CN111353792A - Client portrait system with visual display and data analysis functions - Google Patents
Client portrait system with visual display and data analysis functions Download PDFInfo
- Publication number
- CN111353792A CN111353792A CN202010446110.6A CN202010446110A CN111353792A CN 111353792 A CN111353792 A CN 111353792A CN 202010446110 A CN202010446110 A CN 202010446110A CN 111353792 A CN111353792 A CN 111353792A
- Authority
- CN
- China
- Prior art keywords
- complaint
- customer
- data
- power
- user
- 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
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
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/26—Visual data mining; Browsing structured data
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention aims to provide a customer portrait system with visual display and data analysis, which comprises a user analysis module and a complaint analysis module. The customer analysis module generates a visual result and sends the visual result to the customer manager, the customer manager visits the customer according to the visual result, then the complaint analysis module constructs a customer complaint tendency prediction model, the customer complaint tendency prediction model is applied to a user group without complaint behaviors, target marks of the user group with complaint tendencies are identified, potential complaint risk probability is calculated and is used as precaution reference by the customer manager, further more-fit differentiated and personalized customer service is provided for large customers, the electricity utilization cost benefit management and control of the customers and the customer relationship management level are improved, and the complaint rate of the customers is effectively reduced.
Description
Technical Field
The invention relates to the field of power consumer service systems, in particular to a customer portrait system with visual display and data analysis.
Background
Because the power grid belongs to a huge power supply system, is also a head-end for guaranteeing the life and production power consumption of people, and power users are numerous, each user is required to be better served, various problems of a user end are found in time, which is a very difficult matter, the complaint quantity of the users is increased, power production personnel face power users with huge number and inconsistent power consumption properties, the problems are found at the first time, the lean management level of the customer service of each power supply bureau under the power grid is improved, the high-quality service is provided in a targeted manner, the construction of a power supply service system is enhanced, the power consumption experience of the customers is improved, the satisfaction degree of third-party customers in 2020 is not less than 87 minutes, the first target is kept in the evaluation of the satisfaction degree of public service of government, the management and control of the customers are continuously enhanced, the repeated complaints of the customers 95598 and the complaint quantity of the global 12398 are reduced, the method aims at avoiding 12398 complaints of responsibility all the year around, ensures that important customer complaints do not occur, and needs to judge the condition of complaint risk in advance through data analysis based on huge electricity utilization information, and timely visit customers according to the grasped condition or solve problems for the customers.
Based on the existing problems and the requirement of the power grid province company, a customer portrait system with visual display and data analysis is provided, the overall power utilization general profile and trend view of a large customer are constructed by providing visual display results, sensitive complaint customers are analyzed, users with complaint intentions are predicted, advance prevention and control are realized, and complaint events are effectively reduced.
Disclosure of Invention
The invention provides a customer portrait system with visual display and data analysis, which aims to solve the problems of various defects in the prior art, and mainly utilizes continuously developed and mature analysis dimensions of big data and data analysis and mining technology based on historical data of interaction between customers and a power grid company, such as basic information of big customers, power consumption data, channel interaction data, power failure data and the like of a marketing system, aiming at the composition and variation trend of power consumption and electric quantity concerned by the big customers, the power consumption structure of peak-valley, the power consumption structure distribution of each sub-customer number and metering point, the power consumption condition of force regulation, the power consumption condition of back-supplementing and the like, and develops the power consumption analysis from the level differences of the whole big customers, the single customer number, the metering point and the like, finds the abnormal condition of a user end in time, further makes risk control before customer complaints, reduces customer complaints, and improves the customer service satisfaction.
The method specifically comprises the following steps:
a customer representation system for visual presentation and data analysis, comprising:
and the user analysis module is used for extracting the basic information of the user from the big data platform through the ETL, performing data statistical analysis, generating a visual result and sending the visual result to the client manager, and the client manager visits the user according to the visual result.
Complaint analysis module: the method comprises the steps of obtaining a customer service work order with complaint behaviors, carrying out data sampling at least once on the customer service work order, carrying out data correction, extracting feature keywords by adopting an artificial checking and text mining method, constructing a customer complaint tendency prediction model, applying the customer complaint tendency prediction model to a user group without complaint behaviors, identifying target identifications of the user group with complaint tendencies, and calculating complaint risk probability for customer managers to make pre-risk prevention and control reference.
The basic information of the user is as follows: including power consumption, electricity charge, installation condition and appeal condition.
And (3) performing statistical analysis on the data: the method comprises the step of carrying out data statistical analysis by adopting a statistical method, wherein the data statistical analysis comprises basic attributes, importance levels, power utilization conditions, submission conversion and capacity analysis and industry analysis of customers.
The visualization result is as follows: generating a visualization result based on the data statistical analysis result, including: the method comprises the following steps of power utilization change trend graph, power utilization time period characteristics, submission conversion rate, submission capacity and electronic industry capacity ranking.
The customer complaint tendency prediction model sets a target identifier Y according to the extracted keywords, wherein Y = (Y1, Y2, … yi), and any yi corresponds to a sample set with n feature keywords, is X = (X1, X2, … xn, yi), and serves as a sample X training classifier h (X): and (4) X.
The method for identifying the target identification of the user group with complaint tendency further comprises the following steps: obtaining a customer service worksheet of a user group without complaint behaviors, extracting characteristic keywords by adopting an artificial checking and text mining method to obtain a sampleTraining classifier h (X) by sample X: x sample T complaint propensity prediction:
outputting the target identificationTo obtainWherein, the H (t) is at least one of the target identifiers Y or a new target identifier.
And calculating the weight of each target identification to obtain the complaint risk probability and sending the complaint risk probability to a customer manager.
The keywords of the fault power failure complaint comprise: the number of power failures in day \ week \ month, power failure interval of the last time, power failure duration, power failure time period, emergency repair progress, and power failure and power restoration information caused by external factors, but not limited thereto.
The keywords of the electricity fee error complaint include: the electricity consumption fact is wrong, the electricity charge is abnormal, the electricity meter reading information is wrong, the charging period is disordered to cause the electricity charge to be abnormal, the late payment electricity charge default money is generated due to untimely meter reading, the fee is not returned in time, and the stepped electricity price is questionable, but the method is not limited to the above.
The keywords of the service efficiency complaint include: failure emergency repair is not timely, business expansion overtime, arrearage and power restoration are realized, but the method is not limited to the above.
Before the data correction, the method also comprises the following steps: exploring basic information, power consumption information, electricity fee information, channel interaction information and power failure information of a client, wherein the exploration comprises determining the category of data variables, the correlation among the variables, the missing abnormal condition of the related variables and the distribution inspection of data; and based on the data exploration result, performing data correction, including data cleaning and preprocessing, creating derivative variables and description statistics, screening variables, partitioning data, and merging, supplementing, converting or filtering missing data.
The invention provides a customer portrait system for visual display and data analysis, which mainly comprises a user analysis module and a complaint analysis module, wherein a visual result is obtained for a customer manager to refer to the visit of a user by counting and analyzing mass historical data of a big data platform of a power user. Furthermore, the invention also constructs a customer complaint tendency prediction model for improving the service quality of users and controlling risks before complaints, so as to timely process the user problems of potential complaint risks, effectively reduce the occurrence frequency of complaint events, master the power utilization state, the power utilization condition and the power service level condition of power users in real time for power grid enterprises, and provide guarantee and reference for further providing high-quality power service.
Detailed Description
A customer representation system for visualization and data analysis in accordance with the present invention is described in further detail below with reference to specific embodiments.
The invention provides a client portrait system for visual display and data analysis, which specifically comprises:
the user analysis module is used for extracting basic information of a user from the big data platform through ETL (extract transform load) and performing data statistical analysis by adopting a statistical method, wherein the basic information comprises user types, power consumption, power charge, application conditions and appeal conditions, the statistical analysis comprises the steps of performing multi-dimensional statistical analysis on basic attribute analysis, importance levels, power consumption conditions, application conversion and capacity analysis, industry analysis and the like of the user, finally generating a visual result and sending the visual result to a client manager, and the client manager visits the user according to the visual result, wherein the visual result generated after the statistical analysis comprises the following steps: the presentation form can be selected from a plurality of chart representation forms such as a line chart, a bar chart or a comparison chart of different electronic industries.
The customer manager can visually and clearly know the electricity utilization condition and the demand of the customer through the customer chart information, and then the customer visits according to the analysis result, so that effective electricity utilization suggestions are provided for the customer, and the visiting efficiency and the customer satisfaction are effectively improved.
Complaint analysis module: the method comprises the steps of obtaining a customer service work order with complaint behaviors, and carrying out at least one-time data sampling on the customer service work order, wherein the data sampling method preferably comprises simple random sampling, system sampling, hierarchical sampling, oversampling and the like. And performing data correction on the sampled data, extracting characteristic keywords by adopting an artificial checking and text mining method, constructing a customer complaint tendency prediction model, applying the customer complaint tendency prediction model to a user group without complaint behaviors, identifying a target identifier of the user group with complaint tendency, and calculating potential complaint risk probability for a customer manager to make pre-risk prevention and control reference.
Wherein, still include before the data correction: exploring basic information, power consumption information, electricity fee information, channel interaction information and power failure information of a client, wherein the exploration comprises determining the category of data variables, the correlation among the variables, the missing abnormal condition of the related variables and the distribution inspection of data; and based on the data exploration result, performing data correction, including data cleaning and preprocessing, creating derivative variables and description statistics, screening variables, partitioning data, and merging, supplementing, converting or filtering missing data.
Preferably, the customer complaint tendency prediction model sets a target identifier Y according to the extracted keywords, where Y = (Y1, Y2, … yi), and any yi corresponds to a sample set with n feature keywords, X = (X1, X2, … xn, yi), and trains a classifier h (X) as sample X: and (4) X.
The method for identifying the target identification of the user group with complaint tendency further comprises the following steps: obtaining a customer service worksheet of a user group without complaint behaviors, extracting characteristic keywords by adopting an artificial checking and text mining method to obtain a sampleTraining classifier h (X) by sample X: x sample T complaint propensity prediction:outputting the target identificationTo obtainAnd H (t) is at least one of the target identifications Y or a new target identification, and further calculating the weight of each target identification to obtain the complaint risk probability and sending the complaint risk probability to a customer manager.
The target identification comprises: fault power outage complaints, electricity charge error complaints, or service efficiency complaints.
Preferably, the target identification includes: fault power outage complaints, electricity charge error complaints, or service efficiency complaints.
Wherein, the keyword of trouble power failure complaint includes: the power failure and power restoration information is caused by the daily/week/month power failure times, the power failure interval, the last power failure interval, the power failure duration, the power failure time period, the first-aid repair progress and external factors.
The keywords of the electricity fee error complaint include: the electricity consumption fact is wrong, the electricity charge is abnormal, the electricity meter reading information is wrong, the charging period is disordered, the electricity charge is abnormal, the late payment default money is generated due to untimely meter reading, the fee is not returned in time, and the stepped electricity price is doubtful.
The keywords of the service efficiency complaint include: failure emergency repair is not timely, business expansion is overtime, and power is recovered due to arrearage.
As another preferred embodiment, the summary of the customer service order data extracted in the present invention is shown in Table 1 below,
table name | Description of the invention | Time horizon | Nearly 1 month | Nearly 3 months | In the last 1 year | All amount of |
fw_kfgdxx | Customer service work order information local market account | 201507-201808 | 3099 | 9659 | 44294 | 851536 |
fw_kfgdxx | Account number of customer service work order information province company | 201507-201808 | 36174 | 107327 | 402324 | 1373590 |
hz_khhx_tsmx_gdxx | Work order and user associated left connection (city) | 201507-201808 | 3099 | 9659 | 44357 | 941414 |
hz_khhx_tsmx_gdxx | Work order and user associated left connection (province company) | 201507-201808 | 38999 | 115524 | 452662 | 1561923 |
fw_qxywfjxx | Emergency repair service additional information (city) | 201507-201808 | 0 | 5 | 202 | 35118 |
And associating the extracted customer service work order with the user, wherein after the association, the condition that the key field is empty is as follows:
name of field | Empty (city) | To the ratio (city) | Hollow (province company) | To one percent (province company) |
Filing time | 1230 | 0.14 | 1944 | 0.14 |
Business subclass identification | 1153 | 0.14 | 1172 | 0.09 |
User number | 305606 | 35.89 | 503632 | 36.67 |
Line segment | 308650 | 36.25 | 509553 | 37.10 |
Platform area identification | 321422 | 37.75 | 526947 | 38.36 |
The electricity charge error module data profile is obtained from statistics as follows:
table name | Description of the invention | Time horizon | Nearly 1 month | Nearly 3 months | In the last 1 year | All amount of |
zw_ysdfjl | Record of electricity charge | 201807-201808 | 1287880 | 2725481 | 2725481 | 2725481 |
zw_ysdfjl_his | History table for recording charge | 201201-201806 | 0 | 1278727 | 14720751 | 89078602 |
hz_khhx_tsmx_ysdfjl | Electricity charge record merging table | 201201-201808 | 1287880 | 4004208 | 17446232 | 91804083 |
zw_wyjmx | Breach of security | 199202-201807 | 58387 | 181070 | 589688 | 1556381 |
lc_cbycxx | Abnormal information of meter reading | 201508-201808 | 1977 | 10204 | 278383 | 582730 |
Based on the extracted data result, correction and processing work is carried out.
Preferably, the method comprises the following steps:
1) complaint worksheet data: the data with the service class code of '04' is a complaint work order, 1430 complaint work orders inquired by the account of the provincial company, only 139 work orders inquired by the account of the city department, and the conditions of other key fields are shown in the following table:
name of field | Empty (city) | To the ratio (city) | Hollow (province company) | To one percent (province company) |
Filing time | 1 | 0.72 | 6 | 0.42 |
Business subclass identification | 1 | 0.72 | 1 | 0.07 |
User number | 29 | 20.86 | 479 | 33.50 |
Line segment | 29 | 20.86 | 487 | 34.06 |
Platform area identification | 35 | 25.18 | 499 | 34.90 |
2) Power outage event data:
full data 243446, key field cases are as follows:
numbering | Name of field | Is empty | Ratio (%) |
A | Actual time of power failure | 11046 | 4.54 |
B | Actual end time of power failure | 17369 | 7.13 |
C | Duration of power outage | 43477 | 17.86 |
D | Time to repair a fault | 182629 | 75.02 |
E | Time delay and recovery time | 241644 | 99.26 |
F | Time of arrival at a site | 183298 | 75.29 |
G | Category of power failure | 416 | 0.17 |
The power failure category is 2, namely the power failure data, the total data amount is 113079, and other conditions are as follows:
numbering | Name of field | Is empty | Ratio (%) |
A | Actual time of power failure | 762 | 0.67 |
B | Actual end time of power failure | 4986 | 4.41 |
C | Duration of power outage | 32446 | 28.69 |
D | Time to repair a fault | 52262 | 46.22 |
E | Time delay and recovery time | 112349 | 99.35 |
F | Time of arrival at a site | 52031 | 46.01 |
G | Category of power failure | 0 | 0.00 |
Wherein A, B are all blank 739 in the fault power failure, and are not blank 108070, if 5009 data cannot be calculated by calculating the power failure duration by using the two fields, the condition is better than that of directly calculating by using the power failure duration; the fault repair time and the time reaching the site are seriously lost, so that the calculation of time indexes such as the average time of the fault repair reaching the site is influenced; the delayed power restoration is in an empty-to-air ratio of 99.35%, which shows that the fault power failure delayed power restoration condition is less, and the data is in a normal condition;
preferably, before the model is established, technical knowledge and business knowledge are further combined, rules and trends of customer basic information, power consumption information, electricity fee information, channel interaction information and power failure information in all dimensions are explored, the types of data variables are determined, the correlation among the variables, the missing abnormal conditions of the related variables, data distribution inspection and the like are researched, and the data set can be ensured to meet the requirements for solving business problems.
The method comprises the following specific steps:
1) repair service additional information
A. The time of arrival at the site is 2245 pieces, which accounts for 6.39 percent;
B. the field completion time is 6437 pieces, accounting for 18.33%;
C. the power supply recovery time is 6151 empty, accounting for 17.52%;
D. the power failure time is 2998 empty, accounting for 8.54%;
and the emergency repair service data in the last 1 year, the last 3 months and the last 1 month are few, and if the data quality is not problematic, the indexes of removing and emergency repair service additional information are suggested.
2) Business expansion work order information
A. The printing time of the receipt accepting time is 2261660 pieces which account for 78.77 percent;
B. the filing time is 431949, accounting for 15.05%;
C. 2213221 pieces of power are supplied for ignition, which account for 77.09%;
A. and C, calculating a large number of related indexes of the efficiency of the business expansion for the air influence.
3) Arrearage power outage data
The total amount of power failure event (arrearage power failure) data is 78592, and the data related to the power failure client details through the power failure time identifier is only 71, so that the reference value is not large, and the calculation of the related index of the arrearage power recovery efficiency is influenced. In addition, the arrearage power failure data can be acquired from the power failure event basic condition table, but only 1 record is acquired.
Preferably, the target identifiers of the customer complaint tendency prediction model are divided into failure power failure complaints, power rate error complaints and service efficiency complaints based on the data and the current situation of the complaint service, and three user groups of failure power failure complaints, power rate error complaints and service efficiency complaints are identified through the model.
The target work orders aiming at the three target identifications are as shown in the table:
model (model) | Target work order |
Fault power failure complaint prediction model | Service type is trouble power failure complaint work order |
Electricity charge error complaint prediction model | The service type is the complaint work order of the electric charge error |
Service efficiency complaint prediction model | The service type is the complaint work order of service efficiency (failure emergency repair is not timely, arrearage power recovery is overtime, business expansion transaction is overtime) |
The complaint tendency theme takes a customer service work order as a core, qualitative description keywords of customer complaints can be preliminarily extracted through the content of the customer service work order, the content of the customer service work order is stored in a text form, and the keywords are obtained from a large number of work orders by utilizing an artificial checking and text mining method, and the method specifically comprises the following steps:
further, a customer complaint tendency prediction model is built, the customer complaint tendency prediction model is applied to a user group without complaint behaviors, a target identification of the user group with complaint tendency is identified, potential complaint risk probability is calculated, and a customer manager makes pre-risk prevention and control reference, and the method specifically comprises the following steps: selecting a supervised learning model, wherein the specific mathematical model is as follows:
sample(s)Known target identificationThere are n samples T with characteristics, but the samples T do not know the target markI.e. byNeed to use the samplePredicting target identification of a sample T, using the sampleTraining classifierI.e. byUsing a classifierPredicting sample T, and finally outputting target identification of sample TFurther obtaining a sampleWherein, in the step (A),。
preferably, the potential complaint risk probability is calculated by using, but not limited to, logistic regression, decision tree, random forest, neural network, and support vector machine algorithm.
In conclusion, the customer representation system for visual display and data analysis provided by the invention realizes application of big data, and obtains corresponding chart information by summarizing and analyzing the power consumer information, so that a customer manager can intuitively and clearly know the electricity utilization condition and demand of a customer, and then performs customer visiting work according to an analysis result, thereby providing effective electricity utilization suggestions for large customers, and effectively improving visiting efficiency and customer satisfaction. In order to further improve the quality of electric power service, timely respond to diversified requirements of a user side, construct a customer complaint tendency prediction model, apply the model to a user group without complaint behaviors, identify target identifications of the user group with complaint tendencies, and calculate potential complaint risk probability for customer managers to make pre-risk prevention and control reference. The customer figure system can effectively assist the power customer service center and the customer manager to accurately recognize the customer, master the hot spot of the customer demand and mine the value of the large customer, thereby providing more conformable differentiated and personalized customer service for the large customer. Through three-dimensional multi-dimensional power consumption analysis, a panoramic view of the power consumption of the major customers is provided for the major customers, the power utilization space is further excavated and optimized, the cost-benefit management of the power utilization of the major customer enterprises is improved in an auxiliary mode, and the management level of the customer relationship is improved. A customer complaint tendency prediction model is built, the complaint tendency of a customer is predicted by combining the current business behavior of the customer based on historical complaint sample information, and a relevant strategy is made in advance to reduce the complaint rate of the customer.
While the invention has been described in conjunction with the specific embodiments set forth above, it is evident that many alternatives, modifications, and variations will be apparent to those skilled in the art in light of the foregoing description. Accordingly, it is intended to embrace all such alternatives, modifications, and variations that fall within the spirit and scope of the appended claims.
Claims (10)
1. A client representation system for visual presentation and data analysis, comprising:
the system comprises a user analysis module, a client manager and a data processing module, wherein the user analysis module is used for extracting basic information of a user from a big data platform through ETL (extract transform load) and carrying out data statistical analysis, generating a visual result and sending the visual result to the client manager, and the client manager visits the user according to the visual result;
complaint analysis module: the method comprises the steps of obtaining a customer service work order with complaint behaviors, carrying out data sampling at least once on the customer service work order, carrying out data correction, extracting feature keywords by adopting an artificial checking and text mining method, constructing a customer complaint tendency prediction model, applying the customer complaint tendency prediction model to a user group without complaint behaviors, identifying target identifications of the user group with complaint tendencies, and calculating potential complaint risk probability for customer managers to make pre-risk prevention and control reference.
2. The system of claim 1,
the basic information of the user is as follows: the method comprises the following steps of (1) including user type, power consumption, electric charge, installation and appeal conditions;
and (3) performing statistical analysis on the data: performing data statistical analysis by adopting a statistical method, wherein the data statistical analysis comprises basic attributes, important grades, power utilization conditions, submission conversion and capacity analysis and industry analysis of customers;
the visualization result is as follows: generating a visualization result based on the data statistical analysis result, including: the method comprises the following steps of power utilization change trend graph, power utilization time period characteristics, submission conversion rate, submission capacity and electronic industry capacity ranking.
3. The system of claim 1, further comprising: the customer complaint tendency prediction model sets a target identification Y according to the extracted keyword, wherein Y = (Y)1,y2,…yi) Any of yiCorresponding to a sample set with n feature keywords, X = (X)1,x2,…xn, yi) Training classifier h (X) as sample X: and (4) X.
4. The system of claim 1, wherein the identifying target identities for a population of users with complaint tendencies further comprises: obtaining a customer service worksheet of a user group without complaint behaviors, extracting characteristic keywords by adopting an artificial checking and text mining method to obtain a sampleTraining classifier h (X) by sample X: x sample T complaint propensity prediction:outputting the target identificationTo obtainWherein, the H (t) is at least one of the target identifiers Y or a new target identifier.
5. The system of claim 4, wherein each target identification weight is calculated, and the probability of complaint risk is obtained and sent to a customer manager.
6. The system of claim 5, wherein the target identification comprises: fault power outage complaints, electricity charge error complaints, or service efficiency complaints.
7. The system of claim 6, wherein the keywords of the blackout complaint include: the power failure and power restoration information is caused by the daily/week/month power failure times, the power failure interval, the last power failure interval, the power failure duration, the power failure time period, the first-aid repair progress and external factors.
8. The system according to claim 6, wherein the keywords for the electricity fee error complaint include: the electricity consumption fact is wrong, the electricity charge is abnormal, the electricity meter reading information is wrong, the charging period is disordered, the electricity charge is abnormal, the late payment default money is generated due to untimely meter reading, the fee is not returned in time, and the stepped electricity price is doubtful.
9. The system of claim 6, wherein the keywords for the service efficiency complaint comprise: failure emergency repair is not timely, business expansion is overtime, and power is recovered due to arrearage.
10. The system of claim 1, further comprising, prior to the data modification: exploring basic information, power consumption information, electricity fee information, channel interaction information and power failure information of a client, wherein the exploration comprises determining the category of data variables, the correlation among the variables, the missing abnormal condition of the related variables and the distribution inspection of data; and based on the data exploration result, performing data correction, including data cleaning and preprocessing, creating derivative variables and description statistics, screening variables, partitioning data, and merging, supplementing, converting or filtering missing data.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010446110.6A CN111353792A (en) | 2020-05-25 | 2020-05-25 | Client portrait system with visual display and data analysis functions |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010446110.6A CN111353792A (en) | 2020-05-25 | 2020-05-25 | Client portrait system with visual display and data analysis functions |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111353792A true CN111353792A (en) | 2020-06-30 |
Family
ID=71197760
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010446110.6A Pending CN111353792A (en) | 2020-05-25 | 2020-05-25 | Client portrait system with visual display and data analysis functions |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111353792A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112184486A (en) * | 2020-09-30 | 2021-01-05 | 深圳供电局有限公司 | Power consumer cost management method |
CN112288444A (en) * | 2020-10-23 | 2021-01-29 | 翼果(深圳)科技有限公司 | Cross-border SAAS client analysis method and system based on big data |
CN112395351A (en) * | 2020-11-19 | 2021-02-23 | 平安普惠企业管理有限公司 | Visual identification group complaint risk method, device, computer equipment and medium |
CN112465317A (en) * | 2020-11-16 | 2021-03-09 | 广东电网有限责任公司 | Cadre portrait information analysis system based on big data |
CN112529630A (en) * | 2020-12-17 | 2021-03-19 | 国网河北省电力有限公司营销服务中心 | Construction method of multi-dimensional informatization label system of power consumer |
CN112712270A (en) * | 2020-12-31 | 2021-04-27 | 深圳前海微众银行股份有限公司 | Information processing method, device, equipment and storage medium |
CN113298385A (en) * | 2021-05-26 | 2021-08-24 | 上海晓途网络科技有限公司 | User management method and device, electronic equipment and storage medium |
CN113592297A (en) * | 2021-07-30 | 2021-11-02 | 国家电网有限公司客户服务中心 | Method and system for managing power system-specific customers |
CN113780677A (en) * | 2021-09-26 | 2021-12-10 | 深圳供电局有限公司 | Prediction method and device for potential power repeated appeal user |
CN115131806A (en) * | 2022-06-07 | 2022-09-30 | 福建极推科技有限公司 | Method and system for recognizing OCR (optical character recognition) image information of various certificates based on deep learning |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104113869A (en) * | 2014-06-20 | 2014-10-22 | 北京拓明科技有限公司 | Signaling data-based prediction method and system for potential complaint user |
CN106530127A (en) * | 2016-11-09 | 2017-03-22 | 国网江苏省电力公司南京供电公司 | Complaint early warning and monitoring analysis system based on text mining |
CN107992609A (en) * | 2017-12-15 | 2018-05-04 | 广东电网有限责任公司信息中心 | A kind of complaint tendency determination methods based on Text Classification and decision tree |
CN109165763A (en) * | 2018-06-13 | 2019-01-08 | 广西电网有限责任公司电力科学研究院 | A kind of potential complained appraisal procedure and device of 95598 customer service work order |
CN110163621A (en) * | 2018-02-10 | 2019-08-23 | 广州供电局有限公司 | A kind of electric power customer service big data DSS |
CN110309932A (en) * | 2018-03-21 | 2019-10-08 | 国网浙江省电力公司湖州供电公司 | A kind of power customer resource fine-grained management method based on TEM |
CN110503249A (en) * | 2019-08-07 | 2019-11-26 | 国网河北省电力有限公司 | One kind complaining prediction technique caused by having a power failure |
-
2020
- 2020-05-25 CN CN202010446110.6A patent/CN111353792A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104113869A (en) * | 2014-06-20 | 2014-10-22 | 北京拓明科技有限公司 | Signaling data-based prediction method and system for potential complaint user |
CN106530127A (en) * | 2016-11-09 | 2017-03-22 | 国网江苏省电力公司南京供电公司 | Complaint early warning and monitoring analysis system based on text mining |
CN107992609A (en) * | 2017-12-15 | 2018-05-04 | 广东电网有限责任公司信息中心 | A kind of complaint tendency determination methods based on Text Classification and decision tree |
CN110163621A (en) * | 2018-02-10 | 2019-08-23 | 广州供电局有限公司 | A kind of electric power customer service big data DSS |
CN110309932A (en) * | 2018-03-21 | 2019-10-08 | 国网浙江省电力公司湖州供电公司 | A kind of power customer resource fine-grained management method based on TEM |
CN109165763A (en) * | 2018-06-13 | 2019-01-08 | 广西电网有限责任公司电力科学研究院 | A kind of potential complained appraisal procedure and device of 95598 customer service work order |
CN110503249A (en) * | 2019-08-07 | 2019-11-26 | 国网河北省电力有限公司 | One kind complaining prediction technique caused by having a power failure |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112184486A (en) * | 2020-09-30 | 2021-01-05 | 深圳供电局有限公司 | Power consumer cost management method |
CN112288444A (en) * | 2020-10-23 | 2021-01-29 | 翼果(深圳)科技有限公司 | Cross-border SAAS client analysis method and system based on big data |
CN112465317A (en) * | 2020-11-16 | 2021-03-09 | 广东电网有限责任公司 | Cadre portrait information analysis system based on big data |
CN112395351A (en) * | 2020-11-19 | 2021-02-23 | 平安普惠企业管理有限公司 | Visual identification group complaint risk method, device, computer equipment and medium |
CN112529630A (en) * | 2020-12-17 | 2021-03-19 | 国网河北省电力有限公司营销服务中心 | Construction method of multi-dimensional informatization label system of power consumer |
CN112712270B (en) * | 2020-12-31 | 2024-04-26 | 深圳前海微众银行股份有限公司 | Information processing method, device, equipment and storage medium |
CN112712270A (en) * | 2020-12-31 | 2021-04-27 | 深圳前海微众银行股份有限公司 | Information processing method, device, equipment and storage medium |
CN113298385A (en) * | 2021-05-26 | 2021-08-24 | 上海晓途网络科技有限公司 | User management method and device, electronic equipment and storage medium |
CN113592297B (en) * | 2021-07-30 | 2023-11-24 | 国家电网有限公司客户服务中心 | Dedicated customer management method and system for electric power system |
CN113592297A (en) * | 2021-07-30 | 2021-11-02 | 国家电网有限公司客户服务中心 | Method and system for managing power system-specific customers |
CN113780677A (en) * | 2021-09-26 | 2021-12-10 | 深圳供电局有限公司 | Prediction method and device for potential power repeated appeal user |
CN115131806A (en) * | 2022-06-07 | 2022-09-30 | 福建极推科技有限公司 | Method and system for recognizing OCR (optical character recognition) image information of various certificates based on deep learning |
CN115131806B (en) * | 2022-06-07 | 2023-10-31 | 福建极推科技有限公司 | Method and system for identifying OCR (optical character recognition) image information of various certificates based on deep learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111353792A (en) | Client portrait system with visual display and data analysis functions | |
Cicala | Imperfect markets versus imperfect regulation in US electricity generation | |
CN106780140B (en) | Power credit evaluation method based on big data | |
Joskow et al. | A quantitative analysis of pricing behavior in California's wholesale electricity market during summer 2000 | |
Naganathan et al. | Building energy modeling (BEM) using clustering algorithms and semi-supervised machine learning approaches | |
US20080319923A1 (en) | Investment Analysis and Planning System and Method | |
CN111126776A (en) | Electricity charge risk prevention and control model construction method based on logistic regression algorithm | |
Hallak et al. | Productivity, quality and exporting behavior under minimum quality requirements | |
CN107103548A (en) | The monitoring method and system and risk monitoring and control method and system of network behavior data | |
Yan et al. | A dynamic network analysis of spot electricity prices in the Australian national electricity market | |
CN111626543A (en) | Method and device for processing power related data | |
Datta | Decoupling and demand-side management: Evidence from the US electric industry | |
CN112463760A (en) | Information processing method, information processing device, computer equipment and storage medium | |
Shuai et al. | Analysis and identification of power blackout-sensitive users by using big data in the energy system | |
Nguyen et al. | Governance and export performance in Vietnam | |
CN111626843A (en) | Vintage analysis method based on risk management data | |
CN114757596B (en) | Power business monitoring and early warning method and system and readable storage medium | |
Theodorou et al. | Modeling the return and volatility of the Greek electricity marginal system price | |
Ghasemi et al. | A hierarchical approach to designing an electricity distribution reward‐penalty scheme for service quality improvement | |
de Oliveira et al. | Natural gas power generation in Brazil: New window of opportunity? | |
CN114595909A (en) | Electricity charge recovery risk assessment method and system | |
Sukhareva et al. | Organizational and economic mechanism for creating a digital environment in the energy sector | |
Ye | Mind the Gap: a Case Study of Demand Prediction and Factors Affecting Waiting Time at Uber NYC and Washington DC | |
CN107194771A (en) | A kind of objective credit management method of dynamic corporate client | |
Tian et al. | Research on Scenario-based Intelligent Inspection Mode Based on Big Data Analysis |
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 |
Application publication date: 20200630 |
|
RJ01 | Rejection of invention patent application after publication |