Web page data push method and system

Abstract

The present invention relates to the technical field of data push, and specifically discloses a web page data push method and system, the method comprising obtaining user registration information, determining the user's ability histogram according to the registration information; querying the retrieval process in the storage repository corresponding to the user based on the authority granted by the user, counting the retrieval process based on the ability histogram, and building a user portrait; matching the retrieval process monitored in real time with the user portraits of other users, and determining and displaying push data in the matching process; the present invention determines the user's ability according to the user information, records the user's experience according to the user's retrieval process, and when receiving a new retrieval process, simultaneously obtains matching users according to the two perspectives of ability and experience, at this time, the matching user is equivalent to the predecessor of the current user, and the push data is determined by the subsequent experience of the matching user, the fit is extremely high, and the present invention is suitable for internal communication websites in various fields.

Description

Technical Field

[0001] The present invention relates to the technical field of data push, and in particular to a web page data push method and system. Background Art

[0002] In the process of web page interaction, data retrieval is the most basic function. Whether it is a wide area network or a local area network, the data retrieval process is a common function. In the data retrieval process, the data push link is involved. The data push link in the prior art is a push process based on the search content. The management party predetermines a knowledge graph between keywords. When pushing, it can be directly read. This method has strong universality, but the fit with users is not enough. When facing a large number of users, this method is feasible, but when facing a small number of users, especially users in a field, the fit is not high.

[0003] In actual applications, existing websites have become increasingly specialized, generally serving only a certain type of user, such as photography users, digital users, or other types of users. In this case, the focus of pushing data is on the degree of fit. How to improve the degree of fit of pushed data for local users is the technical problem that the present invention aims to solve. Summary of the invention

[0004] The purpose of the present invention is to provide a web page data push method and system to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A web page data pushing method, the method comprising:

[0007] Acquire the user's registration information, and determine the user's ability histogram according to the registration information; wherein the abscissa of the ability histogram corresponds to a preset ability type, and the ordinate of the ability histogram is an ability score;

[0008] Based on the permissions granted by the user, a search process is queried in the repository corresponding to the user, and the search process is statistically analyzed based on the capability histogram to construct a user portrait; the user portrait is a data set cluster based on the capability histogram;

[0009] Receive search requests input by users, monitor the search process in real time, and update the user's corresponding repository based on the monitoring results;

[0010] The retrieval process monitored in real time is matched with the user portraits of other users, and the push data is determined and displayed during the matching process.

[0011] As a further solution of the present invention: the step of obtaining the user's registration information and determining the user's capability histogram according to the registration information includes:

[0012] Receive the user's registration request, send the information collection template to the user, and receive the filled information entered by the user;

[0013] The filling information is input into a preset information conversion model to determine the capability score; wherein the information conversion model is a string comparison model, the information conversion model contains a preset reference database, the reference database includes information items and scoring items, the information items and the filling information are the comparison parties, similarity is generated during the comparison process, the score weight is determined by the similarity, and the capability score is calculated based on the score weight; the calculation process is: In the formula, Z is the ability score, n is the total number of filled information, α(x) is the preset weight function, is the similarity between the i-th filling information and its corresponding information item, and are the i-th filling information and its corresponding information item respectively; F i is the score corresponding to the i-th filled information;

[0014] Query the ability ranking order corresponding to the information collection template, calculate the ability scores based on the ability ranking order, and create a ability histogram.

[0015] As a further solution of the present invention: the query retrieval process in the repository corresponding to the user based on the permissions granted by the user, and the step of constructing a user portrait based on the capability histogram statistical retrieval process include:

[0016] Receive the permission granted by the user and establish a query and retrieval process in the repository corresponding to the user; the retrieval process is a tag set containing a timestamp;

[0017] Query the required capabilities of each tag, and mark the corresponding capability type in the capability histogram;

[0018] Input each label into the preset transcoding model to obtain the label value;

[0019] Expanding the rectangular unit corresponding to the marked capability type based on the label value, and inserting a cutoff determined by a timestamp at the end of each expansion; wherein the expansion process is to convert the rectangular unit into a rectangular column;

[0020] The expanded capability histogram containing the cutoff character is counted to obtain the user portrait.

[0021] As a further solution of the present invention: the step of receiving a search request input by a user, monitoring the search process in real time, and updating the storage repository corresponding to the user according to the monitoring result includes:

[0022] Receive the search request input by the user, record the search time, and monitor the search process in real time; wherein the monitoring process includes an encryption request port, and when receiving the encryption request input by the user, encrypt the monitored search process;

[0023] Inputting the search process into a part-of-speech recognition model, and extracting search keywords according to the part-of-speech recognition results;

[0024] The extracted search keywords are packaged, the search time is inserted, and the packaged results are input into the user's corresponding repository based on the search time.

[0025] As a further solution of the present invention: the step of matching the real-time monitored retrieval process with the user portraits of other users, and determining and displaying the push data during the matching process includes:

[0026] Reading a user's capability histogram, matching the user's capability histogram with other users' capability histograms, and determining matching users for the user;

[0027] Based on the ability histogram statistics of the matching user, the retrieval process monitored in real time is obtained to obtain the sub-portrait unit;

[0028] Traversing the user portrait of the matched user according to the sub-portrait unit, determining the matching degree, matching time and time span;

[0029] A target user is selected based on the matching degree and the time span, a subsequent tag set is read from the target user based on the matching time, and push data is determined and displayed according to the read subsequent tag set.

[0030] As a further solution of the present invention: the step of reading the capability histogram of the user, matching the capability histogram of the user with the capability histograms of other users, and determining the matching user of the user comprises:

[0031] Read the user's ability histogram and other users' ability histograms;

[0032] Normalize the two ability histograms and convert the ability scores into numerical proportions;

[0033] Calculate the Bhattacharyya coefficient between the normalized ability histograms as similarity, and select a matching user for the user according to the similarity;

[0034] The calculation process of the Bhattacharyya coefficient is as follows:

[0035] Where ρ is the Bhattacharyya coefficient of the two ability histograms, m is the total number of ability types, p(i) and p ′(i) are the numerical proportions of the i-th capability type in the two capability histograms; the numerical proportions are the ratios of the capability score to the sum of all scores in the current capability histogram.

[0036] As a further solution of the present invention: the step of traversing the user portrait of the matched user according to the sub-portrait unit and determining the matching degree, matching time and time span comprises:

[0037] Obtain the tag in the sub-portrait unit, traverse the user portrait of the matched user based on the tag, and determine the tag existence value according to the traversal result; the tag existence value includes 1 and 0, 1 represents existence, and 0 represents non-existence;

[0038] Count the label existence values ​​of all labels in the sub-portrait unit and calculate the matching degree;

[0039] When the matching degree reaches the preset matching threshold, the matching position corresponding to the tag with a tag existence value of 1 is obtained, one of the adjacent cutoffs is read based on the matching position, and the timestamp corresponding to the cutoff is obtained as the matching time;

[0040] Calculate the maximum difference of all matching times as the time span.

[0041] The technical solution of the present invention also provides a web page data push system, the system comprising:

[0042] A capability acquisition module, used to acquire the user's registration information and determine the user's capability histogram according to the registration information; wherein the abscissa of the capability histogram corresponds to the preset capability type, and the ordinate of the capability histogram is the capability score;

[0043] A user portrait construction module is used to query the retrieval process in the repository corresponding to the user based on the permissions granted by the user, and to construct a user portrait based on the capability histogram statistics retrieval process; the user portrait is a data set cluster based on the capability histogram;

[0044] The repository update module is used to receive the search request input by the user, monitor the search process in real time, and update the repository corresponding to the user according to the monitoring results;

[0045] The matching display module is used to match the retrieval process monitored in real time with the user portraits of other users, and to determine and display the push data during the matching process.

[0046] As a further solution of the present invention: the capability acquisition module includes:

[0047] An information receiving unit, used to receive a registration request from a user, send an information collection template to the user, and receive filled information input by the user;

[0048] The scoring determination unit is used to input the filling information into a preset information conversion model to determine the capability score; wherein the information conversion model is a string comparison model, the information conversion model contains a preset reference database, the reference database includes information items and scoring items, the information items and the filling information are the comparison parties, similarity is generated during the comparison process, the score weight is determined by the similarity, and the capability score is calculated based on the score weight; the calculation process is: In the formula, Z is the ability score, n is the total number of filled information, α(x) is the preset weight function, is the similarity between the i-th filling information and its corresponding information item, and are the i-th filling information and its corresponding information item respectively; F i is the score corresponding to the i-th filled information;

[0049] Create an execution unit to query the capability ranking order corresponding to the information collection template, calculate the capability scores based on the capability ranking order, and create a capability histogram.

[0050] As a further solution of the present invention: the user portrait construction module includes:

[0051] A process query unit, used to receive the permissions granted by the user and establish a query and retrieval process in the repository corresponding to the user; the retrieval process is a tag set containing a timestamp;

[0052] A type marking unit, used to query the required capability of each tag and mark the capability type corresponding thereto in the capability histogram;

[0053] A transcoding unit, used for inputting each label into a preset transcoding model to obtain a label value;

[0054] The dimension expansion unit is used to expand the rectangular unit corresponding to the marked capability type based on the label value, and insert a cutoff symbol determined by the timestamp at the end of each dimension expansion; wherein the dimension expansion process is to convert the rectangular unit into a rectangular column;

[0055] The statistical unit is used to count the expanded capability histogram containing the cutoff character to obtain the user portrait.

[0056] Compared with the prior art, the beneficial effects of the present invention are: the present invention determines the user's ability based on user information, records the user's experience based on the user's search process, and when receiving a new search process, obtains matching users from both the ability and experience perspectives. At this time, the matching user is equivalent to the predecessor of the current user, and the push data is determined by the subsequent experience of the matching user. The fit is extremely high, and it is suitable for internal communication websites in various fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0058] Figure 1 This is a flowchart of a web page data push method.

[0059] Figure 2 This is a first sub-flow diagram of the web page data pushing method.

[0060] Figure 3 This is a second sub-flow diagram of the web page data pushing method.

[0061] Figure 4 This is a third sub-flow diagram of the web page data pushing method.

[0062] Figure 5 This is a fourth sub-flow diagram of the web page data pushing method.

[0063] Figure 6 This is a structural block diagram of the web page data push system. DETAILED DESCRIPTION

[0064] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0065] Figure 1 The flowchart of the web page data pushing method is as follows. In an embodiment of the present invention, a web page data pushing method is provided, and the method includes:

[0066] Step S100: obtaining the user's registration information, and determining the user's ability histogram according to the registration information; wherein the abscissa of the ability histogram corresponds to a preset ability type, and the ordinate of the ability histogram is the ability score;

[0067] The technical solution of the present invention occurs during the network interaction process. Before the user interacts with the network, he needs to register an account. The account involves the user's personal information, including age, major and work experience, etc. The personal information can be used to preliminarily determine the user's various abilities, such as the ability to accept numbers and the speed of receiving information. These indicators are pre-set by the management party. After obtaining the ability under each indicator, a graph for representing the user's status can be generated, which is called an ability histogram. Generally, the ability type of the ability histogram is a single digit, such as: logical reasoning ability, calculation ability, spatial thinking ability and art appreciation, etc. These evaluation criteria are not unique, so they are not limited.

[0068] Step S200: querying the retrieval process in the repository corresponding to the user based on the authority granted by the user, and constructing a user portrait based on the capability histogram statistics of the retrieval process; the user portrait is a data set cluster based on the capability histogram;

[0069] The retrieval process belongs to private data. In daily life, many people choose to browse incognito, and this application is a data processing process based on browsing records (retrieval process). If the browsing records cannot be obtained, then the subsequent steps cannot be carried out. Therefore, this application has a pre-permission acquisition step. Only after obtaining the permission granted by the user can the subsequent steps be executed.

[0070] Based on the permissions granted by the user, the user's search process is queried, the search process is converted into a set of multiple keywords, the set is converted into a numerical value, and then filled into the corresponding capability type to obtain the user portrait; this process is equivalent to the expansion of the capability histogram, that is, filling data in the vertical direction of the plane where the capability histogram is located, thereby expanding the original two-dimensional data into three-dimensional data, and converting the rectangles in the capability histogram into rectangular columns, thereby obtaining comprehensive data reflecting the user's capabilities and the search process, which is called a user portrait.

[0071] It should be noted that, considering the privacy of the retrieval process, a storage solution with a higher encryption level needs to be adopted during the storage of the retrieval process.

[0072] Step S300: receiving a search request input by a user, monitoring the search process in real time, and updating the storage repository corresponding to the user according to the monitoring result;

[0073] Step S300 is the application process. Step S100 and step S200 are for all users, and the user faced by step S300 is an independent individual. Every time a search request input by the user is received, the search process will be updated, and accordingly, the user portrait will also be updated.

[0074] Step S400: matching the retrieval process monitored in real time with the user portraits of other users, and determining and displaying the push data during the matching process;

[0075] For each user, during a review behavior, based on his / her short search process, he / she can traverse and match the user profiles of all other users, thereby determining users with similar capabilities and experiences, called matching users. The subsequent search process of matching users (which can be directly obtained from the user profile) is read as a reference for generating push data, which can be more in line with the current user.

[0076] Figure 2 This is a first sub-flow diagram of the webpage data pushing method, wherein the step of obtaining the user's registration information and determining the user's capability histogram according to the registration information includes:

[0077] Step S101: receiving a registration request from a user, sending an information collection template to the user, and receiving filled information input by the user;

[0078] Step S102: input the filled information into a preset information conversion model to determine the capability score; wherein the information conversion model is a string comparison model, the information conversion model contains a preset reference database, the reference database includes information items and score items, the information items and the filled information are the comparison parties, similarity is generated during the comparison process, the score weight is determined by the similarity, and the capability score is calculated based on the score weight; the calculation process is: In the formula, Z is the ability score, n is the total number of filled information, α(x) is the preset weight function, is the similarity between the i-th filling information and its corresponding information item, and are the i-th filling information and its corresponding information item respectively; F i is the score corresponding to the i-th filled information;

[0079] Step S103: query the capability ranking order corresponding to the information collection template, calculate the capability scores according to the capability ranking order, and create a capability histogram.

[0080] The generation logic of the capability histogram is very simple. The administrator first creates an information collection template to obtain various information about the user and then determine the user's capabilities. At the same time, once the information collection template is determined, the arrangement order of each capability type in the capability histogram is also determined.

[0081] Specifically, the information obtained by the information collection template is the text information entered by the user, while generating a capability histogram requires numerical information. Therefore, the filled information needs to be converted into a numerical value, called the capability score. The conversion rules of the conversion process are determined independently by the management party. The conversion rules are generally in tabular form. No matter what kind of filled information is entered, the score can be directly read in the table.

[0082] Figure 3 The second sub-process flowchart of the web page data push method includes the following steps: querying the retrieval process in the repository corresponding to the user based on the permissions granted by the user, and constructing the user portrait based on the capability histogram statistics retrieval process:

[0083] Step S201: receiving the authority granted by the user, and establishing a query retrieval process in the repository corresponding to the user; the retrieval process is a tag set containing a timestamp;

[0084] Step S202: query the required capability of each tag, and mark the capability type corresponding to it in the capability histogram;

[0085] Step S203: input each tag into a preset transcoding model to obtain a tag value;

[0086] Step S204: Expand the rectangular unit corresponding to the marked capability type based on the tag value, and insert a cutoff symbol determined by the timestamp at the end of each expansion; wherein the expansion process is to convert the rectangular unit into a rectangular column;

[0087] Step S205: Count the expanded capability histogram containing the cutoff character to obtain the user portrait;

[0088] The above content makes specific restrictions on the process of constructing user portraits. First, each user has an exclusive repository for storing retrieval processes. When the permissions granted by the user are received, the retrieval process can be queried according to the permissions granted by the user. The retrieval process involved in this application is a collection of multiple tags, and each tag set has a unique time information, called a timestamp.

[0089] There are various types of labels, and the required capabilities are different. For example, when the label is mathematics, the required capabilities include calculation ability and logical ability. The required capabilities of different labels are predetermined by the management party. The required capabilities of each label are queried, and the corresponding rectangles are queried and marked in the capability histogram. On this basis, the label is input into the preset transcoding model to obtain data in numerical form, which is called label value. For the marked rectangle, the label value is used as the height in the vertical direction of the capability histogram to convert the rectangle into a column.

[0090] It should be noted that each retrieval process corresponds to a conversion process, and each conversion process will obtain an expanded three-dimensional data. By inserting a cutoff symbol into the three-dimensional data, the matrix segment corresponding to each retrieval process can be determined.

[0091] Figure 4 The third sub-process flowchart of the web page data push method includes the steps of receiving a search request input by a user, monitoring the search process in real time, and updating the storage repository corresponding to the user according to the monitoring result:

[0092] Step S301: receiving a search request input by a user, recording the search time, and monitoring the search process in real time; wherein the monitoring process includes an encryption request port, and when receiving an encryption request input by a user, encrypting the monitored search process;

[0093] Step S302: input the search process into a part-of-speech recognition model, and extract search keywords according to the part-of-speech recognition results;

[0094] Step S303: Pack the extracted search keywords, insert the search time, and input the packaged results into the storage repository corresponding to the user based on the search time.

[0095] Steps S301 to S303 specifically define the data storage process, which belongs to the micro perspective. Step S300 is the generation process of the retrieval process, which is the basis for building a repository for each user. The specific working principle is as follows:

[0096] Receive the search request input by the user, record the time when the request is sent as the search time, then monitor the search process in real time, extract the search keywords in the search process by the existing part-of-speech recognition model as labels, package all labels corresponding to the same search process, and obtain the label set as data stored in the repository.

[0097] During the storage process, the storage order needs to be determined by the retrieval time to ensure the orderliness of the storage repository.

[0098] Figure 5 This is a fourth sub-process flowchart of the web page data push method, wherein the steps of matching the real-time monitored search process with the user portraits of other users, and determining and displaying the push data during the matching process include:

[0099] Step S401: reading a user's capability histogram, matching the user's capability histogram with other users' capability histograms, and determining matching users of the user;

[0100] Each user has a capability histogram. By comparing the capability histograms of different users, we can determine whether the capabilities of the users are similar.

[0101] Step S402: Counting the retrieval process monitored in real time based on the ability histogram of the matching user to obtain a sub-portrait unit;

[0102] Taking a certain user as an example, when the user's retrieval process is monitored, a user portrait corresponding to this retrieval process is constructed based on the same logic (step S202 to step S205). Since only one section of the retrieval process is analyzed, the amount of data for the obtained user portrait is only one section, so it is called a sub-portrait unit.

[0103] Step S403: traversing the user portrait of the matched user according to the sub-portrait unit to determine the matching degree, matching time and time span;

[0104] It can be seen from the above content that the sub-portrait unit is equivalent to a subset of the user portrait (all three-dimensional data). Therefore, by traversing the user portrait based on the sub-portrait unit, a traversal result can be obtained. The traversal result includes matching degree, matching time and time span. The matching degree is used to characterize the degree of inclusion of the sub-portrait unit in the user portrait; the matching time is used to characterize the timestamp corresponding to the included position; because the sub-portrait unit is a label set, the included position is not unique, and there is a time difference between multiple timestamps. The maximum value of the time difference is called the time span.

[0105] Step S404: selecting a target user based on the matching degree and the time span, reading a subsequent tag set from the target user based on the matching time, and determining and displaying push data according to the read subsequent tag set;

[0106] The user most similar to the current user is selected based on the matching degree and time span, and then the subsequent tag set of the most similar user is read according to the matching time (subsequent retrieval process), and the user's push data is determined based on the subsequent tag set, which is highly consistent with the current user.

[0107] In summary, the matching process of this application is as follows:

[0108] First, the user's ability histogram is compared with the ability histograms of other users to determine users with similar abilities. This is the first level of matching. The determined users with similar abilities are not unique and are collectively referred to as matched users.

[0109] Then, the ability histogram of the matching user is read, and the ability histogram of the matching user (not the current user) is used as a benchmark to count the retrieval process of the current user and obtain the sub-image unit. Since the ability histograms based on the same are the same, the subsequent comparison process of the sub-image unit and the user portrait is more accurate.

[0110] Finally, the appearance position of each label in the sub-image unit in the user portrait (belonging to the same matched user) is obtained in turn, and the cutoff near the appearance position is read to query the time when the matched user performs similar search. Since the number of labels is not unique, the obtained time is also not unique. Generally speaking, the smaller the maximum span of the time corresponding to all labels, the more similar the retrieval process of the matched user is to the retrieval process of the current user.

[0111] The practical significance of the above process is that for a certain user, after the above matching process, users with similar abilities (ability histogram matching) and similar experiences (inspection process matching) can be queried. The push data determined by users with similar abilities and experiences has a high degree of fit with the current user.

[0112] As a preferred embodiment of the technical solution of the present invention, the step of reading the capability histogram of a user, matching the capability histogram of the user with the capability histograms of other users, and determining a matching user of the user comprises:

[0113] Read the user's ability histogram and other users' ability histograms;

[0114] Normalize the two ability histograms and convert the ability scores into numerical proportions;

[0115] Calculate the Bhattacharyya coefficient between the normalized ability histograms as similarity, and select a matching user for the user according to the similarity;

[0116] The calculation process of the Bhattacharyya coefficient is as follows:

[0117] Where ρ is the Bhattacharyya coefficient of the two ability histograms, m is the total number of ability types, p(i) and p ′ (i) are the numerical proportions of the i-th capability type in the two capability histograms; the numerical proportions are the ratios of the capability score to the sum of all scores in the current capability histogram.

[0118] The above content specifically limits the matching process of the ability histogram, and its purpose is to determine two users with similar abilities. From a data perspective, it is to compare the similarity of two ability histograms. This application uses the Bhattacharyya coefficient as an evaluation label. The larger the Bhattacharyya coefficient, the higher the similarity.

[0119] As a preferred embodiment of the technical solution of the present invention, the step of traversing the user portrait of the matching user according to the sub-portrait unit and determining the matching degree, matching time and time span includes:

[0120] Obtain the tag in the sub-portrait unit, traverse the user portrait of the matched user based on the tag, and determine the tag existence value according to the traversal result; the tag existence value includes 1 and 0, 1 represents existence, and 0 represents non-existence;

[0121] Count the label existence values ​​of all labels in the sub-portrait unit and calculate the matching degree;

[0122] When the matching degree reaches the preset matching threshold, the matching position corresponding to the tag with a tag existence value of 1 is obtained, one of the adjacent cutoffs is read based on the matching position, and the timestamp corresponding to the cutoff is obtained as the matching time;

[0123] Calculate the maximum difference of all matching times as the time span.

[0124] The above content provides a specific comparison scheme between sub-portrait units and user portraits. The core of this scheme is the comparison of labels. The number of labels is not unique. If a label in the sub-portrait unit appears in the user portrait, the label existence value is 1. By counting the label existence values ​​of all labels, the matching degree can be calculated.

[0125] During the calculation of the tag existence value, if the tag existence value is 1, the matching position will be read synchronously, that is, the position where the tag appears in the user portrait. By reading the cutoff mark adjacent to the position, a timestamp can be obtained (the cutoff mark is determined by the timestamp, and reversing the timestamp from the cutoff mark is a reverse process and is not complicated) as the matching time. The maximum difference of the matching time is calculated to obtain the time span.

[0126] Figure 6 1 is a structural block diagram of a web page data push system. As a preferred embodiment of the technical solution of the present invention, the present invention also provides a web page data push system. The system 10 includes:

[0127] A capability acquisition module 11 is used to acquire the user's registration information and determine the user's capability histogram according to the registration information; wherein the abscissa of the capability histogram corresponds to the preset capability type, and the ordinate of the capability histogram is the capability score;

[0128] A user portrait construction module 12 is used to query the retrieval process in the repository corresponding to the user based on the permissions granted by the user, and to construct a user portrait based on the capability histogram statistics retrieval process; the user portrait is a data set cluster based on the capability histogram;

[0129] The repository update module 13 is used to receive the search request input by the user, monitor the search process in real time, and update the repository corresponding to the user according to the monitoring result;

[0130] The matching and display module 14 is used to match the retrieval process monitored in real time with the user portraits of other users, and to determine and display the push data during the matching process.

[0131] Furthermore, the capability acquisition module 11 includes:

[0132] An information receiving unit, used to receive a registration request from a user, send an information collection template to the user, and receive filled information input by the user;

[0133] The scoring determination unit is used to input the filling information into a preset information conversion model to determine the capability score; wherein the information conversion model is a string comparison model, the information conversion model contains a preset reference database, the reference database includes information items and scoring items, the information items and the filling information are the comparison parties, similarity is generated during the comparison process, the score weight is determined by the similarity, and the capability score is calculated based on the score weight; the calculation process is: In the formula, Z is the ability score, n is the total number of filled information, α(x) is the preset weight function, is the similarity between the i-th filling information and its corresponding information item, and are the i-th filling information and its corresponding information item respectively; F i is the score corresponding to the i-th filled information;

[0134] Create an execution unit to query the capability ranking order corresponding to the information collection template, calculate the capability scores based on the capability ranking order, and create a capability histogram.

[0135] Specifically, the user portrait building module 12 includes:

[0136] A process query unit, used to receive the permissions granted by the user and establish a query and retrieval process in the repository corresponding to the user; the retrieval process is a tag set containing a timestamp;

[0137] A type marking unit, used to query the required capability of each tag and mark the capability type corresponding thereto in the capability histogram;

[0138] A transcoding unit, used for inputting each label into a preset transcoding model to obtain a label value;

[0139] The dimension expansion unit is used to expand the rectangular unit corresponding to the marked capability type based on the label value, and insert a cutoff symbol determined by the timestamp at the end of each dimension expansion; wherein the dimension expansion process is to convert the rectangular unit into a rectangular column;

[0140] The statistical unit is used to count the expanded capability histogram containing the cutoff character to obtain the user portrait.

[0141] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A web page data push method, characterized in that: The method comprises: Acquire the user's registration information, and determine the user's ability histogram according to the registration information; wherein the abscissa of the ability histogram corresponds to a preset ability type, and the ordinate of the ability histogram is an ability score; Based on the permissions granted by the user, a search process is queried in the repository corresponding to the user, and the search process is statistically analyzed based on the capability histogram to construct a user portrait; the user portrait is a data set cluster based on the capability histogram; Receive a search request input by a user, monitor the search process in real time, and update the repository corresponding to the user according to the monitoring result; wherein the user in the content of obtaining the user's registration information is all users, and the user in the content of receiving the search request input by the user is a certain independent individual; Match the search process monitored in real time with the user profiles of other users, and determine and display the push data during the matching process; The step of obtaining the user's registration information and determining the user's capability histogram according to the registration information comprises: Receive the user's registration request, send the information collection template to the user, and receive the filled information entered by the user; The filling information is input into a preset information conversion model to determine the capability score; wherein the information conversion model is a string comparison model, the information conversion model contains a preset reference database, the reference database includes information items and scoring items, the information items and the filling information are the comparison parties, similarity is generated during the comparison process, the score weight is determined by the similarity, and the capability score is calculated based on the score weight; the calculation process is: In the formula, Z is the ability score, n is the total number of filled information, α(x) is the preset weight function, is the similarity between the i-th filling information and its corresponding information item, and are the i-th filling information and its corresponding information item respectively; F i is the score corresponding to the i-th filled information; Query the ability ranking order corresponding to the information collection template, calculate the ability scores according to the ability ranking order, and create an ability histogram; The steps of querying the retrieval process in the repository corresponding to the user based on the permissions granted by the user and constructing the user portrait based on the capability histogram statistics retrieval process include: Receive the permission granted by the user and establish a query and retrieval process in the repository corresponding to the user; the retrieval process is a tag set containing a timestamp; Query the required capabilities of each tag, and mark the corresponding capability type in the capability histogram; Input each label into the preset transcoding model to obtain the label value; Expanding the rectangular unit corresponding to the marked capability type based on the label value, and inserting a cutoff determined by a timestamp at the end of each expansion; wherein the expansion process is to convert the rectangular unit into a rectangular column; Count the expanded capability histogram containing the cutoff character to obtain the user portrait; The steps of receiving a search request input by a user, monitoring the search process in real time, and updating the storage repository corresponding to the user according to the monitoring result include: Receive the search request input by the user, record the search time, and monitor the search process in real time; wherein the monitoring process includes an encryption request port, and when receiving the encryption request input by the user, encrypt the monitored search process; Inputting the search process into a part-of-speech recognition model, and extracting search keywords according to the part-of-speech recognition results; Packaging the extracted search keywords, inserting the search time, and inputting the packaged results into the user's corresponding repository based on the search time; The step of matching the real-time monitored search process with the user portraits of other users and determining and displaying the push data during the matching process includes: Reading a user's capability histogram, matching the user's capability histogram with other users' capability histograms, and determining matching users for the user; Based on the ability histogram statistics of the matching user, the retrieval process monitored in real time is obtained to obtain the sub-portrait unit; Traversing the user portrait of the matched user according to the sub-portrait unit, determining the matching degree, matching time and time span; Selecting a target user based on the matching degree and the time span, reading a subsequent tag set from the target user based on the matching time, and determining and displaying push data based on the read subsequent tag set; The step of reading the capability histogram of the user, matching the capability histogram of the user with the capability histograms of other users, and determining a matching user of the user comprises: Read the user's ability histogram and other users' ability histograms; Normalize the two ability histograms and convert the ability scores into numerical proportions; Calculate the Bhattacharyya coefficient between the normalized ability histograms as similarity, and select a matching user for the user according to the similarity; The calculation process of the Bhattacharyya coefficient is as follows: Where ρ is the Bhattacharyya coefficient of the two ability histograms, m is the total number of ability types, p(i) and p ′ (i) is the numerical proportion of the i-th ability type in the two ability histograms; the numerical proportion is the ratio of the ability score to the sum of all scores in the current ability histogram; The step of traversing the user portrait of the matched user according to the sub-portrait unit and determining the matching degree, matching time and time span comprises: Obtain the tag in the sub-portrait unit, traverse the user portrait of the matched user based on the tag, and determine the tag existence value according to the traversal result; the tag existence value includes 1 and 0, 1 represents existence, and 0 represents non-existence; Count the label existence values ​​of all labels in the sub-portrait unit and calculate the matching degree; When the matching degree reaches the preset matching threshold, the matching position corresponding to the tag with a tag existence value of 1 is obtained, one of the adjacent cutoffs is read based on the matching position, and the timestamp corresponding to the cutoff is obtained as the matching time; Calculate the maximum difference of all matching times as the time span.

2. A web page data push system, the system being used to execute the web page data push method according to claim 1, characterized in that: The system comprises: A capability acquisition module, used to acquire the user's registration information and determine the user's capability histogram according to the registration information; wherein the abscissa of the capability histogram corresponds to the preset capability type, and the ordinate of the capability histogram is the capability score; A user portrait construction module is used to query the retrieval process in the repository corresponding to the user based on the permissions granted by the user, and to construct a user portrait based on the capability histogram statistics retrieval process; the user portrait is a data set cluster based on the capability histogram; A repository update module is used to receive a search request input by a user, monitor the search process in real time, and update the repository corresponding to the user according to the monitoring result; wherein the user in the content of obtaining the user's registration information is all users, and the user in the content of receiving the search request input by the user is a certain independent individual; The matching display module is used to match the retrieval process monitored in real time with the user portraits of other users, and to determine and display the push data during the matching process.

3. The web page data push system according to claim 2, characterized in that: The capability acquisition module includes: An information receiving unit, used to receive a registration request from a user, send an information collection template to the user, and receive filled information input by the user; The scoring determination unit is used to input the filling information into a preset information conversion model to determine the capability score; wherein the information conversion model is a string comparison model, the information conversion model contains a preset reference database, the reference database includes information items and scoring items, the information items and the filling information are the comparison parties, similarity is generated during the comparison process, the score weight is determined by the similarity, and the capability score is calculated based on the score weight; the calculation process is: In the formula, Z is the ability score, n is the total number of filled information, α(x) is the preset weight function, is the similarity between the i-th filling information and its corresponding information item, and are the i-th filling information and its corresponding information item respectively; F i is the score corresponding to the i-th filled information; Create an execution unit to query the capability ranking order corresponding to the information collection template, calculate the capability scores based on the capability ranking order, and create a capability histogram.

4. The web page data push system according to claim 2, characterized in that: The user portrait building module includes: A process query unit, used to receive the permissions granted by the user and establish a query and retrieval process in the repository corresponding to the user; the retrieval process is a tag set containing a timestamp; A type marking unit, used to query the required capability of each tag and mark the capability type corresponding thereto in the capability histogram; A transcoding unit, used for inputting each label into a preset transcoding model to obtain a label value; A dimension expansion unit is used to expand the rectangular unit corresponding to the marked capability type based on the label value, and insert a cutoff symbol determined by a timestamp at the end of each dimension expansion; wherein the dimension expansion process is to convert the rectangular unit into a rectangular column; The statistical unit is used to count the expanded capability histogram containing the cutoff character to obtain the user portrait.

Images