Intelligent data management system and method based on user portraits

Abstract

The invention discloses an intelligent data management system and method based on a user portrait, and relates to the technical field of data management, and the method comprises the steps: obtaining the occupation of a user and a template selection record, building the user portrait of the user, obtaining the matching degree of each design template, and recommending the design template to the user according to the matching degree; obtaining a marked template corresponding to the user, analyzing the marked template and the user portrait, and extracting a marked user in a template resource platform; feature words corresponding to the attribute features of each template are obtained, keywords in the feature words are extracted, and a keyword set corresponding to the marked user is established based on the keywords; and performing template preference investigation on the user, re-establishing a user portrait of the user, and recommending a design template to the user. According to the method, the user portrait is re-constructed for the user in combination with the historical selection record of the user and the personality factor of the user, the proper design template can be accurately recommended for the user, and the reliability of template recommendation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and specifically to a data intelligent management system and method based on user portraits. Background Art

[0002] A user portrait is usually a virtual user model with typical characteristics abstracted by integrating various types of user data. Based on data, it refines labels such as the core attributes, behavior patterns, and demand preferences of users, helping enterprises or product teams quickly understand target users and providing accurate bases for decision-making. Currently, when recommending design templates to target users, such as personal resumes and electronic business cards, it is usually based on the industry or occupational characteristics of the target users and historical selection records to construct user portraits and perform simple similar-type recommendations, but does not consider the personality factors of the target users, such as style preferences and personal hobbies. As a result, the construction of user portraits ignores the personality factors of the target users, unable to accurately recommend suitable design templates for the target users, making the recommended information lack reliability. Summary of the Invention

[0003] The purpose of the present invention is to provide a data intelligent management system and method based on user portraits to solve the problems raised in the prior art.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: A data intelligent management method based on user portraits, comprising the following steps: Step S100: Obtain the occupation of the user and the template selection records, extract the attribute characteristics of the corresponding design templates in the template selection records, establish a user portrait of the user based on the occupation and attribute characteristics, and obtain the matching degree of each design template to the user according to the user portrait, and recommend design templates to the user according to the matching degree; Step S200: Obtain the marked templates corresponding to the user based on the action behaviors of the user on the template resource platform; and analyze the marked templates and the user portrait to extract the marked users in the template resource platform; Step S300: Obtain the feature words corresponding to the attribute characteristics of each template according to the design templates liked and disliked by the marked users on the template resource platform and the corresponding marked templates; and obtain the degree of preference and deviation of the user for the feature words, and then extract the keywords in the feature words, and establish a keyword set corresponding to the marked users based on the keywords; Step S400: Conduct a template preference survey on the user, and re-establish the user portrait of the user according to the previous user portrait of the user, as well as the survey results and the keyword set, and recommend design templates to the user according to the re-established user portrait.

[0005] Further, step S100 includes: Step S110: Obtain all design templates corresponding to the user's template selection record, extract attribute features of each design template, including style, theme, applicable occupation and applicable scenario; the user's user profile corresponds to several preference tags, including style preference, theme preference, occupation preference and scenario preference; summarize all feature words corresponding to each attribute feature in the template selection record, obtain the number corresponding to each feature word, and use feature words with a number greater than a preset threshold as label words in the user's corresponding preference tag; then obtain all label words corresponding to each preference tag in the user's user profile, and set the initial word weight of each label word to 1; Step S120: Obtain a feature word W of a certain attribute feature d corresponding to a certain design template T in the template resource platform d , and extract all the label words under the preference label corresponding to the attribute feature d in the user portrait, and compare a label word L with the feature word W d The similarity between them is multiplied by the word weight of the label word L to obtain the label word L for the feature word W d The correlation degree of each tag word is obtained, and the maximum correlation degree is used as the correlation degree R between the attribute feature d and the user portrait d ; Then, the correlation degree between each attribute feature in the design template T and the user portrait is obtained, and the weight corresponding to each attribute feature is pre-set to obtain the matching degree of the design template T to the user, and the design template is recommended to the user in order of matching degree from large to small.

[0006] Furthermore, step S200 includes: Step S210: The template resource platform has the function of allowing users to search and filter design templates. The design template recommended in step S100 is used as a recommended template. If a user U uses any of the search or filter functions on the template resource platform to select a design template A and modifies all required content in all editable content of design template A, and the required content is the user's personal information content, and design template A is not a recommended template, then design template A is used as a marked template, thereby obtaining all marked templates. Step S220: Obtain all attribute features of a certain marking template B and the user portrait of the current user. According to all the label words under the preference label, analogously to step S120, obtain the matching degree of the marking template B to the user, and then obtain the matching degree of each marking template to the user, and add them up to get the average value. If the average value is less than the preset numerical threshold, the user U is marked as a marked user.

[0007] Furthermore, step S300 includes: Step S310: The template resource platform has the function of allowing users to mark their likes and dislikes for design templates; obtain all the design templates and corresponding marking templates that a certain user V has marked as having liked, and extract the feature words corresponding to the attribute characteristics of each template as the biased feature words of user V. The same biased feature words are statistically summarized to obtain the number corresponding to each biased feature word, and then the bias degree of user V to a biased feature word i is obtained as follows: X = 1- , where e is a natural constant, K1 is a deflection coefficient, and N i is the number corresponding to the feature word i; and is normalized according to the bias degree of each bias feature word; Step S320: Obtain all the design templates that user V dislikes, and extract the feature words corresponding to the attribute features of each template as the deviation feature words of user V. Statistically summarize the same deviation feature words to obtain the number corresponding to each deviation feature word, and then obtain the deviation degree of user V for a certain deviation feature word j: Y=-(1- ), where e is a natural constant, K2 is a deviation coefficient, and N j is the number corresponding to feature word j; normalization is performed according to the degree of deviation of each feature word; Step S330: If a feature word h is not a biased feature word and a deviation feature word at the same time, the feature word h is used as a keyword; if a feature word h is a biased feature word and a deviation feature word at the same time, the normalized bias degree X of the feature word h is used as the keyword. h and the degree of deviation Y h , get the feature degree Z of feature word h h =X h -|Y h |,|| is to find the absolute value, if |Z h |If the degree is greater than the preset threshold, the feature word h is used as the keyword; and based on all the keywords of user V, a keyword set of user V is established.

[0008] Since the formula y=1-e -x When x is x≥0, y is [0,1), which is a function in which y increases as x increases. In this scheme, the deflection coefficients K1 and K2 are not less than 0. i and N j Refers to the number, which is not less than 0, then the value of the bias degree X is between 0 and 1, and the value of the deviation degree Y is between -1 and 0, then the characteristic degree Z of the feature word h is h The value of Z is between -1 and 1. h When it is close to -1, it means that the user's deviation from the feature word h is relatively large (the degree of deviation X h Small, deviation degree Y h The absolute value of |Y h| is larger), similarly, when Z h When it is close to 1, it indicates that the user has a greater preference for the feature word h. The following step S400 uses the feature degree as the word weight, which is reasonable and in line with the expectation of this solution.

[0009] Use satisfy|Z here h |A degree greater than a preset threshold is used as a condition for filtering keywords because when the bias and deviation degrees of a feature word are similar, it means that this feature word is insignificant to the user. For example, if the user's occupation is also in the favorite template and the disliked template, then this word cannot be used as a keyword for the following decision; the keyword set is obtained based on the design templates that the user likes and dislikes on the template resource platform; the keyword set includes styles and themes that mark the user's likes and dislikes, and is a set that can represent the marked user's characteristics.

[0010] Furthermore, step S400 includes: Step S410: Conduct a template preference survey on all users. The template preference survey includes user style preferences, theme preferences, occupation preferences, and scenario preferences. Set the word weights of all preference keywords in the survey content to 1. If a user is not a tagged user, extract all corresponding label words in the previous user profile, aggregate all preference keywords and label words in the survey content to obtain a final label word, and add the corresponding word weights. Re-establish the user profile of the user based on the final label word. Step S420: If a user is a marked user, extract all keywords in the keyword set of the user, and use the feature degree corresponding to the keyword as the keyword's word weight, and extract all corresponding label words in the previous user portrait, summarize all the preference keywords and all label words in the survey content, and all keywords in the keyword set to obtain the final label words, and add up the corresponding word weights, and re-establish the user portrait of the user based on the final label words; and recommend design templates to the user according to the re-established user portrait.

[0011] A data intelligent management system based on user portraits, including a user portrait building module, a marked user extraction module, a keyword set building module and a design template recommendation module; User profile creation module: This module is used to obtain the user's occupation and template selection records, extract the attribute characteristics of the corresponding design templates in the template selection records, create a user profile based on the occupation and attribute characteristics, and determine the matching degree of each design template to the user based on the user profile, and recommend design templates to the user based on the matching degree; Tag User Extraction Module: It is used to obtain the corresponding tag template for the user based on the user's action behavior on the template resource platform; and analyze the tag template and the user portrait to extract the tagged users in the template resource platform. Keyword Set Establishment Module: It is used to obtain the feature words corresponding to the attribute features of each template according to the design templates liked and disliked by the tagged users on the template resource platform and the corresponding tag templates; and obtain the degree of preference and deviation of the users for the feature words, and then extract the keywords in the feature words, and establish the keyword set corresponding to the tagged users based on the keywords. Design Template Recommendation Module: It is used to conduct a template preference survey on the user, re - establish the user portrait of the user according to the previous user portrait of the user, the survey results and the keyword set, and recommend the design template to the user according to the re - established user portrait.

[0012] Further, the tag user extraction module includes a tag template extraction unit and a tag user extraction unit; Tag Template Extraction Unit: It is used to extract the tag template from the design templates according to the function that the template resource platform has for the user to search and filter the design templates and the obtained recommended templates. Tag User Extraction Unit: It is used to obtain all the attribute features of the tag template and the user portrait of the current user, and obtain the tagged users according to all the tag words under the preference tags.

[0013] Further, the design template recommendation module includes a user portrait re - establishment unit and a design template recommendation unit; User Portrait Re - establishment Unit: It is used to conduct a template preference survey on all users and set the word weights of all preference keywords in the survey content; extract all the corresponding tag words in the previous user portrait of the user to obtain the final tag words, and then re - establish the user portrait of a certain user. Design Template Recommendation Unit: It is used to recommend the design template to the user according to the re - established user portrait of a certain user.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention provides a data intelligent management system and method based on user portraits, including: obtaining the occupation and template selection records of users, establishing user portraits of users, obtaining the matching degrees of each design template, and recommending design templates to users according to the matching degrees; obtaining the marked templates corresponding to users, analyzing the marked templates and user portraits, and extracting marked users in the template resource platform; obtaining the feature words corresponding to the attribute features of each template, extracting the keywords in the feature words, and establishing a keyword set corresponding to the marked users based on the keywords; conducting a template preference survey on users, re - establishing user portraits of users, and recommending design templates to users. By combining the historical selection records of users and the personality factors of users, the present invention reconstructs user portraits for users, can accurately recommend suitable design templates for users, and improves the reliability of template recommendation. Brief Description of the Drawings

[0015] Figure 1 It is a schematic flowchart of a data intelligent management method based on user portraits according to the present invention; Figure 2 It is a structural diagram of a data intelligent management system based on user portraits according to the present invention. Detailed Embodiments

[0016] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0017] Embodiment: As Figure 1 shown, the present invention provides a technical solution for a data intelligent management method based on user portraits, including the following steps: Step S100: Obtain the occupation and template selection records of users, extract the attribute features of the corresponding design templates in the template selection records, establish user portraits of users based on the occupation and attribute features, and obtain the matching degrees of each design template to users according to the user portraits, and recommend design templates to users according to the matching degrees; Step S110: Obtain all the design templates corresponding to the user's template selection records, and extract the attribute features of each design template. The attribute features include style, theme, applicable occupation, and applicable scenario. The user's user profile corresponds to several preference tags, including style preference, theme preference, occupation preference, and scenario preference. Summarize all the feature words corresponding to each attribute feature in the template selection record to obtain the quantity corresponding to each feature word, and use the feature words with a quantity greater than the preset quantity threshold as the tag words in the preference tags corresponding to the user. Furthermore, obtain all the tag words corresponding to each preference tag in the user's user profile, and set the initial word weight of each tag word to 1. Step S120: Obtain the feature word W of a certain attribute feature d corresponding to a certain design template T in the template resource platform. d , and extract all the tag words under the preference tag corresponding to the attribute feature d in the user profile. Multiply the similarity between a certain tag word L and the feature word W d by the word weight of the tag word L to obtain the association degree of the tag word L to the feature word W d . Furthermore, obtain the association degree of each tag word, and use the maximum value of the association degrees as the association degree R between the attribute feature d and the user profile. d . Furthermore, obtain the association degree between each attribute feature in the design template T and the user profile, preset the weights corresponding to each attribute feature, obtain the matching degree of the design template T to the user, and recommend the design templates to the user in descending order of the matching degree.

[0018] Step S‎200: Obtain the marked template corresponding to the user based on the user's action behavior on the template resource platform; and analyze the marked template and the user profile to extract the marked users in the template resource platform. Step S210: The template resource platform has the functions of the user to search for and filter design templates. Use the design templates recommended in Step S100 as the recommended templates. If a certain user U uses any one of the search or filter functions on the template resource platform to select a certain design template A, and modifies all the required content in all the editable content on the design template A, the required content is the user's personal information content, and the design template A is not the recommended template, then use the design template A as the marked template, and thus obtain all the marked templates. Step S220: Obtain all the attribute features of a certain marked template B and the user profile of the current user. According to all the tag words under the preference tag, by analogy with Step S120, obtain the matching degree of the marked template B to the user. Furthermore, obtain the matching degree of each marked template to the user, and add them up and take the average value. If the average value is less than the preset numerical threshold, then use a certain user U as the marked user.

[0019] In this solution, the personality factors are obtained based on the results of the template preference survey of users, and the results obtained from the template preference survey are in line with the users' expectations; marking users in this step refers to those users whose recommended design templates do not meet the users' expectations. When recommending based on the users' personality factors in the following step S400 (that is, based on the results obtained from the template preference survey in this solution), this part of their information (that is, the information related to the design templates they ultimately want) should be taken into account so as to recommend more suitable design templates for users; for the users who are not marked users in step S400, since the recommended design templates meet the users' expectations, when recommending based on the users' personality factors, only the results of the template preference survey need to be considered.

[0020] Step S300: According to the design templates that the marked users like and dislike on the template resource platform, and the corresponding marked templates, obtain the feature words corresponding to the attribute characteristics of each template; and obtain the degree of preference and deviation of the users for the feature words, and then extract the keywords in the feature words, and establish a keyword set corresponding to the marked users based on the keywords. Step S310: The template resource platform has the function of allowing users to mark their likes and dislikes for design templates; obtain all the design templates that a marked user V likes and the corresponding marked templates, and extract the feature words corresponding to the attribute characteristics of each template as the preference feature words of user V. Statistically summarize the same preference feature words to obtain the quantity corresponding to each preference feature word, and then obtain the degree of preference of user V for a certain preference feature word i as: X = 1 - , where e is the natural constant, K1 is the preference coefficient, N i is the quantity corresponding to the feature word i; and perform normalization according to the degree of preference of each preference feature word. Step S320: Obtain all the design templates that user V dislikes, and extract the feature words corresponding to the attribute characteristics of each template as the deviation feature words of user V. Statistically summarize the same deviation feature words to obtain the quantity corresponding to each deviation feature word, and then obtain the degree of deviation of user V for a certain deviation feature word j as: Y = -(1 - ), where e is the natural constant, K2 is the deviation coefficient, N j is the quantity corresponding to the feature word j; perform normalization according to the degree of deviation of each feature word. Step S330: If a certain feature word h is not both a preference feature word and a deviation feature word at the same time, then take the feature word h as a keyword; if a certain feature word h is both a preference feature word and a deviation feature word at the same time, then according to the normalized degree of preference X h and the degree of deviation Y h of the feature word h, obtain the feature degree Z h [[ID=h -|Y h |,|| is to find the absolute value, if |Z h |If the degree is greater than the preset threshold, the feature word h is used as the keyword; and based on all the keywords of user V, a keyword set of user V is established.

[0021] Since the formula y=1-e -x When x is x≥0, y is [0,1), which is a function in which y increases as x increases. In this scheme, the deflection coefficients K1 and K2 are not less than 0. i and N j Refers to the number, which is not less than 0, then the value of the bias degree X is between 0 and 1, and the value of the deviation degree Y is between -1 and 0, then the characteristic degree Z of the feature word h is h The value of Z is between -1 and 1. h When it is close to -1, it means that the user's deviation from the feature word h is relatively large (the degree of deviation X h Small, deviation degree Y h The absolute value of |Y h | is larger), similarly, when Z h When it is close to 1, it indicates that the user has a greater preference for the feature word h. The following step S400 uses the feature degree as the word weight, which is reasonable and in line with the expectation of this solution.

[0022] Step S400: Conduct a template preference survey on the user, re-establish the user profile of the user based on the user's previous user profile, survey results and keyword set, and recommend design templates to the user based on the re-established user profile.

[0023] Step S410: Conduct a template preference survey on all users. The template preference survey includes user style preferences, theme preferences, occupation preferences, and scenario preferences. Set the word weights of all preference keywords in the survey content to 1. If a user is not a tagged user, extract all corresponding label words in the previous user profile, aggregate all preference keywords and label words in the survey content to obtain a final label word, and add the corresponding word weights. Re-establish the user profile of the user based on the final label word. Step S420: If a user is a marked user, extract all keywords in the keyword set of the user, and use the feature degree corresponding to the keyword as the keyword's word weight, and extract all corresponding label words in the previous user portrait, summarize all the preference keywords and all label words in the survey content, and all keywords in the keyword set to obtain the final label words, and add up the corresponding word weights, and re-establish the user portrait of the user based on the final label words; and recommend design templates to the user according to the re-established user portrait.

[0024] According to the established user portrait, the matching degree is obtained based on step S100, and the design template is recommended to the user in descending order of matching degree, which will not be repeated here.

[0025] The present invention also provides a data intelligent management system based on user portraits, such as Figure 2 As shown, it includes: user portrait building module, marked user extraction module, keyword set building module and design template recommendation module; User profile creation module: This module is used to obtain the user's occupation and template selection records, extract the attribute characteristics of the corresponding design templates in the template selection records, create a user profile based on the occupation and attribute characteristics, and determine the matching degree of each design template to the user based on the user profile, and recommend design templates to the user based on the matching degree; Marked user extraction module: used to obtain the corresponding marked template of the user based on the user's action behavior on the template resource platform; and analyze the marked template and user portrait to extract the marked users in the template resource platform; Keyword set building module: This module is used to obtain the feature words corresponding to the attribute features of each template based on the design templates that the user likes and dislikes on the template resource platform and the corresponding tag templates; it also obtains the user's preference and deviation degree for the feature words, and then extracts the keywords from the feature words, and builds the keyword set corresponding to the tagged user based on the keywords; Design template recommendation module: used to conduct template preference surveys on users, re-establish the user profile of the user based on the user's previous user profile, survey results and keyword sets, and recommend design templates to the user based on the re-established user profile.

[0026] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, in any regard, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. A data intelligent management method based on user portraits, characterized in that It includes the following steps: Step S100: Obtain the user's occupation and template selection records, extract the attribute features of the corresponding design templates in the template selection records, establish a user profile for the user based on the occupation and attribute features, obtain the matching degree of each design template to the user according to the user profile, and recommend design templates to the user according to the matching degree; Step S200: Based on the user's action behaviors on the template resource platform, obtain the corresponding marked templates for the user; And analyze the marked templates and the user profile, and extract the marked users in the template resource platform; Step S300: According to the design templates liked and disliked by the marked users on the template resource platform, and the corresponding marked templates, obtain the feature words corresponding to the attribute features of each template; And obtain the degree of preference and deviation of the user for the feature words, then extract the keywords in the feature words, and establish a keyword set corresponding to the marked users based on the keywords; Step S400: Conduct a template preference survey on the user, and re-establish the user profile for the user according to the previous user profile of the user, as well as the survey results and the keyword set, and recommend design templates to the user according to the re-established user profile.

2. The data intelligent management method based on user portraits according to claim 1, wherein, Step S100 includes: Step S110: Obtain all the design templates corresponding to the user's template selection records, extract the attribute features of each design template, and the attribute features include style, theme, applicable occupation, and applicable scenario; the user profile corresponds to several preference labels, and the preference labels include style preference, theme preference, occupation preference, and scenario preference; summarize all the feature words corresponding to each attribute feature in the template selection records to obtain the quantity corresponding to each feature word, and use the feature words with the quantity greater than the preset quantity threshold as the label words in the preference labels corresponding to the user; then obtain all the label words corresponding to each preference label in the user profile, and set the initial word weight of each label word to 1; Step S120: Obtain the feature word W of a certain attribute feature d corresponding to a certain design template T in the template resource platform d , and extract all the tag words under the preference tag corresponding to the attribute feature d in the user portrait. Multiply the similarity between a certain tag word L and the feature word W d by the word weight value of the tag word L to obtain the association degree of the tag word L to the feature word W d . Furthermore, obtain the association degree of each tag word, and take the maximum value of the association degrees as the association degree R between the attribute feature d and the user portrait d ; furthermore, obtain the association degree between each attribute feature in the design template T and the user portrait, preset the weights corresponding to each attribute feature, obtain the matching degree of the design template T to the user, and recommend the design template to the user in descending order of the matching degree.

3. The data intelligent management method based on user portraits according to claim 2, characterized in that Step S200 includes: Step S210: The template resource platform has the functions of the user searching for and screening design templates; use the design templates recommended in step S100 as the recommended templates. If a certain user U selects a certain design template A using any of the search or screening functions on the template resource platform, and modifies all the required content in all the editable content on the design template A, and the required content is the user's personal information content, and the design template A is not a recommended template, then use the design template A as the marked template, and then obtain all the marked templates; Step S220: Obtain all the attribute features of a certain marked template B and the user profile of the current user. According to all the label words under the preference labels, analogize step S120 to obtain the matching degree of the marked template B to the user, and then obtain the matching degree of each marked template to the user, and add them up and take the average value. If the average value is less than the preset numerical threshold, then use the certain user U as the marked user.

4. A data intelligent management method based on user portraits according to claim 1, characterized in that, Step S300 includes: Step S310: The template resource platform has the function for users to mark their preference and dislike for design templates; obtain all the design templates and corresponding marked templates that a certain marked user V likes, and extract the feature words corresponding to the attribute features of each template as the bias feature words of user V. Statistically summarize the same bias feature words to obtain the quantity corresponding to each bias feature word, and further obtain the degree of bias of user V towards a certain bias feature word i as: X = 1 - , where e is the natural constant, K1 is the bias coefficient, and N i is the quantity corresponding to the feature word i; and perform normalization according to the degree of bias of each bias feature word. Step S320: Obtain all the design templates that user V dislikes, and extract the feature words corresponding to the attribute features of each template as the deviation feature words of user V. Statistically summarize the same deviation feature words to obtain the quantity corresponding to each deviation feature word, and further obtain the deviation degree of user V for a certain deviation feature word j as: Y = -(1 - ), where e is the natural constant, K2 is the deviation coefficient, and N j is the quantity corresponding to the feature word j; normalize according to the deviation degree of each feature word; Step S330: If a feature word h is not both a bias feature word and a deviation feature word, then use the feature word h as a keyword; if a feature word h is both a bias feature word and a deviation feature word, then according to the normalized bias degree X h and the deviation degree Y h , obtain the feature degree Z h =X h -|Y h |, where || is to calculate the absolute value. If |Z h | is greater than the preset degree threshold, then use the feature word h as a keyword; and establish a keyword set for user V based on all the keywords of user V.

5. The data intelligent management method based on user portraits according to claim 1, characterized in that Step S400 includes: Step S410: Conduct a template preference survey on all users. The template preference survey includes user style preferences, theme preferences, occupation preferences, and scenario preferences. Set the word weights of all preference keywords in the survey content to 1. If a user is not a tagged user, extract all corresponding label words in the previous user profile, aggregate all preference keywords in the survey content with all the label words to obtain a final label word, and add the corresponding word weights. Re-establish the user profile of the user based on the final label word. Step S420: If a user is a marked user, extract all keywords in the keyword set of the user, and use the feature degree corresponding to the keyword as the word weight of the keyword, and extract all corresponding label words in the previous user portrait, summarize all the preference keywords in the survey content and all the label words, and all the keywords in the keyword set to obtain the final label words, and add the corresponding word weights therein, and re-establish the user portrait of the user according to the final label words; and recommend the design template to the user according to the re-established user portrait.

6. A data intelligent management system for executing a data intelligent management method based on user portraits described in any one of claims 1-5, characterized in that, The system includes a user portrait building module, a marked user extraction module, a keyword set building module and a design template recommendation module; User profile creation module: This module is used to obtain the user's occupation and template selection records, extract the attribute characteristics of the corresponding design templates in the template selection records, create a user profile based on the occupation and attribute characteristics, and determine the matching degree of each design template to the user based on the user profile, and recommend design templates to the user based on the matching degree; Marked user extraction module: used to obtain the corresponding marked template of the user based on the user's action behavior on the template resource platform; and analyze the marked template and user portrait to extract the marked users in the template resource platform; Keyword set building module: used to obtain the feature words corresponding to the attribute features of each template based on the design templates that mark the user's likes and dislikes on the template resource platform and the corresponding marking templates; And get the user's preference and deviation degree for the feature words, then extract the keywords in the feature words, and establish a keyword set corresponding to the marked user based on the keywords; Design template recommendation module: used to conduct template preference surveys on users, re-establish the user profile of the user based on the user's previous user profile, survey results and keyword sets, and recommend design templates to the user based on the re-established user profile.

7. An intelligent data management system according to claim 6, wherein The marked user extraction module includes a marked template extraction unit and a marked user extraction unit; Marking template extraction unit: used for extracting marking templates from design templates according to the functions of searching and filtering design templates by users on the template resource platform and the obtained recommended templates; Marked user extraction unit: used to obtain all attribute features of the marking template and the user portrait of the current user, and obtain the marked user based on all label words under the preference label.

8. A data intelligent management system according to claim 6, characterized in that, The design template recommendation module includes a user portrait reconstruction unit and a design template recommendation unit; User profile re - establishment unit: It is used to conduct a template preference survey for all users, and set the word weights of all preference keywords in the survey content; extract all corresponding tag words from the previous user profile of the user to obtain the final tag words, and then re - establish the user profile of a certain user; Design template recommendation unit: It is used to recommend design templates to users according to the re - established user profile of a certain user.

Images