Intelligent recommendation system and method based on artificial intelligence and electronic equipment

Abstract

The invention provides an intelligent recommendation system and method based on artificial intelligence, and electronic equipment. The system comprises a data acquisition module used for collecting user information from a plurality of data sources; the feature extraction module is used for extracting valuable features from the text data by utilizing a natural language processing technology and a machine learning algorithm; the model training module is used for training the extracted features by adopting a deep neural network model so as to establish a user behavior prediction model; and the recommendation generation module is used for generating personalized recommendation content according to a result of the prediction model and in combination with preference and historical behaviors of the user. According to the system, accurate prediction and personalized recommendation of user behaviors are realized through a deep learning algorithm and a big data analysis technology.

Description

Technical Field

[0001] The present invention relates to the field of computer software technology, and more specifically, to an intelligent recommendation system, method and electronic device based on artificial intelligence. Background Art

[0002] With the rapid development of the Internet and mobile communication technologies, users' demand for personalized services is increasing. Although traditional recommendation systems can provide certain personalized services, their effectiveness is often limited by factors such as data volume and algorithm complexity. Therefore, how to improve the accuracy of recommendation systems and user experience has become the focus of current research. Summary of the invention

[0003] In view of the technical problems existing in the prior art, the present invention provides an intelligent recommendation system, method and electronic device based on artificial intelligence, which realizes accurate prediction of user behavior and personalized recommendation through deep learning algorithm and big data analysis technology.

[0004] According to a first aspect of the present invention, there is provided an artificial intelligence-based intelligent recommendation system, comprising:

[0005] Data collection module: used to collect user information from multiple data sources;

[0006] Feature extraction module: uses natural language processing technology and machine learning algorithms to extract valuable features from text data;

[0007] Model training module: uses a deep neural network model to train the extracted features to establish a user behavior prediction model;

[0008] The recommendation generation module is used to generate personalized recommendation content based on the results of the prediction model and combined with the user's preferences and historical behaviors.

[0009] Based on the above technical solution, the present invention can also make the following improvements.

[0010] Optionally, the data collection module interacts with multiple data sources through an API interface to obtain user information in real time; the data collection module includes a user behavior log collection sub-module and a social media data collection sub-module, which internally includes creating AI models, training AI models, sharing AI models and deploying AI models.

[0011] Optionally, the data collected by the data collection module is divided into structured data and unstructured data, including basic information of the user's age, gender, occupation, interests and hobbies, and geographic location, as well as behavioral data of the user's search keywords, click behaviors, and shopping cart contents.

[0012] Optionally, the feature extraction module adopts word vector representation and semantic analysis technology to improve the accuracy of feature extraction; the feature extraction module includes a sentiment analysis submodule and a topic classification submodule.

[0013] Optionally, the feature extraction module extracts useful feature vectors from the original data for subsequent model training and recommendation generation; wherein the features include: user features, item features, and behavior features; feature extraction methods include PCA, LDA, SVM, and random forest, and feature selection should be based on the characteristics of specific problems and data.

[0014] Optionally, the model training module adopts a method combining convolutional neural network and recurrent neural network to improve the generalization ability of the model; the model training module includes a regularization processing submodule and a hyperparameter tuning submodule.

[0015] Optionally, the recommendation generation module adopts a collaborative filtering algorithm and a content filtering algorithm to further optimize the recommendation results, and the recommendation generation module includes a user feedback analysis submodule and a dynamic adjustment submodule.

[0016] Optionally, the recommendation generation module generates final recommendation results based on the trained model. The recommendation results need to be filtered and ranked to ensure quality. The final recommendation results need to meet the needs of users and be novel and need to combine the explicit preferences and potential preferences of users.

[0017] According to a second aspect of the present invention, there is provided an intelligent recommendation method based on artificial intelligence, comprising the following steps:

[0018] The system interacts with multiple data sources through API interfaces and collects user information from multiple data sources;

[0019] The system uses natural language processing technology and machine learning algorithms to extract valuable features from text data;

[0020] Use a deep neural network model to train the extracted features and build a user behavior prediction model;

[0021] Based on the results of the prediction model, personalized recommendations are generated in combination with user preferences and historical behaviors.

[0022] According to a third aspect of the present invention, there is provided an electronic device comprising a memory and a processor, wherein the processor is configured to implement the steps of an intelligent recommendation method based on artificial intelligence when executing a computer program stored in the memory.

[0023] Technical effects and advantages of the present invention:

[0024] The present invention provides an intelligent recommendation system, method and electronic device based on artificial intelligence. The system realizes accurate prediction of user behavior and personalized recommendation through deep learning algorithm and big data analysis technology. The system not only improves the accuracy of recommendation and user experience, but also has good application prospects and market potential.

[0025] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A schematic diagram of the structure of an intelligent recommendation system based on artificial intelligence provided by an embodiment of the present invention;

[0027] Figure 2 A schematic diagram of the structure of a data acquisition module provided in an embodiment of the present invention;

[0028] Figure 3 A schematic diagram of the structure of a model training module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] It is understandable that based on the defects in the background technology, the embodiment of the present invention proposes an intelligent recommendation system based on artificial intelligence, specifically Figure 1 As shown, including:

[0031] Data collection module, used to collect user information from multiple data sources;

[0032] Feature extraction module, which uses natural language processing (NLP) technology and machine learning algorithms to extract valuable features from text data;

[0033] The model training module uses a deep neural network (DNN) model to train the extracted features to establish a user behavior prediction model.

[0034] The recommendation generation module is used to generate personalized recommendation content based on the results of the prediction model and combined with the user's preferences and historical behaviors.

[0035] Furthermore, if Figure 2 As shown, the data collection module interacts with multiple data sources through an API interface to obtain user information in real time; the data collection module includes a user behavior log collection submodule and a social media data collection submodule, including creating an AI model, training an AI model, sharing an AI model, and deploying an AI model.

[0036] The AI ​​model is created to:

[0037] Demand analysis: First, clarify the application scenarios and target users of the intelligent recommendation system, and understand the types of problems that need to be solved (such as product recommendations, content recommendations, etc.).

[0038] Data collection: Collect relevant data as needed, including but not limited to user behavior data, item information, context information, etc.

[0039] Feature engineering: Process and transform raw data to extract feature vectors that can reflect user preferences or item characteristics.

[0040] Select algorithm: Choose the appropriate machine learning or deep learning algorithm based on the characteristics of the problem. Commonly used recommendation algorithms include collaborative filtering, content-based recommendation, matrix decomposition, neural network, etc.

[0041] Build model architecture: Design and build the AI ​​model structure, such as defining the number of neural network layers, number of nodes, activation functions, etc.

[0042] The training AI model is used to:

[0043] Data preprocessing: cleaning data, handling missing values, standardizing / normalizing numerical features, etc.

[0044] Divide the dataset: Divide the data into training, validation, and test sets to evaluate the model performance.

[0045] Set parameters: Determine hyperparameters such as learning rate, batch size, number of iterations, etc.

[0046] Training process: Use the training set to adjust the model parameters so that the model can minimize the loss function.

[0047] Evaluation and tuning: Use the validation set to evaluate model performance and optimize the model through cross-validation, grid search, etc.

[0048] Test model: Check the performance of the final model on an independent test set to ensure its generalization ability.

[0049] The shared AI model is used to:

[0050] Model saving: Save the trained model in a file format for easy transmission and loading. Common saving formats include HDF5, TensorFlow SavedModel, ONNX, etc.

[0051] Documentation: Provide detailed API documentation and technical descriptions to help others understand how to use the model.

[0052] Version control: Use Git or other tools to manage code and model versions to facilitate tracking changes and collaborative development.

[0053] Release channel: You can choose a public platform (such as GitHub, Kaggle, Model Zoo) or a private repository to share the model.

[0054] The deployed AI model is used to:

[0055] Environment preparation: Configure the server or cloud service environment and install necessary dependent libraries and frameworks.

[0056] Model loading: Load the trained model from the storage location into the memory.

[0057] Interface development: Create interfaces such as RESTful API or gRPC for the model to facilitate front-end application calls.

[0058] Performance optimization: Considering factors such as model inference speed and resource usage, operations such as quantization and pruning may be required.

[0059] Monitoring and maintenance: Establish logging and error reporting mechanisms, regularly check model performance and update the model.

[0060] In the embodiment of the present invention, the system interacts with multiple data sources through an API interface to obtain the user's basic information, browsing history, and purchase history data in real time. The amount of data may reach TB or PB level, and the data types are diverse, including structured data and unstructured data. These data include but are not limited to the user's age, gender, occupation, hobbies, basic information of geographic location, as well as the user's search keywords, click behavior, shopping cart content and other behavioral data.

[0061] The data collection module collects user behavior data, product information and interaction data, and the data source can be a database, cache or external data source. It can use crawler technology to crawl data from the website, use API interface to obtain real-time data and store the data in a database or data lake for subsequent processing.

[0062] Furthermore, the feature extraction module adopts word vector representation and semantic analysis technology to improve the accuracy of feature extraction; the feature extraction module includes a sentiment analysis submodule and a topic classification submodule.

[0063] The feature extraction module extracts useful feature vectors from the raw data for subsequent model training and recommendation generation. Features include: user features, item features, and behavior features. Among them, commonly used feature extraction methods include PCA, LDA, SVM, and random forest. Feature selection should be based on the characteristics of specific problems and data.

[0064] For specific extraction, word vector models and sentence vector models can be used to embed text data or locality sensitive hashing (LSH) models can be used to calculate the similarity of features.

[0065] Furthermore, the model training module adopts a method combining convolutional neural network (CNN) and recurrent neural network (RNN) to improve the generalization ability of the model; the model training module includes a regularization processing submodule and a hyperparameter tuning submodule.

[0066] like Figure 3 As shown, the model training module uses a method combining convolutional neural network (CNN) and recurrent neural network (RNN) to generate common purchase view, social friend view, and user-item view based on user behavior, hobbies, and items; model training includes the following parts:

[0067] 1). Data collection and integration:

[0068] Multi-source data access: Connect to different data sources through API interfaces, such as social media platforms, e-commerce platforms, content websites, etc. These data sources can provide users' browsing history, purchase history, comments, likes, shares and other information.

[0069] Data cleaning and preprocessing: Clean the acquired data to remove noise and invalid information to ensure data quality and consistency. At the same time, preprocess the text data, including word segmentation, stop word removal, stemming or word form restoration, etc., to prepare for subsequent feature extraction.

[0070] 2). Feature extraction and engineering

[0071] Text feature extraction: Apply NLP techniques, such as TF-IDF, Word2Vec, BERT and other pre-trained language models, to convert text data into numerical feature vectors to capture the meaning of words and their contextual relationships.

[0072] User behavior feature construction: Based on the user's interactive behavior, construct features that can reflect the user's interests and preferences, such as click-through rate, dwell time, conversion rate, etc.

[0073] 3). Construction and training of deep neural network models

[0074] Model selection and design: Choose a DNN architecture suitable for recommendation scenarios, such as convolutional neural network (CNN), recurrent neural network (RNN), long short-term memory network (LSTM), gated recurrent unit (GRU), or a more complex Transformer architecture. For serialized user behavior, RNN and its variants are particularly useful; for static features, you can consider using a fully connected layer or a multi-layer perceptron (MLP).

[0075] Model training: Use the collected data set to train the DNN model, adjust the weights to minimize the loss function, and the optimization goal is to improve the prediction accuracy. During the training process, strategies such as cross-validation and early stopping can be used to prevent overfitting.

[0076] 4) User behavior prediction

[0077] Prediction model deployment: Once the model is trained and evaluated to achieve satisfactory performance indicators, it can be deployed to the production environment. At this point, the system can make predictions based on new user data or updated old data.

[0078] Behavior prediction: Use the trained DNN model to predict the user's future behavior, such as predicting the items the user may be interested in, the activities the user may participate in, etc.

[0079] 5). Personalized recommendation generation

[0080] Recommendation strategy formulation: Combine the prediction results with the user's preferences and historical behavior to determine the recommendation strategy. This may involve different methods such as sorting, filtering, and mixing, in order to select the content that best meets the user's needs.

[0081] Recommended content presentation: The last step is to present the selected recommended items to the user in an appropriate form, such as a list, card-style layout, etc., and continuously monitor user feedback to further improve the recommendation effect.

[0082] 6). System iteration and optimization

[0083] Continuous learning: As new data arrives and user preferences change, model parameters are regularly retrained or updated online to keep the system adaptable and fresh.

[0084] Performance monitoring: Set key performance indicators (KPIs) to measure the effectiveness of the recommendation system, such as click-through rate, conversion rate, user satisfaction, etc., and adjust the algorithm and recommendation logic accordingly.

[0085] The model training module uses the extracted features to build and train a recommendation model, which may be a classifier, a regressor, or other machine learning algorithms.

[0086] During the model training process, a large data set is used to validate the model performance and continuously improve its accuracy, which may involve cross-validation to evaluate the performance of different models on the test set.

[0087] The specific implementation includes: using traditional machine learning methods such as GBM tree model or deep learning methods such as embedded methods, adjusting parameters during training to optimize model performance, and performing continuous fine-tuning to ensure that the model can adapt to new environmental changes.

[0088] Furthermore, the recommendation generation module adopts collaborative filtering algorithm and content filtering algorithm to further optimize the recommendation results. The recommendation generation module includes a user feedback analysis submodule and a dynamic adjustment submodule.

[0089] The recommendation generation module generates the final recommendation results based on the trained model. The recommendation results need to be filtered and ranked to ensure quality. The final recommendation results must meet the needs of users and be novel. The recommendation results usually combine the explicit preferences and potential preferences of users.

[0090] The specific implementation includes: using advanced technologies such as generative adversarial networks (GAN) to improve the reliability of the recommendation system, combining the outputs of the content module and the collaboration module for comprehensive scoring, and updating the recommendation list in real time to reflect the latest user behaviors and preferences.

[0091] In summary, the intelligent recommendation system described in the embodiment of the present invention realizes accurate prediction of user behavior and personalized recommendation through deep learning algorithms and big data analysis technology.

[0092] According to a second aspect of the present invention, the present invention provides an artificial intelligence-based intelligent recommendation method, which is used in the artificial intelligence-based intelligent recommendation system described above, and the method comprises the following steps:

[0093] The system interacts with multiple data sources through API interfaces and collects user information from multiple data sources;

[0094] The system uses natural language processing technology and machine learning algorithms to extract valuable features from text data;

[0095] Use a deep neural network model to train the extracted features and build a user behavior prediction model;

[0096] Based on the results of the prediction model, personalized recommendations are generated in combination with user preferences and historical behaviors.

[0097] Furthermore, the specific implementation method specifically includes the following steps:

[0098] The system first interacts with multiple data sources through API interfaces to obtain real-time user basic information, browsing history, purchase history and other data. These data include but are not limited to basic information such as age, gender, occupation, hobbies, geographic location, as well as user search keywords, click behavior, shopping cart content and other behavioral data.

[0099] The system uses natural language processing technology and machine learning algorithms to extract valuable features from text data. For example, sentiment analysis technology can be used to determine the user's emotional inclination towards a product or service; topic classification technology can be used to identify the user's main areas of interest.

[0100] The system uses a deep neural network model to train the extracted features to build a user behavior prediction model. The model can predict the user's future behavior trends based on their interest preferences and historical behaviors. In order to improve the generalization ability of the model, the system also uses regularization processing and hyperparameter tuning technology.

[0101] The system generates personalized recommendations based on the results of the prediction model, combined with the user's preferences and historical behavior. In addition, the system also uses collaborative filtering algorithms and content filtering algorithms to further optimize the recommendation results. For example, through the collaborative filtering algorithm, the system can recommend products that similar users like based on the behavior data of other users; through the content filtering algorithm, the system can recommend products that match the user's interests based on the content attributes of the products.

[0102] According to a third aspect of the present invention, an embodiment of the present invention provides an electronic device, the electronic device comprising: a processor, a communications interface, a memory, and a communication bus, wherein the processor, the communications interface, and the memory communicate with each other via the communication bus. The processor can call logic instructions in the memory to execute the implementation steps of an intelligent recommendation method based on artificial intelligence.

[0103] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0104] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0105] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0106] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0107] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent recommendation system based on artificial intelligence, characterized in that: include: Data collection module: used to collect user information from multiple data sources; Feature extraction module: uses natural language processing technology and machine learning algorithms to extract valuable features from text data; Model training module: uses a deep neural network model to train the extracted features to establish a user behavior prediction model; The recommendation generation module is used to generate personalized recommendation content based on the results of the prediction model and combined with the user's preferences and historical behaviors.

2. The intelligent recommendation system based on artificial intelligence according to claim 1, characterized in that: The data collection module interacts with multiple data sources through an API interface to obtain user information in real time; the data collection module includes a user behavior log collection submodule and a social media data collection submodule, which internally includes creating AI models, training AI models, sharing AI models and deploying AI models.

3. The intelligent recommendation system based on artificial intelligence according to claim 2, characterized in that: The data collected by the data collection module is divided into structured data and unstructured data, including the user's annual basic information and behavior data.

4. The intelligent recommendation system based on artificial intelligence according to claim 1, characterized in that: The feature extraction module adopts word vector representation and semantic analysis technology to improve the accuracy of feature extraction; the feature extraction module includes a sentiment analysis submodule and a topic classification submodule.

5. The intelligent recommendation system based on artificial intelligence according to claim 4, characterized in that: The feature extraction module extracts feature vectors from the original data for subsequent model training and recommendation generation; the extracted features include: user features, item features and behavior features; the feature extraction method includes using PCA, LDA, SVM or random forest model, and feature selection should be based on the characteristics of specific problems and data.

6. The intelligent recommendation system based on artificial intelligence according to claim 1, characterized in that: The model training module combines convolutional neural networks and recurrent neural networks to improve the generalization ability of the model; the model training module includes a regularization processing submodule and a hyperparameter tuning submodule.

7. The intelligent recommendation system based on artificial intelligence according to claim 1, characterized in that: The recommendation generation module adopts collaborative filtering algorithm and content filtering algorithm to further optimize the recommendation results. The recommendation generation module includes a user feedback analysis submodule and a dynamic adjustment submodule.

8. The intelligent recommendation system based on artificial intelligence according to claim 7, characterized in that: The recommendation generation module generates the final recommendation results based on the trained model. The recommendation results need to be filtered and ranked to ensure quality. The final recommendation results must meet the needs of users and be novel and must combine the users' explicit preferences and potential preferences.

9. An artificial intelligence-based intelligent recommendation method, used in an artificial intelligence-based intelligent recommendation system according to any one of claims 1 to 8, characterized in that: The following steps are involved: The system interacts with multiple data sources through API interfaces and collects user information from multiple data sources; The system uses natural language processing technology and machine learning algorithms to extract valuable features from text data; Use a deep neural network model to train the extracted features and build a user behavior prediction model; Based on the results of the prediction model, personalized recommendations are generated in combination with user preferences and historical behaviors.

10. An electronic device, characterized in that: It includes a memory and a processor, and the processor is used to implement the intelligent recommendation method based on artificial intelligence as claimed in claim 9 when executing the computer program stored in the memory.

Images