their search with the recommended term "baby," then the probability of "baby" appearing is 0.231.

[0055] After identifying multiple search results and multiple recommended terms corresponding to the user's input search terms, this application can determine the ranking score for each search result based on the multiple search results, multiple recommended terms, and the user's input search terms. Then, based on the ranking score of each search result, the multiple search results are ranked.

[0056] In some optional embodiments, considering that the determined recommended terms are historical search terms whose prefix information is the user's input search term, this application may optionally fuse the search term with at least two recommended terms to complete the search term, making the fused first fused search term more likely to appear as the recommended term. Furthermore, this application also performs similarity matching between the search term and each recommended term and each search result, and fuses the two types of similarity results according to each search result to obtain a second fused search term corresponding to each search result, making the fused second fused search term more likely to appear as the corresponding search result, i.e., the second fused search term can more closely resemble the presentation format of the corresponding search result.

[0057] Furthermore, the first relevance between the first fused search term and each search result, and the second relevance between the second fused search term and each search result, are calculated. Based on the first and second relevances for each search result, a ranking score is determined. Then, based on the ranking score of each search result, the multiple search results are sorted from highest to lowest. This allows for supplementing shorter search terms entered by the user with predetermined recommended terms, resulting in higher ranking scores for the content the user wants to search for. This ensures that the desired search content is ranked higher in the search results, making it easier for the user to quickly find what they are looking for.

[0058] For example, suppose a user enters the search term "Ri" in the search box of a music software. The related recommended terms determined by the search term "Ri" in the target dictionary corresponding to the user's region, and the probabilities of their occurrence, are: [Rihanxx, 0.222], [ritaccc, 0.148], [riptddd, 0.139], [riverxxyy, 0.083], [ritzzz, 0.083]. Furthermore, the search results determined by the search term "Ri" include: Artist:RIEXX, Artist:RIOZZ, Artist:Rihanxx, etc. A traditional image showing the effect of ranking search results is as follows: Figure 3a As shown in the figure. This application determines the ranking score of each search result based on recommended terms, search terms, and search results, and the effect of ranking at least two search results based on the ranking score of each search result is shown in the figure. Figure 3b As shown. Based on Figure 3b As can be seen, this application can rank the search results "Rihanxx" and "ritzzz" that users want to obtain in the top two positions of all search results, so that users can obtain the search content they want to find more quickly.

[0059] The search result ranking method provided in this application determines multiple recommended terms and multiple search results associated with the search term input by the user, and determines a ranking score for each search result based on the search term, multiple recommended terms, and multiple search results. Then, the multiple search results are ranked according to the ranking score, thereby enabling the desired search content to be ranked at the top of the search results and improving search efficiency.

[0060] Based on the foregoing embodiments, the following is combined with Figure 4 and Figure 5 This application provides further explanation of how the ranking score for each search result is determined based on the search term, at least two search results, and at least two recommended terms. For example... Figure 4 As shown, the above S102 includes the following steps S102-1 to S102-3:

[0061] S102-1, Determine the first fused search term based on the search term and each recommended term.

[0062] The first fused search term can be represented in vector form, that is, the first fused search term is the first fused search term vector.

[0063] In some alternative embodiments, the first fused search term can be determined by determining the occurrence probability of the search term and the occurrence probability of each recommended term, and based on the occurrence probability of the search term and the occurrence probability of each recommended term.

[0064] Given that the probability of the user-inputted search term appearing in this search operation is known (100%), meaning the probability of the user-inputted search term appearing is 1, and since the probability of each recommended term is pre-stored in a dictionary, while searching for at least two recommended terms in the target dictionary based on the user-inputted search term, the probability of each recommended term can be obtained from the target dictionary.

[0065] In some optional embodiments, determining the first fused search term based on the occurrence probability of the search term and the occurrence probability of each recommended term may include, but is not limited to, the following:

[0066] In the first case, the probability of the search term appearing is summed with the probability of each recommended term appearing, and the sum is used as the first fused search term.

[0067] In the second case, the weighted sum of the occurrence probability of the search term and the occurrence probability of each recommended term is calculated, and the resulting weighted sum is used as the first fused search term.

[0068] The above-mentioned weighted sum calculation can be understood as calculating the first product of the occurrence probability of the search term and the first weight, and the second product of each recommended term and its corresponding second weight. Then, the sum of the first and second products is calculated, and the calculated weighted sum is used as the first fused search term.

[0069] Among them, the first weight and each second weight are adjustable parameters, which can be flexibly set according to actual needs, and there are no restrictions on them here.

[0070] The search terms mentioned above specifically refer to the search terms entered by the user.

[0071] S102-2, Based on the first similarity between the search term and each search result, and the second similarity between each recommended term and each search result, determine the second fused search term corresponding to each search result.

[0072] The second fused search term can be represented in vector form, that is, the second fused search term is a vector of the second fused search term.

[0073] Optionally, this application calculates a first similarity between the user-input search term and each search result, and calculates a second similarity between each recommended term and each search result, optionally using a similarity algorithm; alternatively, the search term and each search result can be used as input values ​​to a matching model to determine the first similarity between the search term and each search result, and each recommended term and each search result can be used as input values ​​to a matching model to determine the second similarity between each recommended term and each search result, etc. This application does not impose any restrictions on the implementation method of determining similarity.

[0074] The matching model described above can be understood as any network model that supports determining the similarity between two data points, such as a neural network model or a deep network model, etc., and no restrictions are placed on it here. In this application, when the matching model is a neural network model, the network model can be DeepNet, etc.

[0075] In this application, the aforementioned similarity algorithm may be selected from, but is not limited to, cosine similarity algorithm, Euclidean distance, and Manhattan distance.

[0076] Considering that the user-input search terms, recommended terms, and search results can be text strings, this application optionally converts the user-input search terms, recommended terms, and search results into vector representations to facilitate the determination of similarity between search terms and search results, as well as between recommended terms and search results. Then, a similarity algorithm or matching model is used to determine the first similarity between the search term and each search result, and the second similarity between the recommended term and each search result.

[0077] Furthermore, after calculating the first similarity between the search term and each search result, and the second similarity between each recommended term and each search result, this application may optionally perform a weighted summation or summation of the first and second similarities corresponding to each search result, on a per-search-response basis. Then, the summation is used to determine the second fused search term corresponding to each search result.

[0078] In other words, each search result corresponding to the user's input search term will correspond to a second fusion search term, thereby ensuring that each second fusion search term can be more closely similar to the form of the corresponding search result.

[0079] The weights of the first similarity and the second similarity mentioned above are both adjustable parameters, and can be flexibly set according to actual needs. No restrictions are placed on them here.

[0080] It should be noted that the execution order of the above steps S102-1 and S102-2 may be as follows: step S102-1 is executed first, followed by step S102-2; or step S102-2 is executed first, followed by step S102-1; or steps S102-1 and S102-2 may be executed in parallel. This application does not impose any restrictions on this.

[0081] S102-3, determine the ranking score for each search result based on the first fused search term, each second fused search term, and each search result.

[0082] In some alternative embodiments, the present application determines the ranking score for each search result, which may include the following steps:

[0083] Step 1: Determine the third similarity between the first fused search term and each search result, and the fourth similarity between each search result and the second fused search term corresponding to that search result.

[0084] Step 2: Determine the ranking score for each search result based on the third and fourth similarity scores corresponding to each search result.

[0085] The implementation principle of determining the third and fourth similarity corresponding to each search result in step 1 is the same as or similar to the implementation principle of determining the first and second similarity corresponding to each search result in step S102-2 above. For details, please refer to step S102-2 above. It will not be elaborated on here.

[0086] In some alternative embodiments, when determining the ranking score for each search result in step 2 above, the possible methods include, but are not limited to, the following:

[0087] Method 1 involves inputting the third and fourth similarity scores corresponding to each search result into a scoring model. This model then processes the third and fourth similarity scores of each search result to obtain a ranking score for each search result.

[0088] The scoring model described above can be any network model that supports scoring based on at least two similarities, or a network model trained on a large amount of similarity sample data. This application does not impose any restrictions on this.

[0089] Method 2: According to the preset scoring rules, determine the ranking score of each search result based on the third and fourth similarity scores corresponding to each search result.

[0090] The preset scoring rules can be any scoring calculation method or scoring algorithm, and this application does not impose any restrictions on them. It is understood that the aforementioned preset scoring rules can be used to quantitatively evaluate each search result corresponding to the known third and fourth similarities by calculating these similarities, thereby achieving an evaluation of each search result. The evaluation of each search result can be reflected in a score.

[0091] To more clearly explain steps S102-1 to S102-3, the following will be combined with... Figure 5 Let's illustrate with examples.

[0092] like Figure 5 As shown, assuming the user inputs the search term "xx", the number of recommended terms determined based on the search term "xx" is 5: Recommended term 1 (sug1 in the figure), Recommended term 2 (sug2 in the figure), Recommended term 3 (sug3 in the figure), Recommended term 4 (sug4 in the figure), and Recommended term 5 (sug5 in the figure). The probability of recommended term 1 appearing is probability a, the probability of recommended term 2 appearing is probability b, the probability of recommended term 3 appearing is probability c, the probability of recommended term 4 appearing is probability d, and the probability of recommended term 5 appearing is probability e. Additionally, the number of search results determined based on the search term "xx" is 10: Search result 10, Search result 11, Search result 12, Search result 13, Search result 14, Search result 15, Search result 16, Search result 17, Search result 18, and Search result 19. Based on the occurrence probability of search term xx and the occurrence probabilities of the five recommended terms, the first fused search term (WeightedQuery in the diagram) can be determined. Then, based on the first similarity between search term xx and each search result, and the second similarity between each recommended term and each search result, the second fused search term (Doc Attentioned Query in the diagram) corresponding to each search result can be determined. Each search result corresponds to one second fused search term; that is, the second fused search term is different when the search results are different. Next, the third similarity between the first fused search term and each search result, and the fourth similarity between each search result and its corresponding second fused search term, are determined. The third and fourth similarities for each search result are then weighted and summed to obtain the ranking score for each search result.

[0093] This application supplements the user's input search terms with recommended terms, and determines the ranking score of each search result based on the first fused search term determined by the supplemented search terms, and the second fused search term determined by the user's input search terms, each search result, and each recommended term. This allows users to sort the search content more relevant to the shorter search term to the top when performing a search operation based on the shorter search term, enabling users to quickly find the search content they want.

[0094] In another optional implementation, after determining at least two search results based on the user-input search terms, this application may optionally further include determining a first score for each search result based on the search terms and each search result, and ranking each search result based on the first score and the ranking score determined in the foregoing embodiments. The following is in conjunction with... Figure 6 This application provides a further explanation of the search result ranking method.

[0095] like Figure 6 As shown, the method may include the following steps:

[0096] S201, based on the search term entered by the user, determine at least two recommended terms and at least two search results, wherein the recommended terms are historical search terms that include the search term.

[0097] S202, determine the ranking score for each search result based on the search term, at least two search results, and at least two recommended terms.

[0098] S203, based on the search term and at least two search results, determine the first score for each search result.

[0099] In some optional embodiments, determining a first score for each search result may optionally include: determining a fifth similarity between the search term and each search result, and using each fifth similarity as the first score for each search result. It should be understood that the determination of the fifth similarity between the search term and each search result in this application is the same as the implementation process of determining the first similarity between the search term and each search result in part S102-2 of the foregoing embodiments.

[0100] In some alternative embodiments, the fifth similarity between the search term and each search result can be determined using a similarity algorithm; alternatively, the search term and each search result can be used as input values ​​and fed into a matching model to determine the fifth similarity between the search term and each search result, etc. This application does not impose any restrictions on the implementation method of determining the similarity between the search term and each search result.

[0101] The matching model described above can be understood as any network model that supports determining the similarity between two data points, such as a neural network model or a deep network model, etc., and no restrictions are placed on it here. In this application, when the matching model is a neural network model, the network model can be DeepNet, etc.

[0102] The similarity algorithm mentioned above can be, but is not limited to, cosine similarity algorithm, Euclidean distance and Manhattan distance, etc.

[0103] It should be noted that the above-mentioned S202 and S203 can be executed by executing S202 first and then S203; or, S203 can be executed first and then S202; or, S202 and S203 can be executed in parallel. This application does not impose any restrictions on this.

[0104] S204, rank at least two search results based on the ranking score and the first score of each search result.

[0105] In some alternative embodiments, this application may determine a first comprehensive score for each search result based on the ranking score and the first score of each search result. Then, at least two search results are ranked according to the first comprehensive score of each search result.

[0106] In this application, determining the first overall score for each search result may include, but is not limited to, the following:

[0107] In scenario one, the ranking score and the first score of each search result are summed or weighted and summed, and the resulting sum is used as the first comprehensive score for each search result.

[0108] In the weighted summation, the weight values ​​corresponding to the sorted score and the first score are all adjustable parameters, which can be flexibly adjusted according to the user's search needs.

[0109] In the second scenario, the ranking score and first score of each search result are used as input values ​​and fed into the fusion model. The fusion model then combines the ranking score and first score of each search result to obtain the first comprehensive score of each search result.

[0110] The aforementioned fusion model can be any network model that supports obtaining a total score based on the fusion of at least two scores, or a network model trained on a large amount of score sample data. This application does not impose any restrictions on this.

[0111] In scenario three, the ranking score and the first score of each search result are concatenated, and the concatenated result is input into the scoring model so that the scoring model can determine the first comprehensive score of each search result based on the concatenated result of each search result.

[0112] The scoring model described above can be any network model that supports scoring based on input data, or a network model trained on a large number of score sample data. This application does not impose any restrictions on the scoring model.

[0113] Optionally, the scoring model and the fusion model mentioned above can be the same network model or different network models; no specific restrictions are imposed here.

[0114] The search result ranking method provided in this application determines multiple recommended terms and multiple search results associated with the user's input search term. Based on the search term, the recommended terms, and the search results, a ranking score is determined for each search result. The search results are then ranked according to their ranking scores, thereby prioritizing the user's desired search content and improving search efficiency. Furthermore, by adding a first score to the ranking score of each search result to increase the score of the user's desired search content, it becomes easier for the user to quickly find the desired content from multiple search results.

[0115] As another optional implementation, considering that each search result corresponds to at least one feature information, such as whether the search result is a popular search result, user ratings of the search result, etc., this application further includes, based on the aforementioned embodiments: determining a second score for each search result based on the feature information of each search result. Then, each search result is sorted according to its ranking score and second score. The following is combined with... Figure 7 This application provides an explanation of the search result ranking method.

[0116] like Figure 7 As shown, the method may include the following steps:

[0117] S301, based on the search term entered by the user, determine at least two recommended terms and at least two search results, wherein the recommended terms are historical search terms that include the search term.

[0118] S302, determine the ranking score for each search result based on the search term, at least two search results, and at least two recommended terms.

[0119] S303, determine the second score for each search result based on the feature information of each search result.

[0120] In this application, the characteristic information of the search results includes at least one of the following: whether it is a popular search result, whether it is a major recommended search result, the number of times it has been favorited, and a content quality score. The content quality score refers to the user's rating of the search result content.

[0121] In some optional embodiments, when determining the second score of each search result, the feature information of each search result may be input into a feature fusion model, and the feature fusion model determines the second score of each search result based on the feature information of each search result. Specifically, if any search result has at least two feature information items, the feature information of the search results is concatenated, and the concatenated result is input into the feature fusion model to determine the second score of the search result.

[0122] The feature fusion model can be any network model that supports determining scores based on feature information, or a network model trained on a large amount of feature information sample data. This application does not impose specific restrictions on the feature fusion model. In this application, when the feature fusion model is a network model, the network model can be WidNet, etc.

[0123] For example, when the feature information of a search result includes feature 1, feature 2 and feature 3, the feature information of the search result can be concatenated in the manner of feature 1 + feature 2 + feature 3, or in the manner of feature 1 + feature 3 + feature 2, or in the manner of feature 2 + feature 1 + feature 3, etc. This application does not impose specific restrictions on the concatenation method of multiple feature information of search results.

[0124] It should be noted that S302 and S303 can be executed first and then S303; or S303 can be executed first and then S302; or S302 and S303 can be executed in parallel. This application does not impose any restrictions on this.

[0125] S304, rank at least two search results based on the ranking score and the second score of each search result.

[0126] In some alternative embodiments, this application may determine a second comprehensive score for each search result based on the ranking score and the second score of each search result. Then, at least two search results are ranked according to the second comprehensive score of each search result.

[0127] In this application, the determination of a second comprehensive score for each search result may include, but is not limited to, the following:

[0128] In the first scenario, the ranking score and the second score of each search result are summed or weighted and summed, and the resulting sum is used as the second comprehensive score for each search result.

[0129] In the weighted summation, the weight values ​​corresponding to the sorting score and the second score are adjustable parameters, which can be flexibly adjusted according to the user's search needs.

[0130] In the second scenario, the ranking score and the second score of each search result are used as input values ​​and fed into the fusion model. The fusion model then combines the ranking score and the second score of each search result to obtain a second comprehensive score for each search result.

[0131] The aforementioned fusion model can be any network model that supports obtaining a total score based on the fusion of at least two scores, or a network model trained on a large amount of score sample data. This application does not impose any restrictions on this.

[0132] In scenario three, the ranking score and the second score of each search result are concatenated, and the concatenated result is input into the scoring model so that the scoring model can determine the second comprehensive score of each search result based on the concatenated result of each search result.

[0133] The scoring model described above can be any network model that supports scoring based on input data, or a network model trained on a large number of score sample data. This application does not impose any restrictions on the scoring model.

[0134] In some alternative embodiments, considering that each search result has a ranking score and a first score as determined in the foregoing embodiments, this application may optionally rank at least two search results based on the ranking score, the first score, and the second score of each search result.

[0135] As an optional implementation, this application sorts at least two search results based on a ranking score, a first score, and a second score for each search result. This may include: determining a third comprehensive score for each search result based on its ranking score, first score, and second score; and then sorting the at least two search results based on their third comprehensive score.

[0136] In this application, the determination of a third comprehensive score for each search result may include, but is not limited to, the following:

[0137] In the first scenario, the ranking score, first score, and second score of each search result are summed or weighted, and the resulting sum is used as the third comprehensive score for each search result.

[0138] In the weighted summation, the weight values ​​corresponding to the sorted score, the first score, and the second score are adjustable parameters, which can be flexibly adjusted according to the user's search needs.

[0139] In the second scenario, the ranking score, first score, and second score of each search result are input into the fusion model. The fusion model then combines these scores to obtain a third comprehensive score for each search result.

[0140] The aforementioned fusion model can be any network model that supports obtaining a total score based on the fusion of at least two scores, or a network model trained on a large amount of score sample data. This application does not impose any restrictions on this.

[0141] In the third scenario, the ranking score, first score, and second score of each search result are concatenated, and the concatenated result is input into the scoring model. The scoring model then determines the third comprehensive score for each search result based on the concatenated result.

[0142] The scoring model described above can be any network model that supports scoring based on input data, or a network model trained on a large number of score sample data. This application does not impose any restrictions on the scoring model.

[0143] The search result ranking method provided in this application determines multiple recommended terms and multiple search results associated with the user's input search term. Based on the search term, the recommended terms, and the search results, a ranking score is determined for each search result. The search results are then ranked according to their ranking scores, thereby placing the user's desired search content at the top of the search results and improving search efficiency. Furthermore, by updating the ranking of each search result based on its feature information, the score for the user's desired search content is further improved, making it easier for the user to quickly find the desired content from multiple search results.

[0144] The following is a reference to the appendix. Figure 8 This application describes a search result sorting device proposed in an embodiment. Figure 8 This is a schematic block diagram of a search result sorting device provided in an embodiment of this application.

[0145] like Figure 8 As shown, the search result sorting device 400 includes: a first determining module 410, a second determining module 420, and a result sorting module 430.

[0146] The first determining module 410 is used to determine at least two recommended terms and at least two search results based on the search terms entered by the user.

[0147] The second determining module 420 is used to determine the ranking score of each search result based on the search term, the at least two search results, and the at least two recommended terms.

[0148] The result sorting module 430 is used to sort the at least two search results according to the sorting score.

[0149] In one optional implementation of this application embodiment, the second determining module 420 includes:

[0150] The first determining unit is configured to determine a first fused search term based on the search term and each of the recommended terms;

[0151] The second determining unit is configured to determine a second fused search term corresponding to each search result based on a first similarity between the search term and each search result, and a second similarity between each recommended term and each search result.

[0152] The sorting unit is configured to determine a sorting score for each search result based on the first fused search term, each of the second fused search terms, and each of the search results.

[0153] In one optional implementation of this application, the first determining unit is specifically configured to: determine the occurrence probability of the search term and the occurrence probability of each recommended term; and determine the first fused search term based on the occurrence probability of the search term and the occurrence probability of each recommended term.

[0154] In one optional implementation of this application, the second determining unit is specifically used to: perform a weighted summation of the first similarity and the second similarity corresponding to each search result, and determine the sum value as the second fused search term corresponding to each search result.

[0155] In one optional implementation of this application, the sorting unit is specifically configured to: determine a third similarity between the first fused search term and each search result, and a fourth similarity between each search result and the second fused search term corresponding to the search result; and determine a ranking score for each search result based on the third and fourth similarities corresponding to each search result.

[0156] An optional implementation of this application's embodiments further includes:

[0157] The third determining module is used to determine a first score for each of the search results based on the search term and the at least two search results;

[0158] Accordingly, the result sorting module 430 is specifically used to sort the at least two search results according to the sorting score of each search result and the first score.

[0159] In one optional implementation of this application embodiment, the third determining module is specifically used for:

[0160] A fifth similarity score is determined between the search term and each search result, and each fifth similarity score is used as the first score for the corresponding search result.

[0161] In one optional implementation of this application embodiment, the result sorting module 430 is further used for:

[0162] A first comprehensive score is determined for each search result based on the ranking score and the first score.

[0163] The at least two search results are sorted based on a first comprehensive score for each search result.

[0164] An optional implementation of this application's embodiments further includes:

[0165] The fourth determining module is used to determine a second score for each search result based on the feature information of each search result.

[0166] In one optional implementation of this application, the feature information of the search result includes at least one of the following: whether it is a popular search result, whether it is a major recommended search result, the number of times it has been favorited, and the content evaluation score.

[0167] In one optional implementation of this application embodiment, the fourth determining module is specifically used for:

[0168] The feature information of each search result is input into the feature fusion model, and the feature fusion model determines the second score of each search result based on the feature information of each search result;

[0169] If any of the search results has at least two feature information items, then the feature information items of the search results are concatenated, and the concatenation result is input into the feature fusion model to determine the second score of the search results.

[0170] In one optional implementation of this application embodiment, the result sorting module 430 is specifically used for:

[0171] The at least two search results are sorted based on the ranking score and the second score of each search result; or, the at least two search results are sorted based on the ranking score, the first score, and the second score of each search result.

[0172] In one optional implementation of this application embodiment, the result sorting module 430 is further used for:

[0173] A second comprehensive score is determined for each search result based on the ranking score and the second score.

[0174] The at least two search results are sorted based on a second comprehensive score for each search result.

[0175] In one optional implementation of this application embodiment, the result sorting module 430 is further used for:

[0176] A third comprehensive score is determined for each search result based on the ranking score, the first score, and the second score.

[0177] The at least two search results are sorted based on a third comprehensive score for each search result.

[0178] In one optional implementation of this application embodiment, the first determining module 410 is specifically used for:

[0179] Determine the target dictionary based on the user's location;

[0180] Find at least two recommended terms in the target dictionary that correspond to the search term.

[0181] The search result ranking device provided in this application determines multiple recommended terms and multiple search results associated with the search term input by the user, and determines a ranking score for each search result based on the search term, multiple recommended terms and multiple search results. Then, the multiple search results are ranked according to the ranking score, thereby enabling the search content that the user wants to be ranked at the top of the search results and improving search efficiency.

[0182] It should be understood that the device embodiments and the foregoing method embodiments can correspond to each other, and similar descriptions can be referred to the method embodiments. To avoid repetition, further details will not be provided here. Specifically, Figure 8 The device 400 shown can perform Figure 2 The corresponding method embodiments, and the foregoing and other operations and / or functions of each module in device 400 are respectively implemented to achieve Figure 2 For the sake of brevity, the corresponding processes in each method are not described in detail here.

[0183] The apparatus 400 of this application embodiment has been described above from the perspective of functional modules in conjunction with the accompanying drawings. It should be understood that this functional module can be implemented in hardware, in software instructions, or in a combination of hardware and software modules. Specifically, the steps of the first aspect method embodiment in this application can be completed by integrated logic circuits in the processor's hardware and / or by software instructions. The steps of the first aspect method disclosed in this application embodiment can be directly embodied as being executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. Optionally, the software module can be located in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the first aspect method embodiment described above.

[0184] Figure 9 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Figure 9 As shown, the electronic device 500 may include:

[0185] The system includes a memory 510 and a processor 520. The memory 510 stores computer programs and transfers the program code to the processor 520. In other words, the processor 520 can retrieve and run the computer program from the memory 510 to implement the search result sorting method in this embodiment.

[0186] For example, the processor 520 can be used to execute the above-described search result sorting method embodiment according to instructions in the computer program.

[0187] In some embodiments of this application, the processor 520 may include, but is not limited to:

[0188] General-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0189] In some embodiments of this application, the memory 510 includes, but is not limited to:

[0190] Volatile memory and / or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0191] In some embodiments of this application, the computer program may be divided into one or more modules, which are stored in the memory 510 and executed by the processor 520 to complete the search result sorting method provided in this application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the electronic device.

[0192] like Figure 9 As shown, the electronic device 500 may further include:

[0193] Transceiver 530, which can be connected to processor 520 or memory 510.

[0194] The processor 520 can control the transceiver 530 to communicate with other devices; specifically, it can send information or data to other devices or receive information or data sent by other devices. The transceiver 530 may include a transmitter and a receiver. The transceiver 530 may further include antennas, and the number of antennas may be one or more.

[0195] It should be understood that the various components in the electronic device are connected through a bus system, which includes a data bus, a power bus, a control bus, and a status signal bus.

[0196] This application also provides a computer storage medium storing a computer program thereon, which, when executed by a computer, enables the computer to perform the search result sorting method of the above-described method embodiments.

[0197] This application also provides a computer program product containing program instructions, which, when executed on an electronic device, cause the electronic device to perform the search result sorting method described in the above method embodiments.

[0198] When implemented using software, it can be implemented entirely or partially as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0199] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0200] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0201] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. For example, the functional modules in the various embodiments of this application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0202] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method of ranking search results, characterized by, include: Based on the search terms entered by the user, at least two recommended terms and at least two search results are determined, wherein the recommended terms are historical search terms that include the search terms. Based on the search term, the at least two search results, and the at least two recommended terms, a ranking score is determined for each search result, including determining a first fused search term based on the search term and each recommended term; and determining a second fused search term corresponding to each search result based on a first similarity between the search term and each search result, and a second similarity between each recommended term and each search result. Based on the first fused search term, each of the second fused search terms, and each of the search results, a ranking score is determined for each search result. This determination includes: determining a third similarity between the first fused search term and each search result, and a fourth similarity between each search result and the corresponding second fused search term; and determining a ranking score for each search result based on the third and fourth similarities. The at least two search results are sorted according to the sorting score.

2. The method of claim 1, wherein, The step of determining the first fused search term based on the search term and each of the recommended terms includes: Determine the occurrence probability of the search term and the occurrence probability of each of the recommended terms; The first fused search term is determined based on the occurrence probability of the search term and the occurrence probability of each recommended term.

3. The method of claim 1, wherein, The step of determining the second fused search term corresponding to each search result based on the first similarity between the search term and each search result, and the second similarity between each recommended term and each search result, includes: The first similarity and the second similarity corresponding to each search result are weighted and summed, and the sum is determined as the second fused search term corresponding to each search result.

4. The method of claim 1, wherein, Also includes: Based on the search term and the at least two search results, a first score is determined for each of the search results; Accordingly, sorting the at least two search results includes: The at least two search results are sorted based on the ranking score of each search result and the first score.

5. The method according to claim 4, characterized in that, The step of determining a first score for each search result based on the search term and the at least two search results includes: A fifth similarity score is determined between the search term and each search result, and each fifth similarity score is used as the first score for the corresponding search result.

6. The method according to claim 4, characterized in that, The step of sorting the at least two search results based on the ranking score of each search result and the first score includes: A first comprehensive score is determined for each search result based on the ranking score and the first score. The at least two search results are sorted based on a first comprehensive score for each search result.

7. The method according to claim 1 or 4, characterized in that, Also includes: A second score is determined for each search result based on the feature information of each search result.

8. The method of claim 7, wherein, The feature information of the search results includes at least one of the following: whether it is a popular search result, whether it is a major recommended search result, the number of times it has been favorited, and the content quality score.

9. The method of claim 8, wherein, The step of determining a second score for each search result based on the feature information of each search result includes: The feature information of each search result is input into the feature fusion model, and the feature fusion model determines the second score of each search result based on the feature information of each search result; If any of the search results has at least two feature information items, then the feature information items of the search results are concatenated, and the concatenation result is input into the feature fusion model to determine the second score of the search results.

10. The method of claim 7, wherein, The sorting of the at least two search results includes: The at least two search results are sorted based on the ranking score of each search result and the second score; or, The at least two search results are sorted based on the ranking score, the first score, and the second score of each search result.

11. The method of claim 10, wherein, The process of sorting the at least two search results based on the ranking score of each search result and the second score includes: A second comprehensive score is determined for each search result based on the ranking score and the second score. The at least two search results are sorted based on a second comprehensive score for each search result.

12. The method according to claim 10, characterized in that, Sort the at least two search results according to the ranking score, the first score, and the second score of each search result, including: A third comprehensive score is determined for each search result based on the ranking score, the first score, and the second score. The at least two search results are sorted based on a third comprehensive score for each search result.

13. The method according to claim 1, characterized in that, The step of determining at least two recommended terms based on the user's input search terms includes: Determine the target dictionary based on the user's location; Find at least two recommended terms in the target dictionary that correspond to the search term.

14. A search result ranking apparatus characterized by comprising: include: The first determining module is used to determine at least two recommended terms and at least two search results based on the search terms entered by the user. The second determining module is configured to determine a ranking score for each search result based on the search term, the at least two search results, and the at least two recommended terms, including determining a first fused search term based on the search term and each recommended term; and determining a second fused search term corresponding to each search result based on a first similarity between the search term and each search result, and a second similarity between each recommended term and each search result. Based on the first fused search term, each of the second fused search terms, and each of the search results, a ranking score is determined for each search result. This determination includes: determining a third similarity between the first fused search term and each search result, and a fourth similarity between each search result and the corresponding second fused search term; and determining a ranking score for each search result based on the third and fourth similarities. The result sorting module is used to sort the at least two search results according to the sorting score.

15. An electronic device, comprising: include: A processor and a memory, the memory for storing a computer program, the processor for calling and running the computer program stored in the memory to perform the search result sorting method as described in any one of claims 1 to 13.

16. A computer-readable storage medium, characterized in that, Used to store a computer program that causes a computer to perform the search result ranking method as described in any one of claims 1 to 13.

17. A computer program product comprising program instructions, characterized in that, When the program instructions are executed on an electronic device, the electronic device causes the electronic device to perform the search result sorting method as described in any one of claims 1 to 13.

Images