Search processing method, system, computer-readable storage medium, and computer device

Abstract

The application discloses a search processing method and system, a computer readable storage medium and a computer device. The method is applied to a platform including multiple category systems. The method comprises the following steps: determining multiple search objects under multiple category systems according to a search item; acquiring a first category corresponding to the search item under a complete category system and multiple second categories corresponding to the multiple search objects under the complete category system, wherein the complete category system comprises all category information in the multiple category systems; determining the correlation between the first category and the multiple second categories corresponding to the multiple search objects; and determining the search result of the search item in the platform based on the correlation between the first category and the multiple second categories corresponding to the multiple search objects. The application solves the technical problem of inaccurate prediction result in the related art when predicting the category of user search in the multiple category systems.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically, to a search processing method, system, computer-readable storage medium, and computer device. Background Technology

[0002] In the e-commerce sector, when users enter search terms on relevant pages of an e-commerce platform, the platform will recommend and display products that are highly relevant to the user's search intent.

[0003] In related technologies, a common method to obtain category prediction results is to directly calculate the relevance between the user's search terms and the category, and then recommend and display products based on these results. While this method can achieve category prediction, it is only suitable for category prediction within a single category system, and not for category prediction within multiple category systems. For example, using the aforementioned method of directly calculating category relevance can lead to inaccurate prediction results when performing category prediction on user searches.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a search processing method, system, computer-readable storage medium, and computer device to at least solve the technical problem of inaccurate prediction results when performing category prediction for user searches using multiple category systems in related technologies.

[0006] According to one aspect of the present invention, a search processing method is provided, comprising: a platform including multiple category systems, the method comprising: determining multiple search objects under the multiple category systems based on a search term; obtaining a first category corresponding to the search term under a complete category system, and a second category corresponding to each of the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the relevance between the first category and the second category corresponding to the multiple search objects; and determining the search result of the search term within the platform based on the relevance between the first category and the second category corresponding to the multiple search objects.

[0007] Optionally, obtaining the first category corresponding to the search item under the complete category system includes: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the second semantic vector of the search item in the predetermined vector space; calculating the first vector distance between the first semantic vector and the second semantic vector of the categories included in the complete category system respectively; and selecting the category with the smallest first vector distance from the complete category system as the first category.

[0008] Optionally, obtaining the second category corresponding to the plurality of search objects under the complete category system includes: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the third semantic vector of the search object in the predetermined vector space; calculating the second vector distance between the first semantic vector and the third semantic vector of the categories included in the complete category system; and selecting the category with the smallest second vector distance from the complete category system as the second category.

[0009] Optionally, determining the relevance between the first category and the second category corresponding to the plurality of search objects includes: obtaining a similarity coefficient between the first category and the second category corresponding to the plurality of search objects; and determining the relevance between the first category and the second category corresponding to the plurality of search objects based on the similarity coefficient.

[0010] Optionally, determining the search results for the search item within the platform based on the relevance between the first category and the second category corresponding to the plurality of search objects includes: when there are multiple search results, determining a search results list including the plurality of search results based on the relevance between the first category and the second category corresponding to the plurality of search objects, wherein the plurality of search results included in the search results list are arranged based on the magnitude of relevance.

[0011] Optionally, after determining the search results for the search item within the platform based on the correlation between the first category and the second category corresponding to the plurality of search objects, the method further includes: sending the search results to a client to display the search results on the client's display interface.

[0012] According to another aspect of the present invention, a search processing method is provided for a platform including multiple category systems. The method includes: determining multiple search objects under the multiple category systems based on a search item; selecting multiple target objects corresponding to the search item from the multiple search objects; obtaining a first category corresponding to the search item under a complete category system, and a third category corresponding to the multiple target objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the relevance between the first category and the third category corresponding to the multiple target objects; and determining the search results for the search item within the platform based on the relevance between the first category and the third category corresponding to the multiple target objects.

[0013] Optionally, selecting multiple target objects corresponding to the search item from the multiple search objects includes: determining the possible target categories to which the search item belongs; and selecting multiple target objects corresponding to the search item from the multiple search objects based on the possible target categories.

[0014] According to another aspect of the present invention, a search processing system is provided, comprising: a client and a server, the server being configured to maintain a platform including multiple category systems, wherein the client is configured to display a search box on a search interface of the platform; the client is further configured to receive search terms in response to an input operation on the search box, and to send the search terms to the server in response to a search operation on the search terms; the server is configured to determine multiple search objects under the multiple category systems based on the search terms, obtain a first category corresponding to the search terms under a complete category system, and a second category corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; the server is further configured to determine the correlation between the first category and the second category corresponding to the multiple search objects, and to determine the search results of the search terms within the platform based on the correlation between the first category and the second category corresponding to the multiple search objects, and to send the search results to the client; the client is further configured to display the search results on a search results display interface.

[0015] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform the search processing method described in any one of the above embodiments.

[0016] According to another aspect of the present invention, a computer device is provided, comprising: a memory and a processor, the memory storing a computer program; the processor being configured to execute the computer program stored in the memory, wherein the computer program, when executed, causes the processor to perform the search processing method described in any of the preceding embodiments.

[0017] In this embodiment of the invention, after determining multiple search objects under the multiple category systems based on the search terms, the first category of the search terms under the complete category system and the second category system of the multiple search objects under the complete category system are obtained respectively. Based on the unified complete category system, the category relevance between the search terms and the multiple search objects under the multiple category systems is determined, so that all search objects are compared based on the same standard, thus solving the problem of inaccurate prediction caused by different classification standards in different category systems. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0019] Figure 1 This is a hardware structure block diagram of a computer terminal for implementing a search processing method according to an embodiment of the present invention;

[0020] Figure 2 This is a flowchart of an optional search processing method according to an embodiment of the present invention;

[0021] Figure 3 This is a flowchart of another optional search processing method according to an embodiment of the present invention;

[0022] Figure 4 This is a flowchart of an optional search processing method according to an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of an optional dual-tower deep semantic matching model according to an embodiment of the present invention;

[0024] Figure 6 This is a structural block diagram of a search processing system according to an embodiment of the present invention;

[0025] Figure 7 This is a structural block diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention 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 the invention 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 a non-exclusive inclusion; for example, a process, method, system, product, or apparatus 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 apparatus.

[0028] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0029] Proxima is a vector retrieval engine, a vector search engine used to identify a predetermined number of vectors with high similarity to a given vector. Proxima is a high-performance software library for vector nearest neighbor search. Compared to similar open-source products like Faiss, Proxima excels in stability and performance, providing industry-leading basic method modules that support various application scenarios including images, videos, and faces.

[0030] Deep semantic representation models, also known as deep semantic matching modules, are used to obtain semantic vectors of sentences. For example, DSSM (Deep Structured Semantic Models), as a deep semantic representation model, uses massive click and exposure logs of queries (e.g., search terms) and titles from search engines to represent queries and titles as low-dimensional semantic vectors using deep neural networks. It then calculates the distance between two semantic vectors using cosine distance, ultimately training a semantic similarity model. This model can be used to predict the semantic similarity between two sentences and to obtain the low-dimensional semantic vector representation of a given sentence.

[0031] BERT (Bidirectional Encoder Representation from Transformers) is a pre-trained language representation model based on the bidirectional encoder representation of the Transformer model. It can generate deep bidirectional language representations and can be trained using a large number of unlabeled pre-trained models to obtain semantic information of sentences.

[0032] Contrastive learning focuses on learning the common features among similar instances and distinguishing the differences between dissimilar instances. It contrasts with generative learning, which, represented by methods like autoencoders, generates data from existing data, making it similar to the training data in a holistic or high-level semantic sense. Compared to generative learning, contrastive learning doesn't need to focus on the intricate details of instances; it only needs to learn to distinguish data at the abstract semantic level of the feature space. Therefore, the model and its optimization become simpler, and its generalization ability is stronger.

[0033] Category prediction: Given a search term or product, predict the category to which the search term or product belongs.

[0034] Category relevance refers to the relevance between categories. For example, it could be the relevance between the category corresponding to a search term and the category corresponding to the product that appears as a search result.

[0035] Example 1

[0036] According to an embodiment of the present invention, a method embodiment of a search processing method is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0037] The method embodiment provided in Embodiment 1 of this application can be executed in a mobile terminal, computer terminal or similar computing device. Figure 1 A hardware block diagram of a computer terminal for implementing a search processing method is shown. Figure 1 As shown, the computer terminal 10 may include one or more processors (shown as 102a, 102b, ..., 102n in the figure, and the processors may include, but are not limited to, processing devices such as microprocessors (MCUs) or programmable logic devices (FPGAs), a memory 104 for storing data, and a transmission device for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0038] It should be noted that the aforementioned one or more processors and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be implemented wholly or partially as software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be wholly or partially integrated into any other element in the computer terminal 10. As involved in the embodiments of this application, the data processing circuits serve as processor control (e.g., selection of a variable resistor termination path connected to an interface).

[0039] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the search processing method in this embodiment of the invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby implementing the above-mentioned application vulnerability detection method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0040] The transmission device is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of computer terminal 10. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0041] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10.

[0042] It should be noted here that, in some optional embodiments, the above... Figure 1 The computer device shown may include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that... Figure 1 This is only one instance of a specific particular instance, and is intended to illustrate the types of components that may exist in the aforementioned computer devices.

[0043] It should be noted here that, in some embodiments, the above... Figure 1 The computer device shown has a touch display (also referred to as a "touchscreen" or "touch display screen"). In some embodiments, the above... Figure 1 The computer device shown has a graphical user interface (GUI), through which users can interact with the GUI by touching and / or gesturing on the surface of a touchscreen. The human-computer interaction functions here may include the following: creating web pages, drawing, word processing, creating electronic documents, playing games, video conferencing, instant messaging, sending and receiving emails, call interface, playing digital video, playing digital music, and / or web browsing, etc. Executable instructions for performing the above human-computer interaction functions are configured / stored in one or more processor-executable computer program products or computer-readable storage media.

[0044] Under the aforementioned operating environment, this application provides the following: Figure 2 The flowchart shows one of the optional search processing methods. Figure 2 This is a flowchart of a search processing method according to Embodiment 1 of the present invention. Figure 2 As shown, the search processing method provided in this application embodiment is used in platforms that include multiple category systems, and can be implemented through the following steps:

[0045] Step S202: Based on the search terms, determine multiple search objects under multiple category systems.

[0046] In some alternative embodiments, the search term can be search information entered by the user in the search box on the client's search page. For example, it can be one or more search words, one or more search terms, or one or more search sentences, etc.

[0047] In some alternative embodiments, a platform including multiple category systems refers to a platform that includes multiple category systems, and these different category systems may correspond to different sub-platforms. The classification criteria for these multiple category systems can be different. For example, different e-commerce platforms may have different classifications of the goods they sell. Another example is a unified sales platform integrating multiple e-commerce platforms (such as a shopping guide / cashback platform), where different e-commerce platforms may have different classifications of categories. When the classification criteria for multiple category systems are different, multiple different similar categories may appear for the same search term, and these different similar categories may differ significantly. Therefore, the prediction results may be inaccurate, meaning the platform cannot recommend search objects with high category relevance to the user, thus affecting the user experience. For example, in the case of multiple e-commerce platforms, some e-commerce platforms may classify clothing into underwear and outerwear, while others may classify it into tops and pants, etc. When a user enters "shirt" in their search, the relevant categories can be either outerwear or tops. However, there is a significant difference between outerwear and tops. Therefore, the search results based on these two different categories may be inaccurate.

[0048] In some optional embodiments, multiple search objects under multiple category systems are determined based on the search terms. These multiple search objects can be all search objects included in the platform, i.e., the global search objects corresponding to the platform. These multiple search objects can be the sum of one or more search objects included in each of the multiple category systems. That is, when searching for a target based on a search term, the search is performed within these multiple search objects.

[0049] Step S204: Obtain the first category corresponding to the search item under the complete category system, and the second category corresponding to multiple search objects under the complete category system, wherein the complete category system includes all category information in multiple category systems.

[0050] In some optional embodiments, the aforementioned complete category system is relative to the aforementioned multiple category systems, and is complete relative to one or more of the aforementioned multiple category systems. It should be noted that the aforementioned "complete category system including all category information in multiple category systems" can mean that the complete category system is obtained by directly summarizing the multiple category systems, or it can be obtained after processing and organizing the multiple category systems. Specifically, when directly summarizing the multiple category systems to obtain the complete category system, the categories included in the multiple category systems can be first aggregated to obtain a category set, and then duplicate categories in the category set can be removed to obtain a complete category system that does not include duplicate categories. For example, one category system includes shoes, where the shoe category includes sneakers and high heels; another category system includes footwear, where footwear includes sneakers, casual shoes, high heels, and leather shoes. When obtaining a complete category system using the above direct aggregation method, the above categories of sneakers, high heels, casual shoes, and leather shoes can be directly aggregated into a set. Then, duplicate sneakers in the set can be removed to obtain a complete category system that includes sneakers, high heels, casual shoes, and leather shoes.

[0051] Furthermore, when processing and organizing multiple category systems to obtain a complete category system, it can be the result of processing and organizing multiple category systems. For example, in the above example, high heels can be categorized as flat shoes and wedge heels, and leather shoes can be categorized as cloth shoes, rubber shoes, etc. Therefore, based on the above processing and organization, a complete category system can be obtained that includes athletic shoes, casual shoes, high heels, flat shoes, leather shoes, cloth shoes, and rubber shoes.

[0052] Therefore, this complete category system can be a set of category information summarized from multiple category systems, including all category information from those systems, and possessing a unified classification standard. To a certain extent, this complete category system can be understood as a comprehensive category system with more specific and standardized classifications. Therefore, the categories under the complete category system ultimately obtained from multiple category systems can include categories from the previous multiple category systems, as well as categories not present in the previous multiple category systems.

[0053] In some optional embodiments, when obtaining the first category corresponding to the search term in the complete category system, and the second categories corresponding to multiple search objects in the complete category system, the first category corresponding to the search term can be determined based on the correlation between the search term and the category names of the categories included in the complete category system. Similarly, the second categories corresponding to multiple search objects can be determined based on the correlation between the multiple search objects and the category names of the categories included in the complete category system. Furthermore, the two methods of obtaining the first category corresponding to the search term and the second category corresponding to multiple search objects in the complete category system can be the same or different. Optionally, to ensure the accuracy of subsequent prediction results, the same method can be used to avoid inaccuracies in prediction results caused by differences between different methods.

[0054] It should be noted that, compared to search terms, multiple search objects are relatively fixed within an update period. Therefore, when searching for search terms based on these multiple search objects within that update period, the corresponding categories of these search terms under the complete category system can be determined all at once, without needing to recalculate them each time a search term is received. For multiple search objects within different update periods, it is also possible to update only the categories corresponding to the changed search objects under the complete category system. Changed search objects include: newly added search objects, deleted search objects, modified search objects, etc.

[0055] Step S206: Determine the relevance between the first category and the second category corresponding to multiple search objects.

[0056] In some alternative embodiments, there are various methods to determine the relevance between the first category and the second categories corresponding to multiple search objects. For example, the relevance can be determined by directly comparing the textual similarity between the first category and the second categories corresponding to multiple search objects, in order to determine the relevance more accurately and quantitatively. Alternatively, the relevance can be obtained by calculating the vector similarity between the first category and the second categories corresponding to multiple search objects.

[0057] Step S208: Based on the correlation between the first category and the second category corresponding to multiple search objects, determine the search results of the search item within the platform.

[0058] In some optional embodiments, when determining the search results for a search term within the platform based on the relevance between the first category and the second categories corresponding to multiple search objects, the determined search results can be one or multiple. When there are multiple search results, they can be sorted based on the magnitude of their corresponding relevance, for example, sorted in descending order of relevance, for easier subsequent display.

[0059] Through the above embodiments, after determining multiple search objects under multiple category systems based on the search terms, the first category of the search terms under the complete category system and the second category system of the multiple search objects under the complete category system are obtained respectively. Based on the unified complete category system, the category relevance between the search terms and multiple search objects in the platform is determined, so that all search objects are compared based on the same standard, which solves the problem of inaccurate prediction caused by different classification standards in different category systems.

[0060] In some optional embodiments, obtaining the first category corresponding to the search term under the complete category system includes: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the second semantic vector of the search term in the predetermined vector space; calculating the first vector distance between the first semantic vector and the second semantic vector of the categories included in the complete category system; and selecting the category with the smallest first vector distance from the complete category system as the first category. When obtaining the first semantic vector of the categories included in the complete category system in the predetermined vector space, the semantic vector of the category in the predetermined vector space can be determined based on the category description information, wherein the category description information can be the category name, the category description text, etc. When obtaining the second semantic vector of the search term in the predetermined vector space, the semantic content of the search term is represented based on the semantic vector, so that the search term can more accurately represent its search purpose. Different processing methods can be adopted depending on the form of the search terms. For example, when the search term is a single character, the semantic vector of the single character can be determined directly. When the search term is multiple words, the multiple words can be connected to obtain a combined word, and the semantic vector of the combined word can be determined.

[0061] Furthermore, it should be noted that semantic vectors convert the symbolic representation of text into a vector representation in a semantic space. Vector representation allows for a more accurate determination of the similarity between two compared objects. Therefore, by obtaining the first vector distance between the first semantic vector of the categories included in the complete category system and the second semantic vector of the search term, the category with the smallest first vector distance is determined as the first category (wherein, the category with the smallest first vector distance can be one or more). This method accurately identifies the first category with the highest relevance to the search term, and it is simple and efficient.

[0062] In some optional embodiments, obtaining the second category corresponding to multiple search objects under a complete category system includes: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the third semantic vector of the search object in the predetermined vector space; calculating the second vector distance between the first semantic vector and the third semantic vector of the categories included in the complete category system; and selecting the category with the smallest second vector distance from the complete category system as the second category. When obtaining the third semantic vector of the search result in the predetermined vector space, since the search result is generally an object with relatively detailed descriptions, the semantics of the search object can be accurately and completely reflected from the descriptions of these detailed descriptions. Moreover, as described above, the search object is relatively fixed compared to the search terms; therefore, determining the third semantic vector of the search object in the predetermined vector space does not require frequent updates. Furthermore, similar to determining the first category corresponding to the search term, by obtaining the second vector distance between the first semantic vector of the categories included in the complete category system and the third semantic vector of the search result, the category with the smallest second vector distance is determined as the second category. It can accurately identify the second category that is most relevant to the search terms, and the method is simple and efficient.

[0063] In some optional embodiments, after obtaining the first semantic vector of the categories included in the complete category system in the predetermined vector space, the first semantic vector is stored in the vector retrieval engine. Therefore, when retrieving the first category corresponding to the search item under the complete category system or the second category corresponding to the search result under the complete category system, it is not necessary to repeatedly calculate the first semantic vector of the categories included in the complete category system in the predetermined vector space. This improves computational efficiency.

[0064] In some optional embodiments, determining the relevance between a first category and a second category corresponding to multiple search objects includes: obtaining a similarity coefficient between the first category and the second category corresponding to the multiple search objects; and determining the relevance between the first category and the second category corresponding to the multiple search objects based on the similarity coefficient. In one optional embodiment, there are various methods for determining the relevance between the first category and the second category corresponding to the multiple search objects based on the similarity coefficient. For example, the relevance between the first category and the second category can be calculated using the Jaccard similarity coefficient calculation method. Determining the relevance between the first category and the second category using the similarity coefficient is a simple method and yields highly accurate relevance results.

[0065] In some optional embodiments, after determining the relevance between the first category and the second categories corresponding to multiple search objects, the relevance between the search item and the search results can be determined based on the relevance between the first category and the second categories corresponding to the multiple search objects. By indirectly obtaining the relevance between the search item and the search object based on the relevance between the first category corresponding to the search item and the second category corresponding to the search object, the search objects related to the search item can be determined, and search results that satisfy the predetermined search purpose can be obtained.

[0066] In some optional embodiments, when determining search results for a search term within a predetermined search range based on the relevance between the first category and the second categories corresponding to multiple search objects, if there are multiple search results, a search results list including multiple search results is determined based on the relevance between the first category and the second categories corresponding to the multiple search objects. The multiple search results included in the search results list are arranged according to their relevance. Representing multiple search results in a list format, and with the order of search results in the list representing their respective relevance levels, allows users to select the most needed search results more quickly and efficiently when choosing a desired search result, effectively improving the user experience.

[0067] In some optional embodiments, after determining the search results for the search term within a predetermined search range based on the relevance between the first category and the second categories corresponding to multiple search objects, the method further includes: sending the search results to the client to display the search results on the client's display interface. Arranging the search results based on the magnitude of the relevance between the search term and multiple search results, and displaying them according to the arrangement, allows search results with higher relevance to the user's search term to be displayed first, satisfying the user's search needs and improving the user experience.

[0068] Figure 3 This is a flowchart of another search processing method according to an embodiment of the present invention. Figure 3 As shown, the search processing method provided in this application embodiment is used in a platform that includes multiple category systems. This method can be implemented through the following steps:

[0069] Step S302: Based on the search terms, determine multiple search objects under multiple category systems.

[0070] In some alternative embodiments, the search objects included within the platform may include multiple categories. For example, the search objects included within the platform may include all search objects from one database, or all search objects from multiple databases. These multiple databases correspond to the aforementioned multiple e-commerce platforms; for example, they may be multiple e-commerce platforms with different category classification standards.

[0071] Step S304: Select multiple target objects corresponding to the search item from multiple search objects.

[0072] In some alternative embodiments, there are various methods for selecting multiple target objects corresponding to a search term from multiple search objects. For example, the target objects corresponding to the search term can be obtained through a search engine. Another example is that the target objects corresponding to the search term can be obtained directly through vector similarity calculation.

[0073] There are several methods for obtaining search results corresponding to search terms using vector similarity calculation. For example, one approach is to obtain the semantic vector of the search term and the semantic vectors of multiple target objects in a pre-defined database, then calculate the vector similarity between the semantic vector of the search term and the semantic vectors of the multiple target objects. Objects with a vector similarity exceeding a set threshold, or those with the highest vector similarity, are identified as the target objects corresponding to the search term. Another approach is to obtain keywords or key terms included in the search term, calculate the vector similarity between these keywords or key terms and the vectors of multiple objects in a pre-defined database, and identify objects with a vector similarity exceeding a set threshold, or those with the highest vector similarity, as the target objects corresponding to the search term.

[0074] The target objects obtained using the aforementioned method, corresponding to the search terms, can be determined based on different corresponding conditions. For example, the corresponding conditions may include at least one of the following: the target object includes the text content of the search term; the target object includes the keywords or key words of the search term, that is, the target object includes text that is semantically the same as or similar to the search term. For example, if the search term is "Martin boots," the corresponding target objects obtained may include "Brand A Martin boots," "Material B Martin boots," "Martin boots," "Martin shoes," and so on.

[0075] By selecting multiple target objects corresponding to the search terms from multiple search objects, especially when multiple search objects are updated frequently, it is not necessary to perform relevance analysis on all search objects and search terms. Only some search objects related to this search need to be analyzed. Therefore, it is possible to effectively improve the speed of providing search results and enhance the user experience.

[0076] Step S306: Obtain the first category corresponding to the search item under the complete category system, and the third category corresponding to multiple target objects under the complete category system, wherein the complete category system includes all category information in multiple category systems.

[0077] Step S308: Determine the correlation between the first category and the third category corresponding to multiple target objects.

[0078] Step S310: Based on the correlation between the first category and the third category corresponding to multiple target objects, determine the search results of the search item within the platform.

[0079] In some optional embodiments, after determining multiple search objects under multiple category systems based on the search term, and selecting multiple target objects corresponding to the search term from these multiple search objects, the method described above for determining the relevance of all search objects within a predetermined search space to the search term can be referred to, only determining the relevance between these multiple target objects and the search term. By selecting target objects corresponding to the search term from a wide range of multiple search objects as described above, the time required to determine the relevance between search objects unrelated to the search term and the search term can be greatly reduced.

[0080] Similarly, through the above embodiments, after determining the search term, multiple search objects within a predetermined search range, and selecting multiple target objects corresponding to the search term from the multiple search objects, the first category of the search term under the complete category system and the second category system of the multiple target objects under the complete category system are obtained respectively. Based on the unified complete category system, the category relevance between the search term and the multiple target objects is determined, so that the selected target objects are compared based on the same standard, thus solving the problem of inaccurate prediction caused by different classification standards in different category systems.

[0081] As an optional implementation, when selecting multiple target objects corresponding to a search term from multiple search objects, various methods can be used. For example, the possible target categories to which the search term belongs can be determined first; based on the possible target categories, multiple target objects corresponding to the search term can be selected from multiple search objects. For example, when the search term is "shoes," the corresponding possible target category can be directly determined to be a category related to shoes, such as "shoes" or "boots." This allows the selection of objects included under the "shoes" or "boots" category from multiple search objects as target objects, effectively excluding search objects unrelated to shoes, such as clothing, wallets, and accessories, and effectively avoiding related technical processing for these unrelated search objects.

[0082] Based on the above embodiments and optional embodiments, an optional implementation method is provided, which is described in detail below.

[0083] In e-commerce search scenarios, category prediction can prioritize products from categories relevant to a user's search intent, thus ensuring a better user experience. Typically, category prediction results are obtained by directly calculating the relevance between the user's search terms and the relevant categories, and products are then recommended based on these results. While this method can achieve category prediction, it is only suitable for single-category systems and not for multiple systems. For example, some cashback e-commerce platforms may include multiple sub-platforms, each with its own category system. In this case, the category set of these cashback e-commerce platforms includes categories from multiple category systems. For situations where multiple category systems exist, the current method of directly calculating the relevance between the user's search terms and the relevant categories is insufficient for accurate category prediction. In other words, when performing category prediction for search objects corresponding to multiple category systems, this technology suffers from inaccurate prediction.

[0084] Therefore, this disclosure provides a search processing method that determines the relevance between the first and second categories by obtaining the first category corresponding to the search item in a complete category system and the second category corresponding to the search object in multiple category systems. Thus, by indirectly calculating the indirect category relationship between the search item and the search object through a complete category system, category prediction of the search item is achieved, solving the problem of inaccurate category prediction when predicting search items based on search objects in multiple category systems in related technologies.

[0085] In this embodiment, the following example is used: a complete category system consists of categories from multiple category systems corresponding to multiple e-commerce platforms; the search term is user search information; and the search object is the product title in the e-commerce platform.

[0086] Figure 4 This is a flowchart of an optional search processing method according to an embodiment of the present invention. (Refer to...) Figure 4 As shown, the search processing method includes the following steps:

[0087] Step S402: Using a deep semantic matching model, the product title and the categories under the complete category system are vectorized to obtain the semantic vector of the product title (equivalent to the aforementioned third semantic vector) and the semantic vector of the category (equivalent to the aforementioned first semantic vector).

[0088] Step S404: Use the Proxima vector recall engine to retrieve the category corresponding to the product title (or the title field of the product table) (equivalent to the second category in the aforementioned embodiment).

[0089] Step S406: Vectorize the user search information and use the Proxima vector retrieval engine to retrieve the category corresponding to the user search information (equivalent to the first category in the aforementioned embodiment).

[0090] Step S408: Use the Jaccard similarity coefficient calculation method to calculate the category relevance between the category corresponding to the product title and the category corresponding to the user search information.

[0091] Using the above method, based on a relatively complete and standardized category system, and combined with a deep semantic matching model and the Proxima vector recall engine, the categories corresponding to user search terms and customer products (product titles) are obtained respectively. Then, the category relevance between the categories corresponding to user search terms and customer products is calculated.

[0092] There are various types of deep semantic matching models. In this optional implementation, a deep semantic matching model based on the BERT dual-tower model structure is used to vectorize semantic objects. The BERT dual-tower model is in contrast to the single-tower model. The single-tower model, represented by BERT, is a pre-trained model where semantic similarity is calculated by constructing sentence pairs as input and then performing similarity calculations through feature interactions between sentences; this is also known as an interactive model. The BERT dual-tower model, however, calculates sentence vector representations for each individual sentence and then calculates the distance between the two sentence vectors, thus creating an interaction. A common characteristic of representation-based matching models is that they encode and extract features from the two sentences to be matched separately, and finally perform similarity interaction calculations. This BERT dual-tower model can meet the requirements of a large number of category data and a high amount of information contained in the category names of categories from multiple category systems.

[0093] Figure 5 This is a schematic diagram of an optional dual-tower deep semantic matching model according to an embodiment of the present invention. (Refer to...) Figure 5 As shown, the similarity between the category corresponding to the user's search information and the category corresponding to the product title, as well as the user's search information or product title (statement A) and the corresponding category (statement B) are used as training sample datasets to input into the model for training, thereby obtaining a deep semantic matching model for vectorizing the data to be processed.

[0094] Furthermore, in the deep semantic matching model, to better learn the similarity between categories, user search terms, and product titles, this optional embodiment uses a contrastive learning loss function as the optimization objective to optimize the model. The contrastive learning loss function Loss(U,V) is specifically as follows:

[0095] Loss(U,V)=cosine-sim(U,V) / sum{V'! =V}(cosine-sim(U,V'))

[0096] Where U represents the user search information or product title, V represents the category corresponding to the user search information or product title, and V' represents other categories under the complete category system besides the category corresponding to the user search information or product title.

[0097] In this optional embodiment, both the categories corresponding to the user search terms and the categories corresponding to the customer products include multiple categories. To meet the accuracy requirements of category relevance calculation, this optional embodiment employs the high-performance Jaccard similarity coefficient calculation method to calculate the category relevance between the user search terms and the customer products. Specifically, the calculation of the category relevance J(A,B) between the user search terms and the customer products using the Jaccard similarity coefficient calculation method is as follows:

[0098]

[0099] Where A represents the category corresponding to the user's search, and B represents the category corresponding to the product title.

[0100] This optional implementation method, based on a complete category system containing one or more category systems, calculates the category relevance between product titles and user search terms. This allows for category prediction of user search terms within a category set containing multiple category systems. When using the vector recall method for category prediction, the semantic vectors of each category in the category set are pre-stored in the vector recall engine. That is, when retrieving the category corresponding to the user search or the category corresponding to the product title, it is not necessary to repeatedly calculate the semantic vector of the category. This fully utilizes category information and improves the processing efficiency of the method. The high-performance Jaccard similarity coefficient calculation method is used to calculate the category relevance between user search terms and customer products, improving the accuracy of category relevance calculation.

[0101] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0102] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the related technology, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0103] Example 2

[0104] Embodiments of the present invention also provide a search processing system. Figure 6 This is a structural block diagram of a search processing system according to an embodiment of the present invention. Figure 6 As shown, the search processing system 600 provided in this embodiment includes a client 602 and a server 604, wherein the server 604 is used to maintain a platform including multiple category systems.

[0105] This client 602 is used to display the search box on the platform's search interface;

[0106] The client 602 is also used to receive search terms in response to input to the search box, and to send search terms to the server in response to search for the search terms.

[0107] The server 604 is used to determine multiple search objects under multiple category systems based on the search terms, obtain the first category corresponding to the search terms under the complete category system, and the second category corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in multiple category systems.

[0108] The server 604 is also used to determine the relevance between the first category and the second category corresponding to multiple search objects, and based on the relevance between the first category and the second category corresponding to multiple search objects, to determine the search results of the search items within the platform and send the search results to the client; the client is also used to display the search results on the search results display interface.

[0109] In the above optional embodiments, through interaction between the client and the server, a search box is displayed on the client's search interface. In response to input into the search box, the client receives search terms, and in response to search operations on the search terms, sends the search terms to the server. Based on the search terms, the server determines multiple search objects under multiple category systems, obtains the first category corresponding to the search term in the complete category system, and the second categories corresponding to the multiple search objects in the complete category system. The complete category system includes all category information in multiple category systems. The server determines the relevance between the first category and the second categories corresponding to the multiple search objects, and based on the relevance between the first category and the second categories corresponding to the multiple search objects, determines the search results for the search terms within the platform and sends the search results to the client. The client displays the search results on the search results display interface. Thus, based on the division of labor between the client and the server, the server determines the category relevance between the search terms and the multiple search objects based on a unified complete category system, ensuring that all selected search objects are compared based on the same standard. This solves the problem of inaccurate predictions caused by different classification standards in different category systems. Moreover, through the interaction between the client and the server, the client's display function and the server's computing and processing functions are effectively combined, thereby improving the processing efficiency of search tasks.

[0110] Embodiments of the present invention also provide a computer-readable storage medium. Optionally, in this embodiment, the computer-readable storage medium can be used to store the program code executed by the search processing method provided in Embodiment 1.

[0111] Optionally, in this embodiment, the computer-readable storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0112] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for performing the following steps: for a platform including multiple category systems, the method includes: determining multiple search objects under multiple category systems based on search terms; obtaining a first category corresponding to the search terms under a complete category system, and second categories corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the relevance between the first category and the second categories corresponding to the multiple search objects; and determining the search results for the search terms within the platform based on the relevance between the first category and the second categories corresponding to the multiple search objects.

[0113] Optionally, in this embodiment, the computer-readable storage medium is further configured to store program code for performing the following steps: obtaining the first category corresponding to the search item under the complete category system, including: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the second semantic vector of the search item in the predetermined vector space; calculating the first vector distance between the first semantic vector and the second semantic vector of the categories included in the complete category system respectively; and selecting the category with the smallest first vector distance from the complete category system as the first category.

[0114] Optionally, in this embodiment, the computer-readable storage medium is further configured to store program code for performing the following steps: obtaining the second category corresponding to multiple search objects under a complete category system, including: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the third semantic vector of the search object in the predetermined vector space; calculating the second vector distance between the first semantic vector and the third semantic vector of the categories included in the complete category system; and selecting the category with the smallest second vector distance from the complete category system as the second category.

[0115] Optionally, in this embodiment, the computer-readable storage medium is further configured to store program code for performing the following steps: determining the relevance between a first category and a second category corresponding to multiple search objects, including: obtaining a similarity coefficient between the first category and the second category corresponding to the multiple search objects; and determining the relevance between the first category and the second category corresponding to the multiple search objects based on the similarity coefficient.

[0116] Optionally, in this embodiment, the computer-readable storage medium is further configured to store program code for performing the following steps: determining the search results of the search item within the platform based on the relevance between the first category and the second category corresponding to the multiple search objects, including: when there are multiple search results, determining a search results list including multiple search results based on the relevance between the first category and the second category corresponding to the multiple search objects, wherein the multiple search results included in the search results list are arranged based on the magnitude of relevance.

[0117] Optionally, in this embodiment, the computer-readable storage medium is further configured to store program code for performing the following steps: after determining the search results of the search item within the platform based on the correlation between the first category and the second category corresponding to multiple search objects, the method further includes: sending the search results to the client to display the search results on the client's display interface.

[0118] In this embodiment, a computer-readable storage medium is configured to store program code for performing the following steps for a platform including multiple category systems: determining multiple search objects under multiple category systems based on a search term; selecting multiple target objects corresponding to the search term from the multiple search objects; obtaining a first category corresponding to the search term under a complete category system, and a third category corresponding to each of the multiple target objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the correlation between the first category and the third categories corresponding to the multiple target objects; and determining the search results for the search term within the platform based on the correlation between the first category and the third categories corresponding to the multiple target objects.

[0119] Optionally, in this embodiment, the computer-readable storage medium is further configured to store program code for performing the following steps: selecting multiple target objects corresponding to a search item from multiple search objects, including: determining the possible target category to which the search item belongs; and selecting multiple target objects corresponding to the search item from multiple search objects based on the possible target category.

[0120] Embodiments of the present invention can provide a computer device, which can be any one of a group of computer terminals. Optionally, in this embodiment, the computer terminal can also be replaced by a mobile terminal or other terminal device.

[0121] Optionally, in this embodiment, the computer device described above may be located in at least one of a plurality of network devices in a computer network.

[0122] In this embodiment, the computer device described above can execute program code for the following steps in the application's search processing method: For a platform including multiple category systems, the method includes: determining multiple search objects under the multiple category systems based on the search item; obtaining a first category corresponding to the search item under a complete category system, and second categories corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the relevance between the first category and the second categories corresponding to the multiple search objects; and determining the search results for the search item within the platform based on the relevance between the first category and the second categories corresponding to the multiple search objects.

[0123] Optionally, Figure 7 This is a structural block diagram of a computer device according to an embodiment of the present invention. Figure 7 As shown, the computer device may include one or more (only one is shown in the figure) processors 702, memory 704, etc.

[0124] The memory 704 can be used to store software programs and modules, such as the program instructions / modules corresponding to the search processing method and apparatus in this embodiment of the invention. The processor 702 executes various functional applications and data processing by running the software programs and modules stored in the memory 704, thereby realizing the aforementioned search processing method. The memory 704 may include a high-speed random access memory 704, and may also include non-volatile memory 704, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory 704. In some instances, the memory 704 may further include memory 704 remotely located relative to the processor 702, and these remote memories 704 can be connected to a computer terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0125] The processor 702 can invoke information and application programs stored in the memory 704 via a transmission device to perform the following steps: for a platform including multiple category systems, the method includes: determining multiple search objects under multiple category systems based on a search term; obtaining a first category corresponding to the search term under a complete category system, and second categories corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the relevance between the first category and the second categories corresponding to the multiple search objects; and determining the search results for the search term within the platform based on the relevance between the first category and the second categories corresponding to the multiple search objects.

[0126] Optionally, the processor 702 may also execute program code with the following steps: obtaining the first category corresponding to the search item under the complete category system, including: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the second semantic vector of the search item in the predetermined vector space; calculating the first vector distance between the first semantic vector and the second semantic vector of the categories included in the complete category system respectively; and selecting the category with the smallest first vector distance from the complete category system as the first category.

[0127] Optionally, the processor 702 may also execute program code that performs the following steps: obtaining the second category corresponding to multiple search objects under a complete category system, including: obtaining the first semantic vector of the categories included in the complete category system in a predetermined vector space; obtaining the third semantic vector of the search object in the predetermined vector space; calculating the second vector distance between the first semantic vector and the third semantic vector of the categories included in the complete category system respectively; and selecting the category with the smallest second vector distance from the complete category system as the second category.

[0128] Optionally, the processor 702 may also execute program code that performs the following steps: determining the relevance between the first category and the second category corresponding to the multiple search objects, including: obtaining the similarity coefficient between the first category and the second category corresponding to the multiple search objects; and determining the relevance between the first category and the second category corresponding to the multiple search objects based on the similarity coefficient.

[0129] Optionally, the processor 702 may also execute program code that performs the following steps: determining the search results for the search item within the platform based on the relevance between the first category and the second category corresponding to the multiple search objects, including: when there are multiple search results, determining a search results list including multiple search results based on the relevance between the first category and the second category corresponding to the multiple search objects, wherein the multiple search results included in the search results list are arranged according to the size of the relevance.

[0130] Optionally, the processor 702 may also execute program code that performs the following steps: after determining the search results for the search item within the platform based on the correlation between the first category and the second category corresponding to multiple search objects, it further includes: sending the search results to the client to display the search results on the client's display interface.

[0131] The processor 702 can invoke information and application programs stored in the memory 704 via a transmission device to perform the following steps: for a platform including multiple category systems, the method includes: determining multiple search objects under multiple category systems based on a search item; selecting multiple target objects corresponding to the search item from the multiple search objects; obtaining a first category corresponding to the search item under a complete category system, and a third category corresponding to each of the multiple target objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; determining the relevance between the first category and the third categories corresponding to the multiple target objects; and determining the search results for the search item within the platform based on the relevance between the first category and the third categories corresponding to the multiple target objects.

[0132] Optionally, the processor 702 may also execute program code that performs the following steps: selecting multiple target objects corresponding to the search item from multiple search objects, including: determining the possible target categories to which the search item belongs; and selecting multiple target objects corresponding to the search item from multiple search objects based on the possible target categories.

[0133] In this embodiment of the invention, after determining multiple search objects under multiple category systems based on the search terms, the first category of the search terms under the complete category system and the second category system of the multiple search objects under the complete category system are obtained respectively. Based on the unified complete category system, the category relevance between the search terms and the multiple search objects under the multiple category systems is determined, so that all search objects are compared based on the same standard, thus solving the problem of inaccurate prediction caused by different classification standards in different category systems.

[0134] Those skilled in the art will understand that Figure 7 The structure shown is for illustrative purposes only. The computer terminal can also be a smartphone (such as an Android phone, an iOS phone, etc.), a tablet computer, a mobile internet device (MID), a PAD, and other terminal devices. Figure 7 This does not limit the structure of the aforementioned electronic devices. For example, a computer terminal may also include components that are more... Figure 7 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 7 The different configurations shown.

[0135] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc.

[0136] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0137] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0138] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0139] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0140] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0141] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to related technologies, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned computer-readable storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0142] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A search processing method, characterized in that, For platforms that include multiple category systems, the method includes: Based on the search terms, multiple search objects under the aforementioned multiple category systems are determined; Obtain the first category corresponding to the search item under the complete category system, and the second category corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; Determine the relevance between the first category and the second category corresponding to the plurality of search objects; Based on the correlation between the first category and the second category corresponding to the plurality of search objects, the search results for the search item within the platform are determined; Determining the relevance between the first category and the second category corresponding to the plurality of search objects includes: determining the relevance between the first category and the second category corresponding to the plurality of search objects based on the similarity coefficient between the first category and the second category corresponding to the plurality of search objects.

2. The method according to claim 1, characterized in that, The step of obtaining the first category corresponding to the search item in the complete category system includes: Obtain the first semantic vector of the categories included in the complete category system in a predetermined vector space; Obtain the second semantic vector of the search item in the predetermined vector space; Calculate the first vector distance between the first semantic vector and the second semantic vector of each category included in the complete category system; The category with the smallest first vector distance from the complete category system is selected as the first category.

3. The method according to claim 1, characterized in that, The step of obtaining the second category corresponding to each of the multiple search objects under the complete category system includes: Obtain the first semantic vector of the categories included in the complete category system in a predetermined vector space; Obtain the third semantic vector of the search object in the predetermined vector space; Calculate the second vector distance between the first semantic vector and the third semantic vector of the categories included in the complete category system respectively; The category with the smallest second vector distance from the complete category system is selected as the second category.

4. The method according to claim 1, characterized in that, The step of determining the search results for the search item within the platform based on the relevance between the first category and the second category corresponding to the plurality of search objects includes: When there are multiple search results, a search results list including the multiple search results is determined based on the relevance between the first category and the second category corresponding to the multiple search objects, wherein the multiple search results included in the search results list are arranged according to the magnitude of relevance.

5. The method according to any one of claims 1 to 4, characterized in that, After determining the search results for the search item within the platform based on the relevance between the first category and the second category corresponding to the plurality of search objects, the method further includes: The search results are sent to the client for display on the client's screen.

6. A search processing method, characterized in that, For platforms that include multiple category systems, the method includes: Based on the search terms, multiple search objects under the aforementioned multiple category systems are determined; Select multiple target objects corresponding to the search item from the multiple search objects; Obtain the first category corresponding to the search item under the complete category system, and the third category corresponding to the multiple target objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems; Determine the correlation between the first category and the third category corresponding to the plurality of target objects; Based on the correlation between the first category and the third category corresponding to the plurality of target objects, the search results of the search item within the platform are determined; Determining the correlation between the first category and the third category corresponding to the plurality of target objects includes: determining the correlation between the first category and the third category corresponding to the plurality of target objects based on the similarity coefficient between the first category and the third category corresponding to the plurality of target objects.

7. The method according to claim 6, characterized in that, The step of selecting multiple target objects corresponding to the search item from the multiple search objects includes: Determine the possible target category to which the search item belongs; Based on the possible target categories, select multiple target objects corresponding to the search item from the multiple search objects.

8. A search processing system, characterized in that, include: The client and server, wherein the server is used to maintain a platform including multiple category systems, wherein, The client is used to display a search box on the search interface of the platform; The client is also configured to receive search terms in response to an input operation to the search box, and to send the search terms to the server in response to a search operation to the search terms. The server is configured to determine multiple search objects under the multiple category systems based on the search item, obtain the first category corresponding to the search item under the complete category system, and the second category corresponding to the multiple search objects under the complete category system, wherein the complete category system includes all category information in the multiple category systems. The server is further configured to determine the correlation between the first category and the second category corresponding to the plurality of search objects, and based on the correlation between the first category and the second category corresponding to the plurality of search objects, determine the search results of the search item within the platform, and send the search results to the client; The client is also used to display the search results on the search results display interface; The server is further configured to determine the relevance between the first category and the second category corresponding to the plurality of search objects based on the similarity coefficient between the first category and the second category corresponding to the plurality of search objects.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the search processing method according to any one of claims 1 to 7.

10. A computer device, characterized in that, include: Memory and processor The memory stores computer programs; The processor is configured to execute a computer program stored in the memory, wherein when the computer program is executed, the processor performs the search processing method according to any one of claims 1 to 7.

Images