Service recommendation and commodity object information providing method

Abstract

The invention belongs to the technical field of information. The service recommendation method comprises the following steps: acquiring at least one target service object provided by a target user; according to the target service object, original description information corresponding to the target service object and attribute information of the target user are obtained, and the attribute information of the target user comprises at least one of nationalization information, regional information, school district information and academic record information; performing feature extraction on the attribute information of the target user by using a natural semantic recognition model to obtain attribute information features of the target user; fusing the attribute information features of the target user by using an autoregression learning model to obtain comprehensive attribute features of the target user; by means of the comprehensive attribute features and the original description information, expression preference processing is conducted on the target user, target description information is obtained, and the preference of the location of the target user represents preference expression convenient for the target user to understand; and sending the target description information to the target user.

Description

Technical Field

[0001] The present application belongs to the field of information technology, and in particular relates to a method for recommending services and providing product object information. Background Art

[0002] Improving the learning outcomes of international students and providing targeted tutoring and related services directly impacts the effectiveness of tutoring. Existing technologies primarily rely on one-on-one communication with students to determine their goals, then pairing them with tutors based on those goals. The tutors then engage in in-depth communication with the students, develop a syllabus, organize learning materials, and then provide one-on-one instruction. However, this approach relies on precise communication and the tutor's experience, resulting in low efficiency and accuracy.

[0003] In recent years, with the continuous development of data analysis and artificial intelligence technology, more and more platforms have begun to try to use technology to improve the accuracy of predicting the demand for international student training services during the service process.

[0004] However, the applicant's research found that in current project management, the method of forecasting overseas student demand services can usually only forecast demand for one aspect of service management, and cannot provide a comprehensive forecast report. The accuracy and precision of the forecast needs to be improved. Summary of the Invention

[0005] Based on this, it is necessary to provide a method for service recommendation and providing product object information to address the above technical issues.

[0006] In a first aspect, the present application provides a service recommendation method, comprising:

[0007] Obtain at least one target service object provided by a target user;

[0008] According to the target service object, original description information corresponding to the target service object and attribute information of the target user are obtained, wherein the attribute information of the target user includes at least one of national information, regional information, school district information, and academic performance information;

[0009] Using a natural semantic recognition model, feature extraction is performed on the attribute information of the target user to obtain attribute information features of the target user;

[0010] Using an autoregressive learning model, the attribute information features of the target user are integrated to obtain the comprehensive attribute features of the target user;

[0011] Using the comprehensive attribute features and the original description information, processing the target user's expressed preferences to obtain target description information, wherein the target user's location preference represents a preference expression that is easy for the target user to understand;

[0012] Send the target description information to the target user.

[0013] In some practicable manners, the step of obtaining original description information corresponding to the target service object and attribute information of the target user according to the target service object includes:

[0014] Constructing a description of the target service object according to the target service object to obtain original description information;

[0015] Acquiring historical behavior data of the target user, wherein the historical behavior data includes at least one of country-specific data, regional data, school district data, and academic performance data;

[0016] Determine attribute information of the target user based on the historical behavior data of the target user.

[0017] In some practicable embodiments, the step of extracting features from the target user's attribute information using a natural semantic recognition model to obtain the target user's attribute information features includes:

[0018] Cleaning and formatting the attribute information of the target user to obtain consistent attribute data of the target user;

[0019] Using a natural semantic recognition model, the semantic features of the attribute data of the target user are analyzed and identified to obtain at least one of national information features, regional information features, school district information features and academic performance information features, wherein at least one of the national information features, regional information features, school district information features and academic performance information features constitutes the attribute information features of the target user.

[0020] In some practicable embodiments, the step of processing the target user's expressed preferences using the comprehensive attribute features and the original description information to obtain the target description information includes:

[0021] Obtaining an expression rule for the location of the target user;

[0022] According to the expression rules of the target user's location, the original description information is replaced with words and the sentence structure is adjusted to obtain intermediate description information;

[0023] The intermediate description information is subjected to consistency adjustment and grammatical correction to obtain target description information.

[0024] In some practicable embodiments, the step of processing the target user's expressed preferences using the comprehensive attribute features and the original description information to obtain the target description information includes:

[0025] Obtaining a target page corresponding to the target user's preference expression;

[0026] Determining a number of target display areas based on the target page;

[0027] The same detailed information page shared by multiple services associated with the target service objects is reordered and displayed according to the target display areas.

[0028] In some practicable embodiments, the step of processing the target user's expressed preferences using the comprehensive attribute features and the original description information to obtain the target description information includes:

[0029] When the original description information includes original image information:

[0030] Obtaining image information corresponding to the target user's preference expression;

[0031] Performing feature extraction on the image information to obtain image features;

[0032] According to the image features, the original image information is converted to obtain target image information.

[0033] In some practicable embodiments, the step of processing the target user's expressed preferences using the comprehensive attribute features and the original description information to obtain the target description information includes:

[0034] When the original description information includes original audio information:

[0035] Obtaining audio information corresponding to the target user's preference expression;

[0036] Performing feature extraction on the audio information to obtain audio features;

[0037] The original audio information is converted according to the audio features to obtain target audio information.

[0038] In a second aspect, the present application provides a method for providing product object information, which is applied to the aforementioned service recommendation method. The method for providing product object information includes:

[0039] Acquire attribute information of a target user and a target service object, wherein the attribute information of the target user includes at least one of national information, regional information, school district information, and academic performance information;

[0040] According to the target service object, target description information is obtained, wherein the target description information represents information that is converted according to original description information corresponding to the target service object and is easy for the target user to understand based on the expression preference of the target user.

[0041] In a third aspect, the present application provides a computer storage medium having a computer program stored thereon, which implements the steps of the aforementioned method when executed by a processor.

[0042] In a fourth aspect, the present application provides an electronic device, comprising:

[0043] one or more processors;

[0044] and a memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute the steps of the aforementioned method.

[0045] Beneficial effect: The present application provides a method for service recommendation, which first obtains at least one target service object provided by a specific user; then, based on the target service object, obtains the original description information corresponding to the target service object and the attribute information of the target user, wherein the attribute information of the target user includes at least one of national information, regional information, school district information and academic performance information; uses a natural semantic recognition model to extract features of the attribute information of the target user to obtain the attribute information features of the target user; uses an autoregressive learning model to fuse the attribute information features of the target user to obtain the comprehensive attribute features of the target user; uses the comprehensive attribute features and the original description information to process the expression preferences of the target user to obtain target description information, wherein the location preference of the target user represents a preference expression that is easy for the target user to understand; and sends the target description information to the target user. Through the above method, the original description information corresponding to the target service object is converted in combination with the attribute information of the target user. During the conversion process, the original description information is converted into a language that is easy for the target user to understand using a natural semantic recognition model and an autoregressive learning model, so that the target user can understand it and improve the experience of the target user. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the conventional technology, the following briefly introduces the drawings required for use in the embodiments or the conventional technology descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 A flowchart of a method for providing service recommendations in one embodiment.

[0048] Figure 2 The present invention is a flowchart of a method for providing commodity object information in one embodiment. DETAILED DESCRIPTION

[0049] To facilitate understanding of the present application, the present application will be described more fully below with reference to the accompanying drawings. The accompanying drawings provide embodiments of the present application. However, the present application may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to make the disclosure of the present application more thorough and comprehensive.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein in the specification of this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The term "and / or" as used herein includes any and all couplings of one or more of the associated listed items.

[0051] It will be understood that the terms "first," "second," etc. used herein may be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish a first element from another element.

[0052] The following are some explanations of some terms involved in this application to facilitate understanding of this application:

[0053] BERT, Bidirectional Encoder Representations from Transformers, is a pre-trained language representation model based on the Transformer architecture.

[0054] GPT, Generative Pretrained Transformer, is a Transformer-based pretrained language model developed by OpenAI.

[0055] Autoregressive learning model, Autoregressive Model, referred to as AR model, is a prediction model in statistics and time series analysis, which assumes that the future value of a variable can be predicted by its past value.

[0056] Rich media is often used to describe multimedia content with rich interactivity and sensory experience. This type of information not only includes traditional text and static images, but also covers a wider range of dynamic and interactive elements.

[0057] like Figure 1 As shown, the first aspect of the present application provides a method for service recommendation, the method for service recommendation comprises the following steps:

[0058] S100: Acquire at least one target service object provided by a target user.

[0059] Among them, obtaining at least one target service object provided by the target user can be understood as the server side, that is, the target information recommendation system. The information received by the target information recommendation system can be sent to the target information recommendation system through the receiving server side. After the target information recommendation system processes the message, it returns the processed message to the receiving server side. The target service object can be associated with the minimum inventory unit service information.

[0060] Specifically, the target user may be an overseas student. For example, when this overseas student wants to learn Chinese in China, since he or she does not know Chinese, he or she needs to learn Chinese. When this overseas student wants to learn Chinese, he or she will choose a corresponding type of Chinese to learn based on the purpose of learning. For example, the types of Chinese that overseas students learn may include Mandarin Chinese, dialect Chinese, business Chinese, academic Chinese, and daily Chinese. It is understandable that the target service object indicates the type of Chinese, such as Mandarin Chinese, dialect Chinese, business Chinese, academic Chinese, and daily Chinese mentioned above.

[0061] S200: According to the target service object, original description information corresponding to the target service object and attribute information of the target user are obtained.

[0062] The attribute information of the target user includes at least one of national information, regional information, school district information and academic performance information.

[0063] Specifically, obtaining the original description information corresponding to the target service object and the attribute information of the target user may include the following steps:

[0064] S201: Construct a description of the target service object according to the target service object to obtain original description information.

[0065] It should be noted that the target service object, such as the aforementioned volume, can include multiple Chinese language learning options. In this case, in order to facilitate the target user's understanding of the current target service object learning process and the goals that can be achieved after learning, it is necessary to describe the target service object. For example, the target service object can be understood as a commodity, and the original description information can be understood as a description of the target service object, such as service content, characteristics, and advantages. By combining the descriptive information such as service content, characteristics, and advantages, a comprehensive service description can be constructed using images, text, and videos to ensure the accuracy and appeal of the descriptive information. In this way, the original description information corresponding to the target service object is formed.

[0066] When there are multiple target service objects, there are corresponding multiple original description information.

[0067] It should also be noted that when there are multiple target service objects, each target service object corresponds to one original description information. However, although there are multiple target service objects, some of the original description information of the target service objects is repeated. In this case, if each target service object corresponds to a complete set of original description information, it will take up a large space. Therefore, reducing the space occupied by the original description information includes the following steps:

[0068] Generating N sets of the original description information according to the N target service objects;

[0069] Comparing the N sets of original description information to determine repeated sentences, pictures, and video content;

[0070] Mark repeated sentences, pictures, and video contents to obtain marked sentences, pictures, and videos;

[0071] Build labeled sentence, image, and video databases;

[0072] Mark and replace the content in each set of original description information that is repeated in the original description information in other sets;

[0073] The content of each set of original description information consists of non-repeating sentences, pictures and videos, and tags.

[0074] It should be noted that the above method can effectively reduce the space occupied by each set of original description information. For example, if a section of content in multiple sets of original description information contains "...Learning Chinese well will enable effective communication with over one billion Chinese speakers, Picture 1, Video 2...", then "Learning Chinese well will enable effective communication with over one billion Chinese speakers, Picture 1, Video 2" can be labeled. For example, "Learning Chinese well will enable effective communication with over one billion Chinese speakers" can be labeled as 1-A, Picture 1 as 2-A, and Video 2 as 3-A. In this way, if a section of content in one set of original description information contains "...1-A, 2-A, 3-A...", the corresponding labels can be called in the database to form a complete section of content. It can be understood that in the step of "comparing N sets of the original description information to determine repeated sentences, pictures and video content", when the similarity is greater than a preset threshold, the natural semantic recognition model can be used to extract features from the content of the sentence, and the autoregressive learning model can be used to transform the features so that two sentences with a similarity greater than the preset threshold are adjusted to the same sentence and marked.

[0075] S202: Obtain historical behavior data of the target user.

[0076] The historical behavior data includes at least one of national data, regional data, school district data and academic performance data.

[0077] Specifically, country-specific data can refer to the target user's nationality; regional data can refer to the target user's region of residence; school district data can refer to the target user's school district or school; and academic performance data can refer to the target user's academic performance. The above data can be provided by the target user themselves. Once the target user's historical behavior data is obtained, the target user can be analyzed based on this historical behavior data in subsequent steps.

[0078] S203: Determine attribute information of the target user based on the historical behavior data of the target user.

[0079] It should be noted that after obtaining the historical behavior data of the target user in the aforementioned steps, the attribute information of the target user can be formed through the historical behavior data. This attribute information of the target user will help to analyze the current situation of the target user in subsequent steps so as to provide targeted personalized services, and in subsequent steps, tailor-made descriptive information for the target user to facilitate the target user to understand the descriptive information of the target service object.

[0080] S300 , using a natural semantic recognition model, extracting features from the attribute information of the target user to obtain attribute information features of the target user.

[0081] Specifically, obtaining the attribute information characteristics of the target user includes the following steps:

[0082] S301 , cleaning and formatting the attribute information of the target user to obtain consistent attribute data of the target user.

[0083] S302, using a natural semantic recognition model, analyze and identify the semantic features of the attribute data of the target user to obtain at least one of national information features, regional information features, school district information features and academic performance information features, wherein at least one of the national information features, regional information features, school district information features and academic performance information features constitutes the attribute information features of the target user.

[0084] It should be noted that although the target user's attribute information obtained in the aforementioned steps, such as national information, regional information, school district information, and academic performance information, can be filled in by the target user himself, there may still be noise and irrelevant information, such as irrelevant symbols, blank characters, spelling errors, missing data, etc. Therefore, the target user's attribute information needs to be cleaned and the format converted in order to obtain consistent attribute data of the target user for easy use in subsequent steps.

[0085] It should also be noted that after obtaining the target user's attribute data, a trained natural semantic recognition model, such as BERT, GPT, or other pre-trained deep learning models, is used to analyze and identify the semantic features of the target user's attribute data. Specifically, when the target user's attribute data includes national information, regional information, school district information, and academic performance information, the national information features obtained through analysis and identification can represent the analysis of the target user's nationality or regional background, and the extraction of features related to culture and language habits; regional information features represent the identification of the specific area where the target user lives or studies, and the extraction of features related to local customs, economic conditions, etc.; school district information features can represent the analysis of the target user's school district, and the extraction of features related to educational resources, education level, etc.; and academic performance information features can represent the evaluation of the target user's academic performance, and the extraction of features related to learning habits, academic ability, etc.

[0086] Finally, it should be noted that BERT, GPT or other pre-trained deep learning models are conventional models, and this application does not adjust or improve the models.

[0087] S400 , utilizing an autoregressive learning model to fuse the attribute information features of the target user to obtain comprehensive attribute features of the target user.

[0088] Specifically, the attribute information features of the target user extracted in the previous step are normalized and standardized to eliminate the dimensional effects between different features, so that they can be compared and integrated under the same standards. A suitable conventional autoregressive model is selected to process the data using the autoregressive model and capture the dependencies in the data. The autoregressive model is used to integrate the data into comprehensive attribute features, which are used to represent the characteristics of the target user in different dimensions, and to utilize the target user's behavioral trends and preference changes. Next, the comprehensive attribute features are used to provide personalized service recommendations for the target user, such as learning resource recommendations and tutoring course recommendations.

[0089] Autoregressive learning models can effectively integrate multiple attributes of target users into a comprehensive feature set, supporting personalized services and improving user experience. This approach is particularly suitable for education, training, and consulting services, which require considering multiple aspects of target users' attributes.

[0090] The autoregressive learning model can perform feature connection on at least one of the national information features, the regional information features, the school district information features and the academic performance information, so that all features can be connected without losing any original information. Next, after the features are connected, a long feature vector is formed. Next, the principal component analysis method can be used to reduce the dimension of the long feature vector. Finally, the weighted average algorithm is used to further highlight important features, suppress unimportant features, and improve the accuracy of the results. It should be noted that the autoregressive learning model can also directly use the weighted average algorithm to assign corresponding weights to the national information features, the regional information features, the school district information features and the academic performance information, and perform weighted calculations to obtain a weighted average value, and then use the weighted average value to form a composite attribute feature. The above method is only an exemplary description. You can also use a conventional autoregressive learning model as needed to select other methods to obtain the comprehensive attribute features of the target user.

[0091] S500: Process the target user's expressed preferences using the comprehensive attribute features and the original description information to obtain target description information.

[0092] The location preference of the target user is a preference expression that is easy for the target user to understand.

[0093] Specifically, obtaining target description information includes the following steps:

[0094] S501: Obtain the expression rule of the target user's location.

[0095] It should be noted that the language and cultural expression habits of the target user's region are collected and understood in order to perform appropriate localization processing. For example, the culture, language habits, common vocabulary, grammatical structure, etc. of the target user's region are studied. Using the above method, a set of expression rules are compiled. These rules may include usage preferences of specific vocabulary, regional differences in grammatical structure, sentence expression habits, and regional slang or phrases. It is understood that these rules can be adjusted as needed, and this application does not limit this.

[0096] S502: According to the expression rules of the target user's location, the original description information is replaced with words and the sentence structure is adjusted to obtain intermediate description information.

[0097] Specifically, replace the vocabulary in the original description information with vocabulary that conforms to the preferences of the target region. For example, replace professional terms or uncommon words with more understandable expressions. Adjust the sentence structure to conform to the grammatical habits of the target region. For example, adjusting the sentence structure may include changing the beginning or end of the sentence or using a different sentence structure. Ensure that the description information is culturally appropriate for the target users to avoid cultural conflicts or misunderstandings. For example, natural language processing (NLP) tools can be used to automate this process to improve efficiency and accuracy.

[0098] It should be noted that adjustments to the original description information are made based on the constructed expression rules. Furthermore, the target user's location refers to the target user's long-term residence, not their current location. For example, for an overseas student studying in China, China is their current location, but the target user's location refers to the region where the overseas student has lived long-term, such as their registered residence. This way, adjustments to the original description information better align with the target user's reading and understanding habits.

[0099] S503: Perform consistency adjustment and grammatical correction on the intermediate description information to obtain target description information.

[0100] It's important to note that the intermediate descriptions generated in the previous steps must be coherent and fluent to facilitate reading for the target user. This allows for the addition of transitional vocabulary or adjustments to the order of information. Furthermore, the intermediate descriptions can be grammatically checked to correct grammatical errors, ensure they conform to the language standards of the target region, and verify their contextual appropriateness and comprehension skills and knowledge of the target user. Final language and cultural adjustments are performed to ensure that the original descriptions retain their accuracy and appeal after localization.

[0101] S600: Send the target description information to the target user.

[0102] In one embodiment, S600, sending the target description information to the target user may include the following steps:

[0103] S601, obtaining a target page corresponding to the preference expression of the target user;

[0104] S602, determining a number of target display areas according to the target page;

[0105] S603: Rearrange and display the same detailed information page shared by multiple services associated with the target service objects according to the target display areas.

[0106] Specifically, obtaining the target page corresponding to the target user's preference expression can refer to the layout commonly used by target pages in the country where the target user has lived for a long time, the layout commonly used by target pages in the region where the target user has lived for a long time, or the layout commonly used by target pages in the school district where the target user is located. It should be noted that if there are differences in the layout of national, regional, and school district-specific target pages, a weighted calculation can be performed with the weights of school district, regional, and national in roughly descending order. Based on the weighted calculation results, multiple services associated with target service objects are matched to the target display areas of the target page, and the display order of the target service objects in the target display areas is adjusted. In other words, the multiple services are reordered according to the localized selection preferences corresponding to the national / regional / school district / student performance attribute information, so that when the service selection interface is displayed based on the target page, the display order of the multiple services in the service selection interface is determined according to the reordering results.

[0107] In one embodiment, S600, sending the target description information to the target user may further include the following steps:

[0108] When the original description information includes original image information:

[0109] S604: Obtain image information corresponding to the target user's preference expression.

[0110] S605: Extract features from the image information to obtain image features.

[0111] S606: Convert the original image information according to the image features to obtain target image information.

[0112] Specifically, the target user's image preferences are analyzed. This analysis process can obtain analysis data through the target user's nationalization / regionalization / school district / study performance attribute information in the aforementioned steps, and obtain preferences for the type, style, and content of images through the nationalization / regionalization / school district / study performance attribute information. Among them, for obtaining preferred images, images can be retrieved from a database or obtained from external resources, for example, the network image style corresponding to the target user's nationalization / regionalization / school district, for example, images in the Japanese customary anime style. When the target user's preference for image information is obtained, feature extraction is performed on the image information corresponding to the target user's preference expression to obtain image features. Next, the obtained image features are integrated into the original image information included in the original description information, thereby forming a conversion process for the original image information, and then obtaining the target image information. Finally, the original image information is replaced with the target image information. It should be noted that the extraction of image features is a conventional extraction method, and this application does not limit how to extract image features. It can be understood that, based on the national / regional attribute information of the target user, the composition style, model type and / or atmosphere elements of the original image information are converted and processed to generate target image information that conforms to the localized preferences corresponding to the national / regional / school district / academic performance attribute information.

[0113] In one embodiment, S600, sending the target description information to the target user may further include the following steps:

[0114] When the original description information includes original audio information:

[0115] S607, obtaining audio information corresponding to the target user's preference expression;

[0116] S608, extracting features from the audio information to obtain audio features;

[0117] S609: Convert the original audio information according to the audio features to obtain target audio information.

[0118] It should be noted that the target user's audio preferences are analyzed. This analysis process can obtain analysis data through the target user's nationalization / regionalization / school district / study performance attribute information in the aforementioned steps, and obtain preferences for audio type, style and content through the nationalization / regionalization / school district / study performance attribute information. Among them, for obtaining preferred audio, audio can be retrieved from a database or obtained from external resources, for example, the network audio style corresponding to the target user's nationalization / regionalization / school district, such as background music of the audio. When the target user's preference for audio information is obtained, feature extraction is performed on the audio information corresponding to the target user's preference expression to obtain audio features. Next, the obtained audio features are integrated into the original audio information included in the original description information, thereby forming a conversion process for the original audio information, and then obtaining the target audio information. Finally, the original audio information is replaced with the target audio information. It should be noted that the extraction of audio features is a conventional extraction method, and this application does not limit how to extract audio features. It can be understood that the original audio information is converted and processed according to the nationality / regionalization / school district / academic performance attribute information of the target user to generate target audio information that conforms to the localized preferences corresponding to the nationality / regionalization / school district / academic performance attribute information.

[0119] like Figure 2 As shown, the second aspect of the present application provides a method for providing product object information, which is applied to the aforementioned service recommendation method. The method for providing product object information includes:

[0120] S700: Acquire attribute information of a target user and a target service object.

[0121] The attribute information of the target user includes at least one of national information, regional information, school district information and academic performance information.

[0122] S800: Obtain target description information according to the target service object.

[0123] The target description information represents information that is converted from original description information corresponding to the target service object and is easy for the target user to understand based on the expression preference of the target user.

[0124] It should be noted that the end that provides the commodity object information is the end that receives the service, that is, the receiving service end. Receive target description information of at least one target service object provided by a server for a target user, the target service object being associated with minimum stock keeping unit service information, the target description information being generated by converting original description information corresponding to the service based on localized expression preferences according to the target user's nationality / regionalization / school district / study performance attribute information; the localized expression preferences including one or more of the following: a preference for localized common word classes for service names and / or adjectives included in the original description information, a preference for localized common standards / measurement units for attribute / parameter description information included in the original description information, a preference for localized sorting of user comments included in the original description information, and a preference for localized style of rich media information included in the original description information; then, display the target description information of the at least one target product object through a target page, so that when information about the same service of the same service object is provided to users with different nationality / regionalization / school district / study performance attribute information through the target page, different target description information generated based on different localized expression preferences is provided. In addition, the original description information may also be displayed through images and audio.

[0125] In a third aspect of the present application, a computer storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, it implements the aforementioned service recommendation method and the steps of the aforementioned method for providing commodity object information.

[0126] In a fourth aspect of the present application, an electronic device is provided, comprising:

[0127] one or more processors;

[0128] and a memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute the steps of the aforementioned method.

[0129] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0130] The various embodiments in the present disclosure are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0131] The scope of protection of the present disclosure is not limited to the above-described embodiments. Obviously, those skilled in the art may make various modifications and variations to the present disclosure without departing from the scope and spirit of the present disclosure. If such modifications and variations fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is intended to include such modifications and variations.

Claims

1. A service recommendation method, characterized in that: The methods recommended by the service include: Obtain at least one target service object provided by a target user; According to the target service object, original description information corresponding to the target service object and attribute information of the target user are obtained, wherein the attribute information of the target user includes at least one of national information, regional information, school district information, and academic performance information; Using a natural semantic recognition model, feature extraction is performed on the attribute information of the target user to obtain attribute information features of the target user; Using an autoregressive learning model, the attribute information features of the target user are integrated to obtain the comprehensive attribute features of the target user; Using the comprehensive attribute features and the original description information, processing the target user's expressed preferences to obtain target description information, wherein the target user's location preference represents a preference expression that is easy for the target user to understand; Send the target description information to the target user.

2. The service recommendation method according to claim 1, characterized in that: The step of obtaining original description information corresponding to the target service object and attribute information of the target user according to the target service object includes: According to the target service object, construct a description of the target service object to obtain original description information; Acquiring historical behavior data of the target user, wherein the historical behavior data includes at least one of country-specific data, regional data, school district data, and academic performance data; Determine attribute information of the target user based on the historical behavior data of the target user.

3. The service recommendation method according to claim 1, wherein: The step of extracting features from the target user's attribute information using a natural semantic recognition model to obtain the target user's attribute information features includes: Cleaning and formatting the attribute information of the target user to obtain consistent attribute data of the target user; Using a natural semantic recognition model, the semantic features of the attribute data of the target user are analyzed and identified to obtain at least one of national information features, regional information features, school district information features and academic performance information features, wherein at least one of the national information features, regional information features, school district information features and academic performance information features constitutes the attribute information features of the target user.

4. The service recommendation method according to claim 3, characterized in that: The step of processing the target user's expressed preferences by utilizing the comprehensive attribute features and the original description information to obtain the target description information includes: Obtaining an expression rule for the location of the target user; According to the expression rules of the target user's location, the original description information is replaced with words and the sentence structure is adjusted to obtain intermediate description information; The intermediate description information is subjected to consistency adjustment and grammatical correction to obtain target description information.

5. The service recommendation method according to claim 1, characterized in that: The step of processing the target user's expressed preferences by utilizing the comprehensive attribute features and the original description information to obtain the target description information includes: Obtaining a target page corresponding to the target user's preference expression; Determining a number of target display areas based on the target page; The same detailed information page shared by multiple services associated with the target service objects is reordered and displayed according to the target display areas.

6. The service recommendation method according to claim 1, characterized in that: The step of processing the target user's expressed preferences by utilizing the comprehensive attribute features and the original description information to obtain the target description information includes: When the original description information includes original image information: Obtaining image information corresponding to the target user's preference expression; Performing feature extraction on the image information to obtain image features; According to the image features, the original image information is converted to obtain target image information.

7. The service recommendation method according to claim 1, characterized in that: The step of processing the target user's expressed preferences by utilizing the comprehensive attribute features and the original description information to obtain the target description information includes: When the original description information includes original audio information: Obtaining audio information corresponding to the target user's preference expression; Performing feature extraction on the audio information to obtain audio features; The original audio information is converted according to the audio features to obtain target audio information.

8. A method for providing commodity object information, characterized in that: The service recommendation method according to any one of claims 1 to 7, wherein the method for providing product object information comprises: Acquire attribute information of a target user and a target service object, wherein the attribute information of the target user includes at least one of national information, regional information, school district information, and academic performance information; According to the target service object, target description information is obtained, wherein the target description information represents information that is converted according to original description information corresponding to the target service object and is easy for the target user to understand based on the expression preference of the target user.

9. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

10. An electronic device, characterized in that: include: one or more processors; and a memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute the steps of the method according to any one of claims 1 to 8.

Images