User profiling methods and systems based on big data

Abstract

This invention discloses a user profiling method and system based on big data, relating to the field of data analysis technology. It obtains a quasi-interest array set by analyzing the page browsing trajectory of a target user; retrieves a first set of browsing records from a first sample user group that meets the basic profile of the target user; and performs high-frequency array statistics on the quasi-interest array set based on product type and several advertising element types to obtain an interest array set; retrieves a second set of browsing records from a second sample user group that meets the interest array set to obtain a secondary quasi-interest array set; and performs high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain a secondary interest array set; and updates the basic profile of the target user based on the interest array set and the secondary interest array set. This solves the technical problems of insufficient targeting and individualization of user profiling, and the difficulty in capturing dynamic changes in user interests in real time, leading to errors in the timeliness and accuracy of user profiling.

Description

Technical Field

[0001] This invention relates to the field of data analysis technology, specifically to a user profiling method and system based on big data. Background Technology

[0002] In today's digital age, with the rapid development of internet technology and the widespread application of big data, user profiling technology, as a core technology in fields such as precision marketing and personalized recommendations, has received widespread attention and research. Existing user profiling technologies largely rely on user clustering methods, that is, by performing cluster analysis on a large amount of user behavior data, users with similar behavioral characteristics are grouped together, thereby providing recommendation services for the entire user group.

[0003] However, while user clustering has the advantage of strong generalization ability, providing recommendations based on the common characteristics of a group, its degree of individualization is relatively insufficient. Because users have significantly different interests and needs, recommendations based on group commonalities often fail to accurately match the unique needs of each user, resulting in unsatisfactory recommendation effects. Furthermore, with the dynamic changes in user interests, existing user profiling technologies often struggle to capture these changes in real time, affecting the timeliness and accuracy of user profiles. Summary of the Invention

[0004] This application provides a user profiling method and system based on big data, which addresses the technical problems of insufficient targeting and individualization of user profiles in existing technologies, and the difficulty in capturing dynamic changes in user interests in real time, leading to errors in the timeliness and accuracy of user profiles.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] In a first aspect, the present invention provides a user profiling method based on big data, comprising: analyzing the page browsing trajectory of a target user to obtain a quasi-interest array set, wherein any quasi-interest array includes a product type and several advertising element types; retrieving a first browsing record set of a first sample user group that satisfies the basic profile of the target user; performing high-frequency array statistics on the quasi-interest array set based on the product type and the several advertising element types to obtain an interest array set; retrieving a second browsing record set of a second sample user group that satisfies the interest array set to obtain a secondary quasi-interest array set; performing high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain a secondary interest array set; and updating the basic profile of the target user based on the interest array set and the secondary interest array set.

[0007] Secondly, the present invention provides a user profiling system based on big data. The system includes: an information parsing module, used to parse the page browsing trajectory of a target user to obtain a quasi-interest array set, wherein any quasi-interest array includes a product type and several advertising element types; a first data statistics module, used to retrieve a first set of browsing records of a first sample user group that meets the basic profile of the target user, and perform high-frequency array statistics on the quasi-interest array set based on the product type and the several advertising element types to obtain an interest array set; a second data statistics module, used to retrieve a second set of browsing records of a second sample user group that meets the interest array set to obtain a secondary quasi-interest array set, and perform high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain a secondary interest array set; and a profile updating module, used to update the basic profile of the target user according to the interest array set and the secondary interest array set.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The user profiling method based on big data provided in this application analyzes the page browsing trajectory of target users to obtain a quasi-interest array set. Combined with a hierarchical interest array statistics and update mechanism, it continuously optimizes and updates the basic profile of target users. By comparing with user groups with similar basic profiles, it can more accurately capture individual user interests while maintaining sensitivity to changes in user interests, ensuring the timeliness and accuracy of user profiles. This not only overcomes the shortcomings of insufficient individualization in existing user clustering methods, but also achieves automatic updating and cleaning of user profiles by introducing a timestamp mechanism, further improving the simplification and efficiency of user profiles. Attached Figure Description

[0010] Figure 1 This application provides a flowchart illustrating a user profiling method based on big data.

[0011] Figure 2 This application provides a structural diagram of a user profiling system based on big data.

[0012] Figure labeling: Information parsing module 11, first data statistics module 12, second data statistics module 13, profile update module 14. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0015] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0016] Example 1:

[0017] like Figure 1 As shown, this embodiment of the invention provides a user profiling method based on big data, including:

[0018] S10: Analyze the target user's page browsing history to obtain a set of quasi-interest arrays, where any quasi-interest array includes product type and several advertising element types.

[0019] Specifically, a user profile refers to a set of virtual images or tags constructed using data analysis and mining techniques based on a user's online behavioral data (such as page browsing history, click behavior, search records, etc.) and related attribute information (such as age, gender, region, etc.). This set reflects the user's interests, preferences, and needs, and is used to support applications such as personalized recommendations and precision marketing. Existing technologies rely on user clustering for group recommendations, resulting in strong generalization capabilities but insufficient targeting and individualization of user profiles. Furthermore, the inability to capture dynamic changes in user interests in real time leads to errors in the timeliness and accuracy of user profiles. This solution, by incorporating a hierarchical update mechanism, improves the targeting of user profiles, more accurately capturing individual user interests while maintaining sensitivity to changes in user interests, ensuring the timeliness and accuracy of user profiles. It not only overcomes the shortcomings of insufficient individualization in existing user clustering methods but also introduces a timestamp mechanism to achieve automatic updates and cleanup of user profiles, further improving the conciseness and efficiency of user profiles.

[0020] For example, in the process of building a user profile, the first step is to analyze the target user's page browsing trajectory, aiming to extract feature information with potential interest value from user behavior data. Specifically, a deep analysis of the target user's page browsing trajectory is conducted, and key click windows in the trajectory are extracted to capture the user's interactive behavior within a specific time period. These click windows not only record the pages visited by the user but also implicitly reveal the user's interest in the page content.

[0021] Furthermore, the UI design information within the clicked windows is extracted. The product type describes the category of the products displayed on the page, such as "electronic products" or "clothing." The advertising element type covers a range of elements related to ad display, such as "image ads," "video ads," or "text link ads." Based on this extracted information, the system constructs quasi-interest arrays. Each array contains a specific product type and several associated advertising element types, such as "electronic products - image ads, video ads." This process systematically collects and organizes the interest characteristics exhibited by target users during different page browsing sessions, forming a quasi-interest array set. This lays the foundation for subsequent interest array statistics and user profile updates. For example, if a user browses an electronics product page in multiple consecutive clicked windows, and image and video ads frequently appear on these pages, a quasi-interest array containing "electronic products - image ads, video ads" will be constructed and included in the quasi-interest array set, reflecting the user's strong interest in electronics products and their advertising formats.

[0022] S20: Retrieve the first browsing record set of the first sample user group that meets the basic profile of the target user, and perform high-frequency array statistics by traversing the quasi-interest array set based on the product type and the several advertising element types to obtain the interest array set.

[0023] Furthermore, retrieving the first set of browsing records of the first sample user group that matches the basic profile of the target user, and performing high-frequency array statistics based on product type and advertising element type, is a key step in accurately characterizing user interests. Specifically, based on the basic profile of the target user, a group with similar characteristics to the target user is selected, namely the first sample user group. The basic profile typically includes general profile tags (such as age, gender, region, etc.) and individual profile tags (such as the initial interest array set).

[0024] Subsequently, the system collects the first set of browsing records from sample users, which records their browsing behavior on the internet, including visited pages and clicked advertisements. Based on the product type and advertisement element type parsed from the target user's page browsing trajectory, the system traverses the set of quasi-interest arrays and counts the trigger frequency of each quasi-interest array in the first set of browsing records. The trigger frequency refers to the number or proportion of times a particular quasi-interest array (e.g., "electronic products - image advertisements") appears in the sample user's browsing records. Through high-frequency array statistics, the system can identify interest arrays that frequently appear in the sample user group. These arrays reflect common characteristics that the target user may be interested in. Finally, quasi-interest arrays with trigger frequencies exceeding a preset threshold are aggregated to form an interest array set. This set serves as a concentrated representation of the target user's interest characteristics, providing an important basis for subsequent user profile updates. For example, if the quasi-interest array "electronic products - image advertisements" is found to appear frequently in the first set of browsing records, it is included in the interest array set, indicating that the target user and similar sample users have a high interest in electronic products and their image advertisement formats.

[0025] S30: Retrieve the second browsing record set of the second sample user group that satisfies the interest array set, obtain the secondary quasi-interest array set, and perform high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain the secondary interest array set.

[0026] Next, the next step is to further retrieve a second sample user group that matches the identified interest set and collect their second browsing history set. This is a step to deepen the user interest mining. Specifically, the previously obtained interest set is used first, which reflects the combination of product types and advertising element types that the target users and their similar groups are interested in, such as "electronic products - image ads". Then, based on these interest characteristics, another group of users with similar interest patterns is selected, namely the second sample user group. For this user group, their browsing behavior data is collected to form the second browsing history set, which records the detailed browsing trajectory of these users on the Internet.

[0027] Subsequently, the second set of browsing records was analyzed to extract secondary quasi-interest arrays. These arrays contained interest features similar to but not identical to the initial interest arrays, such as "electronic products - video ads" or "home furnishings - text link ads." To accurately identify potential new interests of the target users, the secondary quasi-interest arrays were compared with the previously obtained quasi-interest arrays, and the non-overlapping portions were extracted—those interest arrays that appeared only in the second sample user group but were not significantly reflected in the initial analysis. Finally, high-frequency array statistics were performed on these non-overlapping quasi-interest arrays. By calculating their trigger frequency in the second set of browsing records, arrays with trigger frequencies exceeding a preset threshold were selected to form the secondary interest array set.

[0028] This process not only enriches the interest dimensions of user profiles but also enhances their sensitivity to changes in user interests. For example, if the non-overlapping quasi-interest array "home furnishings - video ads" is found to appear frequently in the second browsing history set, it is included in the secondary interest array set, indicating that the target user may also show a strong interest in the video ad format for home furnishings, even though this interest was not significantly reflected in the initial analysis.

[0029] S40: Update the target user basic profile based on the interest array set and the secondary interest array set.

[0030] Specifically, the basic profile of the target user is updated based on the acquired interest array and secondary interest array. This process aims to integrate newly discovered user interest features into the existing profile to improve the accuracy of the profile in reflecting the user's true interests. Specifically, it integrates the interest array (e.g., "electronics - image ads") and the secondary interest array (e.g., "home furnishings - video ads"), which represent the interest preferences exhibited by the target user and their similar groups at different stages. Subsequently, these interest features are added as new profile tags to the target user's basic profile, which originally included general profile tags (e.g., age, gender) and individual profile tags (e.g., the initial interest array).

[0031] During the update process, the application of timestamps is emphasized. Newly added interest arrays are marked with their first timestamp, recording the time when they were first identified. When a target user clicks on content related to a particular interest array again, the timestamp of that array is updated using the latest click timestamp to reflect the persistence and variability of user interests. If the time interval between the timestamp of an interest array and the current time exceeds a preset threshold, it is determined that the interest may have faded or become inactive, and thus it is removed from the target user's basic profile, ensuring the timeliness and accuracy of the profile information. For example, if subsequent monitoring reveals that a target user's interest in "electronic products - image ads" remains, while their interest in "home furnishings - video ads" gradually weakens, the former will be retained in the basic profile and its timestamp updated, while the latter will be removed. This dynamically optimizes the user profile, providing more accurate data support for applications such as personalized recommendations.

[0032] In one specific embodiment, parsing the page browsing trajectory of the target user to obtain a quasi-interest array set includes: extracting the first clicked window of the page browsing trajectory; extracting the first UI design information of the first clicked window, wherein the first UI design information includes a first product design type and a first advertising element design type set, constructing a first quasi-interest array; and adding the first quasi-interest array to the quasi-interest array set.

[0033] Optionally, the system first performs time-series slicing on the target user's page browsing trajectory. By setting a reasonable sliding window length (such as 30 seconds or 1 page), the first click window in the trajectory is extracted. This window serves as the basic unit for user behavior analysis, covering the time period and interaction range from page loading to the first click operation.

[0034] Subsequently, UI design information was extracted from the first clicked window, identifying and recording the first product design type (e.g., "smart wearable device") and the first set of advertising element design types (e.g., "dynamic pop-up ad" and "bottom floating ad"). The product design type identifies the product category of the core content displayed on the page, while the set of advertising element design types describes the visual presentation and interaction methods of the page's advertisements.

[0035] Based on the extracted UI design information, a first quasi-interest array is constructed. This array uses product design type as the baseline dimension and advertising element design type set as the feature set, forming a binary structure of "product design type - advertising element design type", such as "smart wearable devices - dynamic pop-up ads, bottom floating ads". Finally, the first quasi-interest array is incorporated into a quasi-interest array set. This set serves as a temporary storage container for users' potential interest characteristics, providing a data foundation for subsequent high-frequency array statistics and profile updates. For example, if a user frequently clicks on windows containing "smart wearable devices", "dynamic pop-up ads", and "bottom floating ads" while browsing technology product pages, the system will construct the corresponding quasi-interest array and add it to the set, thus reflecting the user's interest in technology products and their specific advertising formats.

[0036] In one specific embodiment, the target user basic profile includes general profile tags and individual profile tags. The individual profile tags represent an initial interest array set. A first set of browsing records for a first sample user group that satisfies the target user basic profile is retrieved. Based on the product type and the several advertising element types, the quasi-interest array set is traversed to perform high-frequency array statistics to obtain the interest array set. This includes: retrieving the first set of browsing records for a first sample user group that satisfies the general profile tags and the individual profile tags; extracting the first product design type and the first advertising element design type set from the first quasi-interest array set; calculating the trigger frequency ratio of the first product design type and the first advertising element design type set in the first browsing record set, setting this as the first quasi-interest array confidence level; and adding the first quasi-interest array to the interest array set when the first quasi-interest array confidence level is greater than or equal to a confidence threshold.

[0037] Preferably, in the process of dynamic optimization of user profiles, implementing sample user group screening and high-frequency interest array mining for the basic profile of target users (including general profile tags and individual profile tags) is an important step in accurately identifying users' potential interests.

[0038] First, based on the target users' general profile tags (such as age, occupation, region, etc., e.g., "25-34 years old", "Internet practitioners", "first-tier cities") and individual profile tags (i.e., the initial interest set, e.g., "sports equipment - video ads" and "beauty products - image and text recommendations"), multi-dimensional filtering conditions are constructed to retrieve the first sample user group that meets all the tags, and the first set of browsing records of this user group is collected. This set records the complete browsing trajectory of the sample users on the Internet.

[0039] Subsequently, feature extraction is performed on the first quasi-interest array (e.g., "Sports Equipment - Video Ads") within the target user quasi-interest array set, parsing out its first product design type ("Sports Equipment") and first advertising element design type set ("Video Ads"). Based on the extracted features, the first browsing record set is traversed, and the number of times the first quasi-interest array is triggered in the sample users' browsing records is counted. The proportion of this triggering to the total number of browsing times in the set is calculated and defined as the confidence level of the first quasi-interest array. This indicator is used to quantify the representativeness of the array's interests within the sample user group. For example, if a quasi-interest array is triggered 100 times in the first browsing record set, and the total number of browsing times in the set is 1000, then its confidence level is 10%.

[0040] Finally, the confidence level of the first quasi-interest array is compared with a preset confidence threshold (e.g., 15%). When the confidence level is greater than or equal to the threshold, the array is determined to have significant interest representativeness in the sample user group and is added to the interest array set, thereby dynamically updating the target user's interest feature set. For example, if the confidence level of the quasi-interest array "sports equipment - video ads" reaches 20%, exceeding the preset threshold, the system adds it to the interest array set, indicating that the target user and similar sample users have a high interest in the video ad format for sports equipment, providing data support for subsequent personalized recommendations.

[0041] In one specific embodiment, the proportion of trigger frequencies of the first product design type and the first advertising element design type set in the first browsing record set is statistically analyzed and set as the first quasi-interest array confidence score. This includes: statistically analyzing the first trigger frequency proportion of the first product design type in the first browsing record set; statistically analyzing the second trigger frequency proportion set of the first advertising element design type set in the first browsing record set; and calculating the mean of the first trigger frequency proportion set and the second trigger frequency proportion set, which is set as the first quasi-interest array confidence score.

[0042] For example, in the process of constructing the confidence score of the interest array based on user browsing behavior data, the system first performs multi-dimensional trigger frequency statistics on the first quasi-interest array (such as "sports equipment - video ads, image and text pop-ups") in the target user's quasi-interest array set. Specifically, the system first calculates the percentage of the first trigger frequency for the first product design type ("sports equipment") in the first browsing record set. This percentage is obtained by calculating the proportion of page browsing records containing "sports equipment" to the total number of pages in the first browsing record set. The percentage of the first trigger frequency is used to quantify the breadth of interest coverage of the product type in the sample user group. For example, if there are 300 browsing records involving "sports equipment" in the first browsing record set, and the total number of records in the set is 2000, then the percentage of the first trigger frequency for "sports equipment" is 15%.

[0043] Subsequently, a statistical analysis was conducted on each of the first set of advertising element design types ("video ads, image and text pop-ups"), and the percentage of the second trigger frequency for each advertising element type in the first set of browsing records was calculated to form a second set of trigger frequency percentages. The second set of trigger frequency percentages reflects the interest activation intensity of different advertising formats in the sample user group; for example, the trigger percentage for "video ads" was 10%, and the trigger percentage for "image and text pop-ups" was 8%.

[0044] Finally, the average of the percentages from the first and second trigger frequency sets is used to obtain the confidence level of the first quasi-interest array. This indicator integrates the interest correlation strength between product type and advertising element type. The confidence level is used to assess the interest representativeness of the quasi-interest array in the sample user group. For example, if the trigger percentage of "sports equipment" is 15%, and the trigger percentages of "video ads" and "image and text pop-ups" are 10% and 8% respectively, then the average is calculated as (15%+10%+8%) / 3≈11%. This confidence level will serve as the key criterion for determining whether the first quasi-interest array should be included in the interest array set.

[0045] The above process enables a quantitative assessment of user interest characteristics, providing data support for subsequent profile updates and recommendation strategy optimization.

[0046] In one specific embodiment, the general profile tag includes a general type tag and a general quantitative tag. Retrieving a first set of browsing records for a first sample user group that satisfies both the general profile tag and the individual profile tag includes: configuring a general quantitative attribute deviation threshold; constructing a general profile similarity evaluation function by combining the general type tag and the general quantitative tag; wherein the general profile similarity evaluation function calculates the ratio of the sum of the number of first attributes satisfying the general quantitative attribute deviation threshold in the general quantitative tag deviation between the general quantitative tag and the sample general quantitative tag, and the number of attributes intersecting between the general type tag and the sample general type tag, to the total number of attributes in the general profile tag; constructing an individual profile tag similarity evaluation function; wherein the individual profile tag similarity evaluation function is used to calculate the ratio of the intersection tag and the union tag between the sample individual profile tag and the individual profile tag; and retrieving the general profile similarity evaluation function and the individual profile tag similarity evaluation function to retrieve the first sample user group that simultaneously satisfies both the general similarity threshold and the individual similarity threshold.

[0047] Specifically, in the process of selecting sample user groups driven by user profiles, the general profile tags of target users are first analyzed in a structured manner and broken down into general type tags (such as "age level - youth" and "occupation category - Internet practitioner") and general quantitative tags (such as "age value - 28 years old" and "spending ability score - 75 points"), and a two-layer screening mechanism is constructed based on this.

[0048] In the general profile similarity assessment stage, a general quantitative attribute deviation threshold is pre-configured (e.g., age deviation not exceeding ±3 years, consumption capacity score deviation not exceeding ±10 points). Then, a general profile similarity assessment function is designed. This function achieves sample matching through two steps: First, for each general quantitative tag, the number of attributes whose deviation values ​​between the sample user and the target user are within the threshold range is counted (e.g., if both age and consumption capacity score meet the deviation threshold, the count is 2). Second, the number of intersection attributes of general type tags is calculated (e.g., if both the target user and the sample user belong to "youth" and "internet practitioner," the number of intersection attributes is 2). Finally, the sum of the number of matching attributes of the quantitative tag and the number of intersection attributes of the type tag is divided by the total number of general profile tag attributes (e.g., if the total number of attributes is 5, the number of matching attributes is 4) to obtain the general profile similarity score, thereby quantifying the similarity between the target user and the sample user at the level of general features.

[0049] For example, if the target user's general profile includes five tags: "Age 28 (quantitative)," "Youth (type)," "Internet Practitioner (type)," "First-tier City (type)," and "Spending Power Score 75 (quantitative)," and the sample user profile is "Age 29," "Youth," "Internet Practitioner," "First-tier City," and "Spending Power Score 80," then after calculation, the age deviation of 1 year and the spending power score deviation of 5 points in the quantitative tags both meet the threshold, and the three type tags are completely matched. Then the similarity score is (2+3) / 5=100%.

[0050] In the individual profile tag similarity evaluation stage, a similarity evaluation function based on set operations is constructed for the individual profile tags of the target user (such as the initial interest array set "sports equipment - video ads" and "beauty products - image and text recommendations") and the individual profile tags of the sample user. This function quantifies the degree of overlap between the two at the level of individual interest features by calculating the ratio of the number of intersection tags to the number of union tags between the sample individual profile tags and the target user individual profile tags (e.g., if the target user has 2 interest arrays and the sample user has 3 interest arrays, and one interest array overlaps, then the similarity is 1 / 3≈33.3%).

[0051] Based on the combined results of the two-layer evaluation, the system calls the general profile similarity evaluation function and the individual profile tag similarity evaluation function, sets a general similarity threshold (e.g., similarity score ≥ 80%) and an individual similarity threshold (e.g., similarity score ≥ 50%), retrieves the first sample user group that simultaneously meets the dual threshold constraints, and extracts the first set of browsing records for this user group, providing a data foundation for subsequent high-frequency interest array mining. For example, if a sample user has a similarity score of 90% at the general profile level and a similarity score of 60% at the individual profile level with the target user, both exceeding the preset thresholds, then this user is included in the first sample user group, and their browsing records will be used to analyze the potential interest preferences of the target user.

[0052] In one specific embodiment, retrieving a second sample user group that satisfies the interest array set includes: updating the individual profile tag similarity evaluation function according to the interest array set to obtain an updated individual profile tag similarity evaluation function; and invoking the updated individual profile tag similarity evaluation function to retrieve the second sample user group that satisfies the individual similarity threshold.

[0053] Furthermore, in the user profile-driven dynamic screening mechanism for sample user groups, the individual profile tag similarity evaluation function is iteratively optimized for the constructed interest array set (such as "sports equipment - video ads" and "beauty products - live streaming sales") to improve the matching accuracy between the sample user group and the target user's interest features.

[0054] Specifically, the interest array set is integrated as a core feature into the individual profile tagging system, and the original evaluation function is structurally updated. Through a feature mapping mechanism, each element in the interest array set (e.g., "sports equipment - video ads") is transformed into a quantifiable interest feature vector, and similarity calculation rules between it and the sample user's individual profile tags are defined. For example, a cosine similarity algorithm is used to quantify the angle relationship between the interest arrays in the vector space, or the overlap of the interest tag sets is calculated based on the Jaccard coefficient, thereby constructing an updated individual profile tag similarity evaluation function. This function, while retaining the original individual profile tag (e.g., initial interest array set, browsing preference type) similarity calculation logic, adds an interest array set matching module. It comprehensively evaluates the matching degree between sample users and target users in both basic and dynamic interest dimensions through a weighted summation method. The updated individual profile tag similarity evaluation function refers to this new evaluation model that integrates dynamic interest features. For example, if the target user's individual profile tags include the initial interest array "sports equipment - image and text recommendation" and the newly mined interest array "sports equipment - video advertising", and the sample user's individual profile tags include "sports equipment - video advertising" and "beauty products - image and text recommendation", the updated evaluation function will calculate the similarity of the two sets of tags separately (e.g., by calculating the intersection-union ratio of the tag sets, the similarity is 1 / 2 = 50%), and combine it with preset weight coefficients (e.g., initial interest weight 0.4, dynamic interest weight 0.6) to obtain the final similarity score.

[0055] Subsequently, the system invokes the updated evaluation function to perform a traversal search of the entire user sample database, selecting users whose individual similarity scores are higher than a preset threshold (e.g., similarity ≥ 60%) as the second sample user group. This threshold is determined through cross-validation to ensure that the sample user group maintains high-dimensional consistency with the target user in terms of interest features. For example, if a sample user scores 65% with the target user under the updated individual profile tag similarity evaluation function, exceeding the threshold, then this user is included in the second sample user group, and their subsequent browsing history will be used to further explore the potential interest trends of the target user, thus forming a closed-loop optimization mechanism of "interest feature extraction - evaluation function update - sample user selection".

[0056] In one specific embodiment, updating the target user basic profile based on the interest array set and the secondary interest array set further includes: when the target user basic profile is updated, marking a first timestamp for the newly added first interest array, wherein when the first interest array is clicked again, the first timestamp is updated using the click timestamp; when the time interval between the first timestamp and the current timestamp is greater than or equal to a time interval threshold, the first interest array is deleted from the target user basic profile.

[0057] Specifically, iterative updates are implemented for the basic profile of target users based on interest array sets (such as "sports equipment - video ads" and "digital products - splash screen pop-ups") and secondary interest array sets (such as "outdoor products - news feed images and text" and "home furnishings - short video recommendations"), and a time-sensitive array validity management mechanism is introduced to ensure the timeliness and accuracy of profile data.

[0058] After the system completes the fusion and update of the primary interest set and secondary interest set, it timestamps the first interest set added to the target user's basic profile (such as "Outdoor Products - News Feed Images and Text"). The first timestamp records the moment the set was first included in the profile. For example, when "Outdoor Products - News Feed Images and Text" is added to the profile at 14:00:00 on March 1, 2024, this timestamp is assigned to it. If this first interest set is triggered again in subsequent user actions (such as clicking on similar ads or browsing related pages), the system captures the real-time interaction moment through a click timestamp mechanism and updates its first timestamp accordingly. For example, if the user clicks on "Outdoor Products - News Feed Images and Text" again at 10:30:00 on March 5, 2024, the timestamp of this set is synchronously updated to that moment.

[0059] Meanwhile, a preset time interval threshold (e.g., 30 days) serves as the criterion for determining the validity of interest arrays. By calculating the time interval between the current timestamp and the first timestamp of the first interest array, the system assesses whether it still reflects the user's true interests. If the time interval exceeds the threshold (e.g., the current date is April 2, 2024, while the first timestamp of "Outdoor Products - Information Flow Images and Text" is March 1, 2024, an interval of 32 days), the interest array is deemed invalid and removed from the target user's basic profile through a deletion mechanism to avoid outdated data interfering with the accuracy of the profile. For example, if a user showed interest in "Home Furnishings - Short Video Recommendations" 30 days ago but did not interact further, the system will automatically remove the array on the 31st day; conversely, if a user continuously clicks on "Sports Equipment - Video Ads" within the past 30 days, its timestamp will be continuously updated to ensure that the profile always focuses on the user's active interest areas.

[0060] Through the aforementioned time-sensitive dynamic update mechanism, user profiles have evolved from static feature sets to time-sensitive knowledge bases, providing more valuable data support for scenarios such as personalized recommendations and advertising.

[0061] The user profiling method based on big data provided in this invention has at least the following technical effects:

[0062] 1. By analyzing user page browsing history, a quasi-interest array set is generated, including product type and advertising element type. Based on the target user's basic profile, the browsing records of sample user groups are selected. High-frequency interest arrays are filtered by combining confidence thresholds. Finally, the primary and secondary interest array sets are merged to complete the profile update and form a multi-dimensional interest feature set. This significantly improves the profile's ability to cover users' complex interest preferences and provides richer feature inputs for personalized recommendations.

[0063] 2. A two-layer screening mechanism is constructed using general profile tags and individual profile tags. The general profile similarity evaluation function calculates the basic similarity by quantifying the tag deviation threshold and the number of intersections of type tags, while the individual profile tag similarity evaluation function quantifies the degree of interest overlap by operating on the tag set. The combination of the two achieves accurate matching of sample user groups, ensuring that profile updates are based on user groups with highly similar behaviors and improving data credibility.

[0064] 3. Dynamic validity management of interest arrays is achieved through a timestamp mechanism. New interest arrays are marked with their first timestamp, and interactive behaviors trigger timestamp updates. Arrays that exceed the time interval threshold are automatically deleted, avoiding recommendation bias caused by outdated interest features in the profile. This mechanism ensures that the profile always reflects the user's active interests, significantly improving recommendation accuracy in interest migration scenarios while reducing computational resource consumption.

[0065] Example 2:

[0066] like Figure 2 As shown, based on the same inventive concept as the user profiling method based on big data provided in Embodiment 1, this embodiment of the invention also provides a user profiling system based on big data, the system comprising:

[0067] The information parsing module 11 is used to parse the page browsing trajectory of the target user and obtain a set of quasi-interest arrays, wherein any quasi-interest array includes product type and several advertising element types.

[0068] The first data statistics module 12 is used to retrieve the first browsing record set of the first sample user group that meets the basic profile of the target user, and to perform high-frequency array statistics by traversing the quasi-interest array set based on the product type and the several advertising element types to obtain the interest array set.

[0069] The second data statistics module 13 is used to retrieve the second browsing record set of the second sample user group that satisfies the interest array set, obtain the secondary quasi-interest array set, and perform high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain the secondary interest array set.

[0070] The profile update module 14 is used to update the basic profile of the target user based on the interest array set and the secondary interest array set.

[0071] Furthermore, the information parsing module 11 is also used to perform the following steps:

[0072] Extract the first clicked window from the page browsing trajectory; extract the first UI design information of the first clicked window, wherein the first UI design information includes a first product design type and a first set of advertising element design types, and construct a first quasi-interest array; add the first quasi-interest array to the quasi-interest array set.

[0073] Furthermore, the first data statistics module 12 is also used to perform the following steps:

[0074] Retrieve the first browsing record set of the first sample user group that satisfies the general profile tag and the individual profile tag; extract the first product design type and the first advertising element design type set of the first quasi-interest array set; calculate the trigger frequency ratio of the first product design type and the first advertising element design type set in the first browsing record set, and set it as the first quasi-interest array confidence level; when the first quasi-interest array confidence level is greater than or equal to the confidence level threshold, add the first quasi-interest array to the interest array set.

[0075] Furthermore, the first data statistics module 12 is also used to perform the following steps:

[0076] The first trigger frequency percentage of the first product design type in the first browsing record set is calculated; the second trigger frequency percentage of the first advertising element design type set in the first browsing record set is calculated respectively; the mean of the first trigger frequency percentage and the second trigger frequency percentage set is calculated and set as the confidence level of the first quasi-interest array.

[0077] Furthermore, the first data statistics module 12 is also used to perform the following steps:

[0078] Configure a general quantitative attribute deviation threshold, and construct a general profile similarity evaluation function by combining the general type label and the general quantitative label. The general profile similarity evaluation function calculates the ratio of the sum of the number of first attributes satisfying the general quantitative attribute deviation threshold in the general quantitative label deviation between the general quantitative label and the sample general quantitative label, and the number of attributes intersecting the general type label and the sample general type label, to the total number of attributes in the general profile label. Construct an individual profile label similarity evaluation function, which calculates the ratio of the intersection label to the union label between the sample individual profile label and the individual profile label. Retrieve the general profile similarity evaluation function and the individual profile label similarity evaluation function to find the first sample user group that simultaneously satisfies both the general similarity threshold and the individual similarity threshold.

[0079] Furthermore, the second data statistics module 13 is also used to perform the following steps:

[0080] Based on the interest array set, update the individual profile tag similarity evaluation function to obtain the updated individual profile tag similarity evaluation function; call the updated individual profile tag similarity evaluation function to retrieve the second sample user group that meets the individual similarity threshold.

[0081] Furthermore, the image update module 14 is also used to perform the following steps:

[0082] When the target user's basic profile is updated, a first timestamp is assigned to the newly added first interest array. When the first interest array is clicked again, the first timestamp is updated using the click timestamp. When the time interval between the first timestamp and the current timestamp is greater than or equal to the time interval threshold, the first interest array is deleted from the target user's basic profile.

[0083] Through the foregoing detailed description of the user profiling method based on big data, those skilled in the art can clearly understand the user profiling system based on big data in this embodiment. As the system disclosed in the embodiment corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section description.

[0084] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A user profiling method based on big data, characterized in that, include: Analyze the page browsing history of the target user to obtain a set of quasi-interest arrays, where any quasi-interest array includes product type and several advertising element types; Retrieve the first browsing history set of the first sample user group that meets the basic profile of the target user, and based on the product type and the several advertising element types, traverse the quasi-interest array set to perform high-frequency array statistics to obtain the interest array set; Retrieve the second browsing record set of the second sample user group that satisfies the interest array set, obtain the secondary quasi-interest array set, and perform high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain the secondary interest array set; Update the target user's basic profile based on the interest array set and the secondary interest array set; The target user basic profile includes general profile tags and individual profile tags. The individual profile tags represent an initial interest array set. The first browsing record set of the first sample user group that meets the target user basic profile is retrieved. Based on the product type and the several advertising element types, the quasi-interest array set is traversed to perform high-frequency array statistics to obtain the interest array set, including: Retrieve the first set of browsing records for a first sample user group that satisfies the general profile tags and the individual profile tags; Extract the first product design type and the first advertising element design type set from the first quasi-interest array set of the quasi-interest array set; The proportion of the trigger frequency of the first product design type and the first advertising element design type set in the first browsing record set is statistically analyzed and set as the confidence level of the first quasi-interest array. When the confidence level of the first quasi-interest array is greater than or equal to the confidence threshold, the first quasi-interest array is added to the interest array set. The general profile tags include general type tags and general quantitative tags. Retrieving the first browsing record set of a first sample user group that satisfies the general profile tags and the individual profile tags includes: Configure a general quantitative attribute deviation threshold, and construct a general portrait similarity evaluation function by combining the general type label and the general quantitative label. The general portrait similarity evaluation function calculates the ratio of the sum of the number of first attributes that satisfy the general quantitative attribute deviation threshold in the general quantitative label deviation between the general quantitative label and the sample general quantitative label, and the number of attributes that intersect between the general type label and the sample general type label, to the total number of attributes of the general portrait label. Construct an individual profile label similarity evaluation function, wherein the individual profile label similarity evaluation function is used to calculate the ratio of the intersection label to the union label of the sample individual profile label and the individual profile label; The general profile similarity evaluation function and the individual profile tag similarity evaluation function are invoked to retrieve the first sample user group that simultaneously meets the general similarity threshold and the individual similarity threshold; The process of retrieving a second sample user group that satisfies the interest array set includes: Based on the interest array set, update the individual profile label similarity evaluation function to obtain the updated individual profile label similarity evaluation function; The updated individual profile label similarity evaluation function is invoked to retrieve the second sample user group that meets the individual similarity threshold.

2. The method as described in claim 1, characterized in that, Analyze the target user's page browsing history to obtain a quasi-interest array set, including: Extract the first clicked window from the page browsing trajectory; Extract the first UI design information of the first clicked window, wherein the first UI design information includes a first product design type and a first set of advertising element design types, and construct a first quasi-interest array; Add the first quasi-interest array to the quasi-interest array set.

3. The method as described in claim 1, characterized in that, The trigger frequency ratio of the first product design type and the first advertising element design type set in the first browsing history set is statistically analyzed and set as the first quasi-interest array confidence score, including: The percentage of the first product design type's first trigger frequency in the first browsing history set is statistically analyzed. The proportion of the second trigger frequency of the first advertising element design type set in the first browsing history set is statistically analyzed respectively. Calculate the mean of the first trigger frequency proportion and the second trigger frequency proportion set, and set it as the confidence level of the first quasi-interest array.

4. The method as described in claim 1, characterized in that, Updating the target user's basic profile based on the interest array set and the secondary interest array set further includes: When the target user's basic profile is updated, a first timestamp is marked on the newly added first interest array. When the first interest array is clicked again, the first timestamp is updated using the click timestamp. When the time interval between the first timestamp and the current timestamp is greater than or equal to the time interval threshold, the first interest array is deleted from the target user basic profile.

5. A user profiling system based on big data, characterized in that: The system is used to implement the big data-based user profiling method according to any one of claims 1-4, the system comprising: The information parsing module is used to analyze the page browsing trajectory of the target user and obtain a set of quasi-interest arrays. Each quasi-interest array includes product type and several advertising element types. The first data statistics module is used to retrieve the first browsing record set of the first sample user group that meets the basic profile of the target user, and to perform high-frequency array statistics by traversing the quasi-interest array set based on the product type and the several advertising element types to obtain the interest array set; The second data statistics module is used to retrieve the second browsing record set of the second sample user group that satisfies the interest array set, obtain the secondary quasi-interest array set, and perform high-frequency array statistics on the non-intersecting quasi-interest arrays of the secondary quasi-interest array set and the quasi-interest array set to obtain the secondary interest array set. The profile update module is used to update the basic profile of the target user based on the interest array set and the secondary interest array set.

Images