CN116467517A

CN116467517A - Product recommendation method, product recommendation device, electronic equipment and storage medium

Info

Publication number: CN116467517A
Application number: CN202310430243.8A
Authority: CN
Inventors: 梁昕; 韩勋; 李牧真; 易艳; 王建明; 肖京; 罗玉
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2023-04-12
Filing date: 2023-04-12
Publication date: 2023-07-21

Abstract

The embodiment of the application provides a product recommendation method, a product recommendation device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring target user data, wherein the target user data comprises user basic data and user behavior data of a target user; extracting characteristics of target user data to obtain target user characteristics; carrying out product recall based on a preset product recommendation model and target user characteristics to obtain an initial recommendation list; pushing the initial recommendation list to a target user, and acquiring intention data fed back by the target user according to the initial recommendation list; extracting entity characteristics from the intention data to obtain target intention characteristics; product rearrangement is carried out on the products to be recommended of the initial recommendation list based on the target intention characteristics, and a target recommendation list is obtained; and pushing the target recommendation list to the target user. According to the method and the device for recommending the products, accuracy of product recommendation can be improved.

Description

Product recommendation method, product recommendation device, electronic equipment and storage medium

Technical Field

The application relates to the technical field of artificial intelligence, in particular to a product recommendation method, a product recommendation device, electronic equipment and a storage medium.

Background

The current product recommendation method is often prone to recommending the current hot products to users, the recommendation mode often cannot meet the actual demands of the users, and the accuracy of product recommendation is affected, so that how to improve the accuracy of product recommendation becomes a technical problem to be solved urgently.

Disclosure of Invention

The embodiment of the application mainly aims to provide a product recommending method, a product recommending device, electronic equipment and a storage medium, and aims to improve accuracy of product recommendation.

To achieve the above object, a first aspect of an embodiment of the present application provides a product recommendation method, including:

acquiring target user data, wherein the target user data comprises user basic data and user behavior data of a target user;

extracting the characteristics of the target user data to obtain target user characteristics;

carrying out product recall based on a preset product recommendation model and the target user characteristics to obtain an initial recommendation list;

pushing the initial recommendation list to the target user, and acquiring intention data fed back by the target user according to the initial recommendation list;

extracting entity characteristics from the intention data to obtain target intention characteristics;

Product rearrangement is carried out on the products to be recommended of the initial recommendation list based on the target intention characteristics, and a target recommendation list is obtained;

pushing the target recommendation list to the target user.

In some embodiments, the feature extraction of the target user data to obtain a target user feature includes:

extracting the characteristics of the user basic data to obtain user basic characteristics;

extracting characteristics of the user behavior data to obtain user behavior characteristics;

and carrying out feature fusion on the user behavior feature and the user basic feature to obtain the target user feature.

In some embodiments, the product recall is performed based on the preset product recommendation model and the target user feature to obtain an initial recommendation list, including:

inputting the target user characteristics into the product recommendation model, wherein the product recommendation model comprises an input layer, a coding layer and a prediction layer;

embedding the target user features based on the input layer to obtain user feature embedded vectors;

coding the user characteristic embedded vector based on the coding layer to obtain a user characteristic coding vector;

And carrying out product recall based on the prediction layer, the pre-acquired product features and the user feature code vector to obtain the initial recommendation list, wherein the product features are basic features of the product to be recommended.

In some embodiments, the performing product recall based on the prediction layer, the pre-acquired product features, and the user feature encoding vector to obtain the initial recommendation list includes:

calculating the recommendation scores of the products based on the product features and the user feature code vectors by the prediction layer to obtain the recommendation score of each product to be recommended;

screening the products to be recommended based on the recommendation score and a preset recommendation threshold to obtain candidate products;

and sorting the candidate products according to the recommendation scores to obtain the initial recommendation list.

In some embodiments, the extracting the entity feature from the intention data to obtain the target intention feature includes:

traversing a preset intention database, and performing correlation calculation on the intention data and reference intention characteristics in the intention database to obtain intention correlation;

and screening the reference intention characteristic based on the intention relativity to obtain the target intention characteristic.

In some embodiments, the product rearrangement of the products to be recommended of the initial recommendation list based on the target intention feature to obtain a target recommendation list includes:

calculating predicted intention data of each product to be recommended in the initial recommendation list according to the target intention characteristics;

and carrying out product rearrangement on the products to be recommended of the initial recommendation list based on the predicted intention data to obtain a target recommendation list.

In some embodiments, the calculating the predicted intent data of each product to be recommended in the initial recommendation list according to the target intent feature includes:

calculating the intention probability vector of each product to be recommended in the initial recommendation list based on a preset function and the target intention characteristic;

and obtaining the predicted intention data according to the intention probability vector.

To achieve the above object, a second aspect of the embodiments of the present application proposes a product recommendation device, the device comprising:

the data acquisition module is used for acquiring target user data, wherein the target user data comprises user basic data and user behavior data of a target user;

the feature extraction module is used for extracting features of the target user data to obtain target user features;

The product recall module is used for carrying out product recall based on a preset product recommendation model and the target user characteristics to obtain an initial recommendation list;

the initial recommendation module is used for pushing the initial recommendation list to the target user and acquiring intention data fed back by the target user according to the initial recommendation list;

the entity characteristic extraction module is used for extracting entity characteristics of the intention data to obtain target intention characteristics;

the rearrangement module is used for rearranging products to be recommended of the initial recommendation list based on the target intention characteristics to obtain a target recommendation list;

and the target recommendation module is used for pushing the target recommendation list to the target user.

To achieve the above object, a third aspect of the embodiments of the present application proposes an electronic device, which includes a memory, a processor, where the memory stores a computer program, and the processor implements the method described in the first aspect when executing the computer program.

To achieve the above object, a fourth aspect of the embodiments of the present application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of the first aspect.

According to the product recommending method, the product recommending device, the electronic equipment and the storage medium, target user data are obtained, wherein the target user data comprise user basic data and user behavior data of a target user; the method can obtain the target user characteristics containing important user information more conveniently, so that the target user characteristics can be used for a subsequent recommendation process. And carrying out product recall based on a preset product recommendation model and target user characteristics to obtain an initial recommendation list, and screening and sorting products to be recommended based on the target user characteristics conveniently to obtain the initial recommendation list conforming to the user characteristics of the target user. Pushing the initial recommendation list to a target user, and acquiring intention data fed back by the target user according to the initial recommendation list; the entity characteristic extraction is carried out on the intention data to obtain target intention characteristics, so that the current intention of the target user can be conveniently identified, and the product recommendation can be carried out based on the current intention of the target user. Further, the products to be recommended of the initial recommendation list are rearranged based on the target intention characteristics to obtain a target recommendation list, and the products to be recommended can be conveniently screened and ordered based on the target intention characteristics of the target user to obtain the target recommendation list which accords with the actual intention of the target user. And finally, pushing the target recommendation list to the target user, wherein the method can screen and sort the products to be recommended according to the user characteristics and the intention characteristics of the target user to obtain the target recommendation list which accords with the user characteristics and the current intention of the user, thereby better realizing personalized recommendation of the products and improving the accuracy of product recommendation.

Drawings

FIG. 1 is a flowchart of a product recommendation method provided by an embodiment of the present application;

fig. 2 is a flowchart of step S102 in fig. 1;

fig. 3 is a flowchart of step S103 in fig. 1;

fig. 4 is a flowchart of step S304 in fig. 3;

fig. 5 is a flowchart of step S105 in fig. 1;

fig. 6 is a flowchart of step S106 in fig. 1;

fig. 7 is a flowchart of step S601 in fig. 6;

fig. 8 is a schematic structural diagram of a product recommendation device provided in an embodiment of the present application;

fig. 9 is a schematic hardware structure of an electronic device according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

It should be noted that although functional block division is performed in a device diagram and a logic sequence is shown in a flowchart, in some cases, the steps shown or described may be performed in a different order than the block division in the device, or in the flowchart. The terms first, second and the like in the description and in the claims and in the above-described figures, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the present application.

First, several nouns referred to in this application are parsed:

artificial intelligence (artificial intelligence, AI): is a new technical science for researching and developing theories, methods, technologies and application systems for simulating, extending and expanding the intelligence of people; artificial intelligence is a branch of computer science that attempts to understand the nature of intelligence and to produce a new intelligent machine that can react in a manner similar to human intelligence, research in this field including robotics, language recognition, image recognition, natural language processing, and expert systems. Artificial intelligence can simulate the information process of consciousness and thinking of people. Artificial intelligence is also a theory, method, technique, and application system that utilizes a digital computer or digital computer-controlled machine to simulate, extend, and expand human intelligence, sense the environment, acquire knowledge, and use knowledge to obtain optimal results.

Natural language processing (natural language processing, NLP): NLP is a branch of artificial intelligence that is a interdisciplinary of computer science and linguistics, and is often referred to as computational linguistics, and is processed, understood, and applied to human languages (e.g., chinese, english, etc.). Natural language processing includes parsing, semantic analysis, chapter understanding, and the like. Natural language processing is commonly used in the technical fields of machine translation, handwriting and print character recognition, voice recognition and text-to-speech conversion, information intent recognition, information extraction and filtering, text classification and clustering, public opinion analysis and opinion mining, and the like, and relates to data mining, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research, linguistic research related to language calculation, and the like.

Information extraction (Information Extraction, NER): extracting the fact information of the appointed type of entity, relation, event and the like from the natural language text, and forming the text processing technology of the structured data output. Information extraction is a technique for extracting specific information from text data. Text data is made up of specific units, such as sentences, paragraphs, chapters, and text information is made up of small specific units, such as words, phrases, sentences, paragraphs, or a combination of these specific units. The noun phrase, the name of a person, the name of a place, etc. in the extracted text data are all text information extraction, and of course, the information extracted by the text information extraction technology can be various types of information.

Recall: recall is the first stage of the recommendation system, and mainly according to the characteristics of the users and the commodity parts, a part of articles potentially interested by the users are quickly retrieved from a massive article library and then handed to a sorting link.

Entity: refers to something that is distinguishable and independently present. Such as a person, a city, a plant, etc., a commodity, etc. Worldwide everything has a concrete composition of things, which refers to an entity. The entities are the most basic elements in the knowledge graph, and different relationships exist among different entities.

Embedding (embedding): embedding is a vector representation, which means representing an object, which may be a word, or a commodity, or a movie, etc., with a low-dimensional vector; the nature of this casting vector is such that objects corresponding to vectors that are close in distance have similar meanings. Embedding is essentially a mapping from semantic space to vector space, while maintaining the relation of the original samples in the semantic space as much as possible in the vector space, e.g. two words with close semantics are also located closer together in the vector space. The method can be used for encoding the object by using the low-dimensional vector, can also preserve the meaning of the object, is commonly applied to machine learning, and is used for improving the efficiency by encoding the object into a low-dimensional dense vector and then transmitting the low-dimensional dense vector to DNN in the construction process of a machine learning model.

Coding (encoder): the input sequence is converted into a vector of fixed length.

BERT (Bidirectional Encoder Representation from Transformers) model: the BERT model further increases the generalization capability of the word vector model, fully describes character-level, word-level, sentence-level and even inter-sentence relationship characteristics, and is constructed based on a transducer. There are three types of ebedding in BERT, namely Token ebedding, segment Embedding, position Embedding; wherein Token documents are word vectors, the first word is a CLS Token, which can be used for the subsequent classification task; segment Embeddings is used to distinguish between two sentences, because pre-training does not only LM but also classification tasks with two sentences as input; position Embeddings, here the position word vector is not a trigonometric function in transfor, but BERT is learned through training. However, the BERT directly trains a position embedding to keep the position information, randomly initializes a vector at each position, adds model training, finally obtains an empedding containing the position information, and finally selects direct splicing on the position embedding and word empedding combination mode.

Based on the above, the embodiment of the application provides a product recommendation method, a product recommendation device, electronic equipment and a storage medium, which aim to improve the accuracy of product recommendation.

The product recommendation method, the product recommendation device, the electronic equipment and the storage medium provided by the embodiment of the application are specifically described through the following embodiments, and the product recommendation method in the embodiment of the application is described first.

The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use knowledge to obtain optimal results.

Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.

The embodiment of the application provides a product recommendation method, and relates to the technical field of artificial intelligence. The product recommendation method provided by the embodiment of the application can be applied to the terminal, the server side and software running in the terminal or the server side. In some embodiments, the terminal may be a smart phone, tablet, notebook, desktop, etc.; the server side can be configured as an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, and a cloud server for providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs, basic cloud computing services such as big data and artificial intelligent platforms and the like; the software may be an application or the like that implements the product recommendation method, but is not limited to the above form.

The subject application is operational with numerous general purpose or special purpose computer system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like. The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The application may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

Fig. 1 is an optional flowchart of a product recommendation method provided in an embodiment of the present application, where the method in fig. 1 may include, but is not limited to, steps S101 to S107.

Step S101, obtaining target user data, wherein the target user data comprises user basic data and user behavior data of a target user;

step S102, extracting characteristics of target user data to obtain target user characteristics;

step S103, carrying out product recall based on a preset product recommendation model and target user characteristics to obtain an initial recommendation list;

step S104, pushing the initial recommendation list to a target user, and acquiring intention data fed back by the target user according to the initial recommendation list;

step S105, extracting entity characteristics from the intention data to obtain target intention characteristics;

step S106, product rearrangement is carried out on the products to be recommended of the initial recommendation list based on the target intention characteristics, and a target recommendation list is obtained;

step S107, pushing the target recommendation list to the target user.

Step S101 to step S107 illustrated in the embodiment of the present application, by acquiring target user data, the target user data includes user basic data and user behavior data of a target user; the method can obtain the target user characteristics containing important user information more conveniently, so that the target user characteristics can be used for a subsequent recommendation process. And carrying out product recall based on a preset product recommendation model and target user characteristics to obtain an initial recommendation list, and screening and sorting products to be recommended based on the target user characteristics conveniently to obtain the initial recommendation list conforming to the user characteristics of the target user. Pushing the initial recommendation list to a target user, and acquiring intention data fed back by the target user according to the initial recommendation list; the entity characteristic extraction is carried out on the intention data to obtain target intention characteristics, so that the current intention of the target user can be conveniently identified, and the product recommendation can be carried out based on the current intention of the target user. Further, the products to be recommended of the initial recommendation list are rearranged based on the target intention characteristics to obtain a target recommendation list, and the products to be recommended can be conveniently screened and ordered based on the target intention characteristics of the target user to obtain the target recommendation list which accords with the actual intention of the target user. And finally, pushing the target recommendation list to the target user, wherein the method can screen and sort the products to be recommended according to the user characteristics and the intention characteristics of the target user to obtain the target recommendation list which accords with the user characteristics and the current intention of the user, thereby better realizing personalized recommendation of the products and improving the accuracy of product recommendation.

In step S101 of some embodiments, the target user data may be obtained by means of a web crawler, for example, the web crawler performs data crawling on a preset data source to obtain the target user data, where the preset data source includes a user database or a network platform commonly used by a user, and the target user data includes user basic data and user behavior data of the target user, the user basic data includes age, gender, academic, and the like of the user, and the user behavior data includes activity, potential tags, click records, access records, purchase records, and the like of the target user.

Referring to fig. 2, in some embodiments, step S102 may include, but is not limited to, steps S201 to S203:

step S201, extracting features of the basic data of the user to obtain basic features of the user;

step S202, extracting characteristics of user behavior data to obtain user behavior characteristics;

and step S203, carrying out feature fusion on the user behavior features and the user basic features to obtain target user features.

In step S201 of some embodiments, feature extraction may be performed on the user basic data through a preset BERT model or a named entity extraction algorithm, to obtain user basic features. Taking a BERT model as an example, inputting user basic data into a pre-trained BERT model, performing sentence segmentation on the user basic data through the BERT model to obtain a plurality of user basic word segments, performing data conversion on the user basic word segments based on a preset dictionary, converting the user basic word segments into corresponding numbers according to the mapping relation between the numbers and the words in the preset dictionary, and performing coding processing on the numbers corresponding to the user basic word segments based on an encoder in the BERT model to obtain user basic features, wherein the user basic features comprise feature information such as gender, age and the like of a target user.

In step S202 of some embodiments, feature extraction may be performed on the user behavior data through a preset BERT model or based on a named entity extraction algorithm, so as to obtain user behavior features. Taking a BERT model as an example, inputting user behavior data into a pre-trained BERT model, performing sentence segmentation on the user behavior data through the BERT model to obtain a plurality of user behavior word segments, performing data conversion on the user behavior word segments based on a preset dictionary, converting the user behavior word segments into corresponding numbers according to the mapping relation between the numbers and the words in the preset dictionary, and performing coding processing on the numbers corresponding to the user behavior word segments based on an encoder in the BERT model to obtain user behavior characteristics, wherein the user behavior characteristics comprise characteristic information such as click records, purchase records, browsing records and the like of a target user.

In step S203 of some embodiments, the user behavior feature and the user basic feature may be converted into a vector form, and then vector stitching is performed on the user behavior feature and the user basic feature in the vector form, so as to obtain a target user feature, where the target user feature includes user basic feature information and user behavior feature information.

The steps S201 to S203 can obtain the user basic feature and the user behavior feature characterizing the characteristics of the target user more conveniently by extracting the features of the user basic data and the user behavior data, and perform feature fusion on the user basic feature and the user behavior feature to obtain the target user feature containing important user information, so that the target user feature can be used in the subsequent recommendation process, thereby improving the recommendation accuracy.

Referring to fig. 3, in some embodiments, step S103 may include, but is not limited to, steps S301 to S304:

step S301, inputting target user characteristics into a product recommendation model, wherein the product recommendation model comprises an input layer, a coding layer and a prediction layer;

step S302, embedding the target user characteristics based on the input layer to obtain user characteristic embedded vectors;

step S303, coding the user feature embedded vector based on the coding layer to obtain a user feature coding vector;

step S304, product recall is carried out based on the prediction layer, the pre-acquired product features and the user feature code vector, an initial recommendation list is obtained, and the product features are basic features of the products to be recommended.

In step S301 of some embodiments, a preset product recommendation model may be constructed based on a neural network for deep learning or a random forest algorithm, where the product recommendation model includes an input layer, an encoding layer, and a prediction layer, where the input layer is configured to perform embedding processing on input features, map the input features to a vector space with a fixed dimension, the encoding layer is configured to perform encoding processing on the input features in the fixed vector space, obtain important feature information, obtain feature encoding vectors, and the prediction layer is configured to perform recommendation prediction according to the feature encoding vectors, obtain recommendation scores of each candidate product, and generate a recommendation list according to the recommendation scores. Further, the product recommendation model can be formed by aggregating recommendation models of various types of products, for example, specific recommendation models of various products are respectively built according to various products such as deposit, regular, running, public recruitment, private recruitment and the like in the financial field, then a general recommendation model is built based on common characteristics of the series of products, and model aggregation is carried out on the specific recommendation models and the general recommendation models according to preset model weights to obtain the product recommendation model, wherein model structures of the specific recommendation model and the general recommendation model can be consistent with those of the product recommendation model, namely, the specific recommendation model and the general recommendation model can also be built based on a neural network or a random forest algorithm of deep learning and the like, and the specific recommendation model and the general recommendation model all comprise an input layer, a coding layer and a prediction layer. During model training, specific recommendation models of different products are trained based on specific product characteristics of the products, for example, the products to be recommended can be classified into financial products, fund products and the like, and the specific product characteristics of the financial products comprise purchase amount, equity asset proportion, debt asset proportion and the like; unique product characteristics of the fund product include withdrawal rate, affiliated funds company, affiliated funds manager, and the like. Similarly, the generic recommendation model may be trained based on common product features of different products, including but not limited to product type, product yield, product deadline, product exposure, and product browsing. In addition, a commonly used cross entropy function may be used as a loss function when model training is performed.

When the target user feature is input to the product recommendation model, the target user feature may be input to the product recommendation model by a preset code program or script program, or the like.

In step S302 of some embodiments, the target user features are embedded through the input layer, and mapped to a vector space with a fixed dimension, so as to obtain a user feature embedded vector, where the dimension of the vector space may be set according to the actual service requirement without limitation.

In step S303 of some embodiments, the coding layer encodes the user feature embedded vector to obtain important feature information in the user feature embedded vector, so as to obtain a user feature encoded vector.

In step S304 of some embodiments, product recommendation scores are calculated on the basis of the prediction layer for product features and user feature encoding vectors, so as to obtain recommendation scores of each product to be recommended, the recommendation scores are compared with a preset recommendation threshold, products to be recommended, of which the recommendation scores are higher than the recommendation threshold, are screened out, candidate products are obtained, and a series of candidate products are arranged in descending order according to the recommendation scores of the candidate products, so that an initial recommendation list is obtained.

The steps S301 to S304 can screen and sort the products to be recommended based on the characteristics of the target users more conveniently, so that an initial recommendation list conforming to the user characteristics of the target users is obtained, different initial recommendation lists can be built for different target users, the accuracy of recommendation is improved, and personalized recommendation of the products can be realized.

Referring to fig. 4, in some embodiments, step S304 may include, but is not limited to, steps S401 to S403:

step S401, calculating the recommendation score of the product based on the prediction layer for the product features and the user feature code vectors, and obtaining the recommendation score of each product to be recommended;

step S402, screening the products to be recommended based on the recommended score and a preset recommended threshold value to obtain candidate products;

step S403, sorting the candidate products according to the recommendation scores to obtain an initial recommendation list.

In step S401 of some embodiments, the product features are basic features of the product to be recommended, including a product type, a product name, a product term, and basic product parameters, etc., of the product to be recommended, including financial products, funding products, etc. And converting the product characteristics into vector forms to obtain the product characteristic vector of each product to be recommended. And calculating the similarity of the product feature vector and the user feature code vector by adopting a collaborative filtering algorithm, an alternating least square method or a double-tower algorithm and the like to obtain a vector similarity value, and taking the vector similarity value as the recommendation score of each product to be recommended.

In step S402 of some embodiments, the preset recommendation threshold may be set according to the actual requirement, without limitation. Comparing the recommendation score with a preset recommendation threshold value, wherein the larger the recommendation score is, the more the to-be-recommended products meet the current demands of the target users, so that a plurality of to-be-recommended products with recommendation scores higher than the recommendation threshold value are screened out from the to-be-recommended products according to the size relation of the recommendation score and the recommendation threshold value, the screened to-be-recommended products are used as candidate products, and the candidate products are products which can meet the current demands of the target users.

In step S403 of some embodiments, candidate products are arranged in order of the recommendation scores from high to low according to the recommendation score of each candidate product, so as to obtain an initial recommendation list.

The products to be recommended can be conveniently screened and ordered based on the product features and the target user features through the steps S401 to S403, an initial recommendation list conforming to the user features of the target user is obtained, and product recommendation is carried out to the target user according to the initial recommendation list, so that candidate products in the initial recommendation list can meet the current requirements of the target user, and recommendation accuracy is improved.

In step S104 of some embodiments, the initial recommendation list is pushed to the target user by means of wireless communication or wired communication, so that the target user can browse, click or purchase candidate products in the initial recommendation list, and user operation information is generated, the user operation information can characterize interest preference of the target user on the candidate products, for example, the target user can collect candidate products of interest, mark candidate products of no interest, and the target user can input a screening condition according to actual requirements, for example, the screening condition is that the "yield rate is higher than a certain value", and therefore, user operation information is generated by responding to operation behaviors of the target user, and the user operation information includes intention data of the target user. Further, intention data fed back by the target user according to the initial recommendation list is obtained through wireless communication or wired communication, and the intention data can characterize the intention condition of the target user on candidate products in the initial recommendation list.

Referring to fig. 5, in some embodiments, step S105 may include, but is not limited to, steps S501 to S502:

Step S501, traversing a preset intention database, and carrying out correlation calculation on intention data and reference intention characteristics in the intention database to obtain intention correlation;

step S502, screening the reference intention characteristic based on the intention relativity to obtain the target intention characteristic.

In step S501 of some embodiments, the preset intent database contains a variety of reference intent features, for example, including but not limited to product type, investment direction, investment deadline, and investment return, etc., which may be dynamically adjusted according to business needs. Through traversing a preset intention database, carrying out correlation calculation on the intention data and reference intention features in the intention database by utilizing a collaborative filtering algorithm and the like to obtain intention correlation degrees of the intention data and each reference intention feature, wherein the intention correlation degrees can represent the similarity degree of the intention data and the reference intention features.

In step S502 of some embodiments, the degree of similarity of the intent data to the reference intent feature can be characterized due to the intent correlation. If the intention data has higher intention relativity with a certain reference intention feature, the intention data shows that the user intention in the intention data is closer to the reference intention feature, so that the reference intention feature with the highest intention relativity is selected as a target intention feature by comparing the intention relativity of each reference intention feature.

For example, the reference intention feature includes an investment period of 6 months, an investment period of 1 year, and an investment period of 3 years, wherein the intention correlation between the investment period of 6 months and the intention data is 0.3, the intention correlation between the investment period of 1 year and the intention data is 0.6, and the intention correlation between the investment period of 3 years and the intention data is 0.05, and the reference intention feature indicating that the target user tends to select a product with the investment period of 1 year is the target intention feature.

Through the steps S501 to S502, the reference intention characteristic with higher intention relativity with the intention data of the target user can be conveniently screened out from the preset intention database, and the target intention characteristic is obtained, so that the current intention of the target user can be conveniently identified, the product recommendation can be carried out based on the current intention of the target user, and the accuracy of the product recommendation can be effectively improved.

Referring to fig. 6, in some embodiments, step S106 includes, but is not limited to, steps S601 to S602:

step S601, calculating predicted intention data of each product to be recommended in an initial recommendation list according to target intention characteristics;

Step S602, product rearrangement is carried out on the products to be recommended of the initial recommendation list based on the prediction intention data, and a target recommendation list is obtained.

In step S601 of some embodiments, a collaborative filtering algorithm, an alternating least square method, a double-tower algorithm, or the like is used to calculate a similarity between the target intent feature and the product feature of each product to be recommended in the initial recommendation list, so as to obtain a feature similarity value, and the feature similarity value is used as predicted intent data of each product to be recommended.

In step S602 of some embodiments, according to the predicted intention data of each product to be recommended in the initial recommendation list, the products to be recommended in the initial recommendation list are arranged in the order from high to low according to the predicted intention data, so as to obtain a target recommendation list.

Through the steps S601 to S602, the products to be recommended can be conveniently screened and ordered based on the target intention characteristics of the target user, a target recommendation list conforming to the actual intention of the target user is obtained, and product recommendation is performed to the target user according to the target recommendation list, so that the candidate products in the target recommendation list can meet the current requirements of the target user, and the recommendation accuracy is improved.

Referring to fig. 7, in some embodiments, step S601 may include, but is not limited to, steps S701 to S702:

step S701, calculating the intention probability vector of each product to be recommended in the initial recommendation list based on a preset function and target intention characteristics;

step S702, obtaining the predicted intention data according to the intention probability vector.

In step S701 of some embodiments, the preset function may be a probability function such as a softmax function, and taking the softmax function as an example, a probability distribution situation of the target intention feature on each product to be recommended is created through the softmax function, so as to obtain an intention probability vector of each product to be recommended in the initial recommendation list, and the intention tendency of the target user to each product to be recommended is reflected through the intention probability vector, that is, the larger the intention probability vector of a certain product to be recommended is, the larger the intention tendency of the target user to the certain product to be recommended is.

In step S702 of some embodiments, the intention probability vector of each product to be recommended in the initial recommendation list is used as predicted intention data of each product to be recommended.

Through the steps S701 to S702, the intention probability vector of each product to be recommended in the initial recommendation list can be calculated more conveniently, and the intention condition of the target user for each product to be recommended is determined according to the intention probability vector, so that the product to be recommended which accords with the current intention of the target user can be selected for recommendation based on the intention condition of the target user, and the accuracy of product recommendation is improved.

In step S107 of some embodiments, the target recommendation list is directly pushed to the target user, or a product earlier in the target recommendation list is selected to be pushed to the target user, so that the communication cost is reduced while personalized recommendation is realized.

According to the product recommendation method, target user data are obtained, wherein the target user data comprise user basic data and user behavior data of a target user; the method can obtain the target user characteristics containing important user information more conveniently, so that the target user characteristics can be used for a subsequent recommendation process. And carrying out product recall based on a preset product recommendation model and target user characteristics to obtain an initial recommendation list, and screening and sorting products to be recommended based on the target user characteristics conveniently to obtain the initial recommendation list conforming to the user characteristics of the target user. Pushing the initial recommendation list to a target user, and acquiring intention data fed back by the target user according to the initial recommendation list; the entity characteristic extraction is carried out on the intention data to obtain target intention characteristics, so that the current intention of the target user can be conveniently identified, and the product recommendation can be carried out based on the current intention of the target user. Further, the products to be recommended of the initial recommendation list are rearranged based on the target intention characteristics to obtain a target recommendation list, and the products to be recommended can be conveniently screened and ordered based on the target intention characteristics of the target user to obtain the target recommendation list which accords with the actual intention of the target user. And finally, pushing the target recommendation list to the target user, wherein the method can screen and sort the products to be recommended according to the user characteristics and the intention characteristics of the target user to obtain the target recommendation list which accords with the user characteristics and the current intention of the user, thereby better realizing personalized recommendation of the products and improving the accuracy of product recommendation.

Referring to fig. 8, an embodiment of the present application further provides a product recommendation device, which may implement the product recommendation method, where the device includes:

a data acquisition module 801, configured to acquire target user data, where the target user data includes user basic data and user behavior data of a target user;

a feature extraction module 802, configured to perform feature extraction on the target user data to obtain a target user feature;

the product recall module 803 is configured to perform product recall based on a preset product recommendation model and target user characteristics, so as to obtain an initial recommendation list;

the preliminary recommendation module 804 is configured to push the initial recommendation list to the target user, and obtain intention data fed back by the target user according to the initial recommendation list;

the entity feature extraction module 805 is configured to perform entity feature extraction on the intention data to obtain a target intention feature;

a rearrangement module 806, configured to perform product rearrangement on the product to be recommended in the initial recommendation list based on the target intention feature, to obtain a target recommendation list;

the target recommendation module 807 is configured to push the target recommendation list to the target user.

In some embodiments, feature extraction module 802 includes:

the first feature extraction unit is used for carrying out feature extraction on the basic data of the user to obtain basic features of the user;

The second feature extraction unit is used for carrying out feature extraction on the user behavior data to obtain user behavior features;

and the feature fusion unit is used for carrying out feature fusion on the user behavior features and the user basic features to obtain target user features.

In some embodiments, product recall module 803 includes:

the input unit is used for inputting the target user characteristics into a product recommendation model, and the product recommendation model comprises an input layer, a coding layer and a prediction layer;

the embedding unit is used for carrying out embedding processing on the target user characteristics based on the input layer to obtain user characteristic embedding vectors;

the coding unit is used for coding the user characteristic embedded vector based on the coding layer to obtain a user characteristic coding vector;

and the recall unit is used for carrying out product recall based on the prediction layer, the pre-acquired product characteristics and the user characteristic code vector to obtain an initial recommendation list, wherein the product characteristics are basic characteristics of the product to be recommended.

In some embodiments, the recall unit comprises:

the score calculating subunit is used for calculating the product recommendation score of the product feature and the user feature coding vector based on the prediction layer to obtain the recommendation score of each product to be recommended;

The screening subunit is used for screening the products to be recommended based on the recommendation score and a preset recommendation threshold value to obtain candidate products;

and the sequencing subunit is used for sequencing the candidate products according to the recommendation scores to obtain an initial recommendation list.

In some embodiments, the entity-feature extraction module 805 includes:

the correlation calculation unit is used for traversing a preset intention database, and carrying out correlation calculation on the intention data and reference intention characteristics in the intention database to obtain intention correlation;

and the feature screening unit is used for screening the reference intention features based on the intention correlation degree to obtain target intention features.

In some embodiments, the rearrangement module 806 includes:

the intention calculation unit is used for calculating the predicted intention data of each product to be recommended in the initial recommendation list according to the target intention characteristics;

and the rearrangement unit is used for rearranging the products to be recommended of the initial recommendation list based on the prediction intention data to obtain a target recommendation list.

In some embodiments, the intent calculation unit comprises:

the probability calculation subunit is used for calculating the intention probability vector of each product to be recommended in the initial recommendation list based on a preset function and target intention characteristics;

And the data generation subunit is used for obtaining the predicted intention data according to the intention probability vector.

The specific implementation of the product recommendation device is basically the same as the specific embodiment of the product recommendation method, and is not described herein.

The embodiment of the application also provides electronic equipment, which comprises: the computer-readable storage medium comprises a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for realizing connection communication between the processor and the memory, wherein the program realizes the product recommendation method when being executed by the processor. The electronic equipment can be any intelligent terminal including a tablet personal computer, a vehicle-mounted computer and the like.

Referring to fig. 9, fig. 9 illustrates a hardware structure of an electronic device according to another embodiment, the electronic device includes:

the processor 901 may be implemented by a general purpose CPU (central processing unit), a microprocessor, an application specific integrated circuit (ApplicationSpecificIntegratedCircuit, ASIC), or one or more integrated circuits, etc. for executing related programs to implement the technical solutions provided by the embodiments of the present application;

the memory 902 may be implemented in the form of read-only memory (ReadOnlyMemory, ROM), static storage, dynamic storage, or random access memory (RandomAccessMemory, RAM). The memory 902 may store an operating system and other application programs, and when the technical solutions provided in the embodiments of the present disclosure are implemented by software or firmware, relevant program codes are stored in the memory 902, and the processor 901 invokes the product recommendation method to execute the embodiments of the present disclosure;

An input/output interface 903 for inputting and outputting information;

the communication interface 904 is configured to implement communication interaction between the device and other devices, and may implement communication in a wired manner (e.g. USB, network cable, etc.), or may implement communication in a wireless manner (e.g. mobile network, WIFI, bluetooth, etc.);

a bus 905 that transfers information between the various components of the device (e.g., the processor 901, the memory 902, the input/output interface 903, and the communication interface 904);

wherein the processor 901, the memory 902, the input/output interface 903 and the communication interface 904 are communicatively coupled to each other within the device via a bus 905.

The embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to realize the product recommendation method.

The memory, as a non-transitory computer readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory remotely located relative to the processor, the remote memory being connectable to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The product recommending method, the product recommending device, the electronic equipment and the computer readable storage medium provided by the embodiment of the application are characterized in that target user data are obtained, wherein the target user data comprise user basic data and user behavior data of a target user; the method can obtain the target user characteristics containing important user information more conveniently, so that the target user characteristics can be used for a subsequent recommendation process. And carrying out product recall based on a preset product recommendation model and target user characteristics to obtain an initial recommendation list, and screening and sorting products to be recommended based on the target user characteristics conveniently to obtain the initial recommendation list conforming to the user characteristics of the target user. Pushing the initial recommendation list to a target user, and acquiring intention data fed back by the target user according to the initial recommendation list; the entity characteristic extraction is carried out on the intention data to obtain target intention characteristics, so that the current intention of the target user can be conveniently identified, and the product recommendation can be carried out based on the current intention of the target user. Further, the products to be recommended of the initial recommendation list are rearranged based on the target intention characteristics to obtain a target recommendation list, and the products to be recommended can be conveniently screened and ordered based on the target intention characteristics of the target user to obtain the target recommendation list which accords with the actual intention of the target user. And finally, pushing the target recommendation list to the target user, wherein the method can screen and sort the products to be recommended according to the user characteristics and the intention characteristics of the target user to obtain the target recommendation list which accords with the user characteristics and the current intention of the user, thereby better realizing personalized recommendation of the products and improving the accuracy of product recommendation.

The embodiments described in the embodiments of the present application are for more clearly describing the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application, and as those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

It will be appreciated by those skilled in the art that the solutions shown in fig. 1-7 are not limiting to embodiments of the present application and may include more or fewer steps than shown, or certain steps may be combined, or different steps.

The above described apparatus embodiments are merely illustrative, wherein the units illustrated as separate components may or may not be physically separate, i.e. may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Those of ordinary skill in the art will appreciate that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof.

The terms "first," "second," "third," "fourth," and the like in the description of the present application and in the above-described figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the present application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It should be understood that in this application, "at least one" means one or more, and "a plurality" means two or more. "and/or" for describing the association relationship of the association object, the representation may have three relationships, for example, "a and/or B" may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the above-described division of units is merely a logical function division, and there may be another division manner in actual implementation, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.

The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including multiple instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of the various embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing a program.

Preferred embodiments of the present application are described above with reference to the accompanying drawings, and thus do not limit the scope of the claims of the embodiments of the present application. Any modifications, equivalent substitutions and improvements made by those skilled in the art without departing from the scope and spirit of the embodiments of the present application shall fall within the scope of the claims of the embodiments of the present application.

Claims

1. A method of product recommendation, the method comprising:

pushing the target recommendation list to the target user.

2. The product recommendation method according to claim 1, wherein the feature extraction of the target user data to obtain target user features includes:

3. The product recommendation method according to claim 1, wherein the product recall is performed based on the preset product recommendation model and the target user feature to obtain an initial recommendation list, comprising:

4. The product recommendation method according to claim 3, wherein said performing product recall based on said prediction layer, pre-acquired product features, said user feature encoding vector, to obtain said initial recommendation list, comprises:

5. The product recommendation method according to claim 1, wherein the extracting the physical characteristics of the intention data to obtain target intention characteristics includes:

6. The product recommendation method according to claim 1, wherein the product rearrangement of the products to be recommended of the initial recommendation list based on the target intention feature to obtain a target recommendation list comprises:

7. The product recommendation method according to claim 6, wherein calculating predicted intent data for each product to be recommended in the initial recommendation list based on the target intent features comprises:

8. A product recommendation device, the device comprising:

9. An electronic device comprising a memory storing a computer program and a processor that when executing the computer program implements the product recommendation method of any one of claims 1 to 7.

10. A computer readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the product recommendation method of any one of claims 1 to 7.