CN112052966A - Power customer satisfaction analysis system and method based on site emergency repair work order - Google Patents
Power customer satisfaction analysis system and method based on site emergency repair work order Download PDFInfo
- Publication number
- CN112052966A CN112052966A CN202011013654.XA CN202011013654A CN112052966A CN 112052966 A CN112052966 A CN 112052966A CN 202011013654 A CN202011013654 A CN 202011013654A CN 112052966 A CN112052966 A CN 112052966A
- Authority
- CN
- China
- Prior art keywords
- customer
- repair
- data
- satisfaction
- analysis
- 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.)
- Pending
Links
- 230000008439 repair process Effects 0.000 title claims abstract description 161
- 238000004458 analytical method Methods 0.000 title claims abstract description 73
- 238000000034 method Methods 0.000 title claims description 40
- 238000007781 pre-processing Methods 0.000 claims abstract description 28
- 238000007637 random forest analysis Methods 0.000 claims abstract description 25
- 238000010276 construction Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims description 41
- 230000008569 process Effects 0.000 claims description 23
- 238000011156 evaluation Methods 0.000 claims description 17
- 230000002159 abnormal effect Effects 0.000 claims description 13
- 238000003066 decision tree Methods 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 12
- 238000004140 cleaning Methods 0.000 claims description 9
- 238000007619 statistical method Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 5
- 230000015572 biosynthetic process Effects 0.000 claims description 2
- 238000004141 dimensional analysis Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 239000013589 supplement Substances 0.000 claims 1
- 238000004364 calculation method Methods 0.000 abstract description 5
- 238000005065 mining Methods 0.000 abstract description 3
- 230000008447 perception Effects 0.000 abstract description 3
- 230000006872 improvement Effects 0.000 description 3
- 238000011835 investigation Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 239000000779 smoke Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/01—Customer relationship services
- G06Q30/015—Providing customer assistance, e.g. assisting a customer within a business location or via helpdesk
- G06Q30/016—After-sales
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a customer satisfaction analysis system and a customer satisfaction analysis method based on a field emergency repair work order. The data preprocessing module is used for preprocessing the emergency repair work order; the index system construction module calculates corresponding derived indexes to form a satisfaction analysis system on the basis of an original satisfaction rate calculation formula; the model building module builds a random forest model and perfects a satisfaction analysis system; the analysis decision module comprehensively senses the client appeal according to the model prediction result, and factors influencing the satisfaction degree of the on-site first-aid repair client are analyzed in a multi-dimensional mode, so that data support is provided for decision making. The system provided by the invention adopts the mining analysis based on the on-site emergency repair work order, particularly predicts the unevaluated work order accounting for 30% of the actual service, solves the problem of incomplete perception of customer appeal, supports service management decision, improves the quality of power supply service and improves the customer satisfaction.
Description
Technical Field
The invention belongs to the field of power grid informatization, and particularly relates to a power customer satisfaction analysis system and method based on a field emergency repair work order.
Background
With the development of the new state of the Internet plus, the innovative practice of the new power service of the Internet plus marketing service is continuously promoted, the online channel construction is strengthened, the intelligent upgrade of a business hall is accelerated, and the development trend of providing services for enterprises is achieved. Therefore, new technologies such as big data, cloud computing, internet of things and mobile internet are used for promoting the smart power grid, and changes of quality improvement and efficiency improvement can be brought.
The service is gradually becoming the core competitiveness of enterprises, and the investigation of the service satisfaction degree also becomes an important means for the enterprises to find the service problems and improve the service quality. However, the conventional satisfaction survey is realized by sampling survey, and from the viewpoint of management improvement, the data acquisition period is limited, the time is relatively delayed, and the survey cost is high; meanwhile, the sampling investigation is greatly influenced by the sample, and the risk of deviation of the investigation result and the actual situation exists. How to efficiently and accurately obtain the user satisfaction survey and analyze the reasons of user dissatisfaction becomes the current service problem.
Therefore, in the prior art, more labor cost is needed to realize the evaluation of the customer satisfaction, and along with the increase of data dimensions, the continuous increase of data quantity and the continuous generation of multidimensional data, the overall difficulty and workload of the satisfaction evaluation are greatly increased, and the evaluation and analysis cost is increased.
Disclosure of Invention
In view of the above, the present invention aims to provide a power customer satisfaction analysis system and method based on a field emergency repair work order, so as to solve the problems that the existing satisfaction evaluation cannot efficiently and accurately sense customer appeal, and cannot deeply analyze the true reason of the customer appeal dissatisfaction, resulting in high service cost and poor service quality.
Therefore, the analysis system adopts the following technical scheme: power customer satisfaction analysis system based on-site emergency repair work order includes:
the data preprocessing module is used for preprocessing the emergency repair work order, and comprises data cleaning, data stipulation, data abnormal value repair and data missing value filling;
the index system construction module is used for associating the satisfaction degree evaluation of the power customer with a specific business process according to the preprocessing result of the data preprocessing module, constructing derivative indexes comprehensively describing the on-site emergency repair work quality and efficiency according to the actual business situation and forming a satisfaction degree analysis index system;
the model building module is used for building multi-dimensional width table data of the emergency repair work order according to the satisfaction degree analysis index system and building a random forest model to predict the part which is not evaluated in the customer opinion;
and the analysis decision module is used for comprehensively sensing the client appeal according to the model prediction result, analyzing factors influencing the satisfaction degree of the on-site first-aid repair client in a multi-dimensional manner and providing data support for decision.
The invention relates to a power customer satisfaction analysis method based on a field emergency repair work order, which comprises the following steps:
s301, preprocessing the emergency repair work order, including data cleaning, data specification, data abnormal value repair and data missing value filling;
s302, according to the preprocessing result, the satisfaction evaluation of the power customer is related to a specific business process, a derivative index which comprehensively describes the quality and efficiency of on-site first-aid repair work is constructed according to the actual business situation, and a satisfaction analysis index system is formed;
s303, building multi-dimensional width table data of the emergency repair work order according to a satisfaction degree analysis index system, and building a random forest model to predict an unevaluated part in customer opinions;
s304, comprehensively sensing the customer appeal according to the model prediction result, carrying out multi-dimensional analysis on factors influencing the satisfaction degree of the on-site first-aid repair customer, and providing data support for decision making.
Compared with the prior art, the invention has the following advantages and beneficial effects: mining analysis is carried out based on the on-site first-aid repair work order, and particularly, prediction is carried out on an unevaluated work order which accounts for up to 30% of the actual service; on the basis of an original satisfaction rate calculation formula, corresponding derivative indexes are calculated to form a satisfaction degree analysis system, a random forest model is built, the satisfaction degree analysis system is perfected, the problem that customer appeal perception is not comprehensive is solved, service management decisions are supported, power supply service quality is improved, and customer satisfaction is improved. The unevaluated work order refers to a work order which is not evaluated by a client correspondingly, and due to the fact that 30% of data corresponding to the unevaluated work order is lacked in the prior art, perception of client appeal in the client satisfaction degree analysis method is incomplete, and therefore results of satisfaction degree analysis are not accurate and the satisfaction degree distribution condition of the client cannot be reflected comprehensively.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic diagram of a customer satisfaction analysis system based on a first-aid repair work order according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a Bagging algorithm;
fig. 3 is a flowchart of a customer satisfaction analysis method based on a first-aid repair work order in an embodiment of the invention.
Detailed Description
In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are a part of the embodiments of the present invention, but not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example 1
The embodiment discloses an electric power customer satisfaction analysis system based on a field emergency repair work order, and please refer to fig. 1, the system comprises a data preprocessing module 100, an index system building module 200, a model building module 300, and an analysis decision module 400. The analysis system of the embodiment can evaluate the satisfaction degree of the emergency repair work order, before deeply mining related work order data, exploration and analysis of the data, restoration of abnormal data values and filling of missing data values are performed through the data preprocessing module, the construction of derived indexes is performed through the index system construction module, a data basis is provided for subsequent model construction, and then the client appeal is comprehensively sensed through the analysis decision module.
The analysis system of the embodiment takes a site emergency repair work order as a data base, refers to the content of an emergency repair related work order in a national network business support system, combines 95598 work order tables, an emergency repair process table, a department expansion table, an emergency repair order statistical information table, an emergency repair dispatching information table, an emergency repair team group information table, an unfinished emergency repair table, an emergency repair process detail table and other omnibearing data, and analyzes the site emergency repair work situation from aspects such as the timeliness of emergency repair, the personnel service attitude, the satisfaction degree distribution condition of each city and the like through a data preprocessing module; the index system construction module combines the multiple dimensions to construct derivative indexes on the basis of an original satisfaction rate calculation formula, and comprises the following steps: the method comprises the following steps of (1) extracting and grouping and counting keywords of personnel attitude, power failure information notification timeliness, notification emergency repair speed, arrival speed, emergency repair speed, promised arrival time completion condition and repair reporting content to form a satisfaction degree analysis index system; on the basis, the model building module enables samples with known customer opinions to be obtained according to the following steps of 7:3, dividing the training set into a training set and a testing set, wherein the training set comprises 204005 samples, the testing set comprises 87431 samples, and the training set is used for constructing a random forest regression model containing 5000 decision trees; the random forest regression model is used for predicting the unevaluated work orders accounting for up to 30% of the actual business, a satisfaction degree analysis index system is perfected, the problem that the customer appeals to be imperceptible is solved, service management decisions are supported, the power supply service quality is improved, and the overall satisfaction degree of the customer is improved.
In the implementation process of the embodiment, the data preprocessing module performs data cleaning, data specification, abnormal value processing and missing value processing on the on-site emergency repair work order; the index system construction module is used for linking the satisfaction evaluation of the power customer with a specific business process, constructing a satisfaction analysis index system which comprehensively describes the quality and efficiency of on-site emergency repair work according to the actual business condition, comprehensively measuring the on-site emergency repair work and applying the evaluation index system to the construction of a subsequent model; the model construction module analyzes an index system according to the constructed satisfaction degree and predicts an unevaluated part in the customer opinion by combining with multi-dimensional wide-list data of the emergency repair work order so as to comprehensively sense the customer appeal, further analyze the factors of the customer satisfaction degree and provide corresponding data support for subsequent decision making; and the analysis decision module is used for carrying out multidimensional analysis on the factors of the on-site emergency repair satisfaction according to the model prediction result and carrying out decision support on the related work of the subsequent business department.
The data preprocessing module 100 performs data exploration and analysis, data abnormal value restoration, missing value filling and other processing on the on-site emergency repair work order, and specifically includes client opinion restoration and generalization processing, personnel attitude processing, power failure information notification timeliness processing, notification speed processing, arrival speed processing, restoration speed processing, repair content processing, and missing value restoration processing of each field.
The data abnormal value restoration is mainly used for restoring the customer opinions, and the customer opinions contain blank values and unclear expression attitude. When the opinion of the client is repaired, the method mainly comprises three steps: firstly, associating a 95598 worksheet table with a first-aid repair sheet, and repairing and supplementing customer opinions in the first-aid repair sheet by utilizing the customer opinions in the 95598 worksheet table; secondly, according to the keywords of the customer opinions in the first-aid repair order, including satisfaction, awareness, recognition, comprehension, no evaluation and the like, normalization processing is carried out on the customer opinions, and the keywords after processing are as follows: very satisfactory, generally satisfactory, unsatisfactory, very unsatisfactory, not evaluated; thirdly, according to the business logic mined from the data, when the customer opinions are null and the state of the first-aid repair order is not filed, but the fault content contains the work order with filed characters, the customer opinions are considered to be satisfactory, and the customer opinions in the relevant work order are repaired.
Missing value filling is also called null value restoration, and the phenomenon that important field values are null exists in the data exploration process, including the names of local cities, whether fault points are confirmed, whether important customers are contained, fault information, fault equipment property rights, processing results and whether tripping is detected, and the null values of all the fields are filled by using the values with the maximum contribution degree according to the contribution degree of different values in all indexes to the satisfaction degree. And repairing the null value of the committed arrival time by using the average value.
The index system construction module 200 links the satisfaction evaluation of the power customer with a specific business process according to the preprocessing result of the data preprocessing module, constructs derivative indexes comprehensively describing the on-site emergency repair work quality and efficiency according to the actual business situation, and forms a satisfaction analysis index system; the constructed index system is used for comprehensively measuring the on-site emergency repair work and is applied to the subsequent model construction part. The derivation indexes are respectively introduced as follows:
personnel attitude: for the 95598 worksheet, when the acceptance content or the acceptance opinions have the content related to the personnel attitude, the acceptance content or the acceptance opinions comprise bad personnel attitude, dissonance of a client to the personnel attitude, bad personnel attitude, strong and hard personnel attitude, poor personnel attitude, cold personnel attitude, strong and hard speech, existence of < 35881;, abuse, existence of noise and the like; and constructing a staff attitude derivative index by associating the acceptance content of the cleaning worksheet or the serial number of the application form contained in the acceptance suggestion with the emergency repair worksheet.
Power failure information notification timeliness: in the repair order, when the fault description or repair contents have the relevant contents of power failure information, the relevant contents particularly comprise 'no power for a plurality of users for customer repair', 'no power failure information', 'no on-the-way repair work order'; and constructing a power failure information notification timeliness derivative index by cleaning the fault description and repair content field.
And (3) informing of emergency repair speed: the method mainly reflects the speed of notifying emergency repair by constructing two derived indexes of notifying emergency repair time difference and notifying emergency repair time range, wherein the notifying emergency repair time difference is the difference between the notifying emergency repair time and the order receiving registration time; the notification first-aid repair time range is obtained according to the difference value statistics, and mainly comprises a statistical error, a notification first-aid repair within 0.5 hour, a notification first-aid repair within 0.5-1 hour, a notification first-aid repair within 1-2 hours, a notification first-aid repair within 2-3 hours, a notification first-aid repair within 3-24 hours, a notification first-aid repair within 24-48 hours, a notification first-aid repair within 48-72 hours, and no notification first-aid repair within 72 hours, wherein the statistical error represents that the notification first-aid repair time is earlier than the receipt registration time, possibly belongs to the category of system record errors and abnormal values.
Arrival speed: the method mainly comprises the steps of reacting through two derived indexes of time difference of arriving at an emergency repair site and time range of arriving at the emergency repair site, wherein the time difference of arriving at the emergency repair site is a difference value of recorded arrival time and notified emergency repair time; the time range of arriving at the emergency repair site is obtained according to the difference statistics, and mainly comprises statistical errors, arriving at the site within 0.5 hour, arriving at the site within 0.5-1 hour, arriving at the site within 1-2 hours, arriving at the site within 2-3 hours, arriving at the site within 3-24 hours, arriving at the site within 24-48 hours, arriving at the site within 48-72 hours and not arriving at the site within 72 hours, wherein the statistical errors represent that the recorded arrival time is earlier than the notification emergency repair time, and the statistical errors possibly belong to the category of system recording errors and attributing to abnormal values.
Rush repair speed: the method mainly comprises the steps of reacting through two derived indexes of a repairing process time difference and a repairing process time range, wherein the repairing process time difference is a difference value between recorded repairing time and recorded arrival time; the time spent in the repair process is obtained according to the difference statistics, and mainly comprises statistical errors, the repair success within 0.5 hour, the repair success within 0.5-1 hour, the repair success within 1-2 hours, the repair success within 2-3 hours, the repair success within 3-24 hours, the repair success within 24-48 hours, the repair success within 48-72 hours and the unrepairable success within 72 hours, wherein the statistical errors represent that the recorded repair time is earlier than the recorded arrival time, possibly belong to the category of system recorded errors and the category of abnormal values.
Committed arrival time completion: through the exploration of whether the field is overtime or not, the field has a missing value of nearly 20 percent and can be repaired by whether the promised time range is exceeded or not; whether the time range of the promised arrival site is exceeded or not refers to the difference between the promised arrival time and the time difference of the promised arrival time to the emergency repair site, and the derivative index is used for reflecting whether the emergency repair site arrives according to the promised arrival time or not, and can also reflect the customer opinion to a certain extent.
Extracting keywords of the repair content and performing grouping statistics: extracting keywords of the text information of the repair contents in the repair work order, wherein the keywords comprise a complaint No. 5 key, a default phase, no electricity of one household, no electricity of multiple households, abnormal relays and the like, and grouping and counting the repair contents according to the keyword categories, wherein the keywords comprise the default phase, the quality of electric energy or electric meters, circuit droop, relay state problems, line segments, meter burning, poor contact, tree collision lines, equipment faults, line fault, fire or smoke and other repair contents.
The data dimension of the constructed satisfaction analysis index system comprises the following steps: the method comprises the following steps of name of a local city, whether a power grid fault exists, whether a fault point is confirmed, whether an important customer is included, emergency repair order state, fault information, fault equipment property right, processing result, whether tripping occurs, notification of emergency repair time, time for reaching the emergency repair site, repair process time, promised arrival time, notification of emergency repair time range, time range for reaching the emergency repair site, repair process time range, customer opinion, whether overtime exists, power failure information notification timeliness, work order number and type, keyword extraction and grouping of repair content and personnel attitude.
And the model building module 300 is used for building multi-dimensional wide-table data of the emergency repair work order according to the satisfaction degree analysis index system, and building a random forest model to predict the part which is not evaluated in the customer opinion.
In this embodiment, the random forest model constructed by the training set by the model construction module is an integrated algorithm (Ensemble Learning), and belongs to Bagging type, as shown in fig. 2, by combining a plurality of weak classifiers, the final result is voted or averaged, so that the result of the whole model has higher accuracy and generalization performance, and can obtain good results, mainly due to "random" and "forest", one of which makes it have anti-overfitting capability, and the other makes it more accurate. In the random forest model, a large number of decision trees form a random forest; the random forest constructing process includes the following steps:
step 1: if there are N samples, there is N (N ═ N) samples randomly selected (one sample is randomly selected each time, then returning to continue selection); training a decision tree by using the selected n samples to serve as a weak classifier;
step 2: assuming that a sample has M attributes, when each node of a decision tree needs to be split, randomly selecting M attributes from the M attributes to meet a condition M < < M; then, selecting 1 attribute from the m attributes as the split attribute of the node by adopting a certain strategy (such as information entropy, information gain ratio, a Keyny coefficient and the like);
and step 3: each node in the decision tree formation process is split according to step 2 until no more splits can be made. If the next attribute selected by the node is the attribute that was used just when its parent node split, then the node has reached the leaf node and does not have to continue splitting.
And (4) establishing a large number of decision trees according to the steps 1-3, so that a random forest is formed. And for the new sample, judging the category of the sample by adopting a voting method through a plurality of decision trees.
As can be seen from the above steps, the randomness of the random forest is that the training samples of each tree are random, and the classification attributes of each node in the tree are also randomly selected. With the 2 random guarantees, the random forest will not generate overfitting.
Based on the aforementioned index system and corresponding multidimensional wide table data, because the number of samples which are very satisfactory and unsatisfactory in customer opinions is very small, if multi-classification is carried out, the number of samples is extremely unbalanced, and the model fails, 291436 samples with customer opinions not being empty are selected, samples with customer opinions being unsatisfactory and generally satisfactory are taken as targets when the model is constructed, 23796 samples are total, satisfactory and very satisfactory are classified as non-target objects, and 267640 samples are total. Samples with known customer opinion are divided into a training set and a testing set in a ratio of 7:3, wherein the training set comprises 204005 samples and 87431 samples. And constructing a random forest regression model containing 5000 decision trees by using the training set. To predict the probability of dissatisfaction of the customer opinion in the test set. According to the distribution conditions of very satisfactory, generally satisfactory and unsatisfactory in the samples, the dissatisfaction value of the prediction sample is distinguished by selecting the prediction probability threshold value of 0.275, and the analysis result is shown in table 1:
TABLE 1 analytical results
The hit rate and coverage calculation formulas are explained by combining the confusion matrix and the analysis results of the two types of problems, and the hit rate and the coverage are calculated for the LR model and the random forest respectively, so that the model conclusion and the comparison condition are shown in Table 2:
TABLE 2 model conclusions and comparisons
As can be seen from table 2, the hit rate of the random forest model is 61.82% higher than that of the LR model, and the coverage rate is 57.16% higher than that of the LR model.
The analysis decision module 400 is used for comprehensively sensing the customer appeal according to the model prediction result, deeply analyzing factors influencing the satisfaction degree of the on-site first-aid repair customer in a multi-dimensional manner, and providing data support for the decision of the related work of the subsequent business department.
In the embodiment, the analysis decision module predicts possible customer opinions in the on-site emergency repair work order with empty customer opinions according to the random forest model, calculates a satisfaction degree IV value, and performs multi-dimensional statistical analysis according to a prediction result, wherein the multi-dimensional statistical analysis comprises a city dimension, a notification emergency repair speed distribution category dimension, an arrival speed distribution category dimension, a commitment arrival time category dimension, a repair speed distribution category dimension, an emergency repair order state distribution dimension, a personnel attitude distribution dimension and a power failure notification timeliness dimension, a perfect customer service evaluation system is constructed, the specific dimension and reason where the customer is unsatisfied are efficiently and accurately analyzed, the power supply service quality is purposefully improved, the real appeal of the customer is effectively sensed, the emergency repair service system is perfected, corresponding suggestions are provided according to actual work, and the overall satisfaction degree of the customer is improved.
The multidimensional statistical analysis of the analysis decision module is used for carrying out information statistics on the modeling process of the index data from different angles to obtain a corresponding statistical report. The statistical report comprises a data quality report, a data preprocessing result report, an index system distribution characteristic report, a model result report and an analysis strategy report. The statistical analysis process specifically performs information statistics on the modeling process of the index data from different angles according to various daily report information of the customer satisfaction analysis system based on the on-site emergency repair work order to obtain corresponding statistical reports, so that the statistical requirements of different business departments can be met, and a basis is provided for subsequent information analysis work.
Example 2
The embodiment is a customer satisfaction analysis method based on a field emergency repair work order, and the method is based on the same inventive concept as the customer satisfaction analysis system based on the field emergency repair work order in the embodiment 1. Referring to fig. 3, the method includes the following steps:
s301: and carrying out data preprocessing such as data cleaning, data specification, abnormal value processing, missing value processing and the like on the on-site emergency repair work order, and specifically comprising client opinion restoration and generalization processing, personnel attitude processing, power failure notification timeliness processing, notification speed processing, presence speed processing, restoration speed processing, missing value restoration processing of each field and generalization processing of repair reporting content.
S302: according to the data preprocessing result, the satisfaction evaluation of the power customer is associated with a specific business process, derivative indexes for comprehensively depicting the on-site first-aid repair work quality and efficiency are constructed according to the actual business situation, and a satisfaction analysis index system is formed; the constructed index system is used for comprehensively measuring the on-site emergency repair work and is applied to the subsequent model construction.
S303: and constructing multi-dimensional wide-table data of the emergency repair work order according to a satisfaction degree analysis index system, and constructing a random forest model to predict the part which is not evaluated in the customer opinion.
S304: and comprehensively sensing the client appeal according to the prediction result, deeply analyzing factors influencing the client satisfaction in a multi-dimensional manner, and providing data support for the decision of related work of a business department.
Step S304 also carries out information statistics on the modeling process of the index data from different angles to obtain corresponding statistical reports, wherein the statistical reports comprise a data quality report, a data preprocessing result report, an index system distribution characteristic report, a model result report and an analysis strategy report.
For the description of the customer satisfaction analysis method based on the on-site emergency repair work order in this embodiment, since the implementation of the technical solution corresponds to the customer satisfaction analysis system based on the on-site emergency repair work order in embodiment 1, this embodiment is described more briefly, and for the parts corresponding to the technical features, reference may be made to the description of the customer satisfaction analysis system part based on the on-site emergency repair work order in embodiment 1, which is not described herein again.
In summary, the invention has the following advantages: unified data, namely, unified cleaning, stipulation, exception handling and missing processing are carried out on-site emergency repair work order data, and particularly, the unified data are used for repairing and generalizing customer opinions, processing personnel attitude, processing power failure notification timeliness, processing notification speed, processing on arrival speed, processing on repair content and processing on missing values of all fields; the unified index system is used for associating the evaluation of the satisfaction degree of the power customer with a specific business process, comprehensively depicting the on-site emergency repair work quality and efficiency according to the actual business structure and comprehensively measuring the on-site emergency repair work; the unified model is that after a multi-dimensional wide list of the on-site emergency repair work orders is built according to an index system, the part of the customer opinion which is not evaluated is predicted, the customer appeal and the dissatisfaction factors are comprehensively sensed, the unit of each city is not influenced by the work orders of the part which is not evaluated in the satisfaction degree evaluation, and the model is unified while the calculation standard is unified; and (4) unifying the strategies, namely carrying out decision support on the related work of the subsequent business departments according to the model prediction result, so that the strategies corresponding to the same type of work orders can be classified and processed, the work pressure is reduced, and the related work of the subsequent business departments is improved.
For convenience of description, the above devices are described as being divided into various modules or units by function, respectively. Of course, the functions of the modules and units may be implemented in the same software and/or hardware or more when the present application is implemented.
From the above description of the embodiments, it is clear to those skilled in the art that the present application can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present application may be essentially implemented or portions thereof contributing to the prior art may be embodied in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, an optical disk, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some portions of the embodiments of the present application.
The foregoing represents a preferred embodiment of the present invention, and it will be appreciated by those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and should be considered within the scope of the invention.
Claims (10)
1. Electric power customer satisfaction analytic system based on-the-spot salvagees work order, its characterized in that includes:
the data preprocessing module is used for preprocessing the emergency repair work order, and comprises data cleaning, data stipulation, data abnormal value repair and data missing value filling;
the index system construction module is used for associating the satisfaction degree evaluation of the power customer with a specific business process according to the preprocessing result of the data preprocessing module, constructing derivative indexes comprehensively describing the on-site emergency repair work quality and efficiency according to the actual business situation and forming a satisfaction degree analysis index system;
the model building module is used for building multi-dimensional width table data of the emergency repair work order according to the satisfaction degree analysis index system and building a random forest model to predict the part which is not evaluated in the customer opinion;
and the analysis decision module is used for comprehensively sensing the client appeal according to the model prediction result, analyzing factors influencing the satisfaction degree of the on-site first-aid repair client in a multi-dimensional manner and providing data support for decision.
2. The power customer satisfaction analysis system of claim 1, wherein the preprocessing of the data preprocessing module comprises customer opinion repair processing, personnel attitude processing, outage information notification timeliness processing, notification speed processing, presence speed processing, repair content processing, and missing value repair processing of each field.
3. The power customer satisfaction analysis system of claim 1, wherein the data outlier repair comprises a repair of customer opinion, the repair process comprising: using the first-aid repair order to correlate the worksheet table, and then using the customer opinions in the worksheet table to repair and supplement the customer opinions in the first-aid repair order; normalizing the customer opinions according to the keywords of the customer opinions in the first-aid repair list; according to the business logic mined in the data, when the customer opinions are null and the state of the first-aid repair order is not filed, but the fault content contains the work order with filed characters, the customer opinions are considered to be satisfied, and the customer opinions in the relevant work order are repaired.
4. The power customer satisfaction analysis system of claim 1, wherein the data missing value filling is to fill fields with empty values in the data exploration process, and the empty values of the fields are filled by using the values with the maximum contribution degree according to the contribution degree of different values to the satisfaction degree in each index; repairing the null value of the committed arrival time by using the average value; the fields include the name of the local city to which the fault belongs, whether the fault point is confirmed, whether the important customer is contained, fault information, the property right of the fault equipment, a processing result and whether the fault is tripped.
5. The power customer satisfaction analysis system of claim 1, wherein the satisfaction analysis index system is constructed with data dimensions comprising: the method comprises the following steps of name of a local city, whether a power grid fault exists, whether a fault point is confirmed, whether an important customer is included, emergency repair order state, fault information, fault equipment property right, processing result, whether tripping occurs, notification of emergency repair time, time for reaching the emergency repair site, repair process time, promised arrival time, notification of emergency repair time range, time range for reaching the emergency repair site, repair process time range, customer opinion, whether overtime exists, power failure information notification timeliness, work order number and type, keyword extraction and grouping of repair content and personnel attitude.
6. The power customer satisfaction analysis system of claim 1, wherein the model building module divides a sample with known customer opinions into a training set and a testing set in proportion, constructs a random forest regression model with a plurality of decision trees by using the training set, and predicts the unevaluated work order by using the random forest regression model to perfect a satisfaction analysis index system.
7. The power customer satisfaction analysis system of claim 1, characterized in that the analysis decision module predicts possible customer opinions in the empty site emergency repair work orders according to the random forest model, calculates a satisfaction IV value, performs multi-dimensional statistical analysis according to the prediction result, and comprises a city dimension, a notification emergency repair speed distribution category dimension, an arrival speed distribution category dimension, a commitment arrival time category dimension, a repair speed distribution category dimension, an emergency repair order state distribution dimension, a personnel attitude distribution dimension, and a power failure notification timeliness dimension, constructs a perfect customer service evaluation system, and analyzes specific dimensions and reasons where the customer is unsatisfied.
8. The method for analyzing the satisfaction degree of the power customer based on the on-site emergency repair work order is characterized by comprising the following steps of:
s301, preprocessing the emergency repair work order, including data cleaning, data specification, data abnormal value repair and data missing value filling;
s302, according to the preprocessing result, the satisfaction evaluation of the power customer is related to a specific business process, a derivative index which comprehensively describes the quality and efficiency of on-site first-aid repair work is constructed according to the actual business situation, and a satisfaction analysis index system is formed;
s303, building multi-dimensional width table data of the emergency repair work order according to a satisfaction degree analysis index system, and building a random forest model to predict an unevaluated part in customer opinions;
s304, comprehensively sensing the customer appeal according to the model prediction result, carrying out multi-dimensional analysis on factors influencing the satisfaction degree of the on-site first-aid repair customer, and providing data support for decision making.
9. The method for analyzing the satisfaction degree of the power customer as recited in claim 8, wherein in the step S304, information statistics is further performed on the modeling process of the index data from different angles to obtain corresponding statistical reports, and the statistical reports include a data quality report, a data preprocessing result report, an index system distribution characteristic report, a model result report, and an analysis strategy report.
10. The power customer satisfaction analysis method according to claim 8, wherein the process of constructing the random forest model in step S303 comprises the steps of:
step 1: if N samples exist, N samples are selected randomly and put back, a decision tree is trained by using the selected N samples to serve as a weak classifier, and N < ═ N;
step 2: assuming that a sample has M attributes, when each node of a decision tree needs to be split, randomly selecting M attributes from the M attributes, wherein M < < M; then 1 attribute is selected from the m attributes as the splitting attribute of the node;
and step 3: each node in the decision tree formation process is split according to step 2 until no more splits can be made.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011013654.XA CN112052966A (en) | 2020-09-24 | 2020-09-24 | Power customer satisfaction analysis system and method based on site emergency repair work order |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011013654.XA CN112052966A (en) | 2020-09-24 | 2020-09-24 | Power customer satisfaction analysis system and method based on site emergency repair work order |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112052966A true CN112052966A (en) | 2020-12-08 |
Family
ID=73604436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011013654.XA Pending CN112052966A (en) | 2020-09-24 | 2020-09-24 | Power customer satisfaction analysis system and method based on site emergency repair work order |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112052966A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113449927A (en) * | 2021-07-13 | 2021-09-28 | 广东电网有限责任公司 | Evaluation method, equipment and storage medium based on natural language fault first-aid repair |
| CN113723663A (en) * | 2021-07-12 | 2021-11-30 | 国网冀北电力有限公司计量中心 | Power work order data processing method and device, electronic equipment and storage medium |
| CN113807701A (en) * | 2021-09-18 | 2021-12-17 | 国网福建省电力有限公司 | Power supply service quality analysis method based on information entropy decision tree algorithm |
| CN113837473A (en) * | 2021-09-27 | 2021-12-24 | 佰聆数据股份有限公司 | Charging equipment fault rate analysis system and method based on BP neural network |
| CN114840583A (en) * | 2022-06-24 | 2022-08-02 | 国网浙江省电力有限公司杭州供电公司 | Panoramic index data analysis processing method and system based on block data construction |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105933920A (en) * | 2016-03-31 | 2016-09-07 | 浪潮通信信息系统有限公司 | Method and device for predicting user satisfaction |
| CN108280652A (en) * | 2016-12-31 | 2018-07-13 | 中国移动通信集团辽宁有限公司 | The analysis method and device of user satisfaction |
| CN109885790A (en) * | 2018-12-30 | 2019-06-14 | 贝壳技术有限公司 | The method and apparatus for obtaining satisfaction evaluation data |
| CN110928924A (en) * | 2019-11-28 | 2020-03-27 | 江苏电力信息技术有限公司 | Power system customer satisfaction analyzing and predicting method based on neural network |
| CN111176963A (en) * | 2019-12-13 | 2020-05-19 | 腾讯云计算（北京）有限责任公司 | Service evaluation information processing method and device |
-
2020
- 2020-09-24 CN CN202011013654.XA patent/CN112052966A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105933920A (en) * | 2016-03-31 | 2016-09-07 | 浪潮通信信息系统有限公司 | Method and device for predicting user satisfaction |
| CN108280652A (en) * | 2016-12-31 | 2018-07-13 | 中国移动通信集团辽宁有限公司 | The analysis method and device of user satisfaction |
| CN109885790A (en) * | 2018-12-30 | 2019-06-14 | 贝壳技术有限公司 | The method and apparatus for obtaining satisfaction evaluation data |
| CN110928924A (en) * | 2019-11-28 | 2020-03-27 | 江苏电力信息技术有限公司 | Power system customer satisfaction analyzing and predicting method based on neural network |
| CN111176963A (en) * | 2019-12-13 | 2020-05-19 | 腾讯云计算（北京）有限责任公司 | Service evaluation information processing method and device |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113723663A (en) * | 2021-07-12 | 2021-11-30 | 国网冀北电力有限公司计量中心 | Power work order data processing method and device, electronic equipment and storage medium |
| CN113449927A (en) * | 2021-07-13 | 2021-09-28 | 广东电网有限责任公司 | Evaluation method, equipment and storage medium based on natural language fault first-aid repair |
| CN113449927B (en) * | 2021-07-13 | 2022-09-30 | 广东电网有限责任公司 | Evaluation method, equipment and storage medium based on natural language fault first-aid repair |
| CN113807701A (en) * | 2021-09-18 | 2021-12-17 | 国网福建省电力有限公司 | Power supply service quality analysis method based on information entropy decision tree algorithm |
| CN113837473A (en) * | 2021-09-27 | 2021-12-24 | 佰聆数据股份有限公司 | Charging equipment fault rate analysis system and method based on BP neural network |
| CN114840583A (en) * | 2022-06-24 | 2022-08-02 | 国网浙江省电力有限公司杭州供电公司 | Panoramic index data analysis processing method and system based on block data construction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112052966A (en) | Power customer satisfaction analysis system and method based on site emergency repair work order | |
| García et al. | A retrospective analysis of the impact of the COVID-19 restrictions on energy consumption at a disaggregated level | |
| CN106815709B (en) | Service quick response center support system and method | |
| CN112131441B (en) | Method and system for rapidly identifying abnormal electricity consumption behavior | |
| CN106570778A (en) | Big data-based data integration and line loss analysis and calculation method | |
| CN107862467A (en) | A kind of electric network synthetic data target monitoring method and system based on big data platform | |
| CN110298601A (en) | A kind of real time business air control system of rule-based engine | |
| CN111507013A (en) | Line loss fault positioning implementation method for power system | |
| AU2011204831B2 (en) | Digital analytics platform | |
| CN112419030B (en) | Method, system and equipment for evaluating financial fraud risk | |
| Zhou et al. | Delineating infrastructure failure interdependencies and associated stakeholders through news mining: The case of Hong Kong’s water pipe bursts | |
| CN113449964A (en) | Enterprise financial risk monitoring and early warning system and monitoring and early warning method | |
| KR100713546B1 (en) | Method of technology evaluation | |
| US11657063B2 (en) | Behavioral analytics in information technology infrasturcture incident management systems | |
| CN115860694A (en) | Business expansion process management and control method and system based on instant message technology | |
| CN111767193A (en) | Server data anomaly detection method and device, storage medium and equipment | |
| US20100042446A1 (en) | Systems and methods for providing core property review | |
| Yu et al. | Research on situational perception of power grid business based on user portrait | |
| CN116109364A (en) | Price evaluation method and device | |
| CN115983705A (en) | Evaluation model construction method, computer device and computer-readable storage medium | |
| CN115495587A (en) | Alarm analysis method and device based on knowledge graph | |
| CN114529383A (en) | Method and system for realizing tax payment tracking and tax loss early warning | |
| Yuan et al. | Critical risks identification of Public Private Partnerships in China and the analysis on questionnaire survey | |
| CN113537519A (en) | Method and device for identifying abnormal equipment | |
| TWI634498B (en) | Business operations analysis system and method |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201208 |