CN111507550A - Automatic recommendation method for optimal solution of work order problem - Google Patents

Automatic recommendation method for optimal solution of work order problem Download PDF

Info

Publication number
CN111507550A
CN111507550A CN201910094293.7A CN201910094293A CN111507550A CN 111507550 A CN111507550 A CN 111507550A CN 201910094293 A CN201910094293 A CN 201910094293A CN 111507550 A CN111507550 A CN 111507550A
Authority
CN
China
Prior art keywords
solution
work order
follows
frequency
word
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
Application number
CN201910094293.7A
Other languages
Chinese (zh)
Inventor
张良均
李怡婷
陈世涛
刘名军
赵海林
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Tipdm Intelligent Technology Co ltd
Original Assignee
Guangzhou Tipdm Intelligent Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangzhou Tipdm Intelligent Technology Co ltd filed Critical Guangzhou Tipdm Intelligent Technology Co ltd
Priority to CN201910094293.7A priority Critical patent/CN111507550A/en
Publication of CN111507550A publication Critical patent/CN111507550A/en
Pending legal-status Critical Current

Links

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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • 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

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Educational Administration (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses an automatic recommendation method for historical solutions related to problems in power customer service work order records. The method comprises the following steps: acquiring work order basic data; simplifying the historical solution aiming at the work order problem; then, a solution evaluation index system is constructed by combining business analysis and semantic analysis; further, a solution comprehensive scoring model is constructed to obtain recommendation scores of historical solutions; finally, the recommendation scores of the schemes are sorted and automatically recommended before

Description

Automatic recommendation method for optimal solution of work order problem
Technical Field
The invention relates to the technical field of analysis and management of power customer service work order problems, in particular to an automatic recommendation method for an optimal solution of the work order problem.
Background
The power grid hotline service is a key channel for communication between a power grid company and users and opinion feedback, and the management and management of the power grid company can be improved subtly by acquiring, analyzing and utilizing information such as problems, opinions and the like of the users. The work management system is internally promoted, the electricity selling service and the marketing mode of a power grid company are perfected, the power grid company is more accordant with the market demand, and the market consumption activity is promoted. User experience is improved outwards, the problems existing in the power utilization process of the user are solved, and the user can have good service experience in power purchase, power utilization and power sale consultation. Therefore, the power grid hotline service is an important management mode for power grid business marketing, and obviously, the effect of improving the efficiency of the hotline service is also an important work.
The daily incoming call volume of current electric wire netting hotline service is very big, and the operating pressure of first line seat is increasing day by day, and the expansion of visit volume also can influence the efficiency and the quality of customer service solution problem simultaneously. A plurality of classical problems exist in the work orders accumulated in the history, and a large amount of historical solution experience is not mined and utilized. The problem is interpreted in a manual mode and compared with the conventional problem to obtain a solution, and although the accuracy is high, the efficiency is not satisfactory; on the other hand, a large amount of information is wasted.
In conclusion, the invention provides an automatic recommendation method for an optimal solution of the work order problem, which integrates all solutions accumulated by historical experience of the work order problem; the scheme is processed and simplified in combination with a natural language processing technology, the readability of the solution is improved, and the degree of fit of the model is increased; further, a comprehensive evaluation index system is constructed, wherein the comprehensive evaluation index system comprises solution duration, use frequency, use interval and information quantity; and finally, obtaining the recommendation score of each historical scheme by constructing a solution comprehensive scoring model, thereby selecting the scheme with the front score. The optimal decision for solving the problem is quickly obtained for the first-line seat, the work order conversion rate is reduced, the problem solving rate is improved, and the satisfaction degree of customers is effectively increased.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide an automatic recommendation method for the optimal solution of the work order problem, which is used for realizing the most effective solution in the automatic digging problem and quickly guiding the work of customer service.
In order to achieve the purpose, the invention is realized by the following technical scheme: an automatic recommendation method for an optimal solution of a work order problem comprises the following steps:
(S1): acquiring work order basic data;
(S2): simplifying the historical solution aiming at the work order problem;
(S3): then, a solution evaluation index system is constructed by combining business analysis and semantic analysis;
(S4): further, a solution comprehensive scoring model is constructed to obtain recommendation scores of historical solutions;
(S5): and finally, sequencing the recommended points of the schemes, and automatically recommending the first n optimal schemes.
In the step (S1), the work order basic data required by the analysis of the method includes a problem attribute field and a work order attribute field, where the problem attribute field includes: number, start time, title; the work order attribute fields include: number, title, content, solution, start time, end time, priority.
The work order problem number and the work order number have an association relation, and all corresponding work order numbers and attribute information can be associated by the problem number. The problem is associated with the work order in a one-to-one or one-to-many manner, meaning that a problem may correspond to one or more historical solutions.
In the step (S1), the research object of the method is a work order problem, and after deep excavation, an optimal solution for each problem is automatically recommended.
The step (S2) specifically includes:
(S2.1): identifying the same historical solution record in the work order problem as the same solution;
(S2.2): through exploration on the historical solution texts, a rule base is formulated, redundant text information in the scheme is removed, and the purpose of simplifying the scheme is achieved. The rule base is a regular expression set and represents a character string sequence meeting a specified expression form.
In the step (S3), the evaluation index system of the problem solution includes: the solution duration, the use frequency, the use interval and the information amount. The indexes are as follows:
the solution time is long: referring to the average solving time of the problem of the same solution, the calculation formula is as follows:
Figure 141211DEST_PATH_IMAGE002
wherein the content of the first and second substances,
Figure 748910DEST_PATH_IMAGE004
indicates that the solution exists
Figure DEST_PATH_IMAGE005
Recording a work order;
Figure DEST_PATH_IMAGE007
、
Figure DEST_PATH_IMAGE009
respectively represent the second of the solution
Figure DEST_PATH_IMAGE011
The processing end time and the processing start time of the strip record,
frequency of use: refers to the number of times the solution is used, i.e. the corresponding work order record
Figure 270021DEST_PATH_IMAGE005
It is expressed as follows:
Figure DEST_PATH_IMAGE013
the use interval is as follows: refers to the time interval between the maximum start time in the record of the same solution and the current recommended analysis. Reflecting the freshness and effectiveness of the solution, the formula is calculated as follows:
Figure DEST_PATH_IMAGE015
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE017
representing the system time of the current recommendation analysis;
Figure DEST_PATH_IMAGE019
indicating the maximum start time in the recording of the solution,
information amount: it means that the amount of the solution text information in the record is calculated from the perspective of text semantic analysis. In the method, the more the keywords are, the larger the information amount is, the more important the scheme is, and the formula is calculated as follows:
Figure DEST_PATH_IMAGE021
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE023
a keyword in the solution is represented and,
Figure DEST_PATH_IMAGE025
all the words representing solution text participles.
In the step (S3), after all the indexes in the solution evaluation system are constructed, further standardization is required, the standardization mode adopted by the method is a maximum and minimum standardization method,
the method divides the indexes into two types of high-quality indexes and low-quality indexes. Wherein the high-quality indexes comprise use frequency and information content; the low-priority index has the advantages of long solution time and use interval.
The standardized formula for the high-quality index is as follows:
Figure DEST_PATH_IMAGE027
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE029
is shown as
Figure DEST_PATH_IMAGE031
An index value of the solution;
Figure DEST_PATH_IMAGE033
is shown as
Figure 893638DEST_PATH_IMAGE034
An index normalized value for each solution;
Figure 347753DEST_PATH_IMAGE036
、
Figure 657512DEST_PATH_IMAGE038
represents the maximum value and the minimum value of the solution index value,
the standardized formula for the low-priority index is as follows:
Figure 197078DEST_PATH_IMAGE040
in the step (S4), the best solution comprehensive scoring model is as follows:
Figure 250484DEST_PATH_IMAGE042
wherein the content of the first and second substances,
Figure 90264DEST_PATH_IMAGE044
is shown as
Figure DEST_PATH_IMAGE045
Comprehensive recommendation scores of the solutions;
Figure DEST_PATH_IMAGE047
is shown as
Figure 977449DEST_PATH_IMAGE048
The first solution
Figure 738731DEST_PATH_IMAGE050
An individual index normalized value;
Figure 595829DEST_PATH_IMAGE052
is shown as
Figure DEST_PATH_IMAGE053
The weight of each index can be determined by an expert evaluation method; wherein
Figure 758957DEST_PATH_IMAGE054
When 1, 2, 3, 4, it can respectively represent the solution duration, the use frequency, the use interval, and the information amount; the sum of the weights of all indexes is 1, namely:
Figure 410518DEST_PATH_IMAGE056
。
in the step (S5), after the comprehensive recommendation scores of all the solutions of the problem are obtained, sorting in a descending order according to the recommendation scores from large to small, and selecting the solution before ranking
Figure 434930DEST_PATH_IMAGE058
The solution of (2) is recommended.
In the step (S3), the solution text keyword obtaining step is: word segmentation → stop words → word frequency statistics → keyword selection, which is specifically described as follows:
word segmentation processing: implementing word segmentation of a work order solution text by using a forward maximum matching method to obtain word segmentation corpus of all schemes;
stop words: firstly, a common stop word list is prepared, and stop words in a word material set are filtered out through matching of a participle word material set and the stop word list.
Word frequency statistics: and carrying out word frequency statistics on the processed corpus set to obtain the frequency of each word.
Selecting keywords: and sorting in a descending order according to the statistical frequency of the vocabulary according to the word frequency statistical result, and selecting the vocabulary with higher frequency as the keyword. The keyword can be screened by selecting a suitable frequency threshold according to actual conditions.
The invention can automatically recommend the current best solution from the historical solutions of the work order problem. The benefit of the method is more obvious when thousands of work order solutions are provided for the number of work orders related to long-term problems and key problems.
The invention comprehensively considers the factors of problem information degree, timeliness, use frequency and the like of the solution, so that the automatically recommended scheme is more consistent with the current scene state of the problem. Generally, the first-line processing personnel has a relatively low completion rate for processing complex operation and maintenance problems and needs to be transferred to two or three lines for communication processing. After the effective scheme strategy is automatically obtained, the work order problem can be quickly processed, the first-line problem communication cost is reduced, and the work order circulation rate is reduced. The working efficiency of customer service is improved, and the user experience is improved.
Drawings
FIG. 1 Overall technical flow diagram of the invention
FIG. 2 is a simplified technical flow chart of the scheme of the present invention
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
Referring to fig. 1, the following technical solutions are adopted in the present embodiment: an automatic recommendation method for an optimal solution of a work order problem comprises the following steps:
(S1): acquiring work order basic data;
(S2): simplifying the historical solution aiming at the work order problem;
(S3): then, a solution evaluation index system is constructed by combining business analysis and semantic analysis;
(S4): further, a solution comprehensive scoring model is constructed to obtain recommendation scores of historical solutions;
(S5): and finally, sequencing the recommended points of the schemes, and automatically recommending the first n optimal schemes.
In the step (S1), the work order basic data required by the analysis of the method includes a problem attribute field and a work order attribute field, where the problem attribute field includes: number, start time, title; the work order attribute fields include: number, title, content, solution, start time, end time, priority.
The work order problem number and the work order number have an association relation, and all corresponding work order numbers and attribute information can be associated by the problem number. The problem is associated with the work order in a one-to-one or one-to-many manner, meaning that a problem may correspond to one or more historical solutions.
In the step (S1), the research object of the method is a work order problem, and after deep excavation, an optimal solution for each problem is automatically recommended.
Referring to fig. 2, the step (S2) specifically includes:
(S2.1): identifying the same historical solution record in the work order problem as the same solution;
(S2.2): through exploration on the historical solution texts, a rule base is formulated, redundant text information in the scheme is removed, and the purpose of simplifying the scheme is achieved. The rule base is a regular expression set and represents a character string sequence meeting a specified expression form.
In the step (S3), the evaluation index system of the problem solution includes: the solution duration, the use frequency, the use interval and the information amount. The indexes are as follows:
the solution time is long: refers to the average resolution time of the problem for the same solution. The calculation formula is as follows:
Figure DEST_PATH_IMAGE060
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE062
indicates that the solution exists
Figure 298981DEST_PATH_IMAGE063
Recording a work order;
Figure 316616DEST_PATH_IMAGE065
、
Figure 404657DEST_PATH_IMAGE067
respectively represent the second of the solution
Figure 140532DEST_PATH_IMAGE068
The processing end time and the processing start time of the bar record.
Frequency of use: refers to the number of times the solution is used, i.e. the corresponding work order record. The representation is as follows:
Figure DEST_PATH_IMAGE071
the use interval is as follows: refers to the time interval between the maximum start time in the record of the same solution and the current recommended analysis. Reflecting the freshness and effectiveness of the solution. The formula is calculated as follows:
Figure DEST_PATH_IMAGE073
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE075
representing the system time of the current recommendation analysis;
Figure DEST_PATH_IMAGE077
indicating the maximum start time in the recording of the solution.
Information amount: it means that the amount of the solution text information in the record is calculated from the perspective of text semantic analysis. The more the keywords in the scheme of the method, the larger the information amount is, and the more important the scheme is. The formula is calculated as follows:
Figure DEST_PATH_IMAGE079
wherein the content of the first and second substances,
Figure 277115DEST_PATH_IMAGE080
a keyword in the solution is represented and,
Figure DEST_PATH_IMAGE081
all the words representing solution text participles.
In the step (S3), after all the indexes in the solution evaluation system are constructed, further standardization is required, and the standardization mode adopted by the method is a maximum and minimum standardization method.
The method divides the indexes into two types of high-quality indexes and low-quality indexes. Wherein the high-quality indexes comprise use frequency and information content; the low-priority index has the advantages of long solution time and use interval.
The standardized formula for the high-quality index is as follows:
Figure 149256DEST_PATH_IMAGE082
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE083
is shown as
Figure 877041DEST_PATH_IMAGE034
An index value of the solution;
Figure 631370DEST_PATH_IMAGE084
is shown as
Figure 368382DEST_PATH_IMAGE068
An index normalized value for each solution;
Figure DEST_PATH_IMAGE085
、
Figure DEST_PATH_IMAGE086
represents the maximum value and the minimum value of the solution index value,
the standardized formula for the low-priority index is as follows:
Figure DEST_PATH_IMAGE087
in the step (S4), the best solution comprehensive scoring model is as follows:
Figure DEST_PATH_IMAGE088
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE089
is shown as
Figure DEST_PATH_IMAGE090
Comprehensive recommendation scores of the solutions;
Figure DEST_PATH_IMAGE091
is shown as
Figure DEST_PATH_IMAGE092
The first solution
Figure DEST_PATH_IMAGE093
An individual index normalized value;
Figure DEST_PATH_IMAGE094
is shown as
Figure DEST_PATH_IMAGE095
The weight of each index can be determined by an expert evaluation method; wherein
Figure DEST_PATH_IMAGE096
When 1, 2, 3, 4, it can respectively represent the solution duration, the use frequency, the use interval, and the information amount; the sum of the weights of all indexes is 1, namely:
Figure DEST_PATH_IMAGE097
。
in the step (S5), after the comprehensive recommendation scores of all the solutions of the problem are obtained, sorting in a descending order according to the recommendation scores from large to small, and selecting the solution before ranking
Figure DEST_PATH_IMAGE098
The solution of (2) is recommended.
In the step (S3), the solution text keyword obtaining step is: word segmentation → stop words → word frequency statistics → keyword selection. The concrete description is as follows:
word segmentation processing: implementing word segmentation of a work order solution text by using a forward maximum matching method to obtain word segmentation corpus of all schemes;
stop words: firstly, a common stop word list is prepared, and stop words in a word material set are filtered out through matching of a participle word material set and the stop word list.
Word frequency statistics: and carrying out word frequency statistics on the processed corpus set to obtain the frequency of each word.
Selecting keywords: and sorting in a descending order according to the statistical frequency of the vocabulary according to the word frequency statistical result, and selecting the vocabulary with higher frequency as the keyword. The keyword can be screened by selecting a suitable frequency threshold according to actual conditions.
The above is a detailed description of an embodiment of the method for automatically recommending an optimal solution to a work order problem according to the present invention, and for ease of understanding, a detailed description will be given below of the method for automatically recommending an optimal solution to a work order problem according to the present invention by using a specific example.
(S1): acquiring work order basic data;
the work order basic data required by analysis of the method comprises a problem attribute field and a work order attribute field, wherein the problem attribute field comprises: number, start time, title; the work order attribute fields include: number, title, content, solution, start time, end time, priority.
The work order problem number and the work order number have an association relation, and all corresponding work order numbers and attribute information can be associated by the problem number. The problem is associated with the work order in a one-to-one or one-to-many manner, meaning that a problem may correspond to one or more historical solutions.
After the problem is associated with the work order relationship, the problem of 'abnormal printer use' is selected, and data of part of work order problems and solutions are shown in a table 1:
TABLE 1 problem and History solutions
Figure DEST_PATH_IMAGE099
(S2): simplifying the historical solution aiming at the work order problem;
firstly, recording the same historical solution in the work order problem, and identifying the same solution;
for the example data in Table 1, the results of identifying a unique solution are shown in Table 2:
table 2 identifies unique solutions
Figure DEST_PATH_IMAGE100
(S2.2): through exploration on the historical solution texts, a rule base is formulated, redundant text information in the scheme is removed, and the purpose of simplifying the scheme is achieved. The rule base is a regular expression set and represents a character string sequence with a specified expression form.
The rule generation principle is explained taking the solution of the solution SA _3 as an example.
The original scheme is as follows:
1. recheck the event description results: and through field verification, the event description is substantial. 2. The failure reason is as follows: the printer needs to be provided with a driver. 3. The treatment process comprises the following steps: and installing a printer driver and testing the printing success. 4. User review condition: and confirming to return to normal with the user.
The 1 st and 4 th in the scheme are obviously redundant, corresponding text regular rules are generated according to the two conditions, and then the texts meeting the regular rules are removed, so that the aim of simplifying the scheme is fulfilled. An example of rule generation is as follows:
the simplified rule base is arranged in the above mode.
The effect of eliminating the redundant text in the scheme in table 2 is shown in table 3.
TABLE 3 scheme for eliminating redundant text
Figure DEST_PATH_IMAGE102
(S3): then, a solution evaluation index system is constructed by combining business analysis and semantic analysis;
the evaluation index system of the problem solution comprises: the solution duration, the use frequency, the use interval and the information amount. The indexes are as follows:
the solution time is long: refers to the average resolution time of the problem for the same solution. The calculation formula is as follows:
Figure DEST_PATH_IMAGE104
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE105
indicates that the solution exists
Figure DEST_PATH_IMAGE106
Recording a work order;
Figure DEST_PATH_IMAGE107
、
Figure DEST_PATH_IMAGE108
respectively represent the second of the solution
Figure 577253DEST_PATH_IMAGE090
The processing end time and the processing start time of the bar record.
Frequency of use: refers to the number of times the solution is used, i.e. the corresponding work order record
Figure 7098DEST_PATH_IMAGE105
. The representation is as follows:
Figure DEST_PATH_IMAGE109
the use interval is as follows: refers to the time interval between the maximum start time in the record of the same solution and the current recommended analysis. Reflecting the freshness and effectiveness of the solution. The formula is calculated as follows:
Figure DEST_PATH_IMAGE110
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE111
representing the system time of the current recommendation analysis;
Figure DEST_PATH_IMAGE112
indicating the maximum start time in the recording of the solution.
Information amount: it means that the amount of the solution text information in the record is calculated from the perspective of text semantic analysis. The more the keywords in the scheme of the method, the larger the information amount is, and the more important the scheme is. The formula is calculated as follows:
Figure DEST_PATH_IMAGE113
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE114
a keyword in the solution is represented and,
Figure DEST_PATH_IMAGE115
all the words representing solution text participles.
Through the index structure explanation, the scheme index value table 4 is obtained
TABLE 4 evaluation index of "abnormal Printer use
Figure DEST_PATH_IMAGE117
And standardizing the result by using a high-quality index and low-quality index standardization method. The results are shown in Table 5.
TABLE 5 standardization of evaluation index
Figure DEST_PATH_IMAGE119
By this point, the historical solution evaluation index system of "printer use exception" is constructed and completed, and the model analysis is to be developed.
(S4): further, a solution comprehensive scoring model is constructed to obtain recommendation scores of historical solutions;
the best solution comprehensive scoring model used by the invention is as follows:
Figure DEST_PATH_IMAGE120
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE121
is shown as
Figure DEST_PATH_IMAGE123
Personal solutionA comprehensive recommendation score for the solution;
Figure DEST_PATH_IMAGE124
is shown as
Figure DEST_PATH_IMAGE125
The first solution
Figure 796193DEST_PATH_IMAGE095
An individual index normalized value;
Figure DEST_PATH_IMAGE126
is shown as
Figure DEST_PATH_IMAGE127
The weight of each index can be determined by an expert evaluation method; wherein
Figure 304273DEST_PATH_IMAGE096
When 1, 2, 3, 4, it can respectively represent the solution duration, the use frequency, the use interval, and the information amount; the sum of the weights of all indexes is 1, namely:
Figure DEST_PATH_IMAGE128
。
the weight value obtained by the preliminary expert opinion is
Figure DEST_PATH_IMAGE130
。
The recommendations obtained for each protocol are shown in table 6.
TABLE 6 solution recommendation score
Figure DEST_PATH_IMAGE132
(S5): finally, the recommendation scores of the schemes are sorted and automatically recommended before
Figure DEST_PATH_IMAGE134
And (5) optimizing the scheme.
The rankings obtained by sorting the solution recommendations in descending order are shown in table 7.
Table 7 solution ranking
Figure DEST_PATH_IMAGE136
Selecting a recommended quantity
Figure DEST_PATH_IMAGE138
The best solutions proposed in turn are shown in table 8.
TABLE 8 best case recommendations
Figure DEST_PATH_IMAGE140
From the recommended optimal solution, the first 3 main causes of printer abnormality are: installation drive problems, paper jam problems, set print port problems.
The method of the invention collects all solutions accumulated by historical experience of work order problems; the scheme is processed and simplified in a precise way by combining a natural language processing technology, the readability of the solution is improved, and the degree of fit of the model is increased; further, a comprehensive evaluation index system is constructed, wherein the comprehensive evaluation index system comprises solution duration, use frequency, use interval and information quantity; and finally, obtaining the recommendation score of each historical scheme by constructing a solution comprehensive scoring model, thereby selecting the scheme with the front score. The optimal decision for solving the problem is quickly obtained for the first-line seat, the work order conversion rate is reduced, the problem solving rate is improved, and the satisfaction degree of customers is effectively increased.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.

Claims (9)

1. An automatic recommendation method for an optimal solution of a work order problem is characterized by comprising the following steps: the method comprises the following steps:
(S1): acquiring work order basic data;
(S2): simplifying the historical solution aiming at the work order problem;
(S3): then, a solution evaluation index system is constructed by combining business analysis and semantic analysis;
(S4): further, a solution comprehensive scoring model is constructed to obtain recommendation scores of historical solutions;
(S5): and finally, sequencing the recommended points of the schemes, and automatically recommending the first n optimal schemes.
2. The method as claimed in claim 1, wherein the method comprises: in the step (S1), the work order basic data required by the analysis of the method includes a problem attribute field and a work order attribute field, where the problem attribute field includes: number, start time, title; the work order attribute fields include: number, title, content, solution, start time, end time, priority,
the work order problem numbers and the work order numbers have an association relationship, all corresponding work order numbers and attribute information can be associated by the problem numbers, and the problem and the work order are in one-to-one or one-to-many relationship, which means that one problem corresponds to one or more historical solutions.
3. The method as claimed in claim 1, wherein the method comprises: in the step (S1), the research object of the method is a work order problem, and after deep excavation, an optimal solution for each problem is automatically recommended.
4. The method as claimed in claim 1, wherein the method comprises: the step (S2) specifically includes:
(S2.1): identifying the same historical solution record in the work order problem as the same solution;
(S2.2): through exploring a historical solution text, a rule base is formulated, redundant text information in the scheme is removed, and the purpose of simplifying the scheme is achieved, wherein the rule base is a regular expression set and represents a character string sequence meeting the specified expression form.
5. The method as claimed in claim 1, wherein the method comprises: in the step (S3), the evaluation index system of the problem solution includes: the solution duration, the use frequency, the use interval and the information quantity,
the indexes are as follows:
the solution time is long: referring to the average solving time of the problem of the same solution, the calculation formula is as follows:
Figure 670993DEST_PATH_IMAGE001
wherein the content of the first and second substances,
Figure 559314DEST_PATH_IMAGE002
indicates that the solution exists
Figure 441820DEST_PATH_IMAGE002
Recording a work order;
Figure 794304DEST_PATH_IMAGE003
、
Figure 674535DEST_PATH_IMAGE004
respectively represent the second of the solution
Figure 112469DEST_PATH_IMAGE005
The processing end time and the processing start time of the strip record,
frequency of use: refers to the number of times the solution is used, i.e. the corresponding work order record
Figure 736349DEST_PATH_IMAGE006
，
The representation is as follows:
Figure 943339DEST_PATH_IMAGE007
the use interval is as follows: the time interval from the maximum starting time to the current recommended analysis in the record of the same solution reflects the freshness and effectiveness of the solution, and the formula is calculated as follows:
Figure 994472DEST_PATH_IMAGE008
wherein the content of the first and second substances,
Figure 919703DEST_PATH_IMAGE009
representing the system time of the current recommendation analysis;
Figure 93412DEST_PATH_IMAGE010
indicating the maximum start time in the recording of the solution,
information amount: the method refers to calculating the amount of the text information of the solution in the record from the perspective of text semantic analysis, the more the keywords in the scheme of the method, the larger the information amount is, the more important the scheme is, and the formula is calculated as follows:
Figure 420489DEST_PATH_IMAGE011
wherein the content of the first and second substances,
Figure 439260DEST_PATH_IMAGE012
a keyword in the solution is represented and,
Figure 789470DEST_PATH_IMAGE013
all the words representing solution text participles.
6. The method as claimed in claim 1, wherein the method comprises: in the step (S3), after all the indexes in the solution evaluation system are constructed, further standardization is required, the standardization mode adopted by the method is a maximum and minimum standardization method,
the method divides the indexes into two types of high-quality indexes and low-quality indexes,
wherein the high-quality indexes comprise use frequency and information content; the low-priority indexes comprise the solution duration and the use interval,
the standardized formula for the high-quality index is as follows:
Figure 817469DEST_PATH_IMAGE014
wherein the content of the first and second substances,
Figure 936735DEST_PATH_IMAGE015
is shown as
Figure 391987DEST_PATH_IMAGE016
An index value of the solution;
Figure 26230DEST_PATH_IMAGE017
is shown as
Figure 530024DEST_PATH_IMAGE018
An index normalized value for each solution;
Figure 831692DEST_PATH_IMAGE019
、
Figure 395529DEST_PATH_IMAGE020
represents the maximum value and the minimum value of the solution index value,
the standardized formula for the low-priority index is as follows:
Figure 517069DEST_PATH_IMAGE021
。
7. the method as claimed in claim 1, wherein the method comprises: in the step (S4), the best solution comprehensive scoring model is as follows:
Figure 886870DEST_PATH_IMAGE022
wherein the content of the first and second substances,
Figure 980728DEST_PATH_IMAGE023
is shown as
Figure 777783DEST_PATH_IMAGE024
Comprehensive recommendation scores of the solutions;
Figure 324302DEST_PATH_IMAGE025
is shown as
Figure 232215DEST_PATH_IMAGE026
The first solution
Figure 242896DEST_PATH_IMAGE027
An individual index normalized value;
Figure 148535DEST_PATH_IMAGE028
is shown as
Figure 244667DEST_PATH_IMAGE029
The weight of each index can be determined by an expert evaluation method; wherein
Figure 956271DEST_PATH_IMAGE030
When 1, 2, 3, 4, it can respectively represent the solution duration, the use frequency, the use interval, and the information amount;the sum of the weights of all indexes is 1, namely:
Figure 759142DEST_PATH_IMAGE031
。
8. the method as claimed in claim 1, wherein the method comprises: in the step (S5), after the comprehensive recommendation scores of all the solutions of the problem are obtained, sorting in a descending order according to the recommendation scores from large to small, and selecting the solution before ranking
Figure 898000DEST_PATH_IMAGE032
The solution of (2) is recommended.
9. The method as claimed in claim 1, wherein the method comprises: in the step (S3), the solution text keyword obtaining step is: word segmentation → stop words → word frequency statistics → keyword selection, which is specifically described as follows:
word segmentation processing: implementing word segmentation of a work order solution text by using a forward maximum matching method to obtain word segmentation corpus of all schemes;
stop words: firstly, a common stop word list is required to be prepared, and stop words in a word material set are filtered out through matching of a participle word material set and the stop word list;
word frequency statistics: performing word frequency statistics on the processed corpus to obtain the frequency of each vocabulary;
selecting keywords: according to the word frequency statistical result, sorting in descending order according to the statistical frequency of the words, selecting the words with higher frequency as the keywords, and selecting the appropriate frequency threshold value according to the actual situation to screen the keywords.
CN201910094293.7A 2019-01-30 2019-01-30 Automatic recommendation method for optimal solution of work order problem Pending CN111507550A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910094293.7A CN111507550A (en) 2019-01-30 2019-01-30 Automatic recommendation method for optimal solution of work order problem

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910094293.7A CN111507550A (en) 2019-01-30 2019-01-30 Automatic recommendation method for optimal solution of work order problem

Publications (1)

Publication Number Publication Date
CN111507550A true CN111507550A (en) 2020-08-07

Family

ID=71864586

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910094293.7A Pending CN111507550A (en) 2019-01-30 2019-01-30 Automatic recommendation method for optimal solution of work order problem

Country Status (1)

Country Link
CN (1) CN111507550A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113159625A (en) * 2021-05-12 2021-07-23 厦门立林科技有限公司 Intelligent work order processing method
CN114117029A (en) * 2021-11-24 2022-03-01 国网山东省电力公司信息通信公司 Solution recommendation method and system based on multi-level information enhancement

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011053992A (en) * 2009-09-02 2011-03-17 Nippon Telegr & Teleph Corp <Ntt> Question recommending device, method and program
CN103902731A (en) * 2014-04-16 2014-07-02 国家电网公司 Intelligent information maintenance method based on knowledge base inquiry
CN106095872A (en) * 2016-06-07 2016-11-09 北京高地信息技术有限公司 Answer sort method and device for Intelligent Answer System
CN106095941A (en) * 2016-06-14 2016-11-09 华南师范大学 Recommend method and system based on the solution in big data knowledge storehouse
WO2016179938A1 (en) * 2015-05-14 2016-11-17 百度在线网络技术（北京）有限公司 Method and device for question recommendation
CN108846134A (en) * 2018-07-05 2018-11-20 连云港杰瑞电子有限公司 A kind of O&M scheme recommender system and method based on web crawlers
CN109241413A (en) * 2018-08-20 2019-01-18 安徽智恒信科技有限公司 A kind of the knowledge recommendation method and maintenance work platform of work order

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011053992A (en) * 2009-09-02 2011-03-17 Nippon Telegr & Teleph Corp <Ntt> Question recommending device, method and program
CN103902731A (en) * 2014-04-16 2014-07-02 国家电网公司 Intelligent information maintenance method based on knowledge base inquiry
WO2016179938A1 (en) * 2015-05-14 2016-11-17 百度在线网络技术（北京）有限公司 Method and device for question recommendation
CN106095872A (en) * 2016-06-07 2016-11-09 北京高地信息技术有限公司 Answer sort method and device for Intelligent Answer System
CN106095941A (en) * 2016-06-14 2016-11-09 华南师范大学 Recommend method and system based on the solution in big data knowledge storehouse
CN108846134A (en) * 2018-07-05 2018-11-20 连云港杰瑞电子有限公司 A kind of O&M scheme recommender system and method based on web crawlers
CN109241413A (en) * 2018-08-20 2019-01-18 安徽智恒信科技有限公司 A kind of the knowledge recommendation method and maintenance work platform of work order

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113159625A (en) * 2021-05-12 2021-07-23 厦门立林科技有限公司 Intelligent work order processing method
CN114117029A (en) * 2021-11-24 2022-03-01 国网山东省电力公司信息通信公司 Solution recommendation method and system based on multi-level information enhancement
CN114117029B (en) * 2021-11-24 2023-11-24 国网山东省电力公司信息通信公司 Solution recommendation method and system based on multi-level information enhancement

Similar Documents

Publication Publication Date Title
US10380626B2 (en) System and method for evaluating search queries to identify titles for content production
CN106651424B (en) Power user portrait establishing and analyzing method based on big data technology
CN107992609B (en) Complaint tendency judgment method based on text classification technology and decision tree
US20200143288A1 (en) Training of Chatbots from Corpus of Human-to-Human Chats
CN113704451B (en) Power user appeal screening method and system, electronic device and storage medium
US20160110429A1 (en) Systems and Methods for Social Media Data Mining
WO2019062001A1 (en) Intelligent robotic customer service method, electronic device and computer readable storage medium
KR102396413B1 (en) Social big data analysis report automatic provision system using big data and artificial intelligence
US11004096B2 (en) Buy intent estimation and its applications for social media data
CA2720167A1 (en) An automated media analysis and document management system
CN111127105A (en) User hierarchical model construction method and system, and operation analysis method and system
US11790380B2 (en) Systems and methods for finding an interaction subset within a set of interactions
CN111507550A (en) Automatic recommendation method for optimal solution of work order problem
CN112669113A (en) Product recommendation method and device, storage medium and electronic device
CN117271905A (en) Crowd image-based lateral demand analysis method and system
US20230281228A1 (en) Generating actionable insight information from data sets using an artificial intelligence-based natural language interface
CN112184292A (en) Marketing method and device based on artificial intelligence decision tree
JP7223549B2 (en) Information operation device and information operation method
CN111209394A (en) Text classification processing method and device
Tekin et al. Big data concept in small and medium enterprises: how big data effects productivity
CN114529383A (en) Method and system for realizing tax payment tracking and tax loss early warning
CN109697203A (en) Index unusual fluctuation analysis method and equipment, computer storage medium, computer equipment
CN113269610A (en) Bank product recommendation method and device and storage medium
KR102078541B1 (en) Issue interest based news value evaluation apparatus and method, storage media storing the same
CN117726360B (en) Client behavior driven industrial product sales analysis method and related components

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