Customer information portrait construction processing system

Abstract

The invention relates to the technical field of data analysis, in particular to a customer information portrait construction processing system, which comprises a customer information acquisition module for acquiring customer information and customer browsing data in a monitoring period, a content data acquisition module for acquiring platform content data, and a content feature construction module for constructing the customer information portrait. The system comprises a platform content data preprocessing module used for preprocessing platform content data and analyzing platform content labels, a customer element analysis module used for collecting customer behavior data in a monitoring period, analyzing customer portrait elements and constructing customer features in the monitoring period, and a portrait construction module used for constructing the customer features in the monitoring period. The portrait building module is used for carrying out portrait building on customer characteristics and platform content characteristics in a monitoring period and generating a customer portrait according to a matching result, and the timeliness updating module is used for collecting platform real-time browsing data and analyzing the timeliness of customer contents. According to the invention, the accuracy of customer portrait construction is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a customer information portrait construction and processing system. Background Art

[0002] With the development of big data and artificial intelligence technology, enterprises are paying more and more attention to understanding customer behavior and preferences through data analysis. Traditional customer information management methods usually only record basic transaction information and lack a comprehensive understanding of customers. Therefore, a system that can comprehensively analyze multi-source data and build detailed customer portraits is needed to help enterprises better make personalized services and marketing decisions.

[0003] Chinese patent publication number CN116484109A discloses a customer portrait analysis system and method based on artificial intelligence, the system includes a capture module for capturing several access response web pages of users during the access process; a recording module for recording the operation data performed in the access response web pages and the response data based on the operation data; a first determination module for determining the primary attribute model of the user; a detection and correction module for detecting the browsing time of the user in the response data, and correcting the primary attribute model to form a secondary attribute model; a second determination module for determining the first weight coefficient of the primary attribute model, and determining the second weight coefficient of the secondary attribute model; a construction module for constructing user tags, with the set of user tags as the user portrait; it can be seen that the invention does not analyze the timeliness of customer information, and its construction of customer portraits has a time error, and there is a problem of low efficiency in customer portrait construction. Summary of the invention

[0004] The object of the present invention is to provide a customer information profile construction and processing system to solve at least one of the problems existing in the prior art.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A customer information profile construction and processing system, characterized by comprising: Customer information collection module, used to collect customer information and customer browsing data within the monitoring period; Content data collection module, used to collect platform content data; Content feature construction module, used to pre-process platform content data and analyze platform content tags based on the pre-processing results; The customer factor analysis module is used to collect customer behavior data within the monitoring period, and analyze customer profile factors based on customer information, customer browsing data within the monitoring period, customer behavior data within the monitoring period, and platform content tags, and also to construct customer characteristics within the monitoring period based on the analysis results of customer profile factors; The portrait construction module is used to construct portraits of customer characteristics and platform content characteristics within the monitoring period; The timeliness update module is used to collect real-time browsing data of the platform, analyze the timeliness of customer content based on the real-time browsing data of the platform, and update the portrait construction process of customer characteristics and content characteristics of each platform based on the analysis results.

[0006] Further, the content feature construction module includes a data preprocessing unit, and the data preprocessing unit is used to preprocess the platform content data; The data preprocessing unit performs word segmentation on the platform content text to obtain content phrases; The data preprocessing unit counts the word frequency tf(i,j) of each word in the content phrase.

[0007] Furthermore, the content feature construction module also includes a content label analysis unit, which analyzes the platform content label according to the frequency statistics of each word in the content phrase: if tf(i,j)<TF, the content label analysis unit determines that the word is not the platform content label of the platform content data; if tf(i,j)≥TF, the content label analysis unit determines that the word is the platform content label of the platform content data; wherein TF is the label word frequency threshold.

[0008] Further, the customer factor analysis module includes a customer habit analysis unit, which is used to analyze customer habit weights based on customer information and customer browsing data within a monitoring period; The customer habit analysis unit compares the number of clicks n1 of the customer recommended content with the number of clicks of each preset recommended content, and judges the customer habits based on the comparison result, and then sets the customer habit weight according to the judgment result of the customer habits. The judgment result of the customer habits includes difficult to recommend customers, ordinary recommended customers and easy to recommend customers. When the customer habit is a difficult to recommend customer, the customer habit weight is set to w1; when the customer habit is an ordinary recommended customer, the customer habit weight is set to w2; when the customer habit is an easy to recommend customer, the customer habit weight is set to w3.

[0009] Furthermore, the customer factor analysis module also includes a customer demand analysis unit, which is used to analyze customer demand weights according to customer behavior data within a monitoring period and customer browsing data within a monitoring period; The customer demand analysis unit is used to count the number of times the customer searches for content tags on each platform, and to set the customer demand vector α according to the statistical results; The customer demand analysis unit analyzes the customer demand type according to the customer demand vector α, and sets the customer demand weight according to the analysis result of the customer demand type. The analysis result of the customer demand type includes non-demand guidance customers and demand guidance customers. When the customer demand type is a non-demand guidance customer, the customer demand weight is set to w4; when the customer demand type is a demand guidance customer, the customer demand weight is set to w5.

[0010] Furthermore, the customer factor analysis module further includes a customer interest analysis unit, which is used to analyze customer interest weights according to customer behavior data within a monitoring period; The customer interest analysis unit counts the number of content tags browsed by customers on each platform during the monitoring period, and sorts them in descending order according to the statistical results; The customer interest analysis unit sums the number of tags browsed by customers whose ranking results are first, second and third, to obtain the number of feature tag browsing n2; The customer interest analysis unit compares the feature tag browsing number n2 with the preset number N3 to determine the customer interest weight type, and then sets the customer interest weight according to the customer interest weight type. The analysis results of the customer interest weight type include non-interest-oriented customers and interest-oriented customers, and when the customer interest type is a non-interest-oriented customer, the customer interest weight is set to w6; when the customer interest type is an interest-oriented customer, the customer interest weight is set to w7.

[0011] Furthermore, the customer factor analysis module further includes a customer feature construction unit, which is used to construct customer features within the monitoring period according to the customer habit weight analysis results, the customer interest weight analysis results and the customer demand weight analysis results; The customer feature construction unit is used to calculate the customer feature vector β according to the customer habit weight, the customer interest weight and the customer demand weight.

[0012] Further, the portrait construction module includes a portrait construction unit, which is used to construct a customer portrait according to the customer feature analysis results within the monitoring period; The portrait construction unit is used to calculate the customer recommendation coefficient γ according to the customer characteristics in the monitoring period, and set γ=|β|; The portrait construction unit is used to analyze the content tags of the platform to be recommended according to the content data of the platform to be recommended, and count the number of types of the content tags of the platform to be recommended N4; The portrait construction unit counts the number N5 of platform content tags of customer browsing data within the monitoring period, and calculates the proportion μ of the platform content tags to be recommended.

[0013] Furthermore, the portrait construction unit is used to construct a customer portrait according to the customer recommendation coefficient γ and the proportion μ of the content tags of the platform to be recommended, and the construction results of the customer portrait include easy-to-recommend customers, ordinary customers and difficult-to-recommend customers.

[0014] Furthermore, the timeliness update module is used to collect real-time browsing data of the platform, analyze the real-time browsing content tags of the real-time browsing data of the platform, and count the real-time browsing content tags according to the analysis results; The timeliness update module is used to analyze the timeliness of customer content according to the statistical results of real-time browsing content tags, and to update the portrait construction process of customer characteristics and each platform content characteristics according to the timeliness analysis results of customer content: if g<G, the timeliness update module determines that the timeliness of the customer content is normal and does not update it; if g≥G, the timeliness update module determines that the timeliness of the customer content is abnormal, and updates the second preset recommendation coefficient to Y2'; wherein g is the proportion of real-time browsing content tag types in the platform content tag types of customer browsing data within the monitoring period, and G is the proportion threshold.

[0015] Compared with the prior art, the beneficial effects of the present invention are: the present invention ensures the comprehensiveness, accuracy and timeliness of customer portraits and improves the accuracy of customer portrait construction through comprehensive data collection, refined data preprocessing, multi-dimensional customer factor analysis, comprehensive customer feature construction and real-time timeliness update. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 A schematic diagram of the structure of the customer information profiling processing system in this embodiment.

[0018] Figure 2 This is a schematic diagram of the structure of the content feature building module in this embodiment.

[0019] Figure 3 Schematic diagram of the structure of the customer factor analysis module of this embodiment.

[0020] Figure 4 This is a schematic diagram of the structure of the portrait building module in this embodiment. DETAILED DESCRIPTION

[0021] In order to more clearly illustrate the present invention, the present invention is further described below in conjunction with preferred embodiments and accompanying drawings. Similar components in the accompanying drawings are represented by the same reference numerals. It should be understood by those skilled in the art that the content specifically described below is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present invention.

[0022] It should be noted that, although the terms first, second, third, etc. may be used to describe in the embodiments of the present application, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first.

[0023] Specifically, the customer information portrait construction and processing system described in this embodiment is applied to information push of an Internet data management platform of multiple customers; the Internet data management platform described in this embodiment is specifically a content creation platform.

[0024] It is understood that the information (including but not limited to customer information) and data (including but not limited to customer browsing data) involved in this application are all information and data authorized by the customer or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant national and regional laws, regulations and standards.

[0025] See also Figure 1 As shown, it is a structural diagram of a customer information portrait construction and processing system of this embodiment, including a customer information collection module, which is used to collect customer information and customer browsing data within a monitoring period.

[0026] Specifically, the customer information includes the customer's age and gender; the customer browsing data refers to the content data of multiple platforms browsed by the user during the monitoring period, and the platform content data refers to the content browsed by the customer on the Internet data management platform; it can be understood that the platform content data described in this embodiment includes platform content text; at the same time, this embodiment does not specifically limit the method for obtaining customer browsing data, and those skilled in the art can set it freely, and only need to meet the needs of obtaining customer browsing data, such as obtaining it through the Internet data management platform database of the system application; this embodiment does not specifically limit the value of the duration of the monitoring period, and those skilled in the art can set it freely, and only need to meet the needs of the duration of the monitoring period. In this embodiment, the value of the duration of the monitoring period is set to 2 days.

[0027] Please continue reading Figure 1 As shown, the system also includes a content data acquisition module, which is used to acquire platform content data.

[0028] Please continue reading Figure 1 As shown, the system is also provided with a content feature construction module, which is connected to the content data acquisition module. The content feature construction module is used to analyze the platform content tags according to the platform content data, and to construct the platform content features according to the analysis results of each platform content tag.

[0029] See also Figure 2 As shown, the content feature construction module includes a data preprocessing unit, which is used to preprocess the platform content data to provide raw data for subsequent analysis of platform content tags; The data preprocessing unit performs word segmentation on the platform content text to obtain content phrases; The data preprocessing unit counts the word frequency tf(i,j) of each word in the content phrase, and sets tf(i,j)=d(i,j) / M(j), wherein tf(i,j) represents the word frequency of the i-th word in the j-th platform content data, d(i,j) is the number of occurrences of the i-th word in the j-th platform content data, and M(j) is the total number of words in the content phrase of the j-th platform content data; by segmenting the platform content text and calculating the word frequency of each segmentation result one by one, accurate judgment of the labels of different content data is achieved.

[0030] Please continue reading Figure 2 As shown, the content feature construction module also includes a content label analysis unit, which is connected to the data preprocessing unit. The content label analysis unit analyzes the platform content label according to the frequency statistics of each word in the content phrase, and uses the platform content label to refer to an operation point of the customer, so as to reflect the characteristics behind the customer operation: if tf(i,j)<TF, the content label analysis unit determines that the word is not the platform content label of the platform content data; if tf(i,j)≥TF, the content label analysis unit determines that the word is the platform content label of the platform content data; wherein TF is the label frequency threshold; by performing threshold comparison on the frequency of different words to extract their feature words, the accuracy of platform content label judgment is improved, the diversity of customer portrait element analysis is increased, and the accuracy of customer feature construction is thereby improved.

[0031] Specifically, this embodiment does not specifically limit the value of the label word frequency threshold TF, and those skilled in the art can freely set it as long as the value requirement of TF is met. For example, TF can be set to 2%.

[0032] Please continue reading Figure 1As shown, the system also includes a customer factor analysis module, which is connected to the customer information collection module and the content feature construction module. The customer factor analysis module is used to collect customer behavior data within a monitoring period, and analyze customer portrait elements based on customer information, customer browsing data within a monitoring period, customer behavior data within a monitoring period, and platform content tags, and is also used to construct customer features within the monitoring period based on the analysis results of the customer portrait elements.

[0033] Specifically, the customer profile elements include customer habit weights, customer interest weights and customer demand weights.

[0034] Specifically, the customer behavior data includes the number of clicks on the customer's recommended content and the number of times the customer searches for content tags on each platform, where the number of times the customer searches for content tags on each platform is the number of times the customer searches for different tags through the platform; this embodiment does not specifically limit the method for obtaining customer behavior data, and technicians in this field can set it freely as long as the needs for obtaining customer behavior data are met, such as obtaining it by writing an application in a programming language.

[0035] See also Figure 3 As shown, the customer factor analysis module includes a customer habit analysis unit, which is used to analyze customer habit weights according to customer information and customer browsing data within a monitoring period, and use customer habit weights to quantify the degree to which customers are likely to accept recommended content, thereby achieving accurate judgment of customer habits; The customer habit analysis unit compares the number of clicks n1 of the customer's recommended content with the number of clicks on each preset recommended content, and judges the customer's habits based on the comparison result, and then sets the customer's habit weight based on the judgment result of the customer's habits: if n1<N1, the customer habit analysis unit determines that the customer's habits are not easy to recommend customers, and sets the customer's habit weight to w1, setting w1=1-exp{1-[(N1-n) / N1]}; if N1≤n1<N2, the customer habit analysis unit determines that the customer's habits are ordinary recommended customers, and sets the customer's habit weight to w2, setting w2=1; if n1≥N2, the customer habit analysis unit determines that the customer's habit is easy to recommend, and sets the customer's habit weight to w3, setting w3=exp{(n-N2) / N2}; the customer habit analysis unit uses the number of clicks on the customer's recommended content as the basis for judging the customer's habit, and sets an appropriate threshold to achieve accurate judgment of the customer's habit, so as to quantify the customer's habit of the recommended content, thereby improving the accuracy of the calculation of the customer's habit weight, increasing the diversity of the customer portrait element analysis, and thereby improving the accuracy of the customer feature construction; Wherein, N1 is the number of clicks on the first preset recommended content, N2 is the number of clicks on the second preset recommended content, and N1<N2.

[0036] Specifically, this embodiment does not impose any specific restrictions on the values ​​of the first preset number of recommended content clicks N1 and the second preset number of recommended content clicks N2. Technical personnel in this field can set them freely as long as the value requirements of the first preset number of recommended content clicks N1 and the second preset number of recommended content clicks N2 are met. For example, the first preset number of recommended content clicks N1 can be set to 10, and the second preset number of recommended content clicks N2 can be set to 20.

[0037] Please continue reading Figure 3 As shown, the customer factor analysis module also includes a customer demand analysis unit, which is used to analyze customer demand weights according to customer behavior data and customer browsing data within the monitoring period, and use customer demand weights to quantify the degree of customer demand for content, thereby achieving accurate judgment of customer demand; The customer demand analysis unit is used to count the number of times the customer searches for content tags on each platform, and to set the customer demand vector α according to the statistical results; The customer demand analysis unit analyzes the customer demand weight according to the customer demand vector α: if |α|<A, the customer demand analysis unit determines that the customer demand type is a non-demand-guided customer, and sets the customer demand weight to w4, setting w4=0; if |α|≥A, the customer demand analysis unit determines that the customer demand type is a demand-guided customer, and sets the customer demand weight to w5, setting w5=ln{e+[|α|-A] / A]}; wherein A is the demand vector modulus threshold; the customer demand analysis unit uses the number of content tags searched by the customer on each platform as the data basis for judging the customer demand type, and uses this to set the customer demand vector α, and then judges the customer's demand type, thereby achieving accurate and efficient judgment of the customer demand type, so as to increase the diversity of customer portrait element analysis, and thereby improve the accuracy of customer feature construction.

[0038] It is understandable that this embodiment does not impose any specific limitation on the value of the demand vector modulus threshold A, and those skilled in the art can freely set it as long as the value requirement of the demand vector modulus threshold A is met. For example, the demand vector modulus threshold A can be set to 1.3.

[0039] Specifically, the process of "setting the customer demand vector α according to the statistical results" in this embodiment is as follows: the number of times the customer searches for the content tags of each platform is used as the sub-vector of each dimensional tag. For example, the number of searches for the entity search tag is 12, and the number of searches for the content tags of other platforms is 0, then the sub-vectors of the customer's labels of each dimensionality are (12, 0, 0...), and the customer demand vector α is the vector after Min-Max normalization of the sub-vectors of the labels of each dimensionality (12, 0, 0...); in this embodiment, "|α|" is specifically the result of a modulo operation on the customer demand vector α.

[0040] Please continue reading Figure 3 As shown, the customer factor analysis module also includes a customer interest analysis unit, which is used to analyze customer interest weights according to customer behavior data within a monitoring period, and use customer interest weights to refer to the degree of interest of customers in content with different platform content tags, so as to achieve accuracy in judging whether the customer is interest-oriented; The customer interest analysis unit counts the number of content tags browsed by customers on each platform during the monitoring period, and sorts them in descending order according to the statistical results; The customer interest analysis unit sums the number of tags browsed by customers whose ranking results are first, second and third, to obtain the number of feature tag browsing n2; The customer interest analysis unit compares the number n2 of feature tag browsing with the preset number N3, and analyzes the customer interest weight according to the comparison result: if n2<N3, the customer interest analysis unit determines that the customer interest type is a non-interest-oriented customer, and sets the customer interest weight to w6, setting w6=(N3-n2) / N3; if n2≥N3, the customer interest analysis unit determines that the customer interest type is an interest-oriented customer, and sets the customer interest weight to w7, setting w7=exp{(n2-N3) / N3}; the customer interest analysis unit counts the number of content tags browsed by customers on each platform during the monitoring period, and then sets the customer interest weights in different situations according to the statistical results, so as to quantify the customer's interest in different content, so as to increase the diversity of customer portrait element analysis, and thus improve the accuracy of customer feature construction.

[0041] Specifically, this embodiment does not impose any specific limitation on the value of the preset number N3, and those skilled in the art can freely set it as long as the value requirement of the preset number N3 is met. For example, the preset number N3 can be set to 30% of the number of content tags browsed by customers on each platform.

[0042] Please continue reading Figure 3As shown, the customer factor analysis module also includes a customer feature construction unit, which is connected to the customer habit analysis unit, the customer interest analysis unit and the customer demand analysis unit. The customer feature construction unit is used to construct customer features within the monitoring period according to the customer habit weight analysis results, the customer interest weight analysis results and the customer demand weight analysis results, and quantify the customer's features in the three dimensions of demand, interest and habit; The customer feature construction unit is used to calculate the customer feature vector β according to the customer habit weight, customer interest weight and customer demand weight, and set β=(b1×customer habit weight, b2×customer demand weight, b3×customer interest weight); b1 is the customer habit correction weight, b2 is the customer demand correction weight, b3 is the customer interest correction weight, and b1+b2+b3=1.

[0043] Specifically, this embodiment does not impose any specific limitation on the values ​​of the customer habit correction weight, the customer demand correction weight, and the customer interest correction weight. Those skilled in the art can set them freely as long as the value requirements of the customer habit correction weight, the customer demand correction weight, and the customer interest correction weight are met. This embodiment sets b1=0.3, b2=0.5, and b3=0.2.

[0044] Please continue reading Figure 1 As shown, the system includes a portrait construction module, which is connected to the customer factor analysis module. The portrait construction module is used to generate a customer portrait based on the customer feature analysis results within a monitoring period, and is also used to output the customer portrait to the user.

[0045] See also Figure 4 As shown, the portrait construction module includes a portrait construction unit, which is used to construct a customer portrait according to the customer feature analysis results within the monitoring period, so as to achieve a deep match between the recommended content and the customer, and at the same time, to determine the probability of the customer accepting the recommendation; The portrait construction unit is used to calculate the customer recommendation coefficient γ according to the customer characteristics in the monitoring period, and set γ=|β|; The portrait construction unit is used to analyze the content tags of the platform to be recommended according to the content data of the platform to be recommended, and count the number of types of the content tags of the platform to be recommended N4; The portrait construction unit counts the number N5 of platform content tags of customer browsing data within the monitoring period, and calculates the proportion μ of the platform content tags to be recommended, and sets μ=N4 / N5; The portrait construction unit is used to construct a customer portrait according to the customer recommendation coefficient γ and the proportion μ of the content labels of the platform to be recommended: if γ×exp{μ}<Y1, the portrait construction unit determines that the customer portrait is an easy-to-recommend customer; if Y1≤γ×exp{μ}<Y2, the portrait construction unit determines that the customer portrait is an ordinary customer; if γ×exp{μ}≥Y2, the portrait construction unit determines that the customer portrait is a customer that is not easy to recommend; wherein Y1 is the first preset recommendation coefficient, Y2 is the second preset recommendation coefficient, and Y1<Y2; the portrait construction unit comprehensively analyzes the degree of matching between the content to be recommended and the customer, and obtains a concrete description of the customer's acceptance probability of the recommended content (that is, the construction result of the customer portrait includes easy-to-recommend customers, ordinary customers and not easy-to-recommend customers), thereby improving the accuracy of customer portrait construction.

[0046] Specifically, the "platform content data to be recommended" is obtained through user interactive input, which is the content data that the user needs to push to customers.

[0047] Please continue reading Figure 4 As shown, the portrait construction module also includes an output unit, and the output unit is used to output the customer portrait to the user.

[0048] Specifically, this embodiment does not give a detailed description of the process of "analyzing the content tags of the platform to be recommended based on the content data of the platform to be recommended", and its process is the same as the process of "preprocessing the platform content data, and analyzing the platform content tags according to the preprocessing results"; it can be understood that this embodiment does not make specific limitations on the values ​​of the first preset recommendation coefficient Y1 and the second preset recommendation coefficient Y2, and those skilled in the art can set them freely, as long as the value requirements of the first preset recommendation coefficient Y1 and the second preset recommendation coefficient Y2 are met. For example, the first preset recommendation coefficient Y1 can be set to 1.5, and the second preset recommendation coefficient Y2 can be set to 3.

[0049] Please continue reading Figure 1 As shown, the system is further provided with a timeliness analysis module, and the timeliness update module is connected to the portrait construction module. The timeliness update module is used to collect real-time browsing data of the platform, and analyze the timeliness of the customer content according to the real-time browsing data of the platform to determine the update frequency of the customer's acceptance of the content, and update the portrait construction process of the customer characteristics and the content characteristics of each platform according to the analysis results, so as to improve the accuracy of the customer portrait construction; The platform real-time browsing data is the platform content data browsed by customers counted in real time by the platform; the real-time statistics in this embodiment are specifically counted every 10 minutes; The timeliness update module analyzes the real-time browsing content tags of the real-time browsing data of the platform, and counts the real-time browsing content tags according to the analysis results; The timeliness update module is used to analyze the timeliness of customer content according to the statistical results of real-time browsing content tags, and to update the customer characteristics and the portrait construction process of each platform content characteristics according to the timeliness analysis results of the customer content: if g<G, the timeliness update module determines that the timeliness of the customer content is normal and does not update; if g≥G, the timeliness update module determines that the timeliness of the customer content is abnormal, and updates the second preset recommendation coefficient to Y2', setting Y2'=Y2×(gG) / G; wherein g is the proportion of real-time browsing content tag types in the platform content tag types of customer browsing data within the monitoring period, and G is the ratio threshold; by analyzing the timeliness of user acquisition of content through the timeliness update module, the system has the ability to process and analyze customer data in real time, and realizes the efficient construction of customer portraits.

[0050] Specifically, this embodiment does not impose any specific limitation on the value of the ratio threshold G, and those skilled in the art can freely set it as long as the value requirement of the ratio threshold G is met. For example, the ratio threshold G can be set to 0.3.

[0051] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A customer information profile construction and processing system, characterized in that: include: Customer information collection module, used to collect customer information and customer browsing data within the monitoring period; Content data collection module, used to collect platform content data; Content feature construction module, used to pre-process platform content data and analyze platform content tags based on the pre-processing results; The customer factor analysis module is used to collect customer behavior data within the monitoring period, and analyze customer profile factors based on customer information, customer browsing data within the monitoring period, customer behavior data within the monitoring period, and platform content tags, and also to construct customer characteristics within the monitoring period based on the analysis results of customer profile factors; A portrait building module is used to generate a customer portrait based on the customer feature analysis results within the monitoring period, and also to output the customer portrait to the user; The timeliness update module is used to collect real-time browsing data of the platform, analyze the timeliness of customer content based on the real-time browsing data of the platform, and update the portrait construction process of customer characteristics and content characteristics of each platform based on the analysis results.

2. The customer information profile construction and processing system according to claim 1, characterized in that: The content feature construction module includes a data preprocessing unit, and the data preprocessing unit is used to preprocess the platform content data; The data preprocessing unit performs word segmentation on the platform content text to obtain content phrases; The data preprocessing unit counts the word frequency tf(i,j) of each word in the content phrase.

3. The customer information profile construction and processing system according to claim 2, characterized in that: The content feature construction module also includes a content label analysis unit, which analyzes the platform content label according to the word frequency statistics of each word in the content phrase: if tf(i,j)<TF, the content label analysis unit determines that the word is not the platform content label of the platform content data; if tf(i,j)≥TF, the content label analysis unit determines that the word is the platform content label of the platform content data; wherein TF is a label word frequency threshold.

4. The customer information profile construction and processing system according to claim 3, characterized in that: The customer factor analysis module includes a customer habit analysis unit, which is used to analyze customer habit weights based on customer information and customer browsing data within a monitoring period; The customer habit analysis unit compares the number of clicks n1 of the customer recommended content with the number of clicks of each preset recommended content, and judges the customer habits based on the comparison result, and then sets the customer habit weight according to the judgment result of the customer habits. The judgment result of the customer habits includes difficult to recommend customers, ordinary recommended customers and easy to recommend customers. When the customer habit is a difficult to recommend customer, the customer habit weight is set to w1; when the customer habit is an ordinary recommended customer, the customer habit weight is set to w2; when the customer habit is an easy to recommend customer, the customer habit weight is set to w3.

5. The customer information profile construction and processing system according to claim 4, characterized in that: The customer factor analysis module also includes a customer demand analysis unit, which is used to analyze customer demand weights based on customer behavior data within a monitoring period and customer browsing data within a monitoring period; The customer demand analysis unit is used to count the number of times the customer searches for content tags on each platform, and to set the customer demand vector α according to the statistical results; The customer demand analysis unit analyzes the customer demand type according to the customer demand vector α, and sets the customer demand weight according to the analysis result of the customer demand type. The analysis result of the customer demand type includes non-demand guidance customers and demand guidance customers. When the customer demand type is a non-demand guidance customer, the customer demand weight is set to w4; when the customer demand type is a demand guidance customer, the customer demand weight is set to w5.

6. The customer information profile construction and processing system according to claim 5, characterized in that: The customer factor analysis module also includes a customer interest analysis unit, which is used to analyze customer interest weights according to customer behavior data within a monitoring period; The customer interest analysis unit counts the number of content tags browsed by customers on each platform during the monitoring period, and sorts them in descending order according to the statistical results; The customer interest analysis unit sums the number of tags browsed by customers whose ranking results are first, second and third, to obtain the number of feature tag browsing n2; The customer interest analysis unit compares the feature tag browsing number n2 with the preset number N3 to determine the customer interest weight type, and then sets the customer interest weight according to the customer interest weight type. The analysis results of the customer interest weight type include non-interest-oriented customers and interest-oriented customers, and when the customer interest type is a non-interest-oriented customer, the customer interest weight is set to w6; when the customer interest type is an interest-oriented customer, the customer interest weight is set to w7.

7. The customer information profile construction and processing system according to claim 6, characterized in that: The customer factor analysis module also includes a customer feature construction unit, which is used to construct customer features within the monitoring period according to the customer habit weight analysis results, the customer interest weight analysis results and the customer demand weight analysis results; The customer feature construction unit is used to calculate the customer feature vector β according to the customer habit weight, the customer interest weight and the customer demand weight.

8. The customer information profile construction and processing system according to claim 7, characterized in that: The portrait construction module includes a portrait construction unit, which is used to construct a customer portrait according to the customer feature analysis results within the monitoring period; The portrait construction unit is used to calculate the customer recommendation coefficient γ according to the customer characteristics in the monitoring period, and set γ=|β|; The portrait construction unit is used to analyze the content tags of the platform to be recommended according to the content data of the platform to be recommended, and count the number of types of the content tags of the platform to be recommended N4; The portrait construction unit counts the number N5 of platform content tags of customer browsing data within the monitoring period, and calculates the proportion μ of the platform content tags to be recommended.

9. The customer information profile construction and processing system according to claim 8, characterized in that: The portrait construction unit is used to construct a customer portrait according to the customer recommendation coefficient γ and the proportion μ of the content tags of the platform to be recommended. The construction results of the customer portrait include easy-to-recommend customers, ordinary customers and difficult-to-recommend customers.

10. The customer information profile construction and processing system according to claim 9, characterized in that: The timeliness update module is used to collect real-time browsing data of the platform, analyze the real-time browsing content tags of the real-time browsing data of the platform, and count the real-time browsing content tags according to the analysis results; The timeliness update module is used to analyze the timeliness of customer content according to the statistical results of real-time browsing content tags, and to update the portrait construction process of customer characteristics and each platform content characteristics according to the timeliness analysis results of customer content: if g<G, the timeliness update module determines that the timeliness of the customer content is normal and does not update it; if g≥G, the timeliness update module determines that the timeliness of the customer content is abnormal, and updates the second preset recommendation coefficient to Y2'; wherein g is the proportion of real-time browsing content tag types in the platform content tag types of customer browsing data within the monitoring period, and G is the proportion threshold.

Images