Identification method, recommendation method, display method and computing device

Abstract

Provided in the embodiments of the present disclosure are an identification method, a recommendation method, a display method and a computing device. The identification method comprises: acquiring a plurality of pieces of historical search information; identifying object categories respectively associated with the plurality of pieces of historical search information; extracting, from the plurality of pieces of historical search information, object feature information related to the associated object categories; determining the degree of importance of the object feature information with respect to a target attribution area; and using object feature information with a degree of importance meeting a differentiation condition as differentiated feature information of the target attribution area, wherein the differentiated feature information is used for making a recommendation to a target user belonging to the target attribution area. The technical solution provided in the embodiments of the present disclosure realizes the accurate identification of differentiated requirements of users in different attribution areas.

Description

Identification method, recommendation method, display method and computing device

[0001] This disclosure claims priority to the Chinese patent application filed with the China Patent Office on March 28, 2024, with application number 202410371646.4 and application name “Identification Method, Recommendation Method, Display Method and Computing Device”, the entire contents of which are incorporated by reference into this disclosure. Technical Field

[0002] The embodiments of the present disclosure relate to the field of computer technology, and in particular to an identification method, a recommendation method, a display method, and a computing device. Background Art

[0003] For some online systems that provide objects for users to perform interactive behaviors, such as e-commerce platforms that provide goods for users to purchase, users from different regions have differentiated needs and different preferred characteristics. However, currently, there is no identification method that can accurately identify the differentiated needs of users from different regions. Summary of the Invention

[0004] The embodiments of the present disclosure provide an identification method, a recommendation method, a display method, and a computing device to solve the technical problem in the prior art that the differentiated needs of users in different attribution areas cannot be identified.

[0005] In a first aspect, an embodiment of the present disclosure provides an identification method, including:

[0006] Get multiple historical search information;

[0007] identifying object categories associated with the plurality of historical search information;

[0008] extracting object feature information related to the associated object category from the plurality of historical search information;

[0009] Determining the importance of the object feature information relative to the target attribution area;

[0010] The object feature information whose importance satisfies the differentiation condition is used as the differentiated feature information of the target attribution area.

[0011] In a second aspect, an embodiment of the present disclosure provides a recommendation method, including:

[0012] Obtain target search information input by the target user;

[0013] Determining a target object category associated with the target search information;

[0014] Determining at least one differentiated feature information of the target object category corresponding to the target attribution area of ​​the target user; the at least one differentiated feature information is object feature information whose importance relative to the target attribution area satisfies a differentiated condition; the object feature information is extracted from a plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information;

[0015] The at least one differentiated feature information is recommended to the target user.

[0016] In a third aspect, an embodiment of the present disclosure provides a recommendation method, including:

[0017] Determining at least one object that matches the target user;

[0018] determining at least one object category corresponding to the at least one object;

[0019] Determining at least one differentiated feature information of any object category corresponding to the target attribution area of ​​the target user; the at least one differentiated feature information is object feature information whose importance relative to the target attribution area satisfies a differentiated condition; the object feature information is extracted from a plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information;

[0020] generating object prompt information corresponding to the at least one object in combination with the at least one differential feature information;

[0021] The object prompt information corresponding to the at least one object is sent to the user terminal, so that the object prompt information corresponding to the at least one object is displayed on the object recommendation page.

[0022] In a fourth aspect, an embodiment of the present disclosure provides a display method, including:

[0023] Display the search input page;

[0024] Detecting target search information input by a target user on the search input page;

[0025] Obtaining at least one differentiated feature information of a target object category corresponding to a target attribution area of ​​the target user; the target object category is associated with the target search information; the at least one differentiated feature information is object feature information whose importance relative to the target attribution area satisfies a differentiated condition; the object feature information is extracted from a plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information;

[0026] Displaying the at least one differentiated feature information in the search input page;

[0027] In response to a selection operation on any one of the differentiated feature information, a search request is generated based on the target search information and the selected differentiated feature information.

[0028] In a fifth aspect, an embodiment of the present disclosure provides a display method, including:

[0029] Display object recommendation page;

[0030] Displaying object prompt information corresponding to at least one object matching the target user on the object recommendation page; the object prompt information corresponding to the at least one object is generated in combination with at least one differentiated feature information; the at least one differentiated feature information is object feature information that satisfies a differentiated condition in terms of importance relative to the target user's target region; the object feature information is extracted from multiple historical search information and is related to object categories respectively associated with the multiple historical search information;

[0031] For a target object having any differential feature information, the differential feature information is displayed in the object prompt information of the target object.

[0032] In the sixth aspect, an embodiment of the present disclosure provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the identification method described in the first aspect above or the recommendation method described in the second aspect above or the recommendation method described in the third aspect above or the display method described in the fourth aspect above or the display method described in the fifth aspect above.

[0033] In the seventh aspect, a computer storage medium is provided in an embodiment of the present disclosure, storing a computer program. When the computer program is executed by a computer, it implements the identification method described in the first aspect above, the recommendation method described in the second aspect above, the recommendation method described in the third aspect above, the display method described in the fourth aspect above, or the display method described in the fifth aspect above.

[0034] In the eighth aspect, a computer program product is provided in an embodiment of the present disclosure, including a computer program / instruction. When the computer program / instruction is executed by a computer, it implements the identification method described in the first aspect above, the recommendation method described in the second aspect above, the recommendation method described in the third aspect above, the display method described in the fourth aspect above, or the display method described in the fifth aspect above.

[0035] The disclosed embodiment obtains multiple historical search information, identifies the object categories associated with each of the multiple historical search information, extracts object feature information related to the associated object categories from the multiple historical search information, determines the importance of the object feature information relative to the target attribution area, and uses the object feature information whose importance meets the differentiation conditions as the differentiated feature information of the target attribution area, wherein the differentiated feature information is used to recommend to target users belonging to the target attribution area. By analyzing the historical search information, the differentiated feature information of different attribution areas is determined, and the differentiated needs of users in different attribution areas are accurately identified. The differentiated feature information can be used in recommendation scenarios to achieve personalized recommendations for users in different attribution areas, thereby improving the user experience.

[0036] These and other aspects of the present disclosure will become more apparent from the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0038] FIG1 shows a schematic structural diagram of an embodiment of a processing system provided by the present disclosure;

[0039] FIG2 shows a flow chart of an embodiment of an identification method provided by the present disclosure;

[0040] FIG3 shows a flow chart of an embodiment of a recommendation method provided by the present disclosure;

[0041] FIG4 shows a flow chart of an embodiment of a recommendation method provided by the present disclosure;

[0042] FIG5 shows a flow chart of an embodiment of a display method provided by the present disclosure;

[0043] FIG6 shows a flow chart of an embodiment of a display method provided by the present disclosure;

[0044] FIG7 shows a schematic diagram of scene interaction in a practical application of an embodiment of the present disclosure;

[0045] FIG8 shows a schematic diagram of an interface display of a user terminal in an actual application of an embodiment of the present disclosure;

[0046] 9a and 9b are schematic diagrams showing interface displays of a user terminal in another practical application of an embodiment of the present disclosure;

[0047] FIG10 shows a schematic structural diagram of an embodiment of an identification device provided by the present disclosure;

[0048] FIG11 shows a schematic structural diagram of an embodiment of a recommendation device provided by the present disclosure;

[0049] FIG12 shows a schematic structural diagram of an embodiment of a recommendation device provided by the present disclosure;

[0050] FIG13 shows a schematic structural diagram of an embodiment of a display device provided by the present disclosure;

[0051] FIG14 shows a schematic structural diagram of an embodiment of a display device provided by the present disclosure;

[0052] FIG15 shows a schematic structural diagram of an embodiment of a computing device provided by the present disclosure. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present disclosure.

[0054] In some of the processes described in the specification and claims of the present disclosure and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit "first" and "second" to be different types.

[0055] The technical solutions of the embodiments of the present disclosure can be applied to electronic transaction scenarios that provide objects for user interaction, such as e-commerce scenarios where objects are traded over a network. In e-commerce scenarios, objects referred to herein are also referred to as commodities. In such electronic transaction scenarios, users can exchange objects for virtual resources such as currency.

[0056] Since the technical solution of the embodiment of the present disclosure is applicable to a network virtual environment, the users described generally refer to "virtual users". Real users can register a user account in the server through registration to obtain a user identity in the network environment and interact with objects. These users usually correspond to multiple attribution areas according to their geographical locations. In particular, with the development of international e-commerce, these users belong to different countries. The online systems that currently provide object interaction usually treat users from different attribution areas indiscriminately, resulting in poor user experience and affecting the stickiness of the use of the online system. Therefore, in actual applications, there are differentiated needs for identifying users from different attribution areas and determining the preferred features of users from different attribution areas. However, at present, there is no identification method that can accurately identify users.

[0057] In order to accurately identify the differentiated needs of users in different attribution areas, the inventors proposed the technical solution of the present disclosure after a series of studies. In the embodiment of the present disclosure, multiple historical search information is obtained, the object categories associated with the multiple historical search information are identified, and object feature information related to the associated object categories is extracted from the multiple historical search information. The importance of the object feature information related to the target object category relative to the target attribution area is determined, and the object feature information whose importance meets the differentiation conditions is used as the differentiated feature information of the target attribution area, and the differentiated feature information is recommended to the target users belonging to the target attribution area. The embodiment of the present disclosure determines the differentiated feature information of different attribution areas by analyzing the historical search information, and accurately identifies the differentiated needs of users in different attribution areas. The differentiated feature information can be used in recommendation scenarios to achieve personalized recommendations for users in different attribution areas, thereby improving the user experience.

[0058] The following will be combined with the accompanying drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the embodiments described are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present disclosure.

[0059] It should be noted that the embodiments of the present disclosure may involve the use of user data. In actual applications, user-specific personal data may be used in the scenarios described herein within the scope permitted by applicable laws and regulations, subject to compliance with applicable laws and regulations of the country where the user is located (for example, with the user's explicit consent, effective notification to the user, etc.).

[0060] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0061] The technical solution of the embodiment of the present disclosure can be applied to the processing system shown in Figure 1. In actual applications, the processing system can be an online system that provides objects for users to perform interactive behaviors. The processing system can include a user end 101 and a server end 102.

[0062] The user terminal 101 and the server terminal 102 may be connected via a network. The network provides a medium for a communication link between the user terminal 101 and the server terminal 102. The network may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0063] The user terminal 101 may be a browser, an APP (Application), or a web application such as an H5 (HyperText Markup Language 5, version 5) application, a light application (also known as a mini-program, a lightweight application), or a cloud application. The user terminal 101 may be deployed in an electronic device and may rely on the device to run or on certain apps in the device to run. The electronic device may, for example, have a display screen and support information browsing, such as a personal mobile terminal such as a mobile phone, tablet computer, personal computer, or other client terminal. For ease of understanding, FIG1 mainly uses a device image to represent the user terminal. Various other types of applications may also be configured in the electronic device, such as search applications, instant messaging applications, and the like.

[0064] The server 102 may include one or more servers that provide various services. For example, it may be implemented as a distributed server cluster consisting of multiple servers, or as a single server. In addition, it may be a server for a distributed system, or a server combined with a blockchain. It may also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, or it may be an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.

[0065] The user can interact with the server 102 through the user terminal 101 to receive or send messages, etc. In the application scenario of the embodiment of the present disclosure, for example, the server 102 can obtain the target attribution area of ​​the target user through the user terminal 101, determine the differentiated feature information of the target attribution area, and send the differentiated feature information to the user terminal 101 for the user terminal 101 to display the differentiated feature information, etc.

[0066] The implementation details of the technical solution of the embodiment of the present disclosure are described in detail below.

[0067] FIG2 is a flow chart of an embodiment of an identification method provided by the present disclosure. The technical solution of this embodiment can be executed by a server. The method may include the following steps:

[0068] 201: Get multiple historical search information.

[0069] The historical search information may be historical search information entered by a user on an online system that provides objects for users to interact with. The online system can retrieve objects based on the search information entered by the user. For example, historical search information entered by a user on an e-commerce platform that provides products for users to purchase can be retrieved based on the search information entered by the user. The historical search information may consist of search keywords.

[0070] A plurality of historical search information within a specified time period may be obtained at predetermined intervals. The specified time period may be, for example, within one year.

[0071] 202: Identify object categories associated with multiple pieces of historical search information.

[0072] The object category associated with each historical search information can be identified using a recognition model, which can be a machine learning model trained using sample information and corresponding sample object categories.

[0073] When the object is a product, the object category is the product category.

[0074] 203: Extracting object feature information related to the associated object category from the plurality of historical search information.

[0075] The object feature information is related to the object category associated with the historical search information and can describe the object features under the object category. The object features can include, for example, object attributes.

[0076] 204: Determine the importance of the object feature information relative to the target attribution area.

[0077] The target attribution zone may be any attribution zone. The attribution zone may be determined, for example, based on the user's Internet Protocol (IP) address, the GPS (Global Positioning System) location of the user's smart device, or user-configured zone information. In practical applications, the attribution zone may be, for example, a country or a city.

[0078] 205: Using the object feature information whose importance satisfies the differentiation condition as the differentiated feature information of the target belonging area.

[0079] The differentiation condition may be, for example, that the importance is greater than a specified value or that the importance is the highest, etc. Of course, other methods may also be used to implement it, which will be described in detail in the corresponding embodiments below.

[0080] The differential feature information may be used to make recommendations to target users belonging to the target attribution area. The target user may be any user belonging to the target attribution area.

[0081] In this embodiment, multiple historical search information is obtained, the object categories associated with each of the multiple historical search information are identified, and object feature information related to the associated object categories is extracted from the multiple historical search information. The importance of the object feature information relative to the target attribution area is determined, and feature information whose importance meets differentiation conditions is used as differentiated feature information for the target attribution area. This differentiated feature information can be used to make recommendations to target users belonging to the target attribution area. By analyzing the historical search information, differentiated feature information for different attribution areas is determined, enabling accurate identification of the differentiated needs of users in different attribution areas. This differentiated feature information can be used in recommendation scenarios to achieve personalized recommendations for users in different attribution areas, thereby improving the user experience.

[0082] The differentiated feature information of different attribution areas can be used to make differentiated recommendations to users in different attribution areas, meet the differentiated needs of users in different attribution areas, and improve user experience.

[0083] In some embodiments, the accuracy and efficiency of large models can be leveraged to improve the accuracy and efficiency of extracting object feature information from multiple historical search information. Therefore, extracting object feature information related to the associated object category from multiple historical search information can include: segmenting the multiple historical search information to obtain multiple candidate elements; and using the large model to determine, from the multiple candidate elements, a target candidate element that describes the object attributes corresponding to the associated object category, and using the target candidate element as the object feature information.

[0084] Therefore, the target candidate element may describe the object attribute corresponding to the object category associated with the historical search information and be related to the associated object category.

[0085] Among them, the large model can refer to a machine learning model with a large number of parameters and complex structure, which can process massive data and complete various complex tasks, such as natural language processing, computer vision, speech recognition, etc. It is an AI (Artificial Intelligence, artificial intelligence) model. The large model can be implemented using a large language model (English: Large Language Model, abbreviated as: LLM) or a multimodal large model (English: Multimodal Large Model, abbreviated as: MLM), etc. For example, GPT-3 (Generative Pre-Trained Transformer-3, a third generation generative pre-training model), GPT-4 (Generative Pre-Trained Transforme-4, a fourth generation generative pre-training model), BERT (Bidirectional Encoder Representation from Transformers, a bidirectional encoder model based on Transformers), Turing NLG (Turing Natural language Generation, Turing Natural Language Generation), etc. This disclosure is not limited to this.

[0086] Prompt instructions can be input to the large model to instruct the large model to determine whether the candidate element can describe the object attributes corresponding to the associated object category. Among them, the prompt instruction is a form of input used to prompt or guide the large model to give expected output. It is used to instruct the large model what actions should be taken or what output should be generated when performing a specific task. The prompt information instruction is a natural language input, similar to a command or instruction, to let the large language model know what it needs to do. It can be a question, a sentence or a paragraph, which is used to provide the large model with the context and parameter information of the input information, prompting or guiding the large model to give expected output.

[0087] The associated object category refers to the object category associated with the historical search information corresponding to the candidate element. If so, the candidate element can be determined to be the target candidate element. The prompt instruction can be composed of a prompt template and a candidate element.

[0088] Taking the e-commerce scenario as an example, the prompt template may be, for example, "Give you a batch of data, and judge from the semantics and industry experience whether the keywords in it can be associated with the relevant attributes of the products under the industry category cate_desc, and explain the reasons. If so, the result is set to yes, otherwise it is set to no. The relevant attributes of the products include but are not limited to: product word product, accessory word accessory, audience word audience, brand word brand, color word color, function word function, material word material, time word time, usage word usage, modifier word modify. The data to be judged is: XXXXXX". Multiple candidate elements can be spliced ​​at the end "XXXXXX" of the prompt template to generate a prompt instruction.

[0089] Among them, the keyword in the prompt template refers to the candidate element, and the industry category cate_desc refers to the object category associated with the historical search information of the candidate element. If the result output by the large model is yes, the candidate element is the target candidate element.

[0090] For example, if the object is a commodity, then the object attributes are the commodity attributes, which may include the brand, color, function, model, size, material, style, etc. For example, there are multiple historical search information such as "pure cotton sheets" and "elegant long skirts". The object category associated with "pure cotton sheets" is bedding, which can be divided into two candidate elements: "pure cotton" and "sheets". The large model can determine that "pure cotton", which describes the material attributes of the commodity, is the target candidate element; the object category associated with "elegant long skirt" is skirt, which can be divided into three candidate elements: "elegant", "long", and "skirt". The large model can determine that "elegant", which describes the style attributes of the commodity, and "long", which describes the model attributes of the commodity, are the target candidate elements.

[0091] The segmentation model can be used to segment the multiple historical search information separately, or segment the multiple historical search information separately using word segmentation rules such as special expressions to obtain multiple candidate elements. The segmentation model can be a machine learning model trained using sample text and corresponding sample segmentation results.

[0092] In some embodiments, segmented elements with low search frequencies can be filtered out to improve processing efficiency and accuracy. Therefore, segmenting the plurality of historical search information to obtain a plurality of candidate elements can include: segmenting the plurality of historical search information to obtain a plurality of segmented elements; and filtering, from the plurality of segmented elements, a plurality of candidate elements whose search frequencies exceed a specified number.

[0093] In some embodiments, using a large model to determine a target candidate element used to describe the object attributes corresponding to the associated object category from multiple candidate elements may include: using a large model to determine multiple candidate elements used to describe the object attributes corresponding to the associated object category from multiple candidate elements; calculating the differentiated score of the candidate element relative to its associated object category based on the first occurrence frequency of any candidate element in its associated object category, and / or the first inverse frequency of the object category including the candidate element; and selecting a specified number of target candidate elements in descending order of the differentiated scores corresponding to the multiple candidate elements.

[0094] Among them, the first occurrence frequency of any candidate element in its associated object category can be calculated based on the number of the candidate element in its associated object category and the total number of object feature information associated with the associated object category, and the first inverse frequency of the object category of the candidate element can be calculated based on the total number of object categories associated with multiple historical search information and the number of object categories associated with the candidate element.

[0095] The higher the first occurrence frequency of a candidate element in its associated object category, the more important the candidate element is to the document. However, if the first inverse frequency is low, that is, the candidate element appears frequently in all object categories, then the candidate element may be a common element and has low discrimination ability for the object category.

[0096] The TF-IDF (Term Frequency-Inverse Document Frequency) algorithm can be used to calculate the differential score of the candidate element relative to its associated object category. For example, the following formula can be used:

[0097] The differentiation score of the candidate element relative to its associated object category = (the number of candidate elements in its associated object category / the total number of object feature information associated with the associated object category) * log (the total number of object categories associated with multiple historical search information / the number of object categories associated with the candidate element + 1).

[0098] The first occurrence frequency is the ratio of the number of candidate elements in their associated object categories to the total number of object feature information associated with the associated object categories. The first inverse frequency is log(total number of object categories associated with multiple historical search information / number of object categories associated with the candidate element + 1).

[0099] According to the above formula, the higher the differentiation score of the candidate element relative to its associated object category, the more important it is to the associated object category, and the higher its discrimination for the object category.

[0100] In some embodiments, the above-mentioned large model can be obtained by pre-training, and the method may also include: using the first sample data to train the large model; determining the second sample data; the second sample data includes multiple sample search information and object categories associated with the multiple sample search information; using the large model to identify multiple sample object feature information related to the associated object category from the multiple sample search information; calculating the differentiated score of the sample object feature information relative to its associated object category based on the second occurrence frequency of any sample object feature information in its associated object category, and / or the second inverse frequency in the object category including the sample object feature information; dividing the multiple sample object feature information into multiple groups in order of the differentiated scores from large to small; finding the top N groups that make the model performance meet the performance requirements, and determining the specified value based on the number of sample object feature information in the top N groups, where N is a positive integer.

[0101] The second sample data is used as a test sample to evaluate the large model.

[0102] Among them, model performance can be evaluated by precision and / or recall.

[0103] In some embodiments, finding the top N groups that make the model performance meet the performance requirements, and determining the specified value based on the number of sample object feature information in the top N groups includes: sampling and evaluating whether the sample object feature information in multiple groups is related to the object category; based on the evaluation results, determining the number of true positive examples corresponding to the multiple groups respectively; based on the number of true positive examples, calculating the accuracy and / or recall rate of the model, and finding the specified value of the accuracy and / or recall rate that meets the performance requirements.

[0104] Among them, sampling evaluation can adopt manual evaluation methods to determine whether the sample object feature information is relevant to the object category. If it is relevant and consistent with the recognition result of the large model, the sample object feature information is a true positive example. If it is not relevant and inconsistent with the recognition result of the large model, the sample object feature information is a false positive example. The number of true positive examples can refer to the number of samples where the large model correctly predicts positive examples as positive examples, that is, the number of sample object feature information in any group whose evaluation results are relevant to the object category. The number of false positive examples can refer to the number of samples where the model incorrectly predicts negative examples as positive examples, that is, the number of sample object feature information in any group whose evaluation results are irrelevant to the object category.

[0105] An optional method is to calculate the number of true positives / (number of true positives + number of false positives) as the recall rate, so as to find a recall rate greater than a specified value or a specified value with the highest recall rate based on the recall rate.

[0106] Alternatively, the accuracy of the large model can be considered 100% (percentage sign) or 100% plus a specified error value. Recall can be calculated as: number of true positives / (number of true positives + number of false positives). Therefore, the precision-recall ratio can be calculated as: (2 * precision * recall) / (precision + recall), to find the precision-recall ratio that exceeds the specified value or the highest precision-recall ratio.

[0107] As another optional method, the group number can be used as the horizontal axis and the number of true positive examples as the vertical axis to fit a linear equation; based on the parameters of the linear equation, the accuracy and / or recall of the large model can be calculated; and the specified values ​​of the accuracy and / or recall that meet the performance requirements can be found.

[0108] Since large models can be trained iteratively, the precision and recall can be calculated based on the parameters of a linear equation as follows:

[0109] Among them, it is assumed that the large model is iteratively trained every day, and the feature information of multiple sample objects is divided into 100 groups; k0 and b0 are the parameters of the linear equation corresponding to the initial large model; k and b are the parameters of the linear equation corresponding to the large model iteratively generated every day; x represents the group number, and a represents the weight coefficient, which can be set according to actual conditions.

[0110] Therefore, we can look for x0 that makes the precision-recall ratio greater than the specified value or has the highest precision-recall ratio, and this x0 can be used as the specified value.

[0111] In some embodiments, there are multiple ways to determine the importance of object feature information relative to the target attribution area. As an optional implementation method, determining the importance of object feature information relative to the target attribution area may include: determining the importance of object feature information relative to the target attribution area based on at least one historical interaction behavior performed by the user on the object hit by the object feature information related to the target object category.

[0112] Among them, the importance of the object feature information relative to the target attribution area can be determined in combination with at least one historical interaction behavior performed by users in the target attribution area on objects hit by object feature information related to the target object category, which may include: determining the importance of the object feature information relative to the target attribution area by combining at least one historical interaction behavior performed by users in the target attribution area on objects corresponding to the object feature information related to the target object category, at least one historical interaction behavior performed by users in the target attribution area on objects corresponding to all feature information related to the target object category, at least one historical interaction behavior performed by users in all attribution areas on objects corresponding to the object feature information related to the target object category, and at least one historical interaction behavior performed by users in all attribution areas on objects corresponding to all feature information related to the target object category.

[0113] The target object category may be an object category associated with the object feature information, that is, may be an object category associated with the historical search information from which the object feature information is extracted.

[0114] The target object category can be associated with one or more object feature information.

[0115] The object hit by the object feature information related to the target object category may be an object hit by the object feature information in the historical search results obtained based on the historical search information related to the target object category.

[0116] For example, the object can be a commodity, the object feature information is "elegant," the target object category is dresses, and the feature information associated with the target object category includes "elegant," "business," "retro," "cotton and linen," and "chiffon." The object hit by the feature information associated with the target object category can be an object in the search results for a previous search for "dresses" that contains at least one of the feature information of "elegant," "business," "retro," "cotton and linen," or "chiffon." The object corresponding to the object feature information associated with the target object category can be an object in the search results for a previous search for "dresses" that contains the feature information of "elegant."

[0117] Optionally, determining the importance of the object feature information relative to the target attribution area in combination with at least one historical interaction behavior performed by the user on the object hit by the object feature information related to the target object category may include: calculating a first feature score of the object feature information relative to the target attribution area based on at least one historical interaction behavior performed by the user in the target attribution area on the object corresponding to the object feature information related to the target object category; calculating a second feature score of the object feature information relative to all attribution areas based on at least one historical interaction behavior performed by the users in all attribution areas on the object corresponding to the object feature information related to the target object category; calculating a third feature score of the target object category relative to the target attribution area based on at least one historical interaction behavior performed by the users in the target attribution area on the objects corresponding to all feature information related to the target object category; calculating a fourth feature score of the target object category relative to all attribution areas based on at least one historical interaction behavior performed by the users belonging to all attribution areas on the objects corresponding to all feature information related to the target object category; and determining the importance of the object feature information relative to the target attribution area in combination with at least one of the first feature score, the second feature score, the third feature score, and the fourth feature score.

[0118] Among them, calculating the first feature score of the object feature information relative to the target attribution area based on at least one historical interactive behavior performed by users in the target attribution area on the object corresponding to the object feature information related to the target object category may include: determining the number of first object views and at least one first object operation number based on the browsing behavior and at least one operation behavior performed by users in the target attribution area on the object corresponding to the object feature information related to the target object category; and calculating the first feature score based on the proportion of at least one object operation number relative to the number of first object views.

[0119] Calculating the second feature score of the object feature information relative to all attribution areas based on at least one historical interaction behavior performed by users in all attribution areas on an object corresponding to object feature information related to the target object category may include: determining the number of second object views and the number of at least one second object operation based on the browsing behavior and at least one operation behavior performed by users in all attribution areas on the object corresponding to the object feature information related to the target object category; and calculating the second feature score based on the proportion of the at least one second object operation number relative to the number of third object views.

[0120] Calculating the third feature score of the target object category relative to the target attribution area based on at least one historical interactive behavior performed by users in the target attribution area on objects corresponding to all feature information related to the target object category may include: determining the number of third object views and the number of at least one third object operation based on the browsing behavior and at least one operation behavior performed by users in the target attribution area on objects corresponding to all feature information related to the target object category; and calculating the third feature score based on the proportion of the at least one third object operation number relative to the number of third object views.

[0121] Calculating the fourth feature score of the target object category relative to all attribution areas based on at least one historical interaction behavior performed by users in all attribution areas on objects corresponding to all feature information related to the target object category may include: determining the number of fourth object views and at least one number of fourth object operations based on browsing behavior and at least one operation behavior performed by users in all attribution areas on objects corresponding to all feature information related to the target object category; and calculating the fourth feature score based on the proportion of the at least one fourth object operation number relative to the number of fourth object views.

[0122] The browsing behavior may be a user's click operation on an object, and the object browsing count may be the number of times the user clicks on the object. The at least one operation behavior may include one or more of a communication behavior, a transaction behavior, an add-to-cart behavior, and a favorite behavior. The at least one object operation count may include one or more of the number of times the user communicates with the object, the number of transactions, the add-to-cart behavior, and the number of favorites.

[0123] Furthermore, the object operation count can be updated by combining the weight coefficient corresponding to at least one operation behavior. For example, the object operation count can be simply summed up by adding up the operation counts corresponding to all operation behaviors, or by assigning different weights to the operation counts corresponding to various operation behaviors and then performing the weighted sum to obtain the object operation count.

[0124] For example, the first characteristic score, the second characteristic score, the third characteristic score, and the fourth characteristic score can be calculated according to the following formulas:

[0125] First characteristic score = [log 10 (number of views of the first object)]*(number of communications with the first object + 3*number of transactions with the first object) / number of views of the first object;

[0126] Second characteristic score = [log 10 (number of views of the second object)]*(number of communications with the second object + 3*number of transactions with the second object) / number of views of the second object;

[0127] The third characteristic score = [log10 (number of third-party visits)*(number of third-party communications + 3*number of third-party transactions) / number of third-party visits;

[0128] Fourth characteristic score = [log 10 (number of views of the fourth object)]*(number of communications with the fourth object + 3*number of transactions with the fourth object) / number of views of the fourth object;

[0129] There are multiple implementation methods for determining the importance of the object feature information relative to the target attribution area by combining at least one of the first feature score, the second feature score, the third feature score, and the fourth feature score. One optional implementation method may be to calculate a fifth feature score of the object feature information relative to the remaining attribution areas other than the target attribution area based on the first feature score and the second feature score; calculate a sixth feature score of the target object category relative to the remaining attribution areas other than the target attribution area based on the third feature score and the fourth feature score; and use the relative ratio of a first ratio of the first feature score to the fifth feature score to a second ratio of the third feature score to the sixth feature score as the importance of the object feature information relative to the target attribution area.

[0130] The fifth feature score may be obtained by subtracting the first feature score from the second feature score, and the sixth feature score may be obtained by subtracting the third feature score from the fourth feature score.

[0131] The higher the relative ratio of the first ratio to the second ratio, the more important the object feature information is relative to the target attribution area. The differentiation condition can be set to a predetermined importance value. For example, the predetermined value can be set to 1. If the importance reaches 1, the object feature information can be used as differentiated feature information for the target attribution area.

[0132] Another optional implementation may be to calculate, for any object feature information, the importance of the object feature information relative to the target attribution area based on the ratio of the first feature score to the third feature score and the second inverse frequency.

[0133] For example, the TF-IDF algorithm can be used to calculate the importance of the object feature information relative to the target area, for example, according to the following formula:

[0134] The importance of the object feature information relative to the target attribution zone=(first feature score / third feature score)*log(total number of attribution zones / number of attribution zones including object feature information+1).

[0135] As another optional implementation method, determining the importance of object feature information relative to the target attribution area may include: for any object feature information, determining the third occurrence frequency of the object feature information based on the number of occurrences of the object feature information in the target attribution area and the number of occurrences of all object feature information in the target attribution area; calculating the third inverse frequency corresponding to the object feature information based on the number of attribution areas including the object feature information and the number of all attribution areas; and determining the importance of the object feature information relative to the target attribution area based on the third occurrence frequency and the third inverse frequency.

[0136] For example, the TF-IDF algorithm can be used to calculate the importance of the object feature information relative to the target area, for example, according to the following formula:

[0137] Importance of object feature information relative to target attribution area = (Number of occurrences of object feature information in target attribution area / Number of occurrences of all feature information in target attribution area) * log (Total number of attribution areas / Number of attribution areas including object feature information + 1)

[0138] FIG3 is a flowchart of an embodiment of a recommendation method provided by the present disclosure. The technical solution of this embodiment can be executed by the server. The method may include the following steps:

[0139] 301: Obtain the target search information entered by the target user.

[0140] 302: Determine the target object category associated with the target search information.

[0141] Among them, the recognition model can be used to identify the target object category associated with the target search information.

[0142] 303: Determine at least one differentiated feature information of the target user's target attribution area corresponding to the target object category.

[0143] Among them, at least one differentiated feature information is object feature information whose importance relative to the target attribution area of ​​the target user meets the differentiated condition; the object feature information is extracted from multiple historical search information and is related to the object categories respectively associated with the multiple historical search information.

[0144] Among them, multiple historical search information can be segmented and processed separately to obtain multiple candidate elements; the large model is used to determine the target candidate element used to describe the object attributes corresponding to the associated object category from the multiple candidate elements, and the target candidate element is used as the object feature information.

[0145] The importance of the object feature information relative to the target attribution area may be determined in combination with at least one historical interaction behavior performed by the user on the object hit by the object feature information related to the target object category.

[0146] The method for determining the importance of the object feature information relative to the target attribution area can be seen in the embodiment shown in Figure 2 above, and will not be repeated here.

[0147] 304: Recommend at least one differentiated feature information to the target user.

[0148] In this embodiment, target search information input by a target user is obtained, a target object category associated with the target search information is determined, and at least one differentiated feature information corresponding to the target object category of the target user's target attribution region is determined. The at least one differentiated feature information is object feature information whose importance relative to the target user's target attribution region satisfies a differentiated condition. The object feature information is extracted from multiple historical search information and is associated with the object categories associated with each of the multiple historical search information. Furthermore, the at least one differentiated feature information is recommended to the target user. Based on the target object category associated with the target user's search information and the target attribution region of the target user, at least one differentiated feature information can be identified. When users in different attribution regions search for the same object category, differentiated feature information corresponding to the identified object categories in the different attribution regions can be used to make differentiated recommendations to users in the different attribution regions, thereby meeting the differentiated needs of target users in the different attribution regions. By analyzing the historical search information to determine differentiated feature information for different attribution regions, the differentiated needs of users in the different attribution regions can be accurately identified. When users in different attribution regions search for the same object category, the differentiated feature information for the determined object categories in the different attribution regions can be used to make personalized recommendations to users in the different attribution regions, thereby improving the user experience.

[0149] In some embodiments, the method may further include determining an arrangement order of the at least one differentiated feature information based on the importance of the at least one differentiated feature information to the target attribution area. Accordingly, recommending the at least one differentiated feature information to the target user may include recommending the at least one differentiated feature information to the target user according to the arrangement order.

[0150] For example, the more important the differentiated feature information is, the higher the ranking will be.

[0151] In some embodiments, the differentiated feature information may be used by a user to select and search for objects having the differentiated feature information. Recommending the at least one differentiated feature information to a target user may include: using the at least one differentiated feature information as associated information of the target search information, so that the user terminal displays the at least one differentiated feature information on a search input page for the target search information.

[0152] The user terminal may display at least one differentiated feature information in the search input page of the target search information according to the arrangement order of the at least one differentiated feature information.

[0153] In this case, in response to the target user's selection operation on any differentiated feature information, the object corresponding to the differentiated feature information can be displayed.

[0154] For example, the target search information input by the target user is "dress", the target object category associated with the target search information is dress, the target region of the target user is China, and the corresponding at least one differentiated feature information includes "French", "retro" and "all-cotton", and the importance of at least one differentiated feature information is "French" > "retro" > "all-cotton", then "French", "retro" and "all-cotton" will be displayed in order, and French-style dresses can be displayed in response to the target user's selection operation for "French".

[0155] In some embodiments, in order to allow the target user to quickly notice the object with differential feature information, the differential feature information can be used to generate object prompt information. The method can also include: for any target object with differential feature information, generating object prompt information containing the differential feature information.

[0156] In some embodiments, objects recalled by the target search information may be first determined, and object prompt information may be generated based on the differentiated feature information. Recommending at least one differentiated feature information to the target user may include: determining at least one object recalled by the target search information; generating object prompt information corresponding to the at least one object based on the at least one differentiated feature information; transmitting the object prompt information corresponding to the at least one object to the user terminal, and displaying the object prompt information corresponding to the at least one object on the search results page.

[0157] Furthermore, in some embodiments, after first determining the objects recalled by the target search information, the generated object prompt information can be sorted based on the differentiated feature information possessed by the objects. Recommending at least one differentiated feature information to the target user can include: determining at least one object recalled by the target search information; and determining a sorting order for the at least one object based on whether the at least one differentiated feature information is present. The user terminal can then display the object prompt information corresponding to the at least one object on the search results page according to the sorting order.

[0158] For example, the higher the importance of the differentiated feature information of an object, the higher the object is arranged. For example, if the target user's target region corresponds to at least one differentiated feature information of the target object category including "French", "Vintage", and "Cotton", and the importance of at least one differentiated feature information is "French" > "Vintage" > "Cotton", then the object prompt information of the object with "French" feature, the object prompt information of the object with "Vintage" feature, and the object prompt information of the object with "Cotton" feature can be displayed in order, and finally the object prompt information of the object without any differentiated feature information can be displayed.

[0159] In addition, each differentiated feature information may be scored, and the total score of the differentiated feature information of at least one object recalled by the target search information may be calculated. The higher the total score of the object, the higher the ranking order.

[0160] FIG4 is a flowchart of an embodiment of a recommendation method provided by the present disclosure. The technical solution of this embodiment can be executed by the server. The method may include the following steps:

[0161] 401: Determine at least one object matching the target user.

[0162] The at least one object matched with the target user can be a user preference object determined by analyzing the user's interests and consumption habits based on user behavior data, such as the user's browsing history, search history, shopping cart additions, purchase history, etc. It can also be a user preference object determined based on user characteristics, such as the user's region of origin, gender, age, occupation, income level, family status, etc.

[0163] 402: Determine at least one object category corresponding to at least one object.

[0164] 403: Determine at least one differentiated feature information corresponding to any object category of the target attribution area of ​​the target user.

[0165] The at least one differentiated feature information is object feature information whose importance relative to the target attribution area of ​​the target user satisfies a differentiated condition; the object feature information is extracted from multiple historical search information and is related to object categories respectively associated with the multiple historical search information.

[0166] Among them, multiple historical search information can be segmented and processed separately to obtain multiple candidate elements; the large model is used to determine the target candidate element used to describe the object attributes corresponding to the associated object category from the multiple candidate elements, and the target candidate element is used as the object feature information.

[0167] Among them, multiple historical search information can be segmented and processed separately to obtain multiple candidate elements; the large model is used to determine the target candidate element used to describe the object attributes corresponding to the associated object category from the multiple candidate elements, and the target candidate element is used as the object feature information.

[0168] That is, object feature information can be used to describe object properties.

[0169] The importance of the object feature information relative to the target attribution area may be determined in combination with at least one historical interaction behavior performed by the user on the object hit by the object feature information related to the target object category.

[0170] The method for determining the importance of the object feature information relative to the target attribution area can be seen in the embodiment shown in Figure 2 above, and will not be repeated here.

[0171] 404: Generate object prompt information by combining at least one differential feature information corresponding to any object.

[0172] 405: Send object prompt information corresponding to the at least one object to the user terminal, so as to display the object prompt information corresponding to the at least one object on the object recommendation page.

[0173] In this embodiment, at least one object matching the target user is determined, at least one object category corresponding to the at least one object is determined, and at least one differentiated feature information corresponding to any one of the object categories for the target user's target attribution area is determined. The at least one differentiated feature information is object feature information whose importance relative to the target user's target attribution area satisfies a differentiated condition. The object feature information is extracted from multiple historical search information and is associated with the object categories associated with each of the multiple historical search information. Furthermore, object prompt information is generated based on the at least one differentiated feature information corresponding to any one of the objects, and the object prompt information corresponding to the at least one object is displayed on the object recommendation page. When users in different attribution areas match the same object, differentiated object prompt information can be generated using the differentiated feature information determined for the different attribution areas, thereby recommending differentiated object prompt information to users in different attribution areas, achieving personalized recommendations for users in different attribution areas and improving the user experience.

[0174] In some embodiments, the method may further include: determining an arrangement order of the at least one object based on whether the at least one object has at least one differentiated feature information, so that the user terminal displays object prompt information corresponding to the at least one object on the object result page according to the arrangement order.

[0175] Among them, objects with differentiated feature information are ranked higher than objects without differentiated feature information. In addition, each differentiated feature information can be scored, and the total score of the differentiated feature information possessed by at least one object is calculated. The higher the total score, the higher the object is ranked.

[0176] In some embodiments, the method may further include: determining object attribute information corresponding to at least one differential feature information; and if the target object has multiple differential feature information, determining an arrangement order for the multiple differential feature information based on the importance of the multiple differential feature information relative to the target region. The user terminal may then display the multiple differential feature information and the object attribute information corresponding to the multiple differential feature information in the object prompt information of the target object according to the arrangement order.

[0177] For example, if the target object is a coffee cup, its differentiated feature information is "glass" and "straw type". The object attribute information corresponding to "glass" is "material", and the object attribute information corresponding to "straw type" is "style". If the importance of "glass" to the target belonging area is greater than that of "straw type", the object prompt information will first display "Material: Glass" and then "Style: Straw Type".

[0178] FIG5 is a flowchart of an embodiment of a display method provided by the present disclosure. The technical solution of this embodiment can be executed by a user terminal. The method may include the following steps:

[0179] 501: Serving search input page.

[0180] 502: Detecting the target search information input by the target user on the search input page.

[0181] 503: Obtain at least one differentiated feature information of the target user's target attribution area corresponding to the target object category.

[0182] The target object category is associated with the target search information.

[0183] The at least one differentiated feature information is object feature information whose importance relative to the target attribution area of ​​the target user satisfies a differentiated condition; the object feature information is extracted from multiple historical search information and is related to object categories respectively associated with the multiple historical search information.

[0184] 504: Display at least one differentiated feature information on the search input page.

[0185] In some embodiments, displaying at least one differentiated feature information in the search input page includes: determining an arrangement order of at least one object feature information based on the importance of at least one object feature information and the target attribution area; and displaying at least one differentiated feature information in the search input page according to the arrangement order.

[0186] For example, the higher the importance of the differentiated feature information, the higher the ranking.

[0187] 505 : In response to a selection operation on any piece of differential feature information, a search request is generated based on the target search information and the selected differential feature information.

[0188] In addition, in response to the search request, objects corresponding to the target search information and the differentiated feature information may be displayed.

[0189] In this embodiment, a search input page is provided to detect the target search information input by the target user on the search input page; at least one differentiated feature information corresponding to the target object category of the target user's target attribution area is obtained, wherein the target object category is associated with the target search information, and the at least one differentiated feature information is object feature information whose importance relative to the target attribution area of ​​the target user satisfies a differentiated condition; the object feature information is extracted from multiple historical search information and is related to the object categories respectively associated with the multiple historical search information, and then, at least one differentiated feature information is displayed on the search input page, and in response to a selection operation for any differentiated feature information, a search request is generated based on the target search information and the selected differentiated feature information. The disclosed embodiment utilizes the differentiated feature information of different attribution areas to recommend differentiated object feature information to users in different attribution areas on the search input page when different users in different attribution areas search for the same object category, thereby achieving personalized recommendations for users in different attribution areas, meeting the differentiated needs of users in different attribution areas, and improving the user experience.

[0190] FIG6 is a flowchart of an embodiment of a display method provided by the present disclosure. The technical solution of this embodiment can be executed by a user terminal. The method may include the following steps:

[0191] 601: Provide an object recommendation page.

[0192] 602: Display object prompt information corresponding to at least one object matching the target user on the object recommendation page.

[0193] Among them, object prompt information corresponding to at least one object is generated in combination with at least one differentiated feature information; at least one differentiated feature information is object feature information whose importance relative to the target belonging area of ​​the target user meets the differentiated conditions; the object feature information is extracted from multiple historical search information and is related to the object categories respectively associated with the multiple historical search information.

[0194] The object prompt information corresponding to the at least one object may be displayed on the object recommendation page according to the arrangement order of the at least one object.

[0195] The arrangement order of at least one object can be determined according to the importance of the differentiated feature information of the object relative to the target attribution area. For example, the higher the importance of the differentiated feature information of the object relative to the target attribution area, the higher the arrangement order of the object.

[0196] 603: For a target object having any differential feature information, display the differential feature information in the object prompt information of the target object.

[0197] In this embodiment, an object recommendation page is provided; object prompt information of at least one object is displayed on the object recommendation page according to the arrangement order of at least one object; and for a target object having any differentiated feature information, differentiated feature information is displayed in the object prompt information of the target object, wherein the object prompt information corresponding to at least one object is generated in combination with at least one differentiated feature information; at least one differentiated feature information is object feature information whose importance relative to the target attribution area of ​​the target user meets differentiated conditions; the object feature information is extracted from multiple historical search information and is related to object categories respectively associated with the multiple historical search information. In this embodiment, when users in different attribution areas match the same object, differentiated object prompt information can be generated using the differentiated feature information determined for different attribution areas, thereby recommending differentiated object prompt information to users in different attribution areas, achieving personalized recommendations for users in different attribution areas, and improving user experience.

[0198] In some embodiments, the method may further include: determining object attribute information corresponding to at least one differentiated feature information; in the case where the target object has multiple differentiated feature information, determining the arrangement order of the multiple differentiated feature information according to the importance of the multiple differentiated feature information relative to the target belonging area; and displaying the multiple differentiated feature information and the object attribute information corresponding to the multiple differentiated feature information in the object prompt information of the target object according to the arrangement order.

[0199] In a practical application, the technical solution of the embodiment of the present disclosure can be applied to an e-commerce scenario. For ease of understanding, the technical solution of the embodiment of the present disclosure is introduced below using the e-commerce scenario as an example in combination with the scenario interaction diagram shown in FIG7 .

[0200] In the e-commerce scenario, the online system refers to the e-commerce platform that provides product purchases, so that users can easily understand the products and make purchases.

[0201] The server 701 may be a server in an e-commerce platform, or may be another service node independent of the e-commerce platform.

[0202] The user terminal 702 may provide a search input page, and the user terminal 702 may obtain the target search information input by the target user on the search input page, and send the target search information to the server terminal 701. The server terminal 701 may determine the target product category associated with the target search information, and determine at least one differentiated feature information corresponding to the target product category of the target user's target attribution area. Then, the server terminal 701 may send the at least one differentiated feature information to the user terminal 702. The user terminal 702 may display the at least one differentiated feature information on the search input page. The server terminal may extract product feature information related to the product categories respectively associated with the multiple historical search information from the multiple historical search information, determine the importance of the product feature information relative to the target attribution area, and use the product feature information whose importance meets the differentiated condition as the differentiated feature information of the target attribution area.

[0203] The server 701 may also determine the order of at least one differentiated feature based on the importance of the at least one differentiated feature to the target region. For example, the more important the differentiated feature, the higher the order. The user terminal 702 may display the at least one differentiated feature on the search input page according to the order. The user terminal 702 may also display the product corresponding to any differentiated feature in response to the target user's selection operation for that differentiated feature.

[0204] For ease of understanding, the search input page 800 is displayed in the interface diagram shown in Figure 8. For example, the target search information input by the target user is "dress", the target product category associated with the target search information is dress, the target user's target region is China, and the differentiated feature information determined by the server 701 includes "French", "retro" and "cotton". The importance of the differentiated feature information is "French" > "retro" > "cotton". The user terminal 702 displays "French", "retro" and "cotton" in order in the search input page 800. The user terminal 702 can also display French-style dresses in response to the target user's selection operation for "French", for example.

[0205] Furthermore, the server 701 may determine at least one recalled product based on the target search information and determine a ranking order for the at least one product based on whether the at least one product has at least one differentiating feature information. The client 702 may display product prompt information corresponding to each of the at least one product on the search results page according to the ranking order.

[0206] In addition, when the target product has multiple differentiated feature information, the server 701 can determine the arrangement order of the multiple differentiated feature information based on the importance of the multiple differentiated feature information relative to the target area, and the user terminal 702 can determine the product attribute information corresponding to the multiple differentiated feature information, and display the multiple differentiated feature information and the product attribute information corresponding to the multiple differentiated feature information in the product prompt information of the target product on the search result page.

[0207] The interface diagram shown in Figure 9a displays a search result page 900. For example, if the target search information is "coffee cup" and the target user's target region is the United States, the server can determine at least one coffee cup based on the target search information. The differentiated feature information determined by the server 701 based on the target region includes "glass" and "straw type". The product attribute information corresponding to "glass" is "material", and the product attribute information corresponding to "straw type" is "style". The importance of "glass" to the target region is greater than that of "straw type". The user terminal 702 first displays "Material: Glass" and then displays "Style: Straw Type" in the product prompt information on the search result page.

[0208] The interface diagram shown in Figure 9b displays a search result page 901. For example, if the target search information is "coffee cup" and the target user's target region is India, the server can determine at least one coffee cup based on the target search information. The differentiated feature information determined by the server 701 based on the target region includes "glass" and "straw type". The product attribute information corresponding to "glass" is "material", and the product attribute information corresponding to "straw type" is "style". The importance of "straw type" relative to the target region is greater than that of "glass". The user terminal 702 first displays "Style: Straw type" and then displays "Material: Glass" in the product prompt information on the search result page.

[0209] In this embodiment, the server determines the differentiated feature information of different attribution areas by analyzing historical search information, thereby accurately identifying the differentiated needs of users in different attribution areas. When users in different attribution areas search for the same product category, the server can use the differentiated feature information corresponding to the product category of the determined different attribution areas to make differentiated recommendations to users in different attribution areas. In addition, when users in different attribution areas match at least one product that is the same, the server can use the differentiated feature information of the determined different attribution areas to generate differentiated product prompt information, thereby recommending differentiated product prompt information to users in different attribution areas. The server can also display the product prompt information in different arrangement orders to meet the differentiated needs of users in different attribution areas and improve user experience.

[0210] FIG10 is a schematic diagram of the structure of an embodiment of an identification device provided by an embodiment of the present disclosure, wherein the method and device include:

[0211] The first acquisition module 1001 is used to acquire multiple historical search information;

[0212] Identification module 1002, used to identify object categories associated with multiple historical search information;

[0213] Extraction module 1003, used to extract object feature information related to the associated object category from multiple historical search information;

[0214] The first determining module 1004 is used to determine the importance of the object feature information relative to the target attribution area;

[0215] The second determining module 1005 is configured to use the object feature information whose importance satisfies the differentiation condition as the differentiated feature information of the target attribution area.

[0216] The differentiated feature information can be used to make recommendations to target users in the target attribution area.

[0217] In some embodiments, the extraction module extracts object feature information related to the associated object category from multiple historical search information, which may include: segmenting the multiple historical search information separately to obtain multiple candidate elements; using a large model to determine, from the multiple candidate elements, a target candidate element used to describe the object attributes corresponding to the associated object category, and using the target candidate element as the object feature information.

[0218] Among them, using a large model to determine a target candidate element for describing the object attributes corresponding to the associated object category from multiple candidate elements may include: using a large model to determine multiple candidate elements for describing the object attributes corresponding to the associated object category from multiple candidate elements; calculating the differentiated score of the candidate element relative to its associated object category based on the first occurrence frequency of any candidate element in its associated object category, and / or the first inverse frequency of the object category including the candidate element information; and selecting a specified number of target candidate elements in order from large to small according to the differentiated scores corresponding to the multiple candidate elements.

[0219] In some embodiments, the device can also use the first sample data to train a large model; determine the second sample data; the second sample data includes multiple sample search information and object categories associated with the multiple sample search information; use the large model to identify multiple sample object feature information related to the associated object category from the multiple sample search information; calculate the differentiated score of the sample object feature information relative to its associated object category based on the second occurrence frequency of any sample object feature information in its associated object category, and / or the second inverse frequency in the object category including the sample object feature information; divide the multiple sample object feature information into multiple groups in order of the differentiated scores from large to small; find the top N groups that make the model performance meet the performance requirements, and determine the specified value based on the number of sample object feature information in the top N groups, where N is a positive integer.

[0220] Among them, finding the top N groups that make the model performance meet the performance requirements, and determining the specified value based on the number of sample object feature information of the top N groups can include: sampling and evaluating whether the sample object feature information in multiple groups is related to the object category; based on the evaluation results, determining the number of true positive examples corresponding to the multiple groups respectively; wherein the number of true positive examples is the number of sample object feature information in any group whose evaluation results are related to the object category; fitting and generating a linear equation with one variable using the group number as the horizontal axis and the number of true positive examples as the vertical axis; calculating the accuracy and / or recall rate of the large model based on the parameters of the linear equation; and finding the specified value for the accuracy and / or recall rate that meets the performance requirements.

[0221] In some embodiments, the first determination module determines the importance of object feature information relative to the target attribution area, which may include: determining the importance of object feature information relative to the target attribution area based on at least one historical interaction behavior performed by users in the target attribution area on objects hit by object feature information related to the target object category.

[0222] Among them, the first determination module can calculate a first feature score of the object feature information relative to the target attribution area based on at least one historical interaction behavior performed by users in the target attribution area on objects corresponding to object feature information related to the target object category; calculate a second feature score of the object feature information relative to all attribution areas based on at least one historical interaction behavior performed by users in all attribution areas on objects corresponding to object feature information related to the target object category; calculate a third feature score of the target object category relative to the target attribution area based on at least one historical interaction behavior performed by users in the target attribution area on objects corresponding to all object feature information related to the target object category; calculate a fourth feature score of the target object category relative to all attribution areas based on at least one historical interaction behavior performed by users in all attribution areas on objects corresponding to all object feature information related to the target object category; and determine the importance of the object feature information relative to the target attribution area by combining at least one of the first feature score, the second feature score, the third feature score and the fourth feature score.

[0223] Among them, combining at least one of the first feature score, the second feature score, the third feature score and the fourth feature score to determine the importance of the object feature information relative to the target attribution area may include: calculating the fifth feature score of the object feature information relative to the remaining attribution areas other than the target attribution area based on the first feature score and the second feature score; calculating the sixth feature score of the target object category relative to the remaining attribution areas other than the target attribution area based on the third feature score and the fourth feature score; and taking the relative ratio of the first ratio of the first feature score and the fifth feature score to the second ratio of the third feature score and the sixth feature score as the importance of the object feature information relative to the target attribution area.

[0224] The identification device shown in FIG10 can execute the identification method described in the embodiment shown in FIG2, and its implementation principle and technical effects are not described in detail here. The specific manner in which each module and unit in the identification device in the above embodiment performs operations has been described in detail in the embodiment of the method, and will not be elaborated on here.

[0225] FIG11 is a schematic diagram of a structure of an embodiment of a recommendation device provided by an embodiment of the present disclosure, wherein the method and device include:

[0226] The second acquisition module 1101 is used to acquire target search information input by the target user;

[0227] The third determining module 1102 is used to determine the target object category associated with the target search information;

[0228] The fourth determining module 1103 is configured to determine at least one differentiated feature information of a target object category corresponding to the target attribution area of ​​the target user; wherein the at least one differentiated feature information is object feature information whose importance relative to the target attribution area satisfies a differentiated condition; the object feature information is extracted from the plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information;

[0229] The first recommendation module 1104 is configured to recommend at least one differentiated feature information to a target user.

[0230] In some embodiments, the first recommendation module recommending at least one differentiated feature information to the target user may include: using the at least one differentiated feature information as associated information of the target search information, so that the user terminal displays the at least one differentiated feature information in the search input page of the target search information.

[0231] In some embodiments, the first recommendation module recommending at least one differentiated feature information to the target user may include: determining at least one object recalled by the target search information; determining the arrangement order of the at least one object based on whether the at least one object has at least one differentiated feature information, so that the user terminal can display the object prompt information corresponding to the at least one object in the search results page according to the arrangement order.

[0232] In some embodiments, the device may also generate object prompt information containing differential feature information for a target object having any differential feature information.

[0233] Accordingly, the first recommendation module recommends at least one differentiated feature information to the target user, which may include: determining at least one object recalled by the target search information; generating object prompt information corresponding to at least one object in combination with the at least one differentiated feature information; and sending the object prompt information corresponding to the at least one object to the user end, so as to display the object prompt information corresponding to the at least one object in the search results page.

[0234] In addition, the device may also determine an arrangement order of the at least one differentiated feature information based on the importance of the at least one differentiated feature information to the target attribution area. Accordingly, the first recommendation module recommending the at least one differentiated feature information to the target user may include: recommending the at least one differentiated feature information to the target user in the arrangement order.

[0235] The recommendation device shown in FIG11 can implement the recommendation method described in the embodiment shown in FIG3 , and its implementation principles and technical effects are not further described. The specific manner in which the various modules and units of the recommendation device in the above embodiment perform operations has been described in detail in the embodiment of the method and will not be further elaborated here.

[0236] FIG12 is a schematic diagram of a structure of an embodiment of a recommendation device provided by an embodiment of the present disclosure, wherein the method and device include:

[0237] A fifth determining module 1201 is configured to determine at least one object matching the target user;

[0238] A sixth determining module 1202 is configured to determine at least one object category corresponding to at least one object;

[0239] The seventh determination module 1203 is configured to determine at least one differentiated feature information corresponding to any object category of the target attribution area of ​​the target user; the at least one differentiated feature information is object feature information whose importance relative to the target attribution area satisfies a differentiated condition; the object feature information is extracted from the plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information;

[0240] A first generating module 1204 is configured to generate object prompt information corresponding to at least one object by combining at least one differential feature information;

[0241] The first sending module 1205 is configured to send object prompt information corresponding to at least one object to a user terminal, so as to display the object prompt information corresponding to the at least one object on an object recommendation page.

[0242] The recommendation device shown in FIG12 can implement the recommendation method described in the embodiment shown in FIG4 . The implementation principles and technical effects thereof are not further described. The specific manner in which the various modules and units of the recommendation device in the aforementioned embodiment perform their operations has been described in detail in the embodiment of the method and will not be further elaborated here.

[0243] FIG13 is a schematic structural diagram of an embodiment of a display device provided by an embodiment of the present disclosure, wherein the method and device include:

[0244] The first display module 1301 is used to display the search input page;

[0245] The first detection module 1302 is used to detect the target search information input by the target user on the search input page;

[0246] The third acquisition module 1303 is configured to acquire at least one differentiated feature information of a target object category corresponding to the target attribution area of ​​the target user; wherein the target object category is associated with the target search information; the at least one differentiated feature information is object feature information whose importance relative to the target attribution area satisfies a differentiated condition; the object feature information is extracted from the plurality of historical search information and is associated with the object categories respectively associated with the plurality of historical search information;

[0247] The second display module 1304 is used to display at least one differentiated feature information in the search input page;

[0248] The second generating module 1305 is configured to generate a search request based on the target search information and the selected differentiated feature information in response to a selection operation on any differentiated feature information.

[0249] The display device shown in FIG13 can execute the display method described in the embodiment shown in FIG5, and its implementation principle and technical effects are not described in detail here. The specific manner in which each module and unit of the display device in the above embodiment performs operations has been described in detail in the embodiment of the method, and will not be elaborated on here.

[0250] FIG14 is a schematic structural diagram of an embodiment of a display device provided by an embodiment of the present disclosure, wherein the method and device include:

[0251] The third display module 1401 is used to display the object recommendation page;

[0252] The fourth display module 1402 is configured to display object prompt information corresponding to at least one object matching the target user on the object recommendation page; wherein the object prompt information corresponding to the at least one object is generated in combination with at least one differentiated feature information; the at least one differentiated feature information is object feature information whose importance relative to the target user's target region satisfies a differentiated condition; the object feature information is extracted from multiple historical search information and is related to object categories respectively associated with the multiple historical search information;

[0253] The fifth display module 1403 is configured to display, for a target object having any differentiated feature information, the differentiated feature information in the object prompt information of the target object.

[0254] The display device shown in FIG14 can execute the display method described in the embodiment shown in FIG6 , and its implementation principle and technical effects are not described in detail here. The specific manner in which each module and unit of the display device in the above embodiment performs operations has been described in detail in the embodiment of the method, and will not be elaborated on here.

[0255] The present disclosure also provides a computing device, as shown in FIG15 , which may include a storage component 1501 and a processing component 1502 ;

[0256] The storage component 1501 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component to implement the identification method described in the embodiment shown in Figure 2 or the recommendation method described in the embodiment shown in Figure 3 or Figure 4.

[0257] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0258] In the case where the display component in the computing device is used to implement the display method described in the embodiment shown in FIG. 5 or FIG. 6 , the computing device may further include a display component to perform corresponding display operations.

[0259] The input / output interface provides an interface between the processing component and peripheral interface modules, which may be output devices, input devices, etc. The communication component is configured to facilitate wired or wireless communication between the computing device and other devices.

[0260] The processing component 1502 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0261] The storage component 1501 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0262] The display component may be an electroluminescent (EL) element, a liquid crystal display or a micro display having a similar structure, or a direct retinal display or a similar laser scanning display.

[0263] It should be noted that the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. It can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device.

[0264] It should be noted that when the above-mentioned computing device implements the identification method described in the embodiment shown in Figure 2 or the recommendation method described in the embodiment shown in Figure 3 or Figure 4, it can be a physical device or an elastic computing host provided by a cloud computing platform, etc. It can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the above-mentioned computing device implements the display method described in the embodiment shown in Figure 5 or Figure 6, it can be specifically implemented as an electronic device. The electronic device can refer to a device used by the user and has the computing, Internet access, communication and other functions required by the user, such as a mobile phone, tablet computer, personal computer, wearable device, etc.

[0265] The present disclosure also provides a computer-readable storage medium storing a computer program. When executed by a computer, the computer program can implement the identification method described in the embodiment shown in FIG. 2 , the recommendation method described in the embodiment shown in FIG. 3 or FIG. 4 , or the display method described in the embodiment shown in FIG. 5 or FIG. 6 . The computer-readable medium can be included in the electronic device described in the above embodiments, or it can exist independently and not be incorporated into the electronic device.

[0266] The present disclosure also provides a computer program product, which includes a computer program carried on a computer-readable storage medium. When the computer program is executed by a computer, it can implement the identification method described in the embodiment shown in Figure 2, the recommendation method described in the embodiment shown in Figure 3 or Figure 4, or the display method described in the embodiment shown in Figure 5 or Figure 6. In such an embodiment, the computer program can be downloaded and installed from a network and / or installed from a removable medium. When the computer program is executed by a processor, it performs various functions defined in the system of the present disclosure.

[0267] In the foregoing corresponding embodiments, the computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or device.

[0268] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0269] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0270] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0271] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present disclosure.

Claims

1. A recognition method, wherein: include: Get multiple historical search information; identifying object categories associated with the plurality of historical search information; extracting object feature information related to the associated object category from the plurality of historical search information; Determining the importance of the object feature information relative to the target attribution area; The object feature information whose importance satisfies the differentiation condition is used as the differentiated feature information of the target attribution area.

2. The method according to claim 1, wherein The extracting object feature information related to the associated object category from the plurality of historical search information includes: Segmenting the plurality of historical search information respectively to obtain a plurality of candidate elements; A target candidate element for describing an object attribute corresponding to the associated object category is determined from the plurality of candidate elements using the large model, and the target candidate element is used as object feature information.

3. The method according to claim 2, wherein: The step of determining, from the plurality of candidate elements, a target candidate element for describing an object attribute corresponding to the associated object category using the large model includes: Determining, from the plurality of candidate elements, a plurality of candidate elements for describing object attributes corresponding to the associated object category using the large model; Calculate a differentiation score of any candidate element relative to its associated object category based on a first occurrence frequency of the candidate element in its associated object category and / or a first inverse frequency of the object category including the candidate element information; A specified number of target candidate elements are selected according to the order of the differentiated scores corresponding to the multiple candidate elements from large to small.

4. The method according to claim 3, wherein: Also includes: Using the first sample data to train the large model; determining second sample data; The second sample data includes a plurality of sample search information and object categories associated with the plurality of sample search information respectively; identifying, using the large model, a plurality of sample object feature information related to the associated object category from the plurality of sample search information; Calculating a differentiation score of any sample object feature information relative to its associated object category based on a second occurrence frequency of any sample object feature information in its associated object category and / or a second inverse frequency in the object category including the sample object feature information; Divide the feature information of multiple sample objects into multiple groups according to the order of differentiation scores from large to small; Find the top N groups that make the model performance meet the performance requirements, and determine the specified value based on the number of sample object feature information of the top N groups, where N is a positive integer.

5. The method according to claim 4, wherein The finding of the top N groups for which the model performance meets the performance requirements, and determining the specified value according to the number of sample object feature information of the top N groups includes: Sampling and evaluating the characteristic information of sample objects in the plurality of groups to determine whether they are relevant to the object category; Determining the number of true positive cases corresponding to each of the plurality of groups according to the evaluation results; With the group number as the horizontal axis and the number of true positives as the vertical axis, a linear equation is generated by fitting; the number of true positives is the number of sample object feature information in any group whose evaluation results are related to the object category; Calculating the accuracy and / or recall of the large model based on the parameters of the linear equation; Finds precision and / or recall values ​​that meet specified performance requirements.

6. The method according to any one of claims 1 to 5, wherein: Determining the importance of the object feature information relative to the target attribution area includes: The importance of the object feature information relative to the target attribution area is determined based on at least one historical interaction behavior performed by users in the target attribution area on objects hit by object feature information related to the target object category.

7. The method according to claim 6, wherein: Determining the importance of the object feature information relative to the target attribution area based on at least one historical interaction behavior performed by users in the target attribution area on an object hit by object feature information related to the target object category includes: Calculating a first feature score of the object feature information relative to the target attribution area based on at least one historical interaction behavior performed by a user in the target attribution area on an object corresponding to the object feature information related to the target object category; Calculating a second feature score of the object feature information relative to all attribution areas based on at least one historical interaction behavior performed by users in all attribution areas on an object corresponding to the object feature information related to the target object category; Calculating a third feature score of the target object category relative to the target attribution area based on at least one historical interaction behavior performed by users in the target attribution area on objects corresponding to all object feature information related to the target object category; Calculating a fourth feature score of the target object category relative to all attribution regions based on at least one historical interaction behavior performed by users in all attribution regions on objects corresponding to all object feature information related to the target object category; The importance of the object feature information relative to the target attribution zone is determined by combining at least one of the first feature score, the second feature score, the third feature score, and the fourth feature score.

8. The method according to claim 7, wherein: Determining the importance of the object feature information relative to the target attribution zone by combining at least one of the first feature score, the second feature score, the third feature score, and the fourth feature score includes: Calculating a fifth feature score of the object feature information relative to the remaining attribution areas other than the target attribution area based on the first feature score and the second feature score; Calculating a sixth feature score of the target object category relative to the remaining attribution areas of the non-target attribution areas based on the third feature score and the fourth feature score; A relative ratio of a first ratio of the first feature score to the fifth feature score to a second ratio of the third feature score to the sixth feature score is used as the importance of the object feature information relative to the target attribution area.

9. A recommendation method, wherein: include: Obtain target search information input by the target user; Determining a target object category associated with the target search information; Determining at least one differentiated feature information of the target user's target attribution area corresponding to the target object category; The at least one differentiated feature information is object feature information that satisfies a differentiated condition relative to the importance of the target attribution area; the object feature information is extracted from a plurality of historical search information and is related to object categories respectively associated with the plurality of historical search information; The at least one differentiated feature information is recommended to the target user.

10. The method according to claim 9, wherein: The recommending the at least one differentiated feature information to the target user includes: The at least one differential feature information is used as associated information of the target search information, so that the user terminal can display the at least one differential feature information in a search input page of the target search information.

11. The method according to claim 9, wherein The recommending the at least one differentiated feature information to the target user includes: determining at least one object of the target search information recall; An arrangement order of the at least one object is determined based on whether the at least one object has the at least one differentiated feature information, so that the user terminal displays object prompt information corresponding to the at least one object in the search result page according to the arrangement order.

12. The method according to claim 10, wherein: Also includes: For a target object having any differential feature information, object prompt information including the differential feature information is generated.

13. The method according to any one of claims 9 to 12, wherein: The recommending the at least one differentiated feature information to the target user includes: determining at least one object of the target search information recall; generating object prompt information corresponding to the at least one object in combination with the at least one differential feature information; The object prompt information corresponding to the at least one object is sent to the user terminal, so that the object prompt information corresponding to the at least one object is displayed on the search result page.

14. The method according to any one of claims 9 to 13, wherein: Also includes: determining an arrangement order of the at least one differentiated feature information according to an importance of the at least one differentiated feature information and the target attribution zone; The recommending the at least one differentiated feature information to the target user includes: The at least one differentiated feature information is recommended to the target user according to the arrangement order.

15. A recommendation method, wherein: include: Determining at least one object that matches the target user; determining at least one object category corresponding to the at least one object; Determining at least one differentiated feature information corresponding to any object category of the target attribution area of ​​the target user; The at least one differentiated feature information is object feature information that satisfies a differentiated condition relative to the importance of the target attribution area; the object feature information is extracted from a plurality of historical search information and is related to object categories respectively associated with the plurality of historical search information; generating object prompt information corresponding to the at least one object in combination with the at least one differential feature information; The object prompt information corresponding to the at least one object is sent to the user terminal, so that the object prompt information corresponding to the at least one object is displayed on the object recommendation page.

16. A method of display, wherein: include: Display the search input page; Detecting target search information input by a target user on the search input page; Obtaining at least one differentiated feature information of a target object category corresponding to a target attribution area of ​​the target user; The target object category is associated with the target search information; the at least one differentiated feature information is object feature information that satisfies a differentiated condition relative to the importance of the target attribution area; the object feature information is extracted from a plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information; Displaying the at least one differentiated feature information in the search input page; In response to a selection operation on any one of the differentiated feature information, a search request is generated based on the target search information and the selected differentiated feature information.

17. A method of display, wherein: include: Display object recommendation page; Displaying object prompt information corresponding to at least one object matching the target user on the object recommendation page; The object prompt information corresponding to the at least one object is generated in combination with at least one differential feature information; The at least one differentiated feature information is object feature information that satisfies a differentiated condition relative to the importance of the target attribution area of ​​the target user; the object feature information is extracted from the plurality of historical search information and is related to the object categories respectively associated with the plurality of historical search information; For a target object having any differential feature information, the differential feature information is displayed in the object prompt information of the target object.

18. A computing device, wherein: It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the identification method according to any one of claims 1 to 8 or the recommendation method according to any one of claims 9 to 14 or the recommendation method according to claim 15 or the display method according to claim 16 or the display method according to claim 17.

19. A computer storage medium, wherein: A computer program is stored, and when the computer program is executed by a computer, it implements the identification method according to any one of claims 1 to 8, the recommendation method according to any one of claims 9 to 14, the recommendation method according to claim 15, the display method according to claim 16, or the display method according to claim 17.

20. A computer program product, wherein The invention comprises a computer program / instruction, which, when executed by a computer, implements the identification method according to any one of claims 1 to 8, the recommendation method according to any one of claims 9 to 14, the recommendation method according to claim 15, the display method according to claim 16, or the display method according to claim 17.

Images