CN119691266B - Data processing method applied to feedback platform - Google Patents
Data processing method applied to feedback platform Download PDFInfo
- Publication number
- CN119691266B CN119691266B CN202411736745.4A CN202411736745A CN119691266B CN 119691266 B CN119691266 B CN 119691266B CN 202411736745 A CN202411736745 A CN 202411736745A CN 119691266 B CN119691266 B CN 119691266B
- Authority
- CN
- China
- Prior art keywords
- information
- observation group
- feedback
- matching
- questions
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a data processing method applied to a feedback platform, which comprises the steps of extracting key content of problem information fed back by a user, carrying out weighting algorithm processing on the extracted information, classifying common problems, defining labels for members of an observation group, carrying out intelligent matching on common problems and the labels of the members of the observation group, carrying out problem reply solution by the matched members of the observation group, and converting network comments with the common problems solved into quantifiable satisfaction. The invention has the technical effects of solving the technical problems of low treatment efficiency and weak pertinence of the traditional method and greatly improving the efficiency, accuracy and scientificity of the feedback problem treatment.
Description
Technical Field
The invention relates to the technical field of problem processing platforms, in particular to a data processing method applied to a feedback platform.
Background
The project is a media industry network questioning platform, a user carries out question feedback and exposure through the platform, the platform side transmits the questions to the government functional departments, the functional departments carry out on-line reply processing, and finally the user carries out satisfaction evaluation according to the reply results of the functional departments, wherein the general evaluation results are satisfaction, general and unsatisfied.
The existing scheme provides an online platform for network questioning, lacks intelligent analysis of user feedback problems, has large and small problems reflected by netfriends, is more and more complex, and cannot well extract problem commonalities, namely cannot extract and generalize key contents of problem information fed back by users, so that deep analysis, quick positioning and professional tracking processing cannot be performed on the fed back problems. The traditional method of the online platform for the traditional network questioning has low processing efficiency, low pertinence and low efficiency and accuracy of problem processing.
Disclosure of Invention
The invention aims to provide a data processing method applied to a feedback platform so as to solve the technical problems. The specific technical scheme is as follows:
a data processing method applied to a feedback platform, comprising the following steps:
Extracting key content of problem information fed back by a user, and carrying out weighting algorithm processing on the extracted information to classify out common problems;
defining a label for a member of the observation group;
The problem with commonality is intelligently matched with the label of the member of the observation group, and the problem recovery is carried out by the matched member of the observation group;
Converting the network comments with commonalities after the problems are solved into quantifiable satisfaction;
the method for extracting the key content of the problem information fed back by the user comprises the following steps:
establishing a database aiming at a feedback platform, wherein the database collects feedback problem information of a user, and the feedback problem information comprises picture information and text information;
obtaining a key information list;
purifying the key information in the key information list to obtain the key word information of the problem information fed back by each user, and defining the part of speech of the key word information to form a key word information base;
The method for classifying the commonality problem by carrying out weighting algorithm processing on the extracted information comprises the following steps:
comprehensively considering a plurality of dimensions of the number, the browsing amount, the comment number and the praise number of the user feedback problems, and calculating a heat score for each keyword, wherein the specific formula is as follows:
heat score = (number of questions corresponding to keywords/number of questions corresponding to highest keyword in the current period) ×0.6+ (number of questions browsed corresponding to questions browsed by highest corresponding to questions browsed by quantity) ×0.2+ (number of questions comments corresponding to/number of questions comments highest) ×0.15+ (number of questions praise corresponding to/praise of highest questions) ×0.05,
And selecting the common problem of the main stream according to the heat score.
Further, the data processing method applied to the feedback platform further comprises the following steps:
Processing the extracted information to obtain a burst problem;
the burst problem is intelligently matched with the label of the member of the observation group, and the problem reply is carried out by the matched member of the observation group;
and converting the network comments after the burst problem is solved into quantifiable satisfaction.
Further, the specific method for processing the extracted information to obtain the burst problem comprises the following steps:
Counting the growth rate of the problem corresponding to each keyword, and classifying the problem corresponding to the keyword as a sudden problem if the counted growth rate exceeds a preset rate.
Further, the specific method for counting the growth rate of the problem corresponding to each keyword is as follows:
Counting the number of questions corresponding to each keyword and the release time of each question, calculating the time span of the release time of the questions, and calculating the ratio of the number of the questions to the time span to be used as the growth rate.
Further, after calculating the time span of the issue time of the problems, judging whether the time span is larger than a preset minimum span value, if the time span is larger than the preset minimum span value, further calculating the ratio of the number of the problems to the time span, and taking the ratio as the growth rate.
Further, the key word information of the question information fed back by each user generates key labels of people, things, places, things and organizations, so that management staff can quickly understand and decide whether to adopt the key labels or not.
Further, after the keyword information base is formed, the method for extracting the key content from the problem information fed back by the user further includes:
establishing a department and problem two-stage subject library;
inputting corresponding information into the corresponding theme library for classification in a mode of adding manual correction by a user in the early stage;
and in the later stage, the data input in the earlier stage is used for accumulation to form a bottom word stock, namely, different topic libraries correspond to different keyword information libraries, so that the automatic classification of the questioning and exposing contents of the user is realized, and the optimization and adjustment are continuously carried out in the actual use process.
Further, the method for defining tags for members of an observation group includes:
and (3) data collection:
Collecting basic information of observation group members, professional backgrounds, interests and related experiences, wherein the basic information comprises gender, age and constant regions, the professional backgrounds comprise academic, professional and main research directions, the interests comprise taste types, acquired certificates and competition items, and the related experiences comprise participated items and published papers;
Tag definition:
Constructing a personal portrait tag based on the basic information of the member;
Building an industry specialty tag through professional background, hobbies and related experiences of the members;
Tag verification and optimization:
After the personal portrait tag and the industry special label are defined, verifying and optimizing the personal portrait tag and the industry special label of the member;
Extracting and classifying data information of professional backgrounds, interests and related experiences of members of the observation group, summarizing industry specialty information words capable of reflecting professional capacities and interests of the members, and forming an industry specialty tag based on the industry specialty information words;
Verifying and optimizing the industry specialty label of the member in a mode of comparing and verifying with the information of other observation group members;
the method for intelligently matching the common problems with the labels of the members of the observation group comprises the following steps:
Using text matching technology, matching and analyzing the user feedback problem and the label of the observation group member, including:
Industry mapping matching, namely determining the industry or the field to which the problem belongs according to the problem classification, and searching a label related to the industry or the field in the observation group member label so as to screen members with related industry background or professional knowledge;
matching interest mapping, namely matching and observing interest and hobbies of group members and other labels of professional certificates by using keywords in the questions;
Finally, 1 or more suitable observation group members are selected to track the treatment problem according to the result of the matching degree evaluation.
Further, the method for intelligently matching the common problem with the tag of the member of the observation group further comprises the following steps:
performing matching analysis on the user feedback problem and the tag of the observation group member by using a text matching technology, and performing correlation evaluation and accuracy check sum optimization algorithm on the matching result after obtaining a preliminary matching result;
The correlation evaluation, namely, comprehensively evaluating the correlation between the matched observation group members and the problems according to the experience of the matched observation group members in the field, published papers or factors of research results;
Verifying the matching result by a manual verification or machine learning algorithm to ensure the matching accuracy;
And (3) optimizing the algorithm, namely continuously adjusting member tag information and a keyword information base according to feedback in actual application so as to improve the accuracy and efficiency of matching.
Further, the method for converting the network comments with common problems to quantifiable satisfaction after solving the common problems comprises the following steps:
collecting network comments from a feedback platform and preprocessing data;
According to the preprocessed data, a model for calculating satisfaction is as follows:
Emotion score assignment, namely performing emotion recognition on comment content, using an emotion analysis algorithm to recognize emotion tendencies in the comment, classifying comments into positive, negative or neutral, wherein the positive comment score is higher, the negative comment is deducted, and the neutral comment is not scored;
counting the text word number of the comment content, setting sections with different lengths, wherein each section corresponds to different scores, the shorter comment score is lower, and the longer and detailed comment score is higher;
Identifying related keywords and topics mentioned in the comments by a keyword extraction technology, and grading according to positive and negative surface attributes and importance of the keywords;
Total score satisfaction score = (positive and negative face dimension score 0.5) + (comment detail score 0.3) + (keyword and topic score 0.2). .
The data processing method applied to the feedback platform has the beneficial effects that the data processing method applied to the feedback platform has remarkable effect in the aspect of processing efficiency and quality for solving the mass feedback problem. According to the scheme, the feedback problem can be subjected to depth analysis, quick positioning and professional tracking through a big data technology, so that the technical problems of low processing efficiency and low pertinence of a traditional method are solved, and the efficiency and accuracy of problem processing are greatly improved. Meanwhile, the label of the observation group is constructed, and the problem and an industry expert are automatically docked through the application of an intelligent matching algorithm, so that the innovation mode not only rapidly combines professional knowledge with actual problems, but also remarkably improves the scientificity and the accuracy of problem solving. Therefore, the data processing method applied to the feedback platform of the application shows excellent results in the aspect of processing the mass feedback problem, and makes important contribution to improving the public service quality and efficiency.
Drawings
FIG. 1 is a schematic diagram of a data processing method applied to a feedback platform according to the present invention;
Detailed Description
In order to make the technical scheme of the invention clearer, the invention is further described in detail below with reference to the attached drawings and specific embodiments.
As shown in fig. 1, the data processing method applied to the feedback platform in the present solution includes the following method steps:
step 1, extracting key content of problem information fed back by a user, and carrying out weighting algorithm processing on the extracted information to classify out common problems;
step 2, defining labels for members of the observation group;
step 3, intelligently matching the common problems with the labels of the members of the observation group, and carrying out problem reply solution by the matched members of the observation group;
And 4, converting the network comments with the common problem into quantifiable satisfaction after the problem is solved.
In step 1 of the above method, the specific method for extracting the key content of the problem information fed back by the user is as follows:
Firstly, a database is established aiming at a feedback platform, and the database collects feedback problem information of a user, wherein the feedback problem information comprises picture information and text information. And then, obtaining a key information list according to the picture information and the text information, wherein the key information list is obtained through extracting key information through a three-party capability interface and a media AI model. Then, the key information in the key information list of the upper book is purified to obtain the key information of the problem information fed back by each user, and part-of-speech definition is carried out on the key information to form a key information base, namely, the key information base comprises qualitative key information extracted by the problems fed back by a plurality of users.
The key word information of the problem information fed back by each user generates key labels of people, things, places, things and organizations, so that management staff can quickly understand and decide whether to adopt the problem fed back by the user. The corresponding problems to be adopted and solved can be filtered out rapidly and accurately through the screening of the key labels, and the social problems to be solved by the key labels are corresponding to different management departments, so that the problems corresponding to the processing field of the management departments can be filtered out rapidly and effectively.
As a further method, after the keyword information base is formed, the method for extracting the key content of the problem information fed back by the user further comprises the steps of establishing a department and a problem two-stage subject base. After the department subject database and the problem subject database are established, corresponding information is input into the corresponding subject database to be classified (data accumulation in the early stage) in a mode of manually correcting by using a data processing method applied to a feedback platform in the early stage of user filling, the platform applied to feedback forms a bottom word database by using the data accumulation in the early stage in the later stage of use, different subject databases correspond to different keyword information databases, namely, the department subject database corresponds to the keyword information subject database of related information in the department management field, and the problem subject database corresponds to the keyword information database extracted by the problems fed back by the user and is qualitative. Therefore, the platform applied to feedback can realize automatic classification of the contents of the user question and exposure, namely, when the feedback problem is put forward, the corresponding department subject library and the corresponding problem subject library are directly classified, the management department and the problem direction are quickly locked, the efficiency is high, the classification is accurate, the later-stage accurate solution is convenient, and an accurate solution is provided. In the actual use process, the classification method is continuously optimized and adjusted according to the information changes of the department subject database, the problem subject database and the keyword information database.
In step 1 of the above method, the specific method for classifying the commonality problem includes:
comprehensively considering a plurality of dimensions such as the number, the browsing amount, the comment number, the praise number and the like of the user feedback problems, and calculating a heat score for each keyword, wherein the specific formula is as follows:
heat score = (number of questions corresponding to keywords/number of questions corresponding to highest keyword in the current period) ×0.6+ (number of questions browsed corresponding to questions browsed by highest corresponding to questions browsed by quantity) ×0.2+ (number of questions comments corresponding to/number of questions comments highest) ×0.15+ (number of questions praise corresponding to/praise of highest questions) ×0.05,
The common problem of the main stream is selected according to the heat score, that is, the degree to be solved of the feedback problem is ordered according to the heat score, and the higher the heat score is, the higher the social influence is, the priority and emergency solving are required, so that the solving degree of the problem with larger social influence fed back by the user can be optimized, and the social stability attribute is higher.
The method for defining labels for the members of the observation group in the above step 2 (the observation group members refer to the personnel who can propose professional solutions for the corresponding problems in the platform, and the observation group includes professionals in a plurality of management fields) includes:
and (3) data collection:
basic information, professional background, hobbies and related experience of the observation group member is collected. The basic information comprises gender, age and usual region, the professional background comprises academic, professional and main research directions, and the interests comprise taste types, certificates and competition items;
Tag definition:
based on the basic information of the member, a personal portrait tag is constructed. For example, a member builds a personal portrait tag of "male", "30-35 years old", "eastern China" according to basic information;
Industry specialty tags are constructed based on the members' professional background, hobbies and related experience described above. For example, by extracting and classifying the data information of the professional background, the hobbies and the related experiences of a certain member, summarizing industry specialty information words capable of reflecting the professional ability and the interests of the member, and forming an industry specialty label based on the industry specialty information words;
Tag verification and optimization:
After the personal portrait tag and the industry special label are defined, the personal portrait tag and the industry special label of the member are verified and optimized. Specifically, the personal portrait tags of the members can be verified and optimized in a manner of communication confirmation with the members themselves, and the industry specialty tags of the members can be verified and optimized in a manner of comparison verification with the information of other team members, and the like. If the label is found to be inaccurate or needs to be adjusted, the platform optimizes updating in time.
In step 3 of the above method, the specific method for intelligently matching the common problem with the tag of the member of the observation group is as follows:
and carrying out matching analysis on the problem fed back by the user and the label of the observation group member by using a text matching technology, wherein the specific matching mode comprises the following steps:
Industry mapping matching, namely determining the industry or the field to which the problem belongs according to the problem classification, and searching a label related to the industry or the field in the observation group member label so as to screen members with related industry background or professional knowledge;
Matching interest mapping, namely matching and observing other tags such as interest and hobbies of group members, professional certificates and the like by using keywords in the questions;
Finally, 1 or more suitable observation group members are selected to track the treatment problem according to the result of the matching degree evaluation.
Step 3 of the scheme can accurately and rapidly match problems to observation group members with professional solving capability through the industry mapping matching and interest mapping matching, so that the problems proposed by users are scientifically solved, the problem solving strength and accuracy of the users are improved, and social requirements are met.
Further, the specific method for intelligently matching the common problem with the label of the member of the observation group further comprises the following steps:
performing matching analysis on the user feedback problem and the tag of the observation group member by using a text matching technology, and performing correlation evaluation and accuracy check sum optimization algorithm on the matching result after obtaining a preliminary matching result, wherein the method is specifically as follows;
The correlation evaluation, namely, evaluating the correlation between the matched observation group members and the problems by integrating the experience, published papers or research results and other factors of the matched observation group members in the field;
Verifying the matching result by a manual verification or machine learning algorithm to ensure the matching accuracy;
And (3) optimizing the algorithm, namely continuously adjusting member tag information and a keyword information base according to feedback in actual application so as to improve the accuracy and efficiency of matching.
In step 4 of the above method, the specific method for converting the network comment with the common problem solved into quantifiable satisfaction includes:
network comments are collected from the feedback platform, and data are preprocessed, wherein the preprocessing mode comprises the steps of removing irrelevant characters, punctuation marks, special symbols and the like, and performing word segmentation, stop word removal and the like.
According to the preprocessed data, a model for calculating satisfaction is as follows:
Emotion score assignment, namely performing emotion recognition on comment content, using an emotion analysis algorithm to recognize emotion tendencies in the comment, classifying comments into positive, negative or neutral, wherein the positive comment score is higher, the negative comment is deducted, and the neutral comment is not scored;
counting the text word number of the comment content, setting sections with different lengths, wherein each section corresponds to different scores, the shorter comment score is lower, and the longer and detailed comment score is higher;
Identifying related keywords and topics mentioned in the comments by a keyword extraction technology, and grading according to positive and negative surface attributes and importance of the keywords;
Total score satisfaction score = (positive and negative face dimension score 0.5) + (comment detail score 0.3) + (keyword and topic score 0.2).
According to the satisfaction calculation, the satisfaction condition of the social users on problem solving can be timely and accurately obtained, so that social feedback of the common problem is known according to the satisfaction, a related management department can know the social condition and influence related to the problem more accurately, the management mode of the related management department field of the society is timely adjusted, a better social adjustment function is exerted, and public service quality is improved.
In summary, the data processing method applied to the feedback platform provided by the application achieves remarkable effects in terms of processing efficiency and quality for solving the mass feedback problem. According to the scheme, the feedback problem can be subjected to depth analysis, quick positioning and professional tracking through a big data technology, so that the technical problems of low processing efficiency and low pertinence of a traditional method are solved, and the efficiency and accuracy of problem processing are greatly improved. Meanwhile, the label of the observation group is constructed, and the problem and an industry expert are automatically docked through the application of an intelligent matching algorithm, so that the innovation mode not only rapidly combines professional knowledge with actual problems, but also remarkably improves the scientificity and the accuracy of problem solving. Therefore, the data processing method applied to the feedback platform of the application shows excellent results in the aspect of processing the mass feedback problem, and makes important contribution to improving the public service quality and efficiency.
It can be understood that the identification of the commonality problem takes the thermal value of the keyword as the basis of judgment. And the calculation of the thermal value refers to the number of feedback questions, the browsing amount, the comment number, the praise number and the like. However, in some cases, for the current sudden hot or emergency event, the problem just issued is not widely browsed, reviewed or praised, and the calculated thermal value is smaller and cannot be verified as a common problem, so that the attention is not paid. But these are more of the problems that are reflected and are generally more important.
In order to solve such a problem, as an alternative embodiment, the data processing method applied to the feedback platform further includes:
The extracted information is processed to obtain burst problems, the burst problems are intelligently matched with labels of members of the observation group, problem recovery is carried out by the matched members of the observation group, and network comments after the burst problems are solved are converted into quantifiable satisfaction. Specifically, the processing manner of converting the network comment after the burst problem is solved into quantifiable satisfaction is the same as the statistical manner of the common problem, and is not repeated here.
In the embodiment of the application, the specific method for processing the extracted information to obtain the burst problem is as follows:
Counting the growth rate of the problem corresponding to each keyword, and classifying the problem corresponding to the keyword as a sudden problem if the counted growth rate exceeds a preset rate.
Specifically, the number of questions corresponding to each keyword and the distribution time of each question are counted, and the time span of the distribution time of the questions is calculated. For example, the issue time of the problem closest to the current time is about 3 points 10 minutes of 2022.11.10, and the issue time of the problem farthest from the current time is about 2 points 10 minutes of 2022.11.10, this time span is one hour. The problem number and the time span ratio are then calculated as the growth rate. If the calculated growth rate is too large, it means that a large number of people intensively reflect the problem in a short time, and attention should be paid. In the application, the problem is selected as the burst problem to be processed preferentially. The subsequent processing flow can refer to the processing mode of the common problem. Here, the preset rate may be specifically set as needed, and the present application is not limited thereto.
As a preferred embodiment, after calculating the time span of the issue time of the problems, it is determined whether the time span is greater than a preset minimum span value, and if the time span is greater than the preset minimum span value, the ratio of the number of problems to the time span is further calculated as the growth rate.
It will be appreciated that in some extreme cases, the issue time for two problems to occur is short, and in the above-described calculation, a very large growth rate is calculated even for two problems. However, the burst problem that the present application actually wants to count refers to a problem that is issued by a large number of users in a short time. Only two problems are clearly not expected. Therefore, the application sets a judgment logic and introduces a minimum span value. The calculated time span must meet a value greater than this minimum span and subsequent growth rate calculations are performed.
The foregoing examples illustrate only a few embodiments of the invention and are described in detail herein without thereby limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
Claims (10)
1. A data processing method applied to a feedback platform, comprising the following steps:
Extracting key content of problem information fed back by a user, and carrying out weighting algorithm processing on the extracted information to classify out common problems;
defining a label for a member of the observation group;
The problem with commonality is intelligently matched with the label of the member of the observation group, and the problem recovery is carried out by the matched member of the observation group;
Converting the network comments with commonalities after the problems are solved into quantifiable satisfaction;
the method for extracting the key content of the problem information fed back by the user comprises the following steps:
establishing a database aiming at a feedback platform, wherein the database collects feedback problem information of a user, and the feedback problem information comprises picture information and text information;
obtaining a key information list;
purifying the key information in the key information list to obtain the key word information of the problem information fed back by each user, and defining the part of speech of the key word information to form a key word information base;
The method for classifying the commonality problem by carrying out weighting algorithm processing on the extracted information comprises the following steps:
comprehensively considering a plurality of dimensions of the number, the browsing amount, the comment number and the praise number of the user feedback problems, and calculating a heat score for each keyword, wherein the specific formula is as follows:
heat score = (number of questions corresponding to keywords/number of questions corresponding to highest keyword in the current period) ×0.6+ (number of questions browsed corresponding to questions browsed by highest corresponding to questions browsed by quantity) ×0.2+ (number of questions comments corresponding to/number of questions comments highest) ×0.15+ (number of questions praise corresponding to/praise of highest questions) ×0.05,
And selecting the common problem of the main stream according to the heat score.
2. The method for data processing applied to a feedback platform as claimed in claim 1,
The data processing method applied to the feedback platform further comprises the following steps:
Processing the extracted information to obtain a burst problem;
the burst problem is intelligently matched with the label of the member of the observation group, and the problem reply is carried out by the matched member of the observation group;
and converting the network comments after the burst problem is solved into quantifiable satisfaction.
3. The method for data processing applied to a feedback platform as claimed in claim 2, wherein,
The specific method for processing the extracted information to obtain the burst problem comprises the following steps:
Counting the growth rate of the problem corresponding to each keyword, and classifying the problem corresponding to the keyword as a sudden problem if the counted growth rate exceeds a preset rate.
4. The method for data processing applied to a feedback platform as claimed in claim 3,
The specific method for counting the growth rate of the problem corresponding to each keyword comprises the following steps:
Counting the number of questions corresponding to each keyword and the release time of each question, calculating the time span of the release time of the questions, and calculating the ratio of the number of the questions to the time span to be used as the growth rate.
5. The method for data processing applied to a feedback platform as claimed in claim 4, wherein,
After calculating the time span of the issue time of the problems, judging whether the time span is larger than a preset minimum span value, and if the time span is larger than the preset minimum span value, further calculating the ratio of the number of the problems to the time span, and taking the ratio as the growth rate.
6. The method for data processing applied to a feedback platform as claimed in claim 1,
The key word information of the question information fed back by each user generates key labels of people, things, places, things and organizations so that management staff can quickly understand and decide whether to adopt the key labels or not.
7. The method for data processing applied to a feedback platform as claimed in claim 1,
After the keyword information base is formed, the method for extracting the key content of the problem information fed back by the user further comprises the following steps:
establishing a department and problem two-stage subject library;
inputting corresponding information into the corresponding theme library for classification in a mode of adding manual correction by a user in the early stage;
and in the later stage, the data input in the earlier stage is used for accumulation to form a bottom word stock, namely, different topic libraries correspond to different keyword information libraries, so that the automatic classification of the questioning and exposing contents of the user is realized, and the optimization and adjustment are continuously carried out in the actual use process.
8. The method for data processing applied to a feedback platform as claimed in claim 1,
The method for defining labels for members of an observation group comprises the following steps:
and (3) data collection:
Collecting basic information of observation group members, professional backgrounds, interests and related experiences, wherein the basic information comprises gender, age and constant regions, the professional backgrounds comprise academic, professional and main research directions, the interests comprise taste types, acquired certificates and competition items, and the related experiences comprise participated items and published papers;
Tag definition:
Constructing a personal portrait tag based on the basic information of the member;
Building an industry specialty tag through professional background, hobbies and related experiences of the members;
Tag verification and optimization:
After the personal portrait tag and the industry special label are defined, verifying and optimizing the personal portrait tag and the industry special label of the member;
Extracting and classifying data information of professional backgrounds, interests and related experiences of members of the observation group, summarizing industry specialty information words capable of reflecting professional capacities and interests of the members, and forming an industry specialty tag based on the industry specialty information words;
Verifying and optimizing the industry specialty label of the member in a mode of comparing and verifying with the information of other observation group members;
the method for intelligently matching the common problems with the labels of the members of the observation group comprises the following steps:
Using text matching technology, matching and analyzing the user feedback problem and the label of the observation group member, including:
Industry mapping matching, namely determining the industry or the field to which the problem belongs according to the problem classification, and searching a label related to the industry or the field in the observation group member label so as to screen members with related industry background or professional knowledge;
matching interest mapping, namely matching and observing interest and hobbies of group members and other labels of professional certificates by using keywords in the questions;
Finally, 1 or more suitable observation group members are selected to track the treatment problem according to the result of the matching degree evaluation.
9. The method for data processing applied to a feedback platform as claimed in claim 8,
The method for intelligently matching the common problems with the labels of the members of the observation group further comprises the following steps:
performing matching analysis on the user feedback problem and the tag of the observation group member by using a text matching technology, and performing correlation evaluation and accuracy check sum optimization algorithm on the matching result after obtaining a preliminary matching result;
The correlation evaluation, namely, comprehensively evaluating the correlation between the matched observation group members and the problems according to the experience of the matched observation group members in the field, published papers or factors of research results;
Verifying the matching result by a manual verification or machine learning algorithm to ensure the matching accuracy;
And (3) optimizing the algorithm, namely continuously adjusting member tag information and a keyword information base according to feedback in actual application so as to improve the accuracy and efficiency of matching.
10. The method for data processing applied to a feedback platform as claimed in claim 1,
The method for converting the network comments with common problems into quantifiable satisfaction after solving the common problems comprises the following steps:
collecting network comments from a feedback platform and preprocessing data;
According to the preprocessed data, a model for calculating satisfaction is as follows:
Emotion score assignment, namely performing emotion recognition on comment content, using an emotion analysis algorithm to recognize emotion tendencies in the comment, classifying comments into positive, negative or neutral, wherein the positive comment score is higher, the negative comment is deducted, and the neutral comment is not scored;
counting the text word number of the comment content, setting sections with different lengths, wherein each section corresponds to different scores, the shorter comment score is lower, and the longer and detailed comment score is higher;
Identifying related keywords and topics mentioned in the comments by a keyword extraction technology, and grading according to positive and negative surface attributes and importance of the keywords;
Total score satisfaction score = (positive and negative face dimension score 0.5) + (comment detail score 0.3) + (keyword and topic score 0.2).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411736745.4A CN119691266B (en) | 2024-11-29 | 2024-11-29 | Data processing method applied to feedback platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411736745.4A CN119691266B (en) | 2024-11-29 | 2024-11-29 | Data processing method applied to feedback platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119691266A CN119691266A (en) | 2025-03-25 |
| CN119691266B true CN119691266B (en) | 2025-06-27 |
Family
ID=95037980
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411736745.4A Active CN119691266B (en) | 2024-11-29 | 2024-11-29 | Data processing method applied to feedback platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119691266B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10810357B1 (en) * | 2014-10-15 | 2020-10-20 | Slickjump, Inc. | System and method for selection of meaningful page elements with imprecise coordinate selection for relevant information identification and browsing |
| CN117314470A (en) * | 2023-10-16 | 2023-12-29 | 浪潮软件股份有限公司 | Management subject informatization supervision method and system based on WeChat applet |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118780807A (en) * | 2023-04-07 | 2024-10-15 | 腾讯科技（深圳）有限公司 | A problem feedback method, device and related products |
-
2024
- 2024-11-29 CN CN202411736745.4A patent/CN119691266B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10810357B1 (en) * | 2014-10-15 | 2020-10-20 | Slickjump, Inc. | System and method for selection of meaningful page elements with imprecise coordinate selection for relevant information identification and browsing |
| CN117314470A (en) * | 2023-10-16 | 2023-12-29 | 浪潮软件股份有限公司 | Management subject informatization supervision method and system based on WeChat applet |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119691266A (en) | 2025-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109189901B (en) | Method for automatically discovering new classification and corresponding corpus in intelligent customer service system | |
| CN109783639B (en) | Mediated case intelligent dispatching method and system based on feature extraction | |
| CN117151870A (en) | Portrait behavior analysis method and system based on guest group | |
| CN118246960B (en) | E-commerce platform user feedback analysis method and system based on artificial intelligence | |
| CN114265931B (en) | Consumer policy perception analysis method and system based on big data text mining | |
| CN118939811B (en) | A method for filtering public opinion knowledge graph based on the credibility of news sources | |
| CN117540727B (en) | Subjective question scoring method and system based on ALBERT model and RPA technology | |
| CN119090239B (en) | Intelligent government affair service platform | |
| CN118820985B (en) | Business order analysis and prediction system based on big data | |
| CN112395862A (en) | A Data Mining-Based Environmental Risk Perception Evaluation Method | |
| CN117745048A (en) | Creative flow management method assisted by artificial intelligence | |
| CN118193716A (en) | A self-updating legal case recommendation system based on large models and rule-based reasoning | |
| CN119477243A (en) | An intelligent human resources training recommendation method and system | |
| CN116578703A (en) | Intelligent identification system and method | |
| CN114969560B (en) | Crisis information pushing method and system based on event situation and user cognition | |
| CN118861438B (en) | Engineering information digital consultation management method and system based on Internet of things | |
| CN119807520A (en) | Job recommendation method based on human resources database | |
| CN119691266B (en) | Data processing method applied to feedback platform | |
| CN113762894A (en) | Data processing method and device, electronic equipment and storage medium | |
| CN115146065A (en) | Intelligent information reporting similar content merging method and system | |
| CN114445141A (en) | A method of acquiring customer needs | |
| CN112818215A (en) | Product data processing method, device, equipment and storage medium | |
| WO2025026456A9 (en) | Process and system for providing an environmental, social and governance (esg) report for company | |
| CN117077632B (en) | Automatic generation method for information theme | |
| CN119378759A (en) | An intelligent collection and management system for hot Internet news |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |