Customized gift design scheme intelligent recommendation method and system based on enterprise culture

Abstract

The invention relates to the technical field of intelligent recommendation of design schemes, and particularly discloses an intelligent recommendation method and system for customized gift design schemes based on enterprise culture, and the method comprises the steps: carrying out the semantic analysis of gift customization demands of enterprises through employing a natural language processing technology based on deep learning; the method comprises the following steps of: extracting semantic feature representation of enterprise gift customization requirements, acquiring supplementary description information of enterprise culture from official websites and social media of enterprises, and performing rapid semantic query interaction fusion on the enterprise gift customization requirements and the supplementary description information of the enterprise culture; the semantic feature expression ability of enterprise gift customization requirements is enhanced, and intelligent screening and recommendation of alternative gifts are carried out on the basis of the semantic feature expression ability. According to the method and the device, accurate recommendation of customized gifts can be realized, the gifts are ensured to meet actual requirements of enterprises, and culture and value views of the enterprises can be effectively transmitted.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent recommendation of design solutions, and more specifically, to a method and system for intelligently recommending customized gift design solutions based on corporate culture. Background Art

[0002] In today's highly competitive business environment, companies often express gratitude and appreciation by giving customized gifts to maintain and strengthen relationships with customers, partners, and employees. Customized gifts must not only be practical but, more importantly, reflect the giver's corporate culture and values, leaving a lasting impression on the recipient, thereby strengthening customer relationships, enhancing brand image, and consolidating corporate culture.

[0003] However, most of the gift recommendation systems currently available on the market rely on simple matching recommendations based on user-entered keyword searches or browsing history. They lack a deep understanding of corporate culture and specific gift customization needs, which can easily lead to inaccurate recommendation results and fail to truly meet the cultural connotations and personalized requirements of companies for gifts.

[0004] Therefore, we look forward to an optimized intelligent recommendation method and system for customized gift design solutions based on corporate culture. Summary of the Invention

[0005] This application provides an intelligent recommendation method and system for customized gift design solutions based on corporate culture. In this way, it can achieve accurate recommendation of customized gifts, ensuring that the gifts not only meet the actual needs of the enterprise, but also can effectively convey the enterprise's culture and values.

[0006] First, a method for intelligently recommending customized gift design solutions based on corporate culture is provided, including:

[0007] Obtaining corporate gift customization requirements input by a corporate user object, and obtaining supplementary corporate culture description information from the corporate user object's official website and social media, wherein the corporate gift customization requirements include the cultural values ​​to be conveyed, the budget range, and the purpose of the gift;

[0008] Performing semantic understanding on the corporate gift customization demand to obtain a semantic coding vector of the corporate gift customization demand;

[0009] Performing sentence-granularity semantic feature extraction on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity semantic encoding vectors;

[0010] Based on the sequence of the corporate culture supplement sentence granularity semantic coding vectors, the corporate gift customization demand semantic coding vector is quickly semantically matched and enhanced to obtain a corporate gift customization demand semantic enhancement coding vector;

[0011] Obtain gift characteristic data of the first candidate gift;

[0012] Based on the semantic similarity between the gift characteristic data of the first candidate gift and the semantic enhancement coding vector of the corporate gift customization demand, it is determined whether to use the first candidate gift as a recommendation.

[0013] Secondly, we provide an intelligent recommendation system for customized gift design solutions based on corporate culture, including:

[0014] An enterprise data acquisition module is used to obtain enterprise gift customization requirements input by enterprise user objects and obtain supplementary description information of corporate culture from the official website and social media of the enterprise user objects, wherein the enterprise gift customization requirements include the cultural values ​​to be conveyed, the budget range, and the purpose of the gift;

[0015] A corporate gift customization demand semantic understanding module is used to perform semantic understanding on the corporate gift customization demand to obtain a semantic coding vector of the corporate gift customization demand;

[0016] A corporate culture semantic feature extraction module is used to extract sentence-level semantic features of the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-level semantic encoding vectors;

[0017] A corporate gift customization demand semantic enhancement module is used to perform rapid semantic matching enhancement on the corporate gift customization demand semantic encoding vector based on the sequence of corporate culture supplementary sentence granularity semantic encoding vectors to obtain a corporate gift customization demand semantic enhancement encoding vector;

[0018] A gift characteristic data acquisition module, used to acquire gift characteristic data of the first candidate gift;

[0019] The recommendation scheme determination module is used to determine whether to use the first candidate gift as a recommendation scheme based on the semantic similarity between the gift characteristic data of the first candidate gift and the semantic enhancement coding vector of the corporate gift customization demand.

[0020] Compared with the prior art, this application has at least the following technical effects:

[0021] This application provides a method and system for intelligently recommending customized gift design solutions based on corporate culture. This method uses deep learning-based natural language processing technology to perform semantic analysis on a company's gift customization needs, extracting the semantic feature representation of the company's gift customization needs. At the same time, it obtains supplementary descriptive information about the company's culture from the company's official website and social media. This method enhances the semantic feature expression capability of the company's gift customization needs by quickly and interactively integrating the company's gift customization needs and the supplementary descriptive information about the company's culture. On this basis, it intelligently screens and recommends alternative gifts. In this way, accurate recommendations for customized gifts can be achieved, ensuring that the gifts not only meet the company's actual needs but also effectively convey the company's culture and values. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings of the embodiments of the present application. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0023] Figure 1 This is a schematic flowchart of the intelligent recommendation method for customized gift design solutions based on corporate culture in an embodiment of the present application.

[0024] Figure 2 This is a data flow diagram of the intelligent recommendation method for customized gift design solutions based on corporate culture in an embodiment of the present application.

[0025] Figure 3 This is a schematic flowchart of extracting sentence-granularity semantic features of the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity semantic encoding vectors in the intelligent recommendation method for customized gift design solutions based on corporate culture in an embodiment of the present application.

[0026] Figure 4 In the intelligent recommendation method for customized gift design solutions based on corporate culture in an embodiment of the present application, based on the sequence of the granular semantic coding vectors of the corporate culture supplementary sentences, the corporate gift customization demand semantic coding vector is quickly semantically matched and enhanced to obtain a schematic flowchart of the corporate gift customization demand semantic enhancement coding vector.

[0027] Figure 5 In the intelligent recommendation method for customized gift design solutions based on corporate culture in an embodiment of the present application, a schematic flowchart of screening out a subsequence of corporate culture supplement sentence granularity semantic coding vectors from the sequence of corporate culture supplement sentence granularity semantic coding vectors based on the semantic correlation between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of corporate culture supplement sentence granularity semantic coding vectors.

[0028] Figure 6 In the intelligent recommendation method for customized gift design solutions based on corporate culture in an embodiment of the present application, a schematic flowchart is provided for determining whether to use the first alternative gift as a recommendation solution based on the semantic similarity between the gift characteristic data of the first alternative gift and the semantic enhancement coding vector of the corporate gift customization demand.

[0029] Figure 7 This is a schematic block diagram of an intelligent recommendation system for customized gift design solutions based on corporate culture in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts also fall within the scope of protection of this application.

[0031] In response to the above technical problems, the technical concept of this application is to use deep learning-based natural language processing technology to perform semantic analysis on the company's gift customization needs, extract the semantic feature representation of the company's gift customization needs, and at the same time obtain supplementary descriptive information of the corporate culture from the company's official website and social media. By quickly performing semantic query interaction and fusion of the company's gift customization needs and the supplementary descriptive information of the corporate culture, the semantic feature expression ability of the company's gift customization needs is enhanced, and on this basis, intelligent screening and recommendation of alternative gifts are performed. In this way, accurate recommendations for customized gifts can be achieved, ensuring that the gifts not only meet the actual needs of the company, but also effectively convey the company's culture and values.

[0032] It should be noted that the acquisition and processing of all information or data in this application are carried out in compliance with the relevant national data protection laws and policies and with the authorization given by the owner of the corresponding device.

[0033] like Figure 1 and Figure 2As shown, the intelligent recommendation method for customized gift design solutions based on corporate culture is characterized by including: S1, obtaining corporate gift customization requirements input by a corporate user object, and obtaining corporate culture supplementary description information from the official website and social media of the corporate user object, wherein the corporate gift customization requirements include the cultural values ​​expected to be conveyed, the budget range and the purpose of the gift; S2, performing semantic understanding on the corporate gift customization requirements to obtain a corporate gift customization requirement semantic coding vector; S3, performing sentence-granularity semantic feature extraction on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity semantic coding vectors; S4, based on the sequence of corporate culture supplementary sentence-granularity semantic coding vectors, performing rapid semantic matching and enhancement on the corporate gift customization requirement semantic coding vector to obtain a corporate gift customization requirement semantic enhancement coding vector; S5, obtaining gift characteristic data of a first alternative gift; S6, determining whether to use the first alternative gift as a recommendation scheme based on the semantic similarity between the gift characteristic data of the first alternative gift and the corporate gift customization requirement semantic enhancement coding vector.

[0034] Exemplarily, in step S1, the corporate gift customization requirements input by the corporate user object are obtained, and supplementary description information of the corporate culture is obtained from the official website and social media of the corporate user object, wherein the corporate gift customization requirements include the cultural values ​​expected to be conveyed, the budget range and the purpose of the gift. It should be understood that the corporate gift customization requirements directly reflect the company's basic expectations and purposes for the gift. By understanding the cultural values, budget range and gift purpose that the company wants to convey, it can be ensured that the recommended gifts not only meet the actual needs of the company, but also effectively convey the information and values ​​that the company wants to convey. This is the basis for customized gift design and the prerequisite for ensuring the accuracy and practicality of the recommended plan. In addition, corporate culture, as the soul of the company, includes but is not limited to the company's mission, vision, values, historical stories and employee codes of conduct. Considering that relying solely on the analysis of gift customization requirements is usually unable to accurately grasp the core essence of corporate culture, this application further obtains supplementary description information about corporate culture through public channels such as the company's official website and social media platforms, so as to more accurately understand and convey the corporate image.

[0035] For example, in step S2, semantic understanding of the corporate gift customization requirements is performed to obtain a semantic encoding vector for the corporate gift customization requirements. By converting the corporate gift customization requirements into a semantic encoding vector, the degree of match between the requirements and the alternative gifts can be more accurately measured. This step is crucial to ensuring that the recommendation results not only meet the actual needs of the enterprise but also effectively convey the corporate culture and values. By converting corporate gift customization requirements into semantic encoding vectors, semantic interaction and integration with other data sources such as supplementary descriptive information about corporate culture can be facilitated, thereby enhancing the intelligence of the entire recommendation system.

[0036] In one embodiment, the corporate gift customization demand is semantically understood to obtain a semantic encoding vector of the corporate gift customization demand, including: using a semantic encoder including a Bert model and an LSTM model to semantically understand the corporate gift customization demand to obtain a semantic encoding vector of the corporate gift customization demand. It should be understood that in order to achieve a deep semantic understanding of the corporate gift customization demand, the present application uses a semantic encoder including a Bert model and an LSTM model to semantically encode the corporate gift customization demand. Those of ordinary skill in the art should know that BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained language model that can dynamically adjust the semantic feature representation of each word unit based on text context information, thereby deeply understanding the context of the corporate gift customization demand. LSTM (Long Short-Term Memory) can effectively capture long-term dependencies in a sequence by introducing a "gate" mechanism (forget gate, input gate, output gate), and gradually construct a text feature representation through recursion on time steps. Therefore, by combining the BERT model and the LSTM model, we can fully utilize the advantages of both models, effectively capture the contextual information and semantic structure of the corporate gift customization needs, and more comprehensively understand the semantic content of the corporate gift customization needs, thereby extracting a more accurate semantic feature representation. In a specific implementation, the BERT model can be used to encode the corporate gift customization needs first to obtain the contextual semantic feature representation of each word unit in the text. This is then used as the input of the LSTM model, and through the recursive processing of the LSTM model, the long-term dependencies in the text are further captured, the semantic coherence and accuracy of the text information are enhanced, and finally a semantic encoding vector for the corporate gift customization needs is generated.

[0037] In another embodiment, the Transformer architecture can also be used to semantically understand the corporate gift customization requirements to obtain a semantic encoding vector for the corporate gift customization requirements. In particular, pre-trained language models such as GPT-3 or T5 can directly encode the input corporate gift customization requirements text. This type of model has advantages in handling long-range dependencies and does not need to process information sequentially like LSTM, so it can more efficiently process text data in parallel. By inputting the corporate gift customization requirements into the Transformer model, a semantic encoding vector rich in contextual information can be directly obtained. Specifically, the corporate gift customization requirements text input by the corporate user object is pre-processed, including cleaning and standardization, such as removing irrelevant characters and punctuation. Next, the text is segmented to generate a vocabulary sequence, and special start and end markers are added according to the requirements of the pre-trained Transformer model used (such as BERT, RoBERTa, or T5). Select a pre-trained Transformer model suitable for the task. These models have been pre-trained on large amounts of text data and have rich language representation capabilities. If there is a large amount of data in a specific domain, further fine-tuning can be considered to improve the model's understanding of domain-specific terminology. The preprocessed gift customization request text is then fed into the selected Transformer model. The model outputs contextual word embeddings for each word, a series of vectors that contain rich semantic information. To obtain the semantic encoding vector for the entire sentence or paragraph, the model can be used to generate a single vector by taking the average or maximum of all word embeddings from the final layer of the model output, or by using other aggregation methods (such as weighted summation). Another common approach is to directly use the vector corresponding to the [CLS] token as the semantic encoding vector for the entire input text. This is especially true when using models like BERT, where the [CLS] token is often designed as a summary representation of the entire input sequence. In some cases, additional self-attention layers can be introduced to further strengthen the weighting of key information, thereby generating a more accurate semantic encoding vector. The self-attention mechanism allows the model to learn which parts are more important for the task at hand, thereby enhancing the expressiveness of these parts. In this way, the powerful representational power of the Transformer model can be effectively utilized to understand corporate gift customization requests and convert them into high-quality semantic encoding vectors, providing a solid foundation for subsequent intelligent recommendations.

[0038] In another embodiment, if corporate gift customization requirements include multimedia content such as images and videos in addition to text descriptions, multimodal learning methods can be considered. For example, visual recognition technology (such as convolutional neural networks (CNNs)) can be combined with text analysis technology (such as the aforementioned Transformer model) to extract features from different types of input. These features are then integrated through specific fusion mechanisms (such as attention mechanisms or cross-modal transformers) to generate a semantic encoding vector for the corporate gift customization requirement that integrates multiple information sources. A knowledge graph about corporate culture and gift design is constructed, containing entities (such as cultural values, gift types, etc.) and their relationships. When a new corporate gift customization requirement is received, it is first parsed into nodes in the graph. Graph embedding techniques (such as TransE and TransR) are then used to map the entire requirement into a low-dimensional space to form a semantic encoding vector. This approach can better capture domain-specific concepts and the connections between them. An adaptive encoding framework is designed to dynamically select the most appropriate encoding method based on the specific corporate gift customization requirement. For example, for highly structured requirements, a rule-based approach can be chosen, while for unstructured or semi-structured natural language descriptions, a deep learning model can be used. This flexibility helps improve the robustness and generalization ability of the overall system.

[0039] Exemplarily, in step S3, sentence-granularity semantic feature extraction is performed on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity semantic coding vectors. It should be understood that the expression of corporate culture is often multidimensional and complex, including the company's mission, vision, values, historical stories, and employee code of conduct, etc. By decomposing the corporate culture supplementary description information into sentence-level granularity and semantically encoding each sentence separately, the specific meaning of different cultural elements and their impact on corporate gift customization needs can be captured more finely. When processing the entire corporate culture document, directly encoding it as a whole may result in unnecessary computational overhead, because not all content is relevant to the current gift customization needs. By processing the sentences into sentences and encoding them separately, it is possible to focus on those cultural features that are most relevant to gift customization, thereby improving computational efficiency.

[0040] In one embodiment, Figure 3As shown, sentence-granularity semantic feature extraction is performed on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity semantic coding vectors, including: S31, sentence processing is performed on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity descriptions; S32, using the semantic encoder including the Bert model and the LSTM model to semantically encode each corporate culture supplementary sentence-granularity description in the sequence of corporate culture supplementary sentence-granularity descriptions to obtain a sequence of corporate culture supplementary sentence-granularity semantic coding vectors.

[0041] Exemplarily, in step S31, the corporate culture supplementary description information is sentence-processed to obtain a sequence of granular descriptions of corporate culture supplementary sentences. It should be understood that, considering that when semantically encoding the corporate culture supplementary description information, due to the complexity and multidimensionality of corporate culture, not all parts are directly related to gift customization needs. Therefore, in order to reduce the waste of computing resources, this application adopts a sentence-processing method, by decomposing the corporate culture supplementary description information into multiple sentences, with each sentence representing an independent cultural feature description, to form a sequence of granular descriptions of corporate culture supplementary sentences, so as to facilitate the subsequent semantic encoding process, more accurately identify and extract cultural features related to corporate gift customization needs, so as to reduce unnecessary computing load and improve the accuracy and relevance of recommendations.

[0042] Exemplarily, in step S32, the semantic encoder comprising the Bert model and the LSTM model is used to semantically encode each of the granular descriptions of the corporate culture supplementary sentence in the sequence of granular descriptions of the corporate culture supplementary sentence to obtain a sequence of granular semantic encoding vectors of the corporate culture supplementary sentence. That is, the semantic encoder comprising the Bert model and the LSTM model is also used to semantically encode each of the granular descriptions of the corporate culture supplementary sentence in the sequence of granular descriptions of the corporate culture supplementary sentence to obtain a sequence of granular semantic encoding vectors of the corporate culture supplementary sentence. This identical semantic encoding method helps to reduce semantic deviations introduced by inconsistent semantic encoding processes, thereby improving the efficiency and accuracy of subsequent semantic interaction fusion.

[0043] Exemplarily, in step S4, based on the sequence of the granular semantic coding vectors of the corporate culture supplementary sentences, the corporate gift customization demand semantic coding vector is quickly semantically matched and strengthened to obtain the corporate gift customization demand semantic strengthening coding vector. It should be understood that the corporate gift customization demand directly reflects the specific requirements of the user, while the corporate culture supplementary description provides a broader and deeper corporate background and values. By performing information fusion processing on the two, it helps to ensure that the recommendation system does not rely solely on superficial demand statements, but is able to deeply understand corporate culture and value orientation, thereby making gift recommendations that are more in line with corporate image and culture. In particular, as mentioned above, considering that not all content in the corporate culture supplementary description is directly related to the gift customization demand, in order to improve the efficiency of information fusion, the present application proposes an efficient semantic query matching method, which quickly queries and fuses the cultural feature descriptions that are most relevant to the corporate gift customization demand by analyzing the semantic relevance of the corporate gift customization demand semantic coding vector and the granular semantic coding vectors of each corporate culture supplementary sentence, thereby being able to effectively utilize the corporate culture supplementary description information to strengthen the semantic feature expression of the corporate gift customization demand and improve the accuracy of gift recommendations.

[0044] In one embodiment, Figure 4 As shown, based on the sequence of the corporate culture supplement sentence granularity semantic coding vectors, the corporate gift customization demand semantic coding vector is quickly semantically matched and enhanced to obtain the corporate gift customization demand semantic enhancement coding vector, including: S41, based on the semantic correlation between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of the corporate culture supplement sentence granularity semantic coding vectors, a subsequence of the corporate culture supplement sentence granularity semantic coding vector is screened out from the sequence of the corporate culture supplement sentence granularity semantic coding vectors; S42, cross-domain query matching and fusion are performed on the corporate gift customization demand semantic coding vector and the subsequence of the corporate culture supplement sentence granularity semantic coding vector to obtain the corporate gift customization demand semantic enhancement coding vector.

[0045] For example, Figure 5As shown, in step S41, based on the semantic correlation between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of the corporate culture supplement sentence granularity semantic coding vector, a subsequence of the corporate culture supplement sentence granularity semantic coding vector is screened out from the sequence of the corporate culture supplement sentence granularity semantic coding vector, including: S411, calculating the mutual information between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of the corporate culture supplement sentence granularity semantic coding vector to obtain a sequence of corporate gift customization demand-corporate culture fast semantic matching factors; S412, identifying the first maximum value and the second maximum value from the sequence of corporate gift customization demand-corporate culture fast semantic matching factors; S413, based on the position of the first maximum value and the second maximum value in the sequence of corporate gift customization demand-corporate culture fast semantic matching factors, determining the subsequence of the corporate culture supplement sentence granularity semantic coding vector that is quickly matched from the sequence of the corporate culture supplement sentence granularity semantic coding vector. Specifically, the process can be expressed by the formula:

[0046] v2={v 21 ,v 22 ,...,v 2i ,...,v 2n}

[0047] x1 = sigmoid(v1)

[0048] x 2i =sigmoid(v 2i )

[0049] MI i =H(x1)+H(x 2i )-H(x1,x 2i )

[0050]

[0051] {k sub}={v 2argmax1 ,...,v 2argmax2}

[0052] Among them, v1 represents the semantic encoding vector of the corporate gift customization demand, v2 represents the sequence of the granularity semantic encoding vectors of the corporate culture supplementary sentence, and v 21 、v 22 、v 2i 、v 2j and v 2nThey represent the first, second, i-th, j-th and n-th corporate culture supplement sentence granularity semantic coding vectors in the sequence of the corporate culture supplement sentence granularity semantic coding vectors, n is the number of the corporate culture supplement sentence granularity semantic coding vectors, sigmoid represents the sigmoid normalization function, x1 represents the normalized corporate gift customization demand semantic coding vector, x 2i represents the i-th normalized semantic coding vector of corporate gift customization demand, H(x1) represents the information entropy of the normalized semantic coding vector of corporate gift customization demand, H(x 2i ) represents the information entropy of the i-th normalized semantic encoding vector of corporate gift customization demand, H(x1,x 2i ) represents the joint entropy between the normalized corporate gift customization demand semantic coding vector and the i-th normalized corporate gift customization demand semantic coding vector, MI i represents the i-th corporate gift customization demand-corporate culture fast semantic matching factor, max() represents the maximum value function, MI max1 and MI max2 denote the first maximum value and the second maximum value respectively, argmax(·) denotes the index corresponding to the maximum value, argmax1 and argmax2 denote the position index of the first maximum value and the second maximum value in the sequence of the enterprise gift customization demand-corporate culture fast semantic matching factor respectively, v 2argmax1 and v 2argmax2 Respectively represent the enterprise culture supplement sentence granularity semantic encoding vectors corresponding to the first maximum value and the second maximum value, k sub A subsequence representing the granular semantic encoding vector of the corporate culture supplement sentence.

[0053] That is, first, the mutual information between the granular semantic encoding vectors of each corporate culture supplement sentence and the semantic encoding vector of the corporate gift customization demand is calculated to evaluate the feature correlation between the two and generate a sequence of corporate gift customization demand-corporate culture rapid semantic matching factors. It should be understood that the larger the mutual information value, the higher the correlation; conversely, the smaller the mutual information value, the weaker the correlation. In this way, not only can the corporate culture features most relevant to corporate gift customization demand be effectively identified, but also an objective standard can be provided for subsequent rapid matching in a quantitative manner. Next, the two maximum values ​​are queried from the sequence of corporate gift customization demand-corporate culture rapid semantic matching factors. Based on the position of the two maximum values ​​in the sequence, a subsequence of corporate culture supplement sentence granular features significantly related to corporate gift customization demand is extracted from the corresponding sequence of corporate culture supplement sentence granular semantic encoding vectors to achieve preliminary feature selection and filtering, so as to concentrate resources on processing the most relevant semantic interaction information, reduce unnecessary calculations, and reduce noise interference in corporate culture information.

[0054] Exemplarily, in step S42, a cross-domain query matching fusion is performed on the subsequences of the semantic coding vector of the corporate gift customization demand and the granularity semantic coding vector of the corporate culture supplement sentence to obtain the semantic enhancement coding vector of the corporate gift customization demand, including: performing a linear transformation on the semantic coding vector of the corporate gift customization demand to obtain a query vector and a value vector, and using the subsequence of the granularity semantic coding vector of the corporate culture supplement sentence as a subsequence of the key vector, and inputting the query vector, value vector, and the subsequence of the key vector into a cross-domain query coding module based on a converter structure to obtain the semantic enhancement coding vector of the corporate gift customization demand. Specifically, the process can be expressed as follows:

[0055] v q =v1W q +b q

[0056] v v =v1W v +b v

[0057]

[0058] Among them, W q and W v denote the query embedding matrix and the value embedding matrix respectively, b q and b v Represent the query bias vector and value bias vector respectively, v q and v v denote the query vector and value vector respectively, (·) TRepresents the transpose of a vector, S represents v 2j The length of , softmax(·) represents the normalized exponential function, Represents a matrix multiplication operation, and V represents the semantic reinforcement coding vector of the enterprise gift customization demand.

[0059] That is, based on the semantic coding vector of the corporate gift customization demand, a query vector and a value vector are constructed, and a subsequence of the selected corporate culture supplement sentence granularity semantic coding vector is used as a subsequence of the key vector. Cross-domain query interaction is performed through the converter, and the self-attention mechanism is used to learn the semantic dependency relationship between corporate culture information and corporate gift customization demand information, and generate a semantic reinforcement coding vector for corporate gift customization demand, thereby realizing deep interactive fusion of cross-domain information.

[0060] First, a query vector and a value vector are generated using a linear transformation of the semantic encoding vector of corporate gift customization requirements. This transformation is accomplished using a specific embedding matrix and bias vector. The query vector represents the current requirement characteristics, while the value vector carries the specific content of the requirement. Next, a subsequence of key vectors is selected from the previously selected subsequences of the granular semantic encoding vectors of corporate culture supplement sentences. These key vectors represent the cultural characteristics most relevant to corporate gift customization requirements. This selection process is based on the semantic relevance between the semantic encoding vector of corporate gift customization requirements and the granular semantic encoding vectors of each corporate culture supplement sentence, determined by metrics such as mutual information. The query vector, value vector, and subsequence of key vectors are then input into a cross-domain query encoding module based on a Transformer architecture. The core of this module is the self-attention mechanism, which measures the importance of information at different positions and aggregates it accordingly. Specifically, a series of attention scores are generated by calculating the similarity (typically dot product or scaled dot product) between the query vector and each key vector. These attention scores are normalized using the softmax function to ensure that they can be interpreted as probability distributions, thereby reflecting the relevance of different cultural characteristics to the current gift customization needs. Then, the normalized attention scores are used to perform weighted summation on the corresponding value vectors to generate the final output vector. In summary, the query vector and value vector are constructed based on the semantic encoding vector of the corporate gift customization needs, and the subsequence of the selected corporate culture supplement sentence granularity semantic encoding vector is used as a subsequence of the key vector. Cross-domain query interaction is performed through the converter, and the self-attention mechanism is used to learn the semantic dependency relationship between corporate culture information and corporate gift customization needs information, and generate a semantic reinforcement encoding vector for corporate gift customization needs, thereby achieving deep interactive fusion of cross-domain information.

[0061] For example, in step S5, gift characteristic data of the first candidate gift is obtained. It should be understood that the gift characteristic data provides specific information about the first candidate gift, including but not limited to functional use, material, design style, cultural symbolism, aesthetic value, price range, etc., and is the basis for evaluating whether the gift meets the enterprise's customization requirements.

[0062] In one embodiment, the recommendation system can connect to a database containing a large amount of gift information and query the database to obtain detailed characteristic data for potential gifts. The database typically stores multiple attributes for each gift, such as name, description, image, price, material, and features. SQL queries or other database query languages ​​can efficiently retrieve the required data. Secondly, if the recommendation system is integrated with other gift suppliers or e-commerce platforms, gift characteristic data can be obtained by calling the APIs provided by these platforms. API calls typically require specific parameters, such as gift ID and category, and the returned data may be formatted as JSON or XML, containing detailed gift information. For websites or platforms that do not have API interfaces, web crawlers can be used to automatically capture gift characteristic data. Crawlers access designated web pages, extract relevant information from the pages, and store it in a structured format. It is important to note that when using crawlers to capture data, the relevant website's robots.txt file must be followed to avoid violating the website's terms of use. In some cases, particularly small-scale or specialized scenarios, manual entry of gift characteristic data is feasible.

[0063] Exemplarily, in step S6, based on the semantic similarity between the gift characteristic data of the first alternative gift and the semantic reinforcement coding vector of the corporate gift customization demand, it is determined whether to use the first alternative gift as a recommendation. It should be understood that the gift characteristic data provides specific information about the alternative gifts, including functional use, material, design style, cultural symbolic meaning, aesthetic value, price range, etc. This information is the basis for evaluating whether the gift meets the corporate customization needs. By comparing these characteristic data with the semantic reinforcement coding vector of the corporate gift customization demand, the degree of match between the alternative gifts and the corporate needs can be comprehensively evaluated. Semantic similarity refers to the degree of semantic similarity between two texts or vectors. By calculating the semantic similarity between the gift characteristic data of the first alternative gift and the semantic reinforcement coding vector of the corporate gift customization demand, the correlation between the two can be quantified.

[0064] In one embodiment, Figure 6 As shown, based on the semantic similarity between the gift characteristic data of the first alternative gift and the semantic enhancement coding vector of the corporate gift customization demand, determining whether to use the first alternative gift as a recommendation option includes: S61, using a gift characteristic embedding coding matrix to perform semantic embedding coding on the gift characteristic data of the first alternative gift to obtain a first alternative gift characteristic data embedding coding vector; S62, calculating the semantic matching degree between the corporate gift customization demand semantic enhancement coding vector and the first alternative gift characteristic data embedding coding vector; S63, if the semantic matching degree is greater than or equal to a preset threshold, using the first alternative gift as a recommendation option.

[0065] Exemplarily, in step S61, the gift characteristic embedding coding matrix is ​​used to perform semantic embedding coding on the gift characteristic data of the first alternative gift to obtain the first alternative gift characteristic data embedding coding vector. It should be understood that in order to effectively compare the gift characteristic data of the first alternative gift with the needs of the enterprise, it is necessary to further convert the gift characteristic data into a form that can be understood and processed by the computer. Therefore, the present application uses a gift characteristic embedding coding matrix to perform semantic embedding coding on the gift characteristic data of the first alternative gift to map it to a high-dimensional semantic feature space, and convert it into a numerical vector representation containing rich semantic information, thereby obtaining the first alternative gift characteristic data embedding coding vector. In a specific example of the present application, the gift characteristic embedding coding matrix is ​​generated based on Word2Vec model training.

[0066] Exemplarily, in step S62, the semantic matching degree between the semantic reinforcement coding vector of the enterprise gift customization demand and the embedded coding vector of the first alternative gift characteristic data is calculated, including: calculating the cosine similarity between the semantic reinforcement coding vector of the enterprise gift customization demand and the embedded coding vector of the first alternative gift characteristic data as the semantic matching degree. It should be understood that by calculating the cosine similarity between the semantic reinforcement coding vector of the enterprise gift customization demand and the embedded coding vector of the first alternative gift characteristic data, the semantic matching degree between the two is revealed. The closer the cosine similarity value is to 1, the higher the matching degree between the first alternative gift and the enterprise gift customization demand. Conversely, if the cosine similarity value is low, it means that the matching degree between the first alternative gift and the enterprise demand is not ideal, and additional alternative gifts may need to be considered.

[0067] In the technical solution of the present application, the corporate gift customization demand semantic reinforcement coding vector and the first alternative gift characteristic data embedded coding vector represent the corporate gift definition demand reinforcement semantic coding features and the first alternative gift characteristic data embedded coding features, respectively. Here, considering that the corporate gift customization demand semantic reinforcement coding vector has dimensionality differences and feature order differences relative to the first alternative gift characteristic data embedded coding vector, when calculating the semantic matching degree between the two, the corporate gift customization demand semantic reinforcement coding vector will have a long-distance correspondence deviation relative to the first alternative gift characteristic data embedded coding vector, thereby affecting the calculation accuracy of the semantic matching degree.

[0068] Accordingly, in a preferred embodiment of the present application, before calculating the semantic matching degree between the corporate gift customization demand semantic enhancement coding vector and the first candidate gift characteristic data embedded coding vector, feature distribution modulation is performed on the corporate gift customization demand semantic enhancement coding vector, which includes:

[0069] Arranging the eigenvalues ​​of the corporate gift customization demand semantic reinforcement coding vector in ascending order to form a corporate gift customization demand semantic reinforcement sequence coding vector;

[0070] In response to the absolute value of the difference between the i-th eigenvalue and the i+1-th eigenvalue of the semantic reinforcement sequential encoding vector of the corporate gift customization demand being less than or equal to a distance difference hyperparameter ε, a weighted sum between the i-th eigenvalue and the i+1-th eigenvalue is calculated as the optimized i+1-th eigenvalue;

[0071] Calculate the square root of the sum of squares of all eigenvalues ​​of the semantic reinforcement encoding vector of the corporate gift customization demand:

[0072]

[0073] Among them, v i represents the i-th eigenvalue of the semantic reinforcement coding vector of the enterprise gift customization demand, L represents the length of the semantic reinforcement coding vector of the enterprise gift customization demand, and γ1 represents the square root of the sum of the squares of all eigenvalues ​​of the semantic reinforcement coding vector of the enterprise gift customization demand;

[0074] The square root of the sum of the squares of all eigenvalues ​​of the corporate gift customization demand semantic enhancement coding vector is multiplied by 2 and then divided by the square of the length of the corporate gift customization demand semantic enhancement coding vector to obtain the corporate gift customization demand semantic enhancement space primitive value γ2=2×γ1 / L 2 ;

[0075] In response to the absolute value of the difference between the i-th eigenvalue and the i+1-th eigenvalue of the corporate gift customization demand semantic reinforcement sequential encoding vector being greater than the distance difference hyperparameter ε, multiplying the corporate gift customization demand semantic reinforcement space primitive value by the i-th eigenvalue, and then calculating the weighted subtraction between the product and the i+1-th eigenvalue to obtain the optimized i+1-th eigenvalue;

[0076] On the basis of keeping the first eigenvalue of the semantic reinforcement sequence coding vector of the corporate gift customization demand unchanged, the optimized i+1th eigenvalue is combined to obtain the optimized corporate gift customization demand semantic reinforcement coding vector.

[0077] Accordingly, the high-dimensional feature space primitive representation based on self-inner product fusion of the corporate gift customization demand semantic enhancement coding vector is used to capture the complex structure of the global network interaction of its eigenvalues, thereby reconstructing the interactive response relationship between the eigenvalues ​​of the corporate gift customization demand semantic enhancement coding vector by simulating the scale-based high-dimensional feature space potential primitives, so as to realize the coding reconstruction of the real sequence distribution behavior of the corporate gift customization demand semantic enhancement coding vector under long distance, thereby improving the long-distance correspondence between the corporate gift customization demand semantic enhancement coding vector and the first alternative gift characteristic data embedded coding vector, thereby improving the calculation accuracy of the semantic matching degree.

[0078] For example, in step S63, if the semantic match is greater than or equal to a preset threshold, the first alternative gift is recommended. In other words, by setting a threshold, alternative gifts with a high degree of match to the company's gift customization needs are screened out. If the semantic match between the first alternative gift and the company's needs is greater than or equal to the preset threshold, the alternative gift is deemed to meet the company's gift customization needs and can be recommended to the company as a recommended solution.

[0079] In summary, according to the embodiment of the present application, the intelligent recommendation method for customized gift design solutions based on corporate culture is explained. It uses natural language processing technology based on deep learning to perform semantic analysis on the company's gift customization needs, extracts the semantic feature representation of the company's gift customization needs, and simultaneously obtains supplementary descriptive information of the corporate culture from the company's official website and social media. By performing rapid semantic query interaction and fusion of the company's gift customization needs and the supplementary descriptive information of the corporate culture, the semantic feature expression ability of the company's gift customization needs is enhanced, and on this basis, intelligent screening and recommendation of alternative gifts are performed. In this way, accurate recommendations for customized gifts can be achieved, ensuring that the gifts not only meet the actual needs of the company, but also effectively convey the company's culture and values.

[0080] Figure 7 This is a schematic block diagram of an intelligent recommendation system for customized gift design solutions based on corporate culture according to an embodiment of the present application. Figure 7As shown, the customized gift design scheme intelligent recommendation system 100 based on corporate culture includes: an enterprise data acquisition module 110, which is used to obtain the corporate gift customization requirements input by the corporate user object, and obtain corporate culture supplementary description information from the official website and social media of the corporate user object, wherein the corporate gift customization requirements include the cultural value to be conveyed, the budget range and the purpose of the gift; a corporate gift customization requirement semantic understanding module 120, which is used to perform semantic understanding on the corporate gift customization requirements to obtain the corporate gift customization requirement semantic encoding vector; a corporate culture semantic feature extraction module 130, which is used to perform sentence-granular semantic analysis on the corporate culture supplementary description information. The module 140 is used to extract semantic features to obtain a sequence of granular semantic coding vectors of corporate culture supplement sentences; the module 140 is used to quickly perform semantic matching and enhancement on the semantic coding vectors of corporate gift customization requirements based on the sequence of granular semantic coding vectors of corporate culture supplement sentences to obtain semantic enhancement coding vectors of corporate gift customization requirements; the module 150 is used to obtain gift characteristic data of the first alternative gift; the module 160 is used to determine whether to use the first alternative gift as a recommendation scheme based on the semantic similarity between the gift characteristic data of the first alternative gift and the semantic enhancement coding vector of corporate gift customization requirements.

[0081] In one embodiment, the corporate gift customization demand semantic understanding module is used to: use a semantic encoder including a Bert model and an LSTM model to perform semantic understanding on the corporate gift customization demand to obtain a semantic encoding vector of the corporate gift customization demand.

[0082] Here, those skilled in the art will appreciate that the specific operations of the various modules and units in the above-mentioned customized gift design solution intelligent recommendation system based on corporate culture have been described in the above reference. Figures 1 to 6 The description of the intelligent recommendation method for customized gift design solutions based on corporate culture has been introduced in detail, and therefore, its repeated description will be omitted.

[0083] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0084] It should be understood that the specific examples in this article are only intended to help those skilled in the art better understand the embodiments of the present application, and are not intended to limit the scope of the embodiments of the present application.

[0085] It should also be understood that in the various embodiments of the present application, the size of the serial number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0086] It should also be understood that the various implementation methods described in this specification can be implemented individually or in combination, and the embodiments of the present application are not limited to this.

[0087] Unless otherwise indicated, all technical and scientific terms used in the embodiments of the present application have the same meaning as those generally understood by those skilled in the art in the technical field of the present application. The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit the scope of this application. The term "and / or" used in this application includes any and all combinations of one or more related listed items. The singular forms of "a", "above" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. In addition, the terms "first", "second" etc. are only used for descriptive purposes and are not to be understood as indicating or suggesting relative importance.

[0088] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0089] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0090] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0091] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, 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. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. 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.

[0092] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. An intelligent recommendation method for customized gift design solutions based on corporate culture, characterized in that: include: Obtaining corporate gift customization requirements input by a corporate user object, and obtaining supplementary corporate culture description information from the corporate user object's official website and social media, wherein the corporate gift customization requirements include the cultural values ​​to be conveyed, the budget range, and the purpose of the gift; Performing semantic understanding on the corporate gift customization demand to obtain a semantic coding vector of the corporate gift customization demand; Performing sentence-granularity semantic feature extraction on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-granularity semantic encoding vectors; Based on the sequence of the corporate culture supplement sentence granularity semantic coding vectors, the corporate gift customization demand semantic coding vector is quickly semantically matched and enhanced to obtain a corporate gift customization demand semantic enhancement coding vector; Obtain gift characteristic data of the first candidate gift; Based on the semantic similarity between the gift characteristic data of the first candidate gift and the semantic enhancement coding vector of the corporate gift customization demand, it is determined whether to use the first candidate gift as a recommendation.

2. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 1, characterized in that: Performing semantic understanding on the corporate gift customization demand to obtain a semantic coding vector of the corporate gift customization demand includes: A semantic encoder including a Bert model and an LSTM model is used to perform semantic understanding on the corporate gift customization demand to obtain a semantic encoding vector of the corporate gift customization demand.

3. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 2, characterized in that: The sentence-level semantic feature extraction is performed on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-level semantic encoding vectors, including: Sentence processing is performed on the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence granularity descriptions; The semantic encoder including the Bert model and the LSTM model is used to semantically encode each corporate culture supplement sentence granularity description in the sequence of corporate culture supplement sentence granularity descriptions to obtain a sequence of corporate culture supplement sentence granularity semantic encoding vectors.

4. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 3 is characterized in that: Based on the sequence of the corporate culture supplement sentence granularity semantic coding vectors, the corporate gift customization demand semantic coding vector is quickly semantically matched and enhanced to obtain the corporate gift customization demand semantic enhancement coding vector, including: Based on the semantic correlation between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of corporate culture supplement sentence granularity semantic coding vectors, a subsequence of corporate culture supplement sentence granularity semantic coding vectors is filtered out from the sequence of corporate culture supplement sentence granularity semantic coding vectors; A cross-domain query matching fusion is performed on the subsequences of the corporate gift customization demand semantic coding vector and the corporate culture supplement sentence granularity semantic coding vector to obtain the corporate gift customization demand semantic enhancement coding vector.

5. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 4, characterized in that: Based on the semantic correlation between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of corporate culture supplement sentence granularity semantic coding vectors, a subsequence of corporate culture supplement sentence granularity semantic coding vectors is screened out from the sequence of corporate culture supplement sentence granularity semantic coding vectors, including: Calculating the mutual information between the corporate gift customization demand semantic coding vector and each corporate culture supplement sentence granularity semantic coding vector in the sequence of corporate culture supplement sentence granularity semantic coding vectors to obtain a sequence of corporate gift customization demand-corporate culture fast semantic matching factors; Identifying a first maximum value and a second maximum value from the sequence of the corporate gift customization demand-corporate culture fast semantic matching factors; Based on the positions of the first maximum value and the second maximum value in the sequence of the corporate gift customization requirements-corporate culture fast semantic matching factors, a subsequence of the corporate culture supplement sentence granularity semantic coding vectors that is quickly matched is determined from the sequence of the corporate culture supplement sentence granularity semantic coding vectors.

6. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 5, characterized in that: Performing cross-domain query matching and fusion on subsequences of the corporate gift customization demand semantic coding vector and the corporate culture supplement sentence granularity semantic coding vector to obtain the corporate gift customization demand semantic enhancement coding vector, including: A linear transformation is performed on the semantic coding vector of the corporate gift customization demand to obtain a query vector and a value vector, and a subsequence of the corporate culture supplement sentence granularity semantic coding vector is used as a subsequence of the key vector. The query vector, the value vector and the subsequence of the key vector are input into a cross-domain query coding module based on a converter structure to obtain the semantic enhancement coding vector of the corporate gift customization demand.

7. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 6, characterized in that: Determining whether to use the first candidate gift as a recommendation based on semantic similarity between the gift characteristic data of the first candidate gift and the semantic enhancement coding vector of the corporate gift customization requirement includes: Perform semantic embedding coding on the gift characteristic data of the first candidate gift using a gift characteristic embedding coding matrix to obtain an embedding coding vector for the first candidate gift characteristic data; Calculating the semantic matching degree between the semantic enhancement coding vector of the corporate gift customization demand and the embedded coding vector of the first candidate gift characteristic data; If the semantic matching degree is greater than or equal to a preset threshold, the first candidate gift is recommended.

8. The method for intelligently recommending customized gift design solutions based on corporate culture according to claim 7, characterized in that: Calculating the semantic matching degree between the semantic enhancement coding vector of the corporate gift customization requirement and the embedded coding vector of the first candidate gift characteristic data includes: The cosine similarity between the semantic enhancement coding vector of the corporate gift customization demand and the embedded coding vector of the first candidate gift characteristic data is calculated as the semantic matching degree.

9. An intelligent recommendation system for customized gift design solutions based on corporate culture, characterized in that: include: An enterprise data acquisition module is used to obtain corporate gift customization requirements input by corporate user objects and obtain supplementary description information of corporate culture from the official website and social media of the corporate user objects, wherein the corporate gift customization requirements include the cultural values ​​to be conveyed, the budget range, and the purpose of the gift; A corporate gift customization demand semantic understanding module is used to perform semantic understanding on the corporate gift customization demand to obtain a semantic coding vector of the corporate gift customization demand; A corporate culture semantic feature extraction module is used to extract sentence-level semantic features of the corporate culture supplementary description information to obtain a sequence of corporate culture supplementary sentence-level semantic encoding vectors; A corporate gift customization demand semantic enhancement module is used to perform rapid semantic matching enhancement on the corporate gift customization demand semantic encoding vector based on the sequence of corporate culture supplementary sentence granularity semantic encoding vectors to obtain a corporate gift customization demand semantic enhancement encoding vector; A gift characteristic data acquisition module, used to acquire gift characteristic data of the first candidate gift; The recommendation scheme determination module is used to determine whether to use the first candidate gift as a recommendation scheme based on the semantic similarity between the gift characteristic data of the first candidate gift and the semantic enhancement coding vector of the corporate gift customization demand.

10. The intelligent recommendation system for customized gift design solutions based on corporate culture according to claim 9 is characterized in that: The corporate gift customization demand semantic understanding module is used to: A semantic encoder including a Bert model and an LSTM model is used to perform semantic understanding on the corporate gift customization demand to obtain a semantic encoding vector of the corporate gift customization demand.

Images