Electronic business card associated data analysis system and method based on artificial intelligence

Abstract

The invention discloses an electronic business card associated data analysis system and method based on artificial intelligence, and relates to the technical field of information mining. According to the method, all collected different types of electronic business cards are analyzed and standardized, user information in each electronic business card is normalized, and an electronic business card database is constructed; synthesizing the social relationships of all the users to generate a social relationship network of the users; generating mining elements of each user by utilizing emotional characteristics of each user during communication with different users in history and combining user information in different electronic business cards; potential customers matched with the real-time users are found, and electronic business cards of the potential customers are recommended to the real-time users; and visually displaying the recommendation reason of each recommended potential customer, and optimizing the mining elements of the real-time user by using the electronic business card information corresponding to the key features.

Description

Technical Field

[0001] The present invention relates to the technical field of information mining, and in particular to an electronic business card associated data analysis system and method based on artificial intelligence. Background Art

[0002] With the development of computer and network technology, early business card management software appeared, with the main functions of information storage and simple classification. Compared with traditional paper business cards, electronic business cards are more convenient and have a wider range of dissemination. Gradually, electronic business cards are widely cited. With the development of various modern computer technologies, the research on electronic business cards plays a key role. Among them, the maturity of data mining technology has enabled enterprises to explore how to extract value from customer data, and social network analysis has gradually become a research hotspot; the rise of social media platforms has promoted the use of electronic business cards, and the expansion of user information and network relationships has become a key research direction. The development of big data technology has enabled the processing of massive electronic business card data, and the application of cloud computing has made data storage and analysis flexible and efficient. However, when mining electronic business cards, due to the large amount of data, mining is often only based on the user's own conditional data, without considering the user's own emotional preferences. The mined customers may match the big data conditions but the user does not like them; and the key to mining potential customers lies in the accurate construction of the social relationship network. The existing construction methods all use electronic business card big data, and the constructed social relationship network does not reflect the user's subjective consciousness and is not perfect; in summary, it greatly affects the user's experience; therefore, how to construct a personalized mining method is crucial. Summary of the invention

[0003] The purpose of the present invention is to provide an electronic business card associated data analysis system and method based on artificial intelligence to solve the problems raised in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] An artificial intelligence-based electronic business card associated data analysis method, the method comprising the following steps:

[0006] S100, the user fills in his / her own information in the software to form his / her own electronic business card and transmits it to the cloud for storage, collects the electronic business cards of all users in the cloud, analyzes and standardizes all different types of collected electronic business cards, normalizes the user information in each electronic business card, and constructs an electronic business card database;

[0007] Furthermore, the specific steps of constructing the electronic business card database are:

[0008] S101. Collect all users' electronic business cards in the cloud. Suppose the format of all collected electronic business cards is {G1, G2, G3, ..., G p}, G1, G2, G3, ..., G p represents the 1st, 2nd, 3rd, ..., pth format of all collected user electronic business cards, where p is a positive integer; calculate the number of each electronic business card collected in the cloud, and select the one with the largest number as the standard format of the electronic business card Gb; extract user information from electronic business cards of different formats, and assume that the user information contained in the collected electronic business cards is {X1, X2, X3, ..., X n},

[0009] X1, X2, X3, ..., X n Indicates the first, second, third, ..., nth types of user information contained in the user's electronic business card, where n is a positive integer; re-fills the extracted user information in different formats according to the standard format, and standardizes the electronic business cards of all users collected in the cloud;

[0010] S102, after standardizing the electronic business cards of all users, convert each user information in the electronic business card into vector data Xi using an encoder, and normalize the vector data of each user information in the electronic business card, the formula is:

[0011]

[0012] In the formula, Xi' represents the normalized vector data, Xi represents the vector data of user information, and m represents the dimension of the transformed vector data of user information; the electronic business card database is constructed using the normalized vector data of user information.

[0013] Standardizing and normalizing all electronic business cards collected in the cloud unifies the information format in the electronic business cards. When integrating electronic business card data from multiple sources, this consistency enables the data to be merged and stored quickly and accurately, reducing data errors and duplicate entry problems caused by format differences; and ensuring the accurate transmission of important information in the electronic business card.

[0014] S200, collecting historical communication data between all users, using the communication data and user information in the electronic business card to build the social relationship of each user, and integrating the social relationships of all users to generate a social relationship network of users;

[0015] Furthermore, the specific steps of generating a user social relationship network are:

[0016] S201, collect the communication data between all users in history, take each user as a node, extract the communication ends in the collected communication data as {(a, b)1, (a, b)2, (a, b)3, ..., (a, b) m},

[0017] (a, b)1, (a, b)2, (a, b)3, ..., (a, b) m Indicates the communication ends in the 1st, 2nd, 3rd, ..., mth communication data in the collected history, uses the communication ends in each communication data to form a relationship edge, and finally forms m relationship edges, and uses user nodes and relationship edges to generate an explicit social relationship network;

[0018] S202, extract each type of user information from the electronic business cards of all users, take each user as a node, compare each type of user information from the extracted electronic business cards, compare the electronic business cards of two users, traverse and compare all the user information contained in the two electronic business cards, and when X∩X' exists, it is determined that the two users are connected and a relationship edge is generated; when X∩X' does not exist after traversing all the user information, it is determined that the two users are not connected; X and X' respectively represent the same type of user information contained in the electronic business cards of the two users; compare the electronic business cards of all users by traversing and comparing, set the connection that can generate a relationship edge to 1, and set the connection that cannot generate a relationship edge to 0, and construct the relationship edge matrix of all users as follows:

[0019]

[0020] In the matrix, each row and column represents a user node, and 1 / 0 represents the judgment result of whether two user nodes generate a relationship edge; after judging the relationship edges generated by all users, the relationship edge matrix is ​​output; each user node and the relationship edge matrix are used to construct an implicit social relationship network; the explicit social relationship network and the implicit social relationship network are combined to construct a social relationship network of all users.

[0021] The explicit social relationship network and implicit social relationship network are constructed by using the communication data and electronic business cards of all users in history. The two social relationship networks are combined to generate a comprehensive social relationship network. The generated comprehensive social relationship network connects the relationships between customers subjectively searched by users and customers matched by electronic business cards, ensuring the comprehensiveness of social relationships.

[0022] S300, using a natural language processing algorithm to analyze the language of each user in the history of communication with different electronic business card users, to obtain the emotional characteristics of each user in the history of communication with different users, and to generate mining factors for each user in combination with user information in different electronic business cards;

[0023] Furthermore, the specific steps of combining the user information in different electronic business cards to generate mining elements for each user are as follows:

[0024] S301, collect the communication data of each user in the history when communicating with other different users, extract all the language data of each communication data, extract all the words appearing in the language data, and calculate the importance of each word respectively, the formula is:

[0025]

[0026] In the formula, β represents the importance of each word, C i，k represents the number of the i-th language data in the k-th language data, represents the total number of words appearing in the kth language data; N represents the total number of language data, D i∈k represents the number of language data containing the i-th word; after calculation, the importance of all words is {β1, β2, β3, ..., β Q}, β1, β2, β3,..., β Q represents the importance of the 1st, 2nd, 3rd, ..., Qth words in the communication data, where Q is a positive integer;

[0027] S302, after calculating the importance of all words in the communication data, assign sentiment to each word according to the sentiment dictionary, and comprehensively analyze the importance and sentiment assignment of each word to obtain the total sentiment value of each word, the formula is: E = β i ×em i In the formula, E represents the total sentiment value of each word, β i Indicates the importance of each word, em i Indicates the sentiment value of each word; after calculating the total sentiment value of all words in a communication data, the word with the largest total sentiment value is selected as the sentiment feature of the communication data;

[0028] S303, using the sentiment dictionary to classify the sentiment features of the communication data of all communications into positive and negative; extracting the associated user information between all the communication users of the electronic business cards with positive and negative sentiment features respectively, and comparing the extracted associated user information, and filtering out the associated user information that exists independently in the positive sentiment features and the negative sentiment features respectively; marking the associated user information that exists independently by using the sentiment features, and using the associated user information marked with the positive sentiment features as the mining factor for each user, and using the associated user information marked with the negative sentiment features as the shielding factor for each user.

[0029] Based on the historical communication data of users, the corresponding emotional features are extracted, and the user information in the user's electronic business card is marked using the emotional features to generate mining elements for each user. The mining elements are used to find potential customers to achieve personalized intelligent recommendations. Not only are recommendations made based on the user's own conditions, but they also include the user's preferences during communication, making the recommendation of potential customers more humane and in line with the user's senses.

[0030] S400, when a real-time user logs into the system and uses an electronic business card, mining the electronic business card of each user in the user social relationship network using the mining elements generated by the real-time user, discovering potential customers that match the real-time user, and recommending the electronic business card of the potential customer to the real-time user;

[0031] Furthermore, the specific steps of recommending the potential customer's electronic business card to the real-time user are as follows:

[0032] S401, when a real-time user logs into the system and uses an electronic business card, the electronic business card of each user is mined in the implicit social relationship network using the mining elements generated by the real-time user, and the mining elements are compared with each electronic business card in the implicit social relationship network. When comparing, the mining weight is configured using the total sentiment value of each mining element in S300, and when the mining element exists in the electronic business card and the shielding element does not exist, the user corresponding to the electronic business card is judged to be a potential customer; when the shielding element exists, the electronic business card is shielded;

[0033] When mining potential customers, the weight of each mining factor is calculated as In the formula, R represents the weight of each mining factor, E represents the total sentiment value of each mining factor, and f represents the number of mining factors.

[0034] S402: Use the mining elements generated by the real-time user to mine in the explicit social relationship network, extract the user nodes associated with the real-time user in the explicit social relationship network, and indirectly use the associated user system to mine.

[0035] The mining elements of the user node search for potential customers associated with the user node according to the method in S401, and use the potential customers associated with the user node as potential customers of the real-time user; all potential customers in the invisible and explicit social relationship networks are combined and recommended to the real-time user.

[0036] S500, when recommending potential customers to real-time users, trace the potential customers' mining process to generate recommendation features for each potential customer; use the recommendation features to mark the recommended potential customers, visualize the recommendation reasons for each recommended potential customer, and enable the real-time users to select and communicate with potential customers based on the recommendation features;

[0037] Furthermore, the specific steps for enabling real-time users to select and communicate with potential customers based on the recommended features are as follows:

[0038] S501, when recommending potential customers to real-time users, trace the mining process of potential customers, extract the types of mining elements that exist when each potential customer is mined, and use the mining element with the largest weight among the types of existing mining elements as the recommendation feature for each potential customer;

[0039] S502. Use recommendation features to mark recommended potential customers, visualize the recommendation reasons for each recommended potential customer, and enable real-time users to select and communicate with potential customers based on the recommendation features; provide personalized potential customer recommendation plans.

[0040] Visualizing the recommended features of potential customers can clearly display a variety of potential customers, making it easier for users to understand the reasons for the recommendation and enhancing the transparency of the system;

[0041] S600. When a real-time user selects a potential customer for communication, the real-time user communication data is collected and analyzed to obtain key features of the communication between the real-time user and the potential customer and match them with the potential customer's electronic business card, and the mining factors of the real-time user are optimized using the electronic business card information corresponding to the key features.

[0042] Furthermore, the specific steps of optimizing the mining elements of real-time users by using the electronic business card information corresponding to the key features are as follows:

[0043] S601, when a real-time user selects a potential customer for communication, collect and analyze the real-time user communication data, calculate the total value of the user's sentiment towards each user information of the potential customer in the real-time communication data using the method in S300, and then compare and calculate the total value of the real-time user information sentiment and the weight difference of each mining factor in the mining factor; and use the user information corresponding to the maximum value of the difference as an important feature;

[0044] S602: Using the calculated differences of the important features, the weights of the mining element types corresponding to the important features in the mining elements are adjusted and optimized.

[0045] By continuously learning user behavior and feedback when mining potential customers, the recommendations and analysis results are more in line with user needs, improving the accuracy and personalization of AI, and solving the problem of system recommendations not meeting user expectations.

[0046] An electronic business card associated data analysis system based on artificial intelligence, the electronic business card associated data analysis system includes a data collection module, a database construction module, a social relationship network generation module, a mining element module, a mining module and an optimization module;

[0047] The data collection module is used to collect the communication data and electronic business cards of all users in history;

[0048] The database construction module is used to utilize all user electronic business cards collected in the cloud, analyze and standardize all different types of collected electronic business cards, normalize the user information in each electronic business card, and construct an electronic business card database;

[0049] The social relationship network generation module is used to generate an explicit social relationship network using historical communication data, and to generate an implicit social relationship network using user information in the electronic business card; and to combine the explicit social relationship network and the implicit social relationship network to construct a social relationship network of all users;

[0050] The mining element module is used to analyze the sentiment of each word in the historical user communication data, obtain the sentiment features in each communication data, and generate mining elements in combination with user information;

[0051] The mining module is used to mine potential customers of each user in the social relationship network according to the generated mining elements of each user and visualize them;

[0052] The optimization module is used to analyze the real-time communication data between users and potential customers, calculate important features, and use the important features to adjust and optimize the mining factors.

[0053] The social relationship network generation module includes explicit social relationship network units and implicit social relationship network units;

[0054] The explicit social relationship network unit is used to generate relationship edges from the communication data of each user in history, and to build an explicit social relationship network in combination with user nodes;

[0055] The implicit social relationship network is used to generate relationship edges according to the user information in all electronic business cards, and to construct the implicit social relationship network in combination with user nodes.

[0056] The mining module includes a mining unit and a visualization display unit;

[0057] The mining unit is used to mine potential customers of the user in the social relationship network according to the generated mining elements of each user;

[0058] The visual display unit is used to mark the recommended potential customers using the recommendation features, and visually display the recommendation reason for each recommended potential customer.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. The present invention marks the user information in the user's electronic business card by using emotional features to generate mining elements for each user; the mining elements are used to find potential customers to achieve personalized intelligent recommendations, which not only make recommendations based on the user's own conditions, but also include the user's preferences during communication, making the recommendation of potential customers more humane and in line with the user's senses.

[0061] 2. The present invention utilizes the communication data and electronic business cards of all users in history to construct explicit social relationship networks and implicit social relationship networks respectively, and combines the two social relationship networks to generate a comprehensive social relationship network. The generated comprehensive social relationship network simultaneously connects the relationships between customers subjectively searched by users and customers matched by electronic business cards, thereby ensuring the comprehensiveness of social relationships. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a module distribution diagram of an electronic business card associated data analysis system based on artificial intelligence of the present invention;

[0063] Figure 2 A schematic diagram of the steps of an electronic business card associated data analysis method based on artificial intelligence of the present invention;

[0064] Figure 3 A social relationship network diagram of an electronic business card associated data analysis method based on artificial intelligence in the present invention. DETAILED DESCRIPTION

[0065] 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.

[0066] Example: Figure 1-Figure 3 As shown, the present invention provides a technical solution.

[0067] An artificial intelligence-based electronic business card associated data analysis method, the method comprising the following steps:

[0068] S100, the user fills in his / her own information in the software to form his / her own electronic business card and transmits it to the cloud for storage, collects the electronic business cards of all users in the cloud, analyzes and standardizes all different types of collected electronic business cards, normalizes the user information in each electronic business card, and constructs an electronic business card database;

[0069] The specific steps to build an electronic business card database are:

[0070] S101. Collect all users' electronic business cards in the cloud. Suppose the format of all collected electronic business cards is {G1, G2, G3, ..., G p}, G1, G2, G3, ..., G p represents the 1st, 2nd, 3rd, ..., pth format of all collected user electronic business cards, where p is a positive integer; calculate the number of each electronic business card collected in the cloud, and select the one with the largest number as the standard format of the electronic business card Gb; extract user information from electronic business cards of different formats, and assume that the user information contained in the collected electronic business cards is {X1, X2, X3, ..., X n},

[0071] X1, X2, X3, ..., X n Indicates the first, second, third, ..., nth types of user information contained in the user's electronic business card, where n is a positive integer; re-fills the extracted user information in different formats according to the standard format, and standardizes the electronic business cards of all users collected in the cloud;

[0072] S102, after standardizing the electronic business cards of all users, convert each user information in the electronic business card into vector data Xi using an encoder, and normalize the vector data of each user information in the electronic business card, the formula is:

[0073]

[0074] In the formula, Xi' represents the normalized vector data, Xi represents the vector data of user information, and m represents the dimension of the transformed vector data of user information; the electronic business card database is constructed using the normalized vector data of user information.

[0075] Standardizing and normalizing all electronic business cards collected in the cloud unifies the information format in the electronic business cards. When integrating electronic business card data from multiple sources, this consistency enables the data to be merged and stored quickly and accurately, reducing data errors and duplicate entry problems caused by format differences; and ensuring the accurate transmission of important information in the electronic business card.

[0076] S200, collecting historical communication data between all users, using the communication data and user information in the electronic business card to build the social relationship of each user, and integrating the social relationships of all users to generate a social relationship network of users;

[0077] The specific steps to generate a user social relationship network are:

[0078] S201, collect the communication data between all users in history, take each user as a node, extract the communication ends in the collected communication data as {(a, b)1, (a, b)2, (a, b)3, ..., (a, b) m},

[0079] (a, b)1, (a, b)2, (a, b)3, ..., (a, b) m Indicates the communication ends in the 1st, 2nd, 3rd, ..., mth communication data in the collected history, uses the communication ends in each communication data to form a relationship edge, and finally forms m relationship edges, and uses user nodes and relationship edges to generate an explicit social relationship network;

[0080] S202, extract each type of user information from the electronic business cards of all users, take each user as a node, compare each type of user information from the extracted electronic business cards, compare the electronic business cards of two users, traverse and compare all the user information contained in the two electronic business cards, and when X∩X' exists, it is determined that the two users are connected and a relationship edge is generated; when X∩X' does not exist after traversing all the user information, it is determined that the two users are not connected; X and X' respectively represent the same type of user information contained in the electronic business cards of the two users; compare the electronic business cards of all users by traversing and comparing, set the connection that can generate a relationship edge to 1, and set the connection that cannot generate a relationship edge to 0, and construct a relationship edge matrix for all users as follows:

[0081]

[0082] In the matrix, each row and column represents a user node, and 1 / 0 represents the judgment result of whether two user nodes generate a relationship edge; after judging the relationship edges generated by all users, the relationship edge matrix is ​​output; each user node and the relationship edge matrix are used to construct an implicit social relationship network; the explicit social relationship network and the implicit social relationship network are combined to construct a social relationship network of all users.

[0083] The explicit social relationship network and implicit social relationship network are constructed by using the communication data and electronic business cards of all users in history. The two social relationship networks are combined to generate a comprehensive social relationship network. The generated comprehensive social relationship network connects the relationships between customers subjectively searched by users and customers matched by electronic business cards, ensuring the comprehensiveness of social relationships.

[0084] S300, using a natural language processing algorithm to analyze the language of each user in the history of communication with different electronic business card users, to obtain the emotional characteristics of each user in the history of communication with different users, and to generate mining factors for each user in combination with user information in different electronic business cards;

[0085] The specific steps of combining user information in different electronic business cards to generate mining elements for each user are:

[0086] S301, collect the communication data of each user in the history when communicating with other different users, extract all the language data of each communication data, extract all the words appearing in the language data, and calculate the importance of each word respectively, the formula is:

[0087]

[0088] In the formula, β represents the importance of each word, C i，k represents the number of the i-th language data in the k-th language data, represents the total number of words appearing in the kth language data; N represents the total number of language data, D i∈k represents the number of language data containing the i-th word; after calculation, the importance of all words is {β1, β2, β3, ..., β Q}, β1, β2, β3,..., β Q represents the importance of the 1st, 2nd, 3rd, ..., Qth words in the communication data, where Q is a positive integer;

[0089] S302, after calculating the importance of all words in the communication data, assign sentiment to each word according to the sentiment dictionary, and comprehensively analyze the importance and sentiment assignment of each word to obtain the total sentiment value of each word, the formula is: E = β i ×em i In the formula, E represents the total sentiment value of each word, β i Indicates the importance of each word, em i Indicates the sentiment value of each word; after calculating the total sentiment value of all words in a communication data, the word with the largest total sentiment value is selected as the sentiment feature of the communication data;

[0090] S303, using the sentiment dictionary to classify the sentiment features of the communication data of all communications into positive and negative; extracting the associated user information between all the communication users of the electronic business cards with positive and negative sentiment features respectively, and comparing the extracted associated user information, and filtering out the associated user information that exists independently in the positive sentiment features and the negative sentiment features respectively; marking the associated user information that exists independently by using the sentiment features, and using the associated user information marked with the positive sentiment features as the mining factor for each user, and using the associated user information marked with the negative sentiment features as the shielding factor for each user.

[0091] Based on the historical communication data of users, the corresponding emotional features are extracted, and the user information in the user's electronic business card is marked using the emotional features to generate mining elements for each user. The mining elements are used to find potential customers to achieve personalized intelligent recommendations. Not only are recommendations made based on the user's own conditions, but they also include the user's preferences during communication, making the recommendation of potential customers more humane and in line with the user's senses.

[0092] S400, when a real-time user logs into the system and uses an electronic business card, mining the electronic business card of each user in the user social relationship network using the mining elements generated by the real-time user, discovering potential customers that match the real-time user, and recommending the electronic business card of the potential customer to the real-time user;

[0093] The specific steps to recommend potential customers' electronic business cards to real-time users are:

[0094] S401, when a real-time user logs into the system and uses an electronic business card, the electronic business card of each user is mined in the implicit social relationship network using the mining elements generated by the real-time user, and the mining elements are compared with each electronic business card in the implicit social relationship network. When comparing, the mining weight is configured using the total sentiment value of each mining element in S300, and when the mining element exists in the electronic business card and the shielding element does not exist, the user corresponding to the electronic business card is judged to be a potential customer; when the shielding element exists, the electronic business card is shielded;

[0095] When mining potential customers, the weight of each mining factor is calculated as In the formula, R represents the weight of each mining factor, E represents the total sentiment value of each mining factor, and f represents the number of mining factors.

[0096] S402: Use the mining elements generated by the real-time user to mine in the explicit social relationship network, extract the user nodes associated with the real-time user in the explicit social relationship network, and indirectly use the associated user system to mine.

[0097] The mining elements of the user node search for potential customers associated with the user node according to the method in S401, and use the potential customers associated with the user node as potential customers of the real-time user; all potential customers in the invisible and explicit social relationship networks are combined and recommended to the real-time user.

[0098] S500, when recommending potential customers to real-time users, trace the potential customers' mining process to generate recommendation features for each potential customer; use the recommendation features to mark the recommended potential customers, visualize the recommendation reasons for each recommended potential customer, and enable the real-time users to select and communicate with potential customers based on the recommendation features;

[0099] The specific steps to enable real-time users to select and communicate with potential customers based on recommended features are:

[0100] S501, when recommending potential customers to real-time users, trace the mining process of potential customers, extract the types of mining elements that exist when each potential customer is mined, and use the mining element with the largest weight among the types of existing mining elements as the recommendation feature for each potential customer;

[0101] S502. Use recommendation features to mark recommended potential customers, visualize the recommendation reasons for each recommended potential customer, and enable real-time users to select and communicate with potential customers based on the recommendation features; provide personalized potential customer recommendation plans.

[0102] Visualizing the recommended features of potential customers can clearly display a variety of potential customers, making it easier for users to understand the reasons for the recommendation and enhancing the transparency of the system;

[0103] S600. When a real-time user selects a potential customer for communication, the real-time user communication data is collected and analyzed to obtain key features of the communication between the real-time user and the potential customer and match them with the potential customer's electronic business card, and the mining factors of the real-time user are optimized using the electronic business card information corresponding to the key features.

[0104] The specific steps for optimizing the mining elements of real-time users by using the electronic business card information corresponding to the key features are as follows:

[0105] S601, when a real-time user selects a potential customer for communication, collect and analyze the real-time user communication data, calculate the total value of the user's sentiment towards each user information of the potential customer in the real-time communication data using the method in S300, and then compare and calculate the total value of the real-time user information sentiment and the weight difference of each mining factor in the mining factor; and use the user information corresponding to the maximum value of the difference as an important feature;

[0106] S602: Using the calculated differences of the important features, the weights of the mining element types corresponding to the important features in the mining elements are adjusted and optimized.

[0107] By continuously learning user behavior and feedback when mining potential customers, the recommendations and analysis results are more in line with user needs, improving the accuracy and personalization of AI, and solving the problem of system recommendations not meeting user expectations.

[0108] An electronic business card associated data analysis system based on artificial intelligence, the electronic business card associated data analysis system includes a data collection module, a database construction module, a social relationship network generation module, a mining element module, a mining module and an optimization module;

[0109] The data collection module is used to collect the communication data and electronic business cards of all users in history;

[0110] The database construction module is used to utilize all user electronic business cards collected in the cloud, analyze and standardize all different types of collected electronic business cards, normalize the user information in each electronic business card, and construct an electronic business card database;

[0111] The social relationship network generation module is used to generate an explicit social relationship network using historical communication data, and to generate an implicit social relationship network using user information in the electronic business card; and to combine the explicit social relationship network and the implicit social relationship network to construct a social relationship network of all users;

[0112] The mining element module is used to analyze the sentiment of each word in the historical user communication data, obtain the sentiment features in each communication data, and generate mining elements in combination with user information;

[0113] The mining module is used to mine potential customers of each user in the social relationship network according to the generated mining elements of each user and visualize them;

[0114] The optimization module is used to analyze the real-time communication data between users and potential customers, calculate important features, and use the important features to adjust and optimize the mining factors.

[0115] The social relationship network generation module includes explicit social relationship network units and implicit social relationship network units;

[0116] The explicit social relationship network unit is used to generate relationship edges from the communication data of each user in history, and to build an explicit social relationship network in combination with user nodes;

[0117] The implicit social relationship network is used to generate relationship edges according to the user information in all electronic business cards, and to construct the implicit social relationship network in combination with user nodes.

[0118] The mining module includes a mining unit and a visualization display unit;

[0119] The mining unit is used to mine potential customers of the user in the social relationship network according to the generated mining elements of each user;

[0120] The visual display unit is used to mark the recommended potential customers using the recommendation features, and visually display the recommendation reason for each recommended potential customer.

[0121] Example: In a certain APP, the electronic business cards and historical communication data of all users are analyzed to obtain a social relationship network, where the total number of users is 5; first, the explicit social relationship network is obtained based on the historical communication data, such as Figure 3 As shown; the relationship edge matrix obtained after analyzing the electronic business card and comparing the user information is: The implicit social relationship network formed by Figure 3 As shown in Figure 2, the two social relationship networks are combined to form a comprehensive social relationship network. Figure 3 As shown;

[0122] Assume that according to historical communication data, the mining factors of user 1 are young, 25-35 years old, and above the position of director;

[0123] Mining elements are used to find potential customers in the generated social network for recommendation.

[0124] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An electronic business card related data analysis method based on artificial intelligence, characterized by: The method comprises the following steps: S100, the user fills in his / her own information in the software to form his / her own electronic business card and transmits it to the cloud for storage, collects the electronic business cards of all users in the cloud, analyzes and standardizes all different types of collected electronic business cards, normalizes the user information in each electronic business card, and constructs an electronic business card database; S200, collecting historical communication data between all users, using the communication data and user information in the electronic business card to build the social relationship of each user, and integrating the social relationships of all users to generate a social relationship network of users; S300, using a natural language processing algorithm to analyze the language of each user in the history of communication with different electronic business card users, to obtain the emotional characteristics of each user in the history of communication with different users, and to generate mining factors for each user in combination with user information in different electronic business cards; S400, when a real-time user logs into the system and uses an electronic business card, mining the electronic business card of each user in the user social relationship network using the mining elements generated by the real-time user, discovering potential customers that match the real-time user, and recommending the electronic business card of the potential customer to the real-time user; S500, when recommending potential customers to real-time users, trace the potential customers' mining process to generate recommendation features for each potential customer; use the recommendation features to mark the recommended potential customers, visualize the recommendation reasons for each recommended potential customer, and enable the real-time users to select and communicate with potential customers based on the recommendation features; S600. When a real-time user selects a potential customer for communication, the real-time user communication data is collected and analyzed to obtain key features of the communication between the real-time user and the potential customer and match them with the potential customer's electronic business card, and the mining factors of the real-time user are optimized using the electronic business card information corresponding to the key features.

2. The method for analyzing electronic business card related data based on artificial intelligence according to claim 1, characterized in that: The specific steps of constructing the electronic business card database in S100 are: S101. Collect all users' electronic business cards in the cloud. Suppose the format of all collected electronic business cards is {G1, G2, G3, ..., G p }, G1, G2, G3, ..., G p represents the 1st, 2nd, 3rd, ..., pth format of all collected user electronic business cards, where p is a positive integer; calculate the number of each electronic business card collected in the cloud, and select the one with the largest number as the standard format of the electronic business card Gb; extract user information from electronic business cards of different formats, and assume that the user information contained in the collected electronic business cards is {X1, X2, X3, ..., X n }, X1, X2, X3, ..., X n Indicates the first, second, third, ..., nth types of user information contained in the user's electronic business card, where n is a positive integer; re-fills the extracted user information in different formats according to the standard format, and standardizes the electronic business cards of all users collected in the cloud; S102, after standardizing the electronic business cards of all users, convert each user information in the electronic business card into vector data Xi using an encoder, and normalize the vector data of each user information in the electronic business card, the formula is: In the formula, Xi' represents the normalized vector data, Xi represents the vector data of user information, and m represents the dimension of the transformed vector data of user information; the electronic business card database is constructed using the normalized vector data of user information.

3. The method for analyzing electronic business card related data based on artificial intelligence according to claim 1, characterized in that: The specific steps of generating the user social relationship network in S200 are: S201, collect the communication data between all users in history, take each user as a node, extract the communication ends in the collected communication data as {(a, b)1, (a, b)2, (a, b)3, ..., (a, b) m }, (a, b)1, (a, b)2, (a, b)3, ..., (a, b) m Indicates the communication ends in the 1st, 2nd, 3rd, ..., mth communication data in the collected history, uses the communication ends in each communication data to form a relationship edge, and finally forms m relationship edges, and uses user nodes and relationship edges to generate an explicit social relationship network; A202, extract each user information in the electronic business cards of all users, take each user as a node, compare each user information in the extracted electronic business cards, compare the electronic business cards of two users, traverse and compare all user information contained in the two electronic business cards, when X∩X' exists, it is judged that the two users are connected, and a relationship edge is generated; when X∩X' does not exist after traversing all user information, it is judged that the two users are not connected; X and X' respectively represent the same type of user information contained in the electronic business cards of the two users; compare the electronic business cards of all users by traversing and comparing, set the connection that can generate a relationship edge to 1, and set the connection that cannot generate a relationship edge to 0, and construct the relationship edge matrix of all users as follows: In the matrix, each row and column represents a user node, and 1 / 0 represents the judgment result of whether two user nodes generate a relationship edge. After judging the relationship edges generated by all users, the relationship edge matrix is ​​output. Each user node and the relationship edge matrix are used to construct a hidden social relationship network. Combine explicit social relationship networks and implicit social relationship networks to construct the social relationship networks of all users.

4. The method for analyzing electronic business card related data based on artificial intelligence according to claim 1, characterized in that: The specific steps of generating mining elements for each user by combining user information in different electronic business cards in S300 are: S301, collect the communication data of each user in the history when communicating with other different users, extract all the language data of each communication data, extract all the words appearing in the language data, and calculate the importance of each word respectively, the formula is: In the formula, β represents the importance of each word, C i，k represents the number of the i-th language data in the k-th language data, represents the total number of words appearing in the kth language data; N represents the total number of language data, D i∈k represents the number of language data containing the i-th word; after calculation, the importance of all words is {β1, β2, β3, ..., β Q }, β1, β2, β3,..., β Q represents the importance of the 1st, 2nd, 3rd, ..., Qth words in the communication data, where Q is a positive integer; S302, after calculating the importance of all words in the communication data, assign sentiment to each word according to the sentiment dictionary, and comprehensively analyze the importance and sentiment assignment of each word to obtain the total sentiment value of each word, the formula is: E = β i ×em i In the formula, E represents the total sentiment value of each word, β i Indicates the importance of each word, em i Indicates the sentiment assignment of each word; After calculating the total sentiment value of all words in a communication data, the one with the largest total sentiment value is selected as the sentiment feature of the communication data; S303, using the sentiment dictionary to classify the sentiment characteristics of the communication data of all the communications into positive and negative; extracting the associated user information between all the communication users of the electronic business cards with positive and negative sentiment characteristics, and comparing the extracted associated user information, and respectively screening out the associated user information that exists alone in the positive sentiment characteristics and the negative sentiment characteristics; The emotional features are used to mark the associated user information that exists separately, and the associated user information marked with positive emotional features is used as the mining factor of each user, and the associated user information marked with negative emotional features is used as the shielding factor of each user.

5. The method for analyzing electronic business card related data based on artificial intelligence according to claim 1, characterized in that: The specific steps of recommending the potential customer's electronic business card to the real-time user in S400 are: S401, when a real-time user logs into the system and uses an electronic business card, the electronic business card of each user is mined in the implicit social relationship network using the mining elements generated by the real-time user, and the mining elements are compared with each electronic business card in the implicit social relationship network. When comparing, the mining weight is configured using the total sentiment value of each mining element in S300, and when the mining element exists in the electronic business card and the shielding element does not exist, the user corresponding to the electronic business card is judged to be a potential customer; when the shielding element exists, the electronic business card is shielded; S402. Use the mining elements generated by the real-time user to mine in the explicit social relationship network, extract the user nodes associated with the real-time user in the explicit social relationship network, indirectly use the mining elements of the associated user nodes to find potential customers of the associated user nodes according to the method in S401, and use the potential customers of the associated user nodes as the potential customers of the real-time user; combine all potential customers in the implicit and explicit social relationship networks and recommend them to the real-time user.

6. The method for analyzing electronic business card related data based on artificial intelligence according to claim 1, characterized in that: The specific steps of enabling the real-time user to select and communicate with potential customers according to the recommendation features in S500 are: S501, when recommending potential customers to real-time users, trace the mining process of potential customers, extract the types of mining elements that exist when each potential customer is mined, and use the mining element with the largest weight among the types of existing mining elements as the recommendation feature for each potential customer; S502. Use recommendation features to mark recommended potential customers, visualize the recommendation reasons for each recommended potential customer, and enable real-time users to select and communicate with potential customers based on the recommendation features; provide personalized potential customer recommendation plans.

7. The method for analyzing electronic business card related data based on artificial intelligence according to claim 1, characterized in that: The specific steps of optimizing the mining elements of real-time users by using the electronic business card information corresponding to the key features in S600 are as follows: S601, when a real-time user selects a potential customer for communication, collect and analyze the real-time user communication data, calculate the total value of the user's sentiment towards each user information of the potential customer in the real-time communication data using the method in S300, and then compare and calculate the total value of the real-time user information sentiment and the weight difference of each mining factor in the mining factor; and use the user information corresponding to the maximum value of the difference as an important feature; S602: Using the calculated differences of the important features, the weights of the mining element types corresponding to the important features in the mining elements are adjusted and optimized.

8. An artificial intelligence-based electronic business card related data analysis system, characterized by: The electronic business card associated data analysis system includes a data collection module, a database construction module, a social relationship network generation module, a mining element module, a mining module and an optimization module; The data collection module is used to collect the communication data and electronic business cards of all users in history; The database construction module is used to utilize all user electronic business cards collected in the cloud, analyze and standardize all different types of collected electronic business cards, normalize the user information in each electronic business card, and construct an electronic business card database; The social relationship network generation module is used to generate an explicit social relationship network using historical communication data and to generate an implicit social relationship network using user information in an electronic business card; Combine explicit social relationship networks and implicit social relationship networks to construct the social relationship networks of all users; The mining element module is used to analyze the sentiment of each word in the historical user communication data, obtain the sentiment features in each communication data, and generate mining elements in combination with user information; The mining module is used to mine potential customers of each user in the social relationship network according to the generated mining elements of each user and visualize them; The optimization module is used to analyze the real-time communication data between users and potential customers, calculate important features, and use the important features to adjust and optimize the mining factors.

9. The electronic business card related data analysis system based on artificial intelligence according to claim 8, characterized in that: The social relationship network generation module includes an explicit social relationship network unit and an implicit social relationship network unit; The explicit social relationship network unit is used to generate relationship edges from the communication data of each user in history, and to build an explicit social relationship network in combination with user nodes; The implicit social relationship network is used to generate relationship edges according to the user information in all electronic business cards, and to construct the implicit social relationship network in combination with user nodes.

10. The electronic business card related data analysis system based on artificial intelligence according to claim 8, characterized in that: The mining module includes a mining unit and a visual display unit; The mining unit is used to mine potential customers of the user in the social relationship network according to the generated mining elements of each user; The visual display unit is used to mark the recommended potential customers using the recommendation features, and visually display the recommendation reason for each recommended potential customer.

Images