CN117669866A - Customer call service monitoring method and system of three-party data platform - Google Patents
Customer call service monitoring method and system of three-party data platform Download PDFInfo
- Publication number
- CN117669866A CN117669866A CN202311462764.8A CN202311462764A CN117669866A CN 117669866 A CN117669866 A CN 117669866A CN 202311462764 A CN202311462764 A CN 202311462764A CN 117669866 A CN117669866 A CN 117669866A
- Authority
- CN
- China
- Prior art keywords
- call
- grading
- display
- score
- client
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 169
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000002159 abnormal effect Effects 0.000 claims abstract description 58
- 235000019580 granularity Nutrition 0.000 claims description 22
- 230000007246 mechanism Effects 0.000 claims description 11
- 230000005856 abnormality Effects 0.000 claims description 10
- 230000000694 effects Effects 0.000 claims description 5
- 230000009471 action Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 description 16
- 230000009286 beneficial effect Effects 0.000 description 10
- 230000035945 sensitivity Effects 0.000 description 4
- 230000004807 localization Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 230000003442 weekly effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method and a system for monitoring customer call service of a three-party data platform, comprising the following steps: collecting data source sub-division of calling data of a client by a three-party data platform; calculating the grading value of the data source sub-score under each grading type; designing a dynamic monitoring display for the client call based on the grading values under each grading type; positioning and calling an abnormal accident based on the monitoring display; the method ensures the effective monitoring of the call of the client, and is convenient for finding out the grading value of each grading type which does not meet the preset call requirement in time and the corresponding abnormal call data of the client.
Description
Technical Field
The invention relates to the technical field of data information monitoring, in particular to a method and a system for monitoring customer calling service of a three-party data platform.
Background
In the background of big data age, the data resource is an important resource for the internal operation management of enterprises, in the field of financial credit, the data is an important resource for the enterprises to construct a personal credit risk control system, the quality and diversity of the data determine the accuracy of risk management decisions to a certain extent, the data resource comprises internal data and external third party data, and at present, in the market, credit institutions commonly combine the called external grading data products to carry out credit decision management. The dependency of the enterprise's risk decision system on external data represents the necessity of monitoring the data products of the three-party data platform. The mechanism needs to regularly comb the calling condition of external data and timely make multiple disc analysis, and part of the mechanism has the dilemma of shortage of internal personnel resources and insufficient capability of monitoring technicians, and needs to provide technical support by a three-party data platform and output a calling service monitoring report.
Disclosure of Invention
The invention provides a method and a system for monitoring customer call service of a three-party data platform, which ensure effective monitoring of customer call and quickly locate abnormal customer call data.
A customer call service monitoring method of a three-party data platform comprises the following steps:
s1: collecting data source sub-division of calling data of a client by a three-party data platform;
s2: calculating the grading value of the data source sub-score under each grading type;
wherein each grading type comprises grading calling quantity, grading searching quantity, grading non-searching quantity, grading average, grading maximum value, grading minimum value, grading box distribution and grading group stability index;
s3: designing a dynamic monitoring display for the client call based on the grading values under each grading type;
comprising the following steps: determining a correlation between the score types based on the characteristics between the score types;
determining to call dynamic monitoring display for the client according to the grading values under each grading type by utilizing the correlation;
s4: and positioning and calling the abnormal accident based on the monitoring display.
Preferably, after S1, further comprising:
performing traversal analysis on the data source sub-division to extract the data source sub-division with abnormality;
mapping the data source sub-division with the abnormality to obtain the normal data source sub-division.
Preferably, in S2, calculating a score value of the data source sub-score under each score type includes:
dividing the data source sub-division according to different time granularities to obtain corresponding data source sub-division under different time granularities;
calculating the grading values of the corresponding data source sub-divisions under each grading type under the different time granularities;
based on the different time granularities and the corresponding values thereof, establishing a corresponding relation between the time granularities and the scoring values under each scoring type, and based on the corresponding relation, obtaining the scoring values of the data source sub-division under each scoring type under different time granularities.
Preferably, in S2, calculating a score value of the data source sub-score under each score type includes:
the scoring types comprise scoring call quantity, scoring search quantity, scoring non-search quantity, scoring equipartition, scoring maximum value, scoring minimum value, scoring box distribution and scoring population stability index;
wherein, the grading call volume: dividing the call total amount, call uncorrupted amount and call obtained amount; calling a total amount calculation client mechanism to send out an application calling total amount to a data product of the three-party data platform; invoking the unsearched calculation without returning the normal score to the total amount of invocation of the client mechanism; the calculation of the call query returns the normal score to the call total amount of the client mechanism;
score equipartition: calculating a normal score average value except for a special score value, and calling and returning to the client;
score maximum: calculating the maximum value of normal scores except for special scores, which are returned to the call of the client;
score minimum: calculating a minimum value of normal scores except for a special score, which is returned to the call of the client;
grading and box distribution: calculating call returns to clients, and normally scoring call distribution under each sub-box except for special scores;
scoring population stability index: and measuring an index of the deviation between the actual sub-component and the preset sub-component.
Preferably, in S3, the step of designing a dynamic monitoring display for the client call based on the scoring values under each scoring type includes:
determining a correlation between the score types based on the characteristics between the score types;
and determining to call dynamic monitoring display for the client according to the grading values under the grading types by utilizing the correlation.
Preferably, determining the correlation between the score types based on the features between the score types includes:
determining a first relation among the scoring types according to the semantic features of the scoring types;
acquiring the acquisition modes of the score types, acquiring necessary dimension parameters from the acquisition modes, and acquiring the roles of the necessary dimension parameters in the corresponding score types;
determining a second relationship between the respective scoring types based on the role of the necessary dimension parameters in the different scoring types;
the correlation relationship is determined based on the first relationship and the second relationship.
Preferably, determining, by using the correlation, to invoke dynamic monitoring display on the client according to the score values under the score types, including:
setting an index typesetting relationship for the display of the scoring type based on the index action in the correlation;
setting score value range classification under each score type and corresponding index value standard quantity in each score type based on score value information of the historical data source sub-score under each score type;
classifying the scoring values under each scoring type based on the value ranges, and obtaining the actual number of index values under each value range;
acquiring the quantity difference between the actual quantity of the index value and the standard quantity of the index value, and establishing a fluctuation curve of the quantity difference along with the change of the value range;
determining the fluctuation amplitude between the value range and the quantity difference based on the fluctuation curve;
judging whether the fluctuation amplitude is in a preset fluctuation range or not;
if yes, determining the time display dimension of each grading type and the display sequence of the time display dimension according to the history monitoring display of the history data source sub-division;
otherwise, extracting a target value range which does not meet the preset fluctuation range from the value range, and classifying the target value range according to the grading type to obtain a target index value range and a corresponding data difference value;
based on the correlation, determining the relation characteristic of the target index value range, and based on the relation characteristic and combining the data difference value, matching time display dimensions for each grading type and the display sequence of the time display dimensions;
and designing the dynamic monitoring display for the client according to a preset chart display format based on the index typesetting relation, the time display dimension and the display sequence.
Preferably, after S3, the method further includes evaluating and adjusting the monitoring display, specifically:
acquiring client information corresponding to the monitoring display, acquiring historical monitoring display corresponding to the client from the client information, and acquiring abnormal accident positioning information under the historical monitoring display;
based on the abnormal accident positioning information, determining abnormal accident positioning efficiency and abnormal accident positioning accuracy under the history monitoring display;
based on the abnormal accident positioning efficiency and the abnormal accident positioning accuracy, sequencing the history monitoring display according to the comprehensive display effect to obtain a history monitoring display sequence;
extracting features of each history monitoring display in the history monitoring display sequence to obtain display features, and classifying the display features according to display attributes to obtain display attribute features;
selecting optimal display attribute features from the display attribute features based on the feature differences of the display attribute features under each history monitoring display in the history monitoring display sequence and the sequence corresponding to the history monitoring display sequence;
acquiring actual display attribute characteristics of the monitoring display, and comparing the actual display attribute characteristics with optimal display attribute characteristics according to display attributes to obtain feature differences;
judging whether the monitoring display is in an acceptable difference range or not based on the characteristic difference;
if yes, determining that the monitoring display is qualified;
otherwise, determining that the monitoring display is unqualified;
after determining that the monitoring display is unqualified, extracting display attribute characteristics to be adjusted from the actual display attribute characteristics based on the characteristic difference;
and adjusting the display attribute characteristics to be adjusted based on the optimal display attribute characteristics, and obtaining the adjusted monitoring display according to the adjusted display attribute characteristics.
Preferably, in S4, based on the monitoring display, locating the calling abnormal accident includes:
determining call data of the clients based on the monitoring display, classifying the call data according to a call object to obtain a plurality of groups of call sub-data, and setting attribute abnormal values for each attribute of each group of call sub-data based on the call object;
determining an object call outlier D of the call object according to the call object and the attribute outlier thereof;
judging whether the object calling abnormal value D is larger than a preset object abnormal value, if so, taking the calling object as a key object, otherwise, taking the calling object as a secondary object;
calculating the total call outlier R of the call data of the client based on the key object, the secondary object and the corresponding object call outlier;
judging whether the overall calling outlier R is larger than a preset outlier;
if yes, determining that the call data of the client is abnormal, and positioning a data source corresponding to the call data based on the call data to serve as abnormal accident data;
otherwise, determining that the call data of the client is not abnormal, and not positioning abnormal accidents.
A customer call service monitoring system for a three-way data platform, comprising:
the sub-division acquisition module is used for acquiring data source sub-division of the data called by the client by the three-party data platform;
the dimension value calculation module is used for calculating the grading value of the data source sub-division under each grading type;
wherein each grading type comprises grading calling quantity, grading searching quantity, grading non-searching quantity, grading average, grading maximum value, grading minimum value, grading box distribution and grading group stability index;
the monitoring display module is used for designing dynamic monitoring display for the client call based on the grading values under each grading type;
comprising the following steps: determining a correlation between the score types based on the characteristics between the score types;
determining to call dynamic monitoring display for the client according to the grading values under each grading type by utilizing the correlation;
and the abnormality positioning module is used for positioning and calling abnormal accidents based on the monitoring display.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a flowchart of a method for monitoring customer call service of a three-party data platform according to an embodiment of the present invention;
FIG. 2 is another flow chart of a method for monitoring customer call service of a three-party data platform according to an embodiment of the present invention;
fig. 3 is a block diagram of a client call service monitoring system of a three-party data platform according to an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.
Example 1
The embodiment of the invention provides a client call service monitoring method of a three-party data platform, which is shown in fig. 1 and comprises the following steps:
s1: collecting data source sub-division of calling data of a client by a three-party data platform;
s2: calculating the grading value of the data source sub-score under each grading type;
s3: designing a dynamic monitoring display for the client call based on the grading values under each grading type;
s4: and positioning and calling the abnormal accident based on the monitoring display.
In this embodiment, the respective score types include a score call amount, a score found amount, a score not found amount, a score average, a score maximum, a score minimum, a score bin distribution, and a score population stability index.
In this embodiment, the design of the dynamic monitoring display for the client call is specifically to design an optimal dynamic monitoring display for the client call according to the characteristics and the values of each scoring type, so as to realize effective monitoring for the client call.
In this embodiment, locating the call exception accident specifically refers to searching out the score values under each score type which does not meet the preset call requirement and the corresponding exception client call data according to the monitoring display.
The beneficial effects of above-mentioned design scheme are: according to the invention, intelligent monitoring is carried out on the client call data through the three-party data platform, so that the dilemma of shortage of internal personnel resources and insufficient capability of monitoring technicians is avoided, meanwhile, the comprehensiveness and accuracy of monitoring on the client call service are ensured by acquiring the grading values of the client call data under each grading type and acquiring the grading values of a plurality of grading types, then, the optimal client call dynamic monitoring display is designed through the grading values under each grading type, the effective monitoring on the client call is ensured, and the grading values under each grading type which do not meet the preset call requirement and the abnormal client call data corresponding to the grading values are conveniently found in time.
Example 2
Based on embodiment 1, the embodiment of the invention provides a method for monitoring customer call service of a three-party data platform, which further comprises, after step 1:
performing traversal analysis on the data source sub-division to extract the data source sub-division with abnormality;
mapping the data source sub-division with the abnormality to obtain the normal data source sub-division.
In this embodiment, the active data source sub-score is naturally at most in a large number of special cases other than normal scoring, such as scoring timeout, guest group miss, etc., so all special cases need to be filled in with a specified value.
The beneficial effects of above-mentioned design scheme are: the data source sub-division with abnormality is supplemented, so that the integrity and the accuracy of the data source sub-division are ensured, and the quality of the data source sub-division is ensured.
Example 3
Based on embodiment 1, the embodiment of the invention provides a client call service monitoring method of a three-party data platform, in S2, calculating the scoring value of the data source sub-score under each scoring type comprises the following steps:
dividing the data source sub-division according to different time granularities to obtain corresponding data source sub-division under different time granularities;
calculating the grading values of the corresponding data source sub-divisions under each grading type under the different time granularities;
based on the different time granularities and the corresponding values thereof, establishing a corresponding relation between the time granularities and the scoring values under each scoring type, and based on the corresponding relation, obtaining the scoring values of the data source sub-division under each scoring type under different time granularities.
In this embodiment, the different time granularities are e.g. daily, weekly, monthly.
The beneficial effects of above-mentioned design scheme are: the data source sub-division is divided according to different time granularities, so that the grading values of the data source sub-division under different time granularities and under various grading types are obtained, the grading type values are obtained under different time granularities, comprehensive numerical values are provided for monitoring and analyzing the data, and the accuracy of data monitoring is ensured.
Example 4
Based on embodiment 1, the embodiment of the invention provides a client call service monitoring method of a three-party data platform, in S2, calculating the scoring value of the data source sub-score under each scoring type comprises the following steps:
the scoring types comprise scoring call quantity, scoring search quantity, scoring non-search quantity, scoring equipartition, scoring maximum value, scoring minimum value, scoring box distribution and scoring population stability index;
wherein, the grading call volume: dividing the call total amount, call uncorrupted amount and call obtained amount; calling a total amount calculation client mechanism to send out an application calling total amount to a data product of the three-party data platform; invoking the unsearched calculation without returning the normal score to the total amount of invocation of the client mechanism; the calculation of the call query returns the normal score to the call total amount of the client mechanism;
score equipartition: calculating a normal score average value except for a special score value, and calling and returning to the client;
score maximum: calculating the maximum value of normal scores except for special scores, which are returned to the call of the client;
score minimum: calculating a minimum value of normal scores except for a special score, which is returned to the call of the client;
grading and box distribution: calculating call returns to clients, and normally scoring call distribution under each sub-box except for special scores;
scoring population stability index: and measuring an index of the deviation between the actual sub-component and the preset sub-component.
In this embodiment, the calculation formula of the score group stability index is as follows:
wherein S is 0 Representing the preset sub-division, N represents the number of data source sub-divisions, S i Representing the actual subdivision of the ith data source subdivision.
The beneficial effects of above-mentioned design scheme are: and further analyzing the data source sub-division according to the grading type to determine the value of the data source sub-division in each dimension, thereby providing a basis for monitoring the calling condition of the user.
Example 5
Based on embodiment 1, the embodiment of the invention provides a method for monitoring a customer call service of a three-party data platform, as shown in fig. 2, in S3, dynamic monitoring display for the customer call is designed based on the grading values under each grading type, which comprises the following steps:
s301: determining a correlation between the score types based on the characteristics between the score types;
s302: and determining to call dynamic monitoring display for the client according to the grading values under the grading types by utilizing the correlation.
In this embodiment, the correlation between the score types, for example, the value of the first score type is always greater than the value of the second score type, and the value of the third score type is always between the value of the first score type and the value of the second score type.
In this embodiment, the monitoring display includes, for example, determination of the monitoring expression, arrangement distribution of the monitoring expression, and the like.
The beneficial effects of above-mentioned design scheme are: according to the relation and the value among the scoring types, the dynamic monitoring display is called for the client by reasonable design, so that the monitoring display reflects the service calling condition of the client, and the service calling condition of the client is convenient to know.
Example 6
Based on embodiment 5, the embodiment of the invention provides a method for monitoring customer call service of a three-party data platform, which determines the correlation between each scoring type based on the characteristics between each scoring type, and comprises the following steps:
determining a first relation among the scoring types according to the semantic features of the scoring types;
acquiring the acquisition modes of the score types, acquiring necessary dimension parameters from the acquisition modes, and acquiring the roles of the necessary dimension parameters in the corresponding score types;
determining a second relationship between the respective scoring types based on the role of the necessary dimension parameters in the different scoring types;
the correlation relationship is determined based on the first relationship and the second relationship.
In this embodiment, the first relationship, for example, determines that the sum of the scored query volume and the scored query volume is the scored call volume; score maximum value is greater than score minimum value, etc.
In this embodiment, the second relation, for example, the maximum value of the normal score is taken as a value of the score maximum value, and the maximum value of the normal score is taken as a value of score equipartition.
The beneficial effects of above-mentioned design scheme are: and through analysis according to the acquisition mode and semantic features of each scoring type, the comprehensiveness and the accuracy of the correlation between each scoring type are ensured to be determined, and a basis is provided for design monitoring display.
Example 7
Based on embodiment 5, the embodiment of the invention provides a method for monitoring a customer call service of a three-party data platform, which comprises the steps of:
setting an index typesetting relationship for the display of the scoring type based on the index action in the correlation;
setting score value range classification under each score type and corresponding index value standard quantity in each score type based on score value information of the historical data source sub-score under each score type;
classifying the scoring values under each scoring type based on the value ranges, and obtaining the actual number of index values under each value range;
acquiring the quantity difference between the actual quantity of the index value and the standard quantity of the index value, and establishing a fluctuation curve of the quantity difference along with the change of the value range;
determining the fluctuation amplitude between the value range and the quantity difference based on the fluctuation curve;
judging whether the fluctuation amplitude is in a preset fluctuation range or not;
if yes, determining the time display dimension of each grading type and the display sequence of the time display dimension according to the history monitoring display of the history data source sub-division;
otherwise, extracting a target value range which does not meet the preset fluctuation range from the value range, and classifying the target value range according to the grading type to obtain a target index value range and a corresponding data difference value;
based on the correlation, determining the relation characteristic of the target index value range, and based on the relation characteristic and combining the data difference value, matching time display dimensions for each grading type and the display sequence of the time display dimensions;
and designing the dynamic monitoring display for the client according to a preset chart display format based on the index typesetting relation, the time display dimension and the display sequence.
In this embodiment, the index typesetting relationship is, for example, that the correlation between the first score type and the second score type is the largest, and then the first score type and the second score type are processed in the monitoring display to be close to positions, so that the user can conveniently watch and compare the first score type and the second score type, and the satisfaction degree of the user is improved.
In this embodiment, the time display dimension may be a month, week, day, etc. dimension.
In this embodiment, according to the history monitoring display of the history data source sub-division, determining the time display dimension of each score type, and the display order of the time display dimension, specifically, the display order of the current time display dimension, is consistent with the history time display dimension and the display order in the history monitoring display.
In this embodiment, when the relationship feature of the target index value range is determined, for example, that the data difference value of the target index value range under the first score type and the data difference value of the target index value range under the second score type are in the first relationship feature, the priority order of the time display dimensions of the first score type and the second score type is determined to be month, year, and day, otherwise, the priority order of the time display dimensions of the first score type and the second score type is determined to be month, day, and year.
The beneficial effects of above-mentioned design scheme are: according to the correlation between the scoring types, firstly, the display setting index typesetting relation of the scoring types is determined, so that the viewing and comparison of users are facilitated, the satisfaction degree of the users is improved, secondly, the time display dimension and the corresponding display sequence of the scoring types are determined by combining the historical data source sub-division, and the optimal client calling dynamic monitoring display is designed from the aspects of typesetting and time dimension display, so that the monitoring key points of the client calling of the users are facilitated, the effective monitoring of the calling by the users is ensured, and the scoring values of the scoring types which do not meet the preset calling requirements and the abnormal client calling data corresponding to the scoring values are found in time.
Example 8
Based on embodiment 1, the embodiment of the invention provides a method for monitoring service call by a client of a three-party data platform, and after S3, the method further comprises the steps of evaluating and adjusting the monitoring display, specifically:
acquiring client information corresponding to the monitoring display, acquiring historical monitoring display corresponding to the client from the client information, and acquiring abnormal accident positioning information under the historical monitoring display;
based on the abnormal accident positioning information, determining abnormal accident positioning efficiency and abnormal accident positioning accuracy under the history monitoring display;
based on the abnormal accident positioning efficiency and the abnormal accident positioning accuracy, sequencing the history monitoring display according to the comprehensive display effect to obtain a history monitoring display sequence;
extracting features of each history monitoring display in the history monitoring display sequence to obtain display features, and classifying the display features according to display attributes to obtain display attribute features;
selecting optimal display attribute features from the display attribute features based on the feature differences of the display attribute features under each history monitoring display in the history monitoring display sequence and the sequence corresponding to the history monitoring display sequence;
acquiring actual display attribute characteristics of the monitoring display, and comparing the actual display attribute characteristics with optimal display attribute characteristics according to display attributes to obtain feature differences;
judging whether the monitoring display is in an acceptable difference range or not based on the characteristic difference;
if yes, determining that the monitoring display is qualified;
otherwise, determining that the monitoring display is unqualified;
after determining that the monitoring display is unqualified, extracting display attribute characteristics to be adjusted from the actual display attribute characteristics based on the characteristic difference;
and adjusting the display attribute characteristics to be adjusted based on the optimal display attribute characteristics, and obtaining the adjusted monitoring display according to the adjusted display attribute characteristics.
In this embodiment, the comprehensive effect is obtained by combining analysis of both the abnormal event localization efficiency and the abnormal event localization accuracy.
In this embodiment, the smaller the feature variance, the greater the probability that the monitoring shows in the range of acceptable variances.
In this embodiment, the display attributes include, for example, typesetting, time dimension selection, chart format, and the like.
The beneficial effects of above-mentioned design scheme are: the optimal monitoring display characteristics conforming to the client are determined according to the history monitoring display corresponding to the client and the corresponding comprehensive display effect, the current monitoring display is evaluated and adjusted according to the optimal monitoring display characteristics, the obtained monitoring display is ensured to have pertinence to the client, the client can check and analyze the monitoring display more smoothly, and therefore accuracy and efficiency of locating abnormal call data are improved.
Example 9
Based on embodiment 1, the embodiment of the invention provides a method for monitoring customer calling service of a three-party data platform, which is characterized in that in S4, based on the monitoring display, abnormal calling accidents are positioned, and the method comprises the following steps:
determining call data of the clients based on the monitoring display, classifying the call data according to a call object to obtain a plurality of groups of call sub-data, and setting attribute abnormal values for each attribute of each group of call sub-data based on the call object;
determining an object call outlier D of the call object according to the call object and the attribute outlier thereof;
wherein τ represents a call sensitive value corresponding to the call object, the value is (0.01,0.10), n represents the attribute number of the call sub-data, S i An attribute anomaly value of the ith attribute, a value of (0, 1), e represents a natural constant, a value of 2.72, G i Data amount representing i-th attribute correspondence data, G A Representing the data quantity of the calling sub-data corresponding to the calling object, and tau i The attribute sensitivity value of the i-th attribute is (0.01, 0.05);
judging whether the object calling abnormal value D is larger than a preset object abnormal value, if so, taking the calling object as a key object, otherwise, taking the calling object as a secondary object;
calculating the total call outlier R of the call data of the client based on the key object, the secondary object and the corresponding object call outlier;
wherein, gamma 1 The weight value of the key object is represented as (0.5, 1.0), and gamma 2 The weight value of the secondary object is (0.1, 0.5), m represents the number of the key objects, h represents the number of the secondary objects, and B a Representing the total data quantity corresponding to the key object, B j Represents the data quantity corresponding to the jth key object, B b Representing the total data volume corresponding to the secondary object, B r Representing the data amount corresponding to the r-th secondary object, D j Object call outlier representing jth key object, D r An object call outlier representing the r-th secondary object;
judging whether the overall calling outlier R is larger than a preset outlier;
if yes, determining that the call data of the client is abnormal, and positioning a data source corresponding to the call data based on the call data to serve as abnormal accident data;
otherwise, determining that the call data of the client is not abnormal, and not positioning abnormal accidents.
In this embodiment, the more sensitive the call of the call object, the greater the corresponding call sensitivity value, and the less likely it is that an exception will occur.
In this embodiment, the greater the contribution of the attribute to the call, the greater the corresponding attribute sensitivity value, and the greater the likelihood of an anomaly.
In this embodiment, each attribute of the calling sub-data includes, for example, a scoring calling amount, a scoring searching amount, a scoring non-searching amount, a scoring equipartition, a scoring maximum value, a scoring minimum value, a scoring box distribution and a scoring group stability index, where the abnormal values of the corresponding attributes are standardized in advance and are in the range of (0, 1).
The beneficial effects of above-mentioned design scheme are: the method mainly comprises the steps of determining call data corresponding to a client in monitoring display, classifying and analyzing the call data through a call object, comprehensively determining the object call outlier of the call object by combining the attribute outlier of the classified call sub-data under each attribute with the data quantity corresponding to each attribute and the call sensitivity value corresponding to the call object, ensuring the accuracy of the obtained object call outlier, further grouping the call object by using the object call outlier, providing a basis for calculating the overall call outlier of the call data, considering the data quantity of the call object in the process of calculating the overall call outlier of the call data, ensuring that the determination of the overall call outlier is more accurate, avoiding the situation of erroneous judgment and ensuring the accuracy of positioning abnormal accident data.
Example 10
A customer call service monitoring system of a three-party data platform, as shown in fig. 3, comprising:
the sub-division acquisition module is used for acquiring data source sub-division of the data called by the client by the three-party data platform;
the dimension value calculation module is used for calculating the grading value of the data source sub-division under each grading type;
wherein each grading type comprises grading calling quantity, grading searching quantity, grading non-searching quantity, grading average, grading maximum value, grading minimum value, grading box distribution and grading group stability index;
the monitoring display module is used for designing dynamic monitoring display for the client call based on the grading values under each grading type;
comprising the following steps: determining a correlation between the score types based on the characteristics between the score types;
determining to call dynamic monitoring display for the client according to the grading values under each grading type by utilizing the correlation;
and the abnormality positioning module is used for positioning and calling abnormal accidents based on the monitoring display.
In this embodiment, the respective score types include a score call amount, a score found amount, a score not found amount, a score average, a score maximum, a score minimum, a score bin distribution, and a score population stability index.
In this embodiment, the design of the dynamic monitoring display for the client call is specifically to design an optimal dynamic monitoring display for the client call according to the characteristics and the values of each scoring type, so as to realize effective monitoring for the client call.
In this embodiment, locating the call exception accident specifically refers to searching out the score values under each score type which does not meet the preset call requirement and the corresponding exception client call data according to the monitoring display.
The beneficial effects of above-mentioned design scheme are: according to the invention, intelligent monitoring is carried out on the client call data through the three-party data platform, so that the dilemma of shortage of internal personnel resources and insufficient capability of monitoring technicians is avoided, meanwhile, the comprehensiveness and accuracy of monitoring on the client call service are ensured by acquiring the grading values of the client call data under each grading type and acquiring the grading values of a plurality of grading types, then, the optimal client call dynamic monitoring display is designed through the grading values under each grading type, the effective monitoring on the client call is ensured, and the grading values under each grading type which do not meet the preset call requirement and the abnormal client call data corresponding to the grading values are conveniently found in time.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (9)
1. A customer call service monitoring method for a three-way data platform, comprising:
s1: collecting data source sub-division of calling data of a client by a three-party data platform;
s2: calculating the grading value of the data source sub-score under each grading type;
wherein each grading type comprises grading calling quantity, grading searching quantity, grading non-searching quantity, grading average, grading maximum value, grading minimum value, grading box distribution and grading group stability index;
s3: designing a dynamic monitoring display for the client call based on the grading values under each grading type;
comprising the following steps: determining a correlation between the score types based on the characteristics between the score types;
determining to call dynamic monitoring display for the client according to the grading values under each grading type by utilizing the correlation;
s4: and positioning and calling the abnormal accident based on the monitoring display.
2. The method for monitoring customer call services of a three-party data platform according to claim 1, further comprising, after S1:
performing traversal analysis on the data source sub-division to extract the data source sub-division with abnormality;
mapping the data source sub-division with the abnormality to obtain the normal data source sub-division.
3. The method for monitoring customer call services of a three-way data platform according to claim 1, wherein in S2, calculating the score value of the data source sub-score under each score type comprises:
dividing the data source sub-division according to different time granularities to obtain corresponding data source sub-division under different time granularities;
calculating the grading values of the corresponding data source sub-divisions under each grading type under the different time granularities;
based on the different time granularities and the corresponding values thereof, establishing a corresponding relation between the time granularities and the scoring values under each scoring type, and based on the corresponding relation, obtaining the scoring values of the data source sub-division under each scoring type under different time granularities.
4. The method for monitoring customer call services of a three-way data platform according to claim 1, wherein in S2, calculating the score value of the data source sub-score under each score type comprises:
the scoring types comprise scoring call quantity, scoring search quantity, scoring non-search quantity, scoring equipartition, scoring maximum value, scoring minimum value, scoring box distribution and scoring population stability index;
wherein, the grading call volume: dividing the call total amount, call uncorrupted amount and call obtained amount; calling a total amount calculation client mechanism to send out an application calling total amount to a data product of the three-party data platform; invoking the unsearched calculation without returning the normal score to the total amount of invocation of the client mechanism; the calculation of the call query returns the normal score to the call total amount of the client mechanism;
score equipartition: calculating a normal score average value except for a special score value, and calling and returning to the client;
score maximum: calculating the maximum value of normal scores except for special scores, which are returned to the call of the client;
score minimum: calculating a minimum value of normal scores except for a special score, which is returned to the call of the client;
grading and box distribution: calculating call returns to clients, and normally scoring call distribution under each sub-box except for special scores;
scoring population stability index: and measuring an index of the deviation between the actual sub-component and the preset sub-component.
5. The method of claim 1, wherein determining the correlation between the score types based on the characteristics between the score types comprises:
determining a first relation among the scoring types according to the semantic features of the scoring types;
acquiring the acquisition modes of the score types, acquiring necessary dimension parameters from the acquisition modes, and acquiring the roles of the necessary dimension parameters in the corresponding score types;
determining a second relationship between the respective scoring types based on the role of the necessary dimension parameters in the different scoring types;
the correlation relationship is determined based on the first relationship and the second relationship.
6. The method for monitoring the call service of a client of a three-party data platform according to claim 1, wherein determining a dynamic monitoring display for the call of the client according to the score values under the respective score types by using the correlation comprises:
setting an index typesetting relationship for the display of the scoring type based on the index action in the correlation;
setting score value range classification under each score type and corresponding index value standard quantity in each score type based on score value information of the historical data source sub-score under each score type;
classifying the scoring values under each scoring type based on the value ranges, and obtaining the actual number of index values under each value range;
acquiring the quantity difference between the actual quantity of the index value and the standard quantity of the index value, and establishing a fluctuation curve of the quantity difference along with the change of the value range;
determining the fluctuation amplitude between the value range and the quantity difference based on the fluctuation curve;
judging whether the fluctuation amplitude is in a preset fluctuation range or not;
if yes, determining the time display dimension of each grading type and the display sequence of the time display dimension according to the history monitoring display of the history data source sub-division;
otherwise, extracting a target value range which does not meet the preset fluctuation range from the value range, and classifying the target value range according to the grading type to obtain a target index value range and a corresponding data difference value;
based on the correlation, determining the relation characteristic of the target index value range, and based on the relation characteristic and combining the data difference value, matching time display dimensions for each grading type and the display sequence of the time display dimensions;
and designing the dynamic monitoring display for the client according to a preset chart display format based on the index typesetting relation, the time display dimension and the display sequence.
7. The method for monitoring customer call service of a three-party data platform according to claim 1, wherein after S3, further comprising evaluating and adjusting the monitoring display, specifically:
acquiring client information corresponding to the monitoring display, acquiring historical monitoring display corresponding to the client from the client information, and acquiring abnormal accident positioning information under the historical monitoring display;
based on the abnormal accident positioning information, determining abnormal accident positioning efficiency and abnormal accident positioning accuracy under the history monitoring display;
based on the abnormal accident positioning efficiency and the abnormal accident positioning accuracy, sequencing the history monitoring display according to the comprehensive display effect to obtain a history monitoring display sequence;
extracting features of each history monitoring display in the history monitoring display sequence to obtain display features, and classifying the display features according to display attributes to obtain display attribute features;
selecting optimal display attribute features from the display attribute features based on the feature differences of the display attribute features under each history monitoring display in the history monitoring display sequence and the sequence corresponding to the history monitoring display sequence;
acquiring actual display attribute characteristics of the monitoring display, and comparing the actual display attribute characteristics with optimal display attribute characteristics according to display attributes to obtain feature differences;
judging whether the monitoring display is in an acceptable difference range or not based on the characteristic difference;
if yes, determining that the monitoring display is qualified;
otherwise, determining that the monitoring display is unqualified;
after determining that the monitoring display is unqualified, extracting display attribute characteristics to be adjusted from the actual display attribute characteristics based on the characteristic difference;
and adjusting the display attribute characteristics to be adjusted based on the optimal display attribute characteristics, and obtaining the adjusted monitoring display according to the adjusted display attribute characteristics.
8. The method for monitoring customer call services of a three-way data platform according to claim 1, wherein in S4, locating call exceptions based on the monitoring display comprises:
determining call data of the clients based on the monitoring display, classifying the call data according to a call object to obtain a plurality of groups of call sub-data, and setting attribute abnormal values for each attribute of each group of call sub-data based on the call object;
determining an object call outlier D of the call object according to the call object and the attribute outlier thereof;
judging whether the object calling abnormal value D is larger than a preset object abnormal value, if so, taking the calling object as a key object, otherwise, taking the calling object as a secondary object;
calculating the total call outlier R of the call data of the client based on the key object, the secondary object and the corresponding object call outlier;
judging whether the overall calling outlier R is larger than a preset outlier;
if yes, determining that the call data of the client is abnormal, and positioning a data source corresponding to the call data based on the call data to serve as abnormal accident data;
otherwise, determining that the call data of the client is not abnormal, and not positioning abnormal accidents.
9. A customer call service monitoring system for a three-way data platform, comprising:
the sub-division acquisition module is used for acquiring data source sub-division of the data called by the client by the three-party data platform;
the dimension value calculation module is used for calculating the grading value of the data source sub-division under each grading type;
wherein each grading type comprises grading calling quantity, grading searching quantity, grading non-searching quantity, grading average, grading maximum value, grading minimum value, grading box distribution and grading group stability index;
the monitoring display module is used for designing dynamic monitoring display for the client call based on the grading values under each grading type;
comprising the following steps: determining a correlation between the score types based on the characteristics between the score types;
determining to call dynamic monitoring display for the client according to the grading values under each grading type by utilizing the correlation;
and the abnormality positioning module is used for positioning and calling abnormal accidents based on the monitoring display.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311462764.8A CN117669866A (en) | 2023-11-06 | 2023-11-06 | Customer call service monitoring method and system of three-party data platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311462764.8A CN117669866A (en) | 2023-11-06 | 2023-11-06 | Customer call service monitoring method and system of three-party data platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117669866A true CN117669866A (en) | 2024-03-08 |
Family
ID=90085374
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311462764.8A Pending CN117669866A (en) | 2023-11-06 | 2023-11-06 | Customer call service monitoring method and system of three-party data platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117669866A (en) |
-
2023
- 2023-11-06 CN CN202311462764.8A patent/CN117669866A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5845285A (en) | Computer system and method of data analysis | |
| US7627620B2 (en) | Data-centric automatic data mining | |
| US6574585B2 (en) | Method for improving robustness of weighted estimates in a statistical survey analysis | |
| US20020128884A1 (en) | Computer method for using sample data to predict future population and domain behaviors | |
| EP4358005A1 (en) | Information recommendation method and apparatus based on data interaction, and device and storage medium | |
| CN110472114A (en) | Abnormal data method for early warning, device, computer equipment and storage medium | |
| CN116644184B (en) | Human resource information management system based on data clustering | |
| CN112860769A (en) | Energy planning data management system | |
| CN110570097A (en) | business personnel risk identification method and device based on big data and storage medium | |
| CN115409395A (en) | Quality acceptance inspection method and system for hydraulic construction engineering | |
| CN116384889A (en) | Intelligent analysis method for information big data based on natural language processing technology | |
| CN112330227A (en) | Urban management capability evaluation method and system based on digital urban management service construction | |
| US20240144405A1 (en) | Method for information interaction, device, and storage medium | |
| CN110689211A (en) | Method and device for evaluating website service capability | |
| US20240127143A1 (en) | Method, device and storage medium for information processing based on data interaction | |
| CN107958346A (en) | The recognition methods of abnormal behaviour and device | |
| CN112950359B (en) | User identification method and device | |
| CN112862593B (en) | Credit scoring card model training method, device and system and computer storage medium | |
| Maldonado et al. | Big data and financial auditing in Portugal | |
| CN117669866A (en) | Customer call service monitoring method and system of three-party data platform | |
| Zhang et al. | Dynamic time warp-based clustering: Application of machine learning algorithms to simulation input modelling | |
| CN112506930B (en) | Data insight system based on machine learning technology | |
| CN115392351A (en) | Risk user identification method and device, electronic equipment and storage medium | |
| KR20230103025A (en) | Method, Apparatus, and System for provision of corporate credit analysis and rating information | |
| CN114493851A (en) | Risk processing method and device |
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 |