KR102298907B1 - Method And Apparatus For Monitoring Communication Service Failure - Google Patents

Abstract

The present invention generates a plurality of time objects and extracts a plurality of keywords from the first Voice of Customer (VoC) data provided from the customer management server to calculate the first number of keywords and the first extended keyword ratio for each time object and keyword and a calculator for generating a current time object by receiving the second VoC data from the customer terminal, extracting a plurality of keywords from the second VoC data, and calculating a second extended keyword ratio for each keyword in the current visual object; A threshold learning unit that receives the first number of keywords and the first extended keyword ratio and calculates and learns a threshold value for each time object and keyword based on them; Provided is a communication service failure monitoring device including a failure detection unit for receiving a second extended keyword ratio and comparing each of the threshold values in the time object corresponding to the current time object to detect the communication service failure of the customer terminal.

Description

Communication service failure monitoring device and method {Method And Apparatus For Monitoring Communication Service Failure}

The present invention relates to a communication service failure monitoring apparatus and method.

In providing communication services using a network, if a communication service failure occurs, post-processing to recover the service failure after quickly detecting the failure, analyzing the cause, and predicting the failure before the failure occurs Prevention is of the utmost importance.

On the other hand, a customer recognizes a communication service failure and receives a Voice of Customer (VoC) through the customer center, but there is a limit to the network maintenance person's ability to monitor various VoCs in real time. And, there is a limit in analyzing the VoC to identify the cause, symptom, and action of service failure.

In addition, when many VoCs are accumulated due to such a limitation, there is a problem in that customer inconvenience is prolonged.

In addition, the network maintenance manager cannot determine in which section of the customer terminal, home network, or telecommunication company network the service failure occurred, so he has to be dispatched to the site. There is a problem that a lot of money is consumed in handling service failures due to on-site dispatch.

The present invention can detect and preemptively prevent quality issues that can escalate to communication service failures, minimize customer inconvenience by providing a quality issue monitoring function to a customer terminal, and provide an automatic notification function when a quality issue occurs. An object of the present invention is to provide a communication service failure monitoring apparatus and method that can increase the operator's work efficiency by providing it.

Another object of the present invention is to provide a communication service failure monitoring apparatus and method capable of detecting quality issues that change over time by updating a threshold based on new VoC data at regular intervals.

The technical problems to be achieved in the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned above can be clearly understood by those of ordinary skill in the art to which the present invention belongs from the description below. There will be.

In order to solve the above problems, the present invention generates a plurality of time objects and extracts a plurality of keywords from the first VoC (Voice of Customer) data provided from the customer management server to extract the first keywords for each time object and keyword. Calculate the number of cases and the first extended keyword ratio, receive the second VoC data from the customer terminal, create a current time object, extract a plurality of keywords from the second VoC data, and calculate the second extended keyword ratio for each keyword in the current visual object. A calculation unit that calculates, a threshold learning unit that receives the number of first keywords and a first extended keyword ratio from the calculation unit, and calculates and learns a threshold value for each time object and keyword based on them, and receives the threshold value from the threshold learning unit , communication service failure comprising a failure detection unit receiving the second extended keyword ratio from the calculator and comparing the second extended keyword ratio with a threshold value in the time object corresponding to the current time object, respectively, to detect the communication service failure of the customer terminal monitoring device is provided.

Here, the threshold learning unit may graph the number of first keywords with respect to the first extended keyword ratio and calculate the first extended keyword ratio, which is an inflection point at which the slope of the graph changes, as the threshold value.

In addition, the failure detection unit may detect a communication service failure of the customer terminal by extracting a keyword that exceeds the threshold of the second extended keyword ratio.

In addition, the failure detection unit may determine at least one of a type, cause, symptom, and action of a communication service failure of the customer terminal by combining keywords exceeding the threshold.

In addition, it may further include a notification unit for notifying at least one of the type, cause, symptom, and action of the determined communication service failure of the customer terminal.

Also, the calculator may generate a plurality of time objects with a predetermined length of time at regular intervals for 24 hours a day.

In addition, the calculator may receive the second VoC data from the customer terminal, extract a plurality of keywords, and calculate the number of second keywords for each keyword.

In addition, the threshold learning unit may receive the second number of keywords and the second extended keyword ratio from the calculator at regular intervals to update the threshold for each time object and keyword.

In addition, the steps of extracting a plurality of keywords from the first Voice of Customer (VoC) data provided from the customer management server, generating a plurality of time objects, and the first keywords for each time object and keyword from the first VoC data Calculating the number of cases and the first extended keyword ratio, receiving the first number of keyword cases and the first extended keyword ratio and calculating and learning thresholds for each time object and keyword based on them, and second VoC data from the customer terminal generating a current time object by receiving It provides a communication service failure monitoring method comprising the step of detecting a communication service failure of the customer terminal by comparing each with a threshold value in the time object corresponding to the current time object.

Here, the step of learning the threshold is a step of calculating the first extended keyword ratio, which is an inflection point at which the slope of the graph changes, as a threshold by graphing the number of first keywords with respect to the first extended keyword ratio.

In addition, the step of detecting the communication service failure is a step of detecting the communication service failure of the customer terminal by extracting a keyword exceeding the threshold of the second extended keyword ratio.

In addition, the method may further include updating the threshold for each time object and keyword by receiving the second number of keywords and the second extended keyword ratio at regular intervals.

According to the present invention, it is possible to detect and preemptively prevent quality issues that can escalate to communication service failures, to minimize customer inconvenience by providing a quality issue monitoring function to a customer terminal, and to automatically notify when a quality issue occurs By providing this, it is possible to increase the work efficiency of the operator.

In addition, according to the present invention, by updating the threshold based on new VoC data at regular intervals, quality issues that change over time can also be detected.

The effects obtainable in the present invention are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those of ordinary skill in the art to which the present invention belongs from the following description. will be.

1 is a block diagram of a communication service failure monitoring system according to an embodiment of the present invention.
2 is a block diagram of a communication service failure monitoring apparatus according to an embodiment of the present invention.
3 is a detailed block diagram of the calculator of FIG. 2 .
4 is a diagram illustrating a threshold value learned by the threshold learning unit of FIG. 2 for each time object and keyword.
FIG. 5 is a graph used to calculate a threshold in the threshold learning unit of FIG. 2 .
6 is a first table constituting VoC data.
7 is a second table in which the number of keywords, the number of VoCs, and the ratio of extended keywords are reconstructed by time and keyword.
8 is a third table in which threshold values are configured for each time object and keyword.
9 is a flowchart of a method for learning a threshold value in a communication service failure monitoring method according to an embodiment of the present invention.
10 is a flowchart of a method for detecting a communication service failure of a customer terminal in a communication service failure monitoring method according to an embodiment of the present invention.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. DETAILED DESCRIPTION The detailed description set forth below in conjunction with the appended drawings is intended to describe exemplary embodiments of the present invention and is not intended to represent the only embodiments in which the present invention may be practiced. In order to clearly explain the present invention in the drawings, parts not related to the description may be omitted, and the same reference numerals may be used for the same or similar components throughout the specification.

In an embodiment of the present invention, expressions such as “or” and “at least one” may indicate one of the words listed together, or a combination of two or more. For example, “A or B” or “at least one of A and B” may include only one of A or B, or both A and B.

1 is a block diagram of a communication service failure monitoring system according to an embodiment of the present invention.

Referring to FIG. 1 , the communication service failure monitoring system may include a customer management server 10 , a customer terminal 20 , and a communication service failure monitoring apparatus 100 .

Here, the customer management server 10 , the customer terminal 20 , and the communication service failure monitoring apparatus 100 exemplarily illustrate components that can be controlled by the communication service failure monitoring system.

Each component of the communication service failure monitoring system is generally connected through a network. For example, the communication service failure monitoring apparatus 100 may be connected to the customer management server 10 and the customer terminal 20 at the same time or at a time interval.

Here, the network refers to a connection structure capable of exchanging information between each node, such as terminals and servers, and includes a local area network (LAN), a wide area network (WAN), and the Internet (WWW). : World Wide Web), wired and wireless data networks, telephone networks, and wired and wireless television networks. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi, Bluetooth communication, infrared communication, ultrasound Communication, Visible Light Communication (VLC), and LiFi (LiFi), and the like are included, but are not limited thereto.

The customer management server 10 may store the first Voice of Customer (VoC) data previously received by the communication terminal 20 and transmit it to the communication service failure monitoring apparatus 100 .

The customer terminal 20 transmits second Voice of Customer (VoC) data including at least one of failure information and quality information to the customer management server 20 and the communication service failure monitoring device 100 to be diagnosed with a communication service failure. can be transmitted

Here, the first and second VoC data may be failure report data regarding a failure or quality occurring in connection with a service provided to a customer through the customer terminal 20 .

2 is a block diagram of a communication service failure monitoring apparatus according to an embodiment of the present invention, FIG. 3 is a detailed block diagram of the calculation unit of FIG. 2, and FIG. 4 is a time object and a threshold value for learning by the threshold learning unit of FIG. It is a diagram showing each keyword, and FIG. 5 is a graph used to calculate a threshold in the threshold learning unit of FIG. 2 .

As shown in FIG. 2 , the communication service failure monitoring apparatus 100 according to an embodiment of the present invention may include a calculation unit 110 , a threshold learning unit 120 , and a failure detection unit 130 .

As shown in FIG. 3 , the calculator 110 may include a keyword extractor 111 , a time object generator 112 , and an extended keyword ratio calculator 113 .

Here, the keyword extraction unit 111 extracts a plurality of keywords related to communication service failure or quality from the first Voice of Customer (VoC) data provided from the customer management server 10 .

Meanwhile, although not shown in the drawings, the calculator 110 may further include a preprocessor (not shown) for removing stopwords that may be included in the keyword extracted by the keyword extractor 111 . Here, the stop word means a word that is not used, and specifically means a word that has no meaning, such as an article, a preposition, a proposition, and a conjunction.

The time object generating unit 112 generates a plurality of time objects with a predetermined length of time at predetermined intervals for 24 hours a day. For example, a total of 288 time objects can be created with a length of 1 hour every 5 minutes.

The extended keyword ratio calculator 113 calculates the first number of keywords and the first keyword ratio for each time object and keyword. Here, the first keyword ratio may be defined as the number of keywords to the number of VoCs.

The extended keyword ratio calculator 113 calculates the first extended keyword ratio based on the number of first keywords and the number of VoCs for each time object and keyword. That is, the first extended keyword ratio may be calculated by Equation 1 below.

<Equation 1>

×ln(VoC 건수)extended keyword ratio=

×ln(Number of VoCs)

where ln is a natural logarithmic function.

Meanwhile, the calculator 110 may calculate the number of second keywords for each keyword by extracting a plurality of keywords from the second VoC data currently provided from the customer terminal 20 in the same manner as the number of first keywords described above.

Also, the calculator 110 generates a current time object corresponding to the time at which the second VoC data is provided, and extracts a plurality of keywords from the second VoC data in the same manner as in the above-described first extended keyword ratio to extract the current time. A second extended keyword ratio may be calculated for each keyword in the object.

Referring to FIG. 4 , the threshold learning unit 120 receives the first number of keywords and the first extended keyword ratio from the calculation unit 110 , and calculates and learns thresholds for each visual object and keyword based on them.

Specifically, as shown in FIG. 5 , the threshold learning unit 120 graphs the number of first keywords with respect to the first extended keyword ratio for each visual object and keyword, and the first extended keyword is an inflection point at which the slope of the graph changes. The ratio can be calculated as a critical value. Such a threshold may be different for each visual object and keyword.

Here, it can be seen that the slopes of the left and right graphs are different based on the inflection point (θ). Based on this inflection point (θ), the left side can be determined as normal and the right side as abnormal. It can be set as a threshold that is a criterion for determining communication service failure.

Meanwhile, the reason for using the first extended keyword ratio instead of the first keyword ratio in calculating the threshold is to remove the influence of the number of VoCs that vary for each keyword on the threshold.

Specifically, when the threshold is calculated using the first keyword ratio, the number of first keywords may be underestimated in a relatively large number of VoCs, and the number of first keywords may be overestimated in a relatively small number of VoCs. In order to remove the influence of the number of cases on the threshold, a threshold point is calculated using the first extended keyword ratio.

The failure detecting unit 130 receives the threshold value from the threshold learning unit 120 and the second extended keyword ratio from the calculator 110 , and sets the second extended keyword ratio as a threshold value in the time object corresponding to the current time object. Detects a communication service failure of the customer terminal 20 by comparing with each.

Specifically, the failure detection unit 130 may detect a communication service failure of the customer terminal 20 by extracting a keyword exceeding the threshold value with the second extended keyword ratio, and combining keywords exceeding the threshold value for the customer terminal 20 ) can determine at least one of the type, cause, symptom, and action of a communication service failure.

Such communication service failures may include data access impossible, video call connection impossible after firmware update in a specific terminal, long-term reboot failure, critical error code generation, quality threshold exceeded, and the like.

In addition, the communication service failure monitoring apparatus 100 according to the embodiment of the present invention notifies at least one of the type, cause, symptom and action of the communication service failure of the customer terminal 20 determined by the failure detection unit 130 . It may further include a notification unit (not shown).

The threshold learning unit 120 may receive the second number of keywords and the second extended keyword ratio from the calculator 110 at regular intervals to update the threshold for each visual object and keyword.

In this way, the communication service failure monitoring apparatus 100 according to the embodiment of the present invention can detect and preemptively prevent quality issues that may be escalated to communication service failures, and provide a quality issue monitoring function to the customer terminal 20 . By providing this, customer inconvenience can be minimized and the operator's work efficiency can be increased by providing an automatic notification function when a quality issue occurs. In addition, by updating the threshold based on new VoC data at regular intervals, quality issues that change over time can also be detected.

6 is a first table constituting VoC data, FIG. 7 is a second table in which the number of keywords, VoC cases, and extended keyword ratio are reconstructed by time and keyword, and FIG. 8 is a third table in which thresholds are configured for each time object and keyword It is a table.

Referring to FIG. 6 , the first table may include a time field, a VoC identification number field, a model name field, a subdivision field, a keyword field, and a keyword number field.

Here, by using the first table, it is possible to check the customer terminal that has received the VoC, the reception time of the VoC, the model name of the terminal that has received the VoC, the subcategory, the keywords of the VoC, and the number of cases.

Such a first table may be stored in the customer management server 10 and provided to the communication service failure monitoring apparatus 100 or may be generated in the customer terminal 20 and provided to the communication service failure monitoring apparatus 100 .

Referring to FIG. 7 , the second table may include a time field, a keyword field, a number of keywords field, a VoC number field, and an extended keyword ratio field.

Here, by using the second table, the number of keywords, the number of VoCs, and the ratio of extended keywords can be checked by time and keyword.

Such a second table may be reconfigured using the first table provided from the customer management server 10 or the customer terminal 20 in the communication service failure monitoring apparatus 100 or an external server.

Referring to FIG. 8 , the third table may include a time object field, a keyword field, a category field, and a threshold field.

Here, by using the third table, it is possible to detect a communication service failure of the customer terminal 20 by extracting a keyword exceeding the threshold from the current visual object.

Such a third table may be stored by the threshold learning unit 120 and provided to the failure detection unit 130 .

9 is a flowchart of a method for learning a threshold in a communication service failure monitoring method according to an embodiment of the present invention, and FIG. 10 is a method for detecting a communication service failure of a customer terminal in a communication service failure monitoring method according to an embodiment of the present invention. It is a flow chart of the method.

Hereinafter, a communication service failure monitoring method according to an embodiment of the present invention will be described with reference to FIGS. 1 to 10, but the description of the same content as the communication service failure monitoring apparatus according to the embodiment of the present invention will be omitted.

As shown in FIG. 9 , the method for learning the threshold includes the steps of extracting a plurality of keywords from the first VoC data (S11), generating a plurality of time objects (S12), and first for each time object and keyword. It may include calculating the number of keywords and the first extended keyword ratio (S13) and calculating and learning a threshold value for each time object and keyword (S14).

In the step of extracting a plurality of keywords from the first VoC data ( S11 ), a plurality of keywords related to communication service failure or quality is extracted from the first Voice of Customer (VoC) data provided from the customer management server 10 .

In the step of generating a plurality of time objects (S12), a plurality of time objects are created with a certain length of time at regular intervals for 24 hours a day. For example, a total of 288 time objects can be created with a length of 1 hour every 5 minutes.

In the step of calculating the first number of keywords and the first extended keyword ratio for each time object and keyword (S13), first, the first number of keywords and the first keyword ratio for each time object and keyword are calculated. Here, the first keyword ratio may be defined as the number of keywords to the number of VoCs. Next, a first extended keyword ratio is calculated based on the number of first keywords and the number of VoCs for each time object and keyword. That is, the first extended keyword ratio may be calculated by Equation 1 above.

In the step of calculating and learning the threshold value for each time object and keyword ( S14 ), the threshold value is calculated and learned for each visual object and keyword based on the first number of keywords and the first extended keyword ratio.

As shown in FIG. 10 , the method of detecting a communication service failure of a customer terminal includes the steps of receiving second VoC data and generating a current time object ( S21 ), and extracting a plurality of keywords from the second VoC data. (S22), calculating the number of second keywords and the second extended keyword ratio for each keyword in the current visual object (S23), comparing the second extended keyword ratio with a threshold value corresponding to the current time object (S24), and the second It may include extracting a keyword with a ratio of two extended keywords exceeding a threshold (S25).

In the step S21 of receiving the second VoC data and generating the current time object, a current time object corresponding to the time at which the second VoC data is provided is generated,

In the step of extracting a plurality of keywords from the second VoC data ( S22 ), a plurality of keywords are extracted from the second VoC data provided from the customer terminal 20 in the same manner as the number of first keywords described above, and second keywords for each keyword are extracted. number can be calculated.

In the step S23 of calculating the number of second keywords and the second extended keyword ratio for each keyword in the current visual object, the second extended keyword ratio can be calculated for each keyword in the current time object in the same way as the above-described first extended keyword ratio. have.

In the step of comparing the second extended keyword ratio with a threshold value corresponding to the current time object (S24), the communication service of the customer terminal 20 by comparing the second extended keyword ratio with a threshold value in the time object corresponding to the current time object, respectively detect an obstacle

In the step of extracting the keyword with the second extended keyword ratio exceeding the threshold value (S25), the communication service failure of the customer terminal 20 can be detected by extracting the keyword with the second extended keyword ratio exceeding the threshold value, and the threshold value is exceeded. By combining the keywords, it is possible to determine at least one of a type, cause, symptom, and action of a communication service failure of the customer terminal 20 .

In addition, the communication service failure monitoring method according to an embodiment of the present invention may further include the step of notifying at least one of the type, cause, symptom, and action of the determined communication service failure of the customer terminal 20 .

In addition, the communication service failure monitoring method according to an embodiment of the present invention may receive the second number of keywords and the second extended keyword ratio at regular intervals to update the threshold for each visual object and keyword.

As such, the communication service failure monitoring method according to an embodiment of the present invention can detect and preemptively prevent quality issues that can be extended to communication service failures, and by providing a quality issue monitoring function to the customer terminal 20 , Customer inconvenience can be minimized and operator's work efficiency can be increased by providing an automatic notification function when a quality issue occurs. In addition, by updating the threshold based on new VoC data at regular intervals, quality issues that change over time can also be detected.

Embodiments of the present invention disclosed in the present specification and drawings are merely provided for specific examples in order to easily explain the technical contents of the present invention and help the understanding of the present invention, and are not intended to limit the scope of the present invention. Therefore, the scope of the present invention should be construed as including all changes or modifications derived based on the technical spirit of the present invention in addition to the embodiments disclosed herein are included in the scope of the present invention.

10: Customer Management Server
20: customer terminal
100: communication service failure monitoring device
110: output unit
120: threshold learning unit
130: failure detection unit

Claims (13)

A plurality of time objects are generated and a plurality of keywords are extracted from the first Voice of Customer (VoC) data provided from the customer management server, and the number of first keywords and a first extended keyword ratio are calculated for each time object and keyword, and the customer a calculator for generating a current time object by receiving second VoC data from the terminal, extracting a plurality of keywords from the second VoC data, and calculating a second extended keyword ratio for each keyword in the current time object;
a threshold learning unit that receives the number of first keywords and a first extended keyword ratio from the calculation unit, and calculates and learns a threshold value for each time object and keyword based on them; and
The threshold value is provided from the threshold learning unit, the second extended keyword ratio is provided from the calculator, and the second extended keyword ratio is compared with the threshold value in the time object corresponding to the current time object. Comprising a failure detection unit for detecting a communication service failure of the customer terminal,
The threshold learning unit
Graphing the number of first keywords with respect to the first extended keyword ratio to calculate the first extended keyword ratio, which is an inflection point at which the slope of the graph changes, as the threshold value
Communication service failure monitoring device.
delete The method of claim 1,
The failure detection unit
Detecting a communication service failure of the customer terminal by extracting a keyword that the second extended keyword ratio exceeds the threshold
Communication service failure monitoring device.
4. The method of claim 3,
The failure detection unit
Combining keywords exceeding the threshold to determine at least one of a type, cause, symptom, and action of a communication service failure of the customer terminal
Communication service failure monitoring device.
5. The method of claim 4,
Further comprising a notification unit for informing at least one of the determined type, cause, symptom, and action of the communication service failure of the customer terminal
Communication service failure monitoring device.
The method of claim 1,
The first and second extended keyword ratios are
Calculated by the formula below
<Equation>
extended keyword ratio=

×ln(Number of VoCs)
Communication service failure monitoring device.
The method of claim 1,
the calculation unit
Creating the plurality of time objects with a certain length of time at regular intervals for 24 hours a day
Communication service failure monitoring device.
The method of claim 1,
the calculation unit
receiving the second VoC data from the customer terminal, extracting a plurality of keywords, and calculating the number of second keywords for each keyword
Communication service failure monitoring device.
9. The method of claim 8,
The threshold learning unit
receiving the second number of keywords and the second extended keyword ratio from the calculator at regular intervals to update the threshold value for each time object and keyword
Communication service failure monitoring device.
extracting a plurality of keywords from the first Voice of Customer (VoC) data provided from the customer management server;
creating a plurality of time objects;
calculating a first number of keywords and a first extended keyword ratio for each time object and keyword in the first VoC data;
receiving the first number of keywords and a first extended keyword ratio, and calculating and learning a threshold value for each time object and keyword based on them;
generating a current time object by receiving second VoC data from a customer terminal;
extracting a plurality of keywords from the second VoC data and calculating a second keyword count and a second extended keyword ratio for each keyword in the current time object; and
Comprising the step of detecting the communication service failure of the customer terminal by comparing the second extended keyword ratio with the threshold value in the time object corresponding to the current time object, respectively,
The step of learning the threshold is
Graphing the number of first keywords with respect to the first extended keyword ratio and calculating the first extended keyword ratio, which is an inflection point at which the slope of the graph changes, as the threshold value
Communication service failure monitoring method.
delete 11. The method of claim 10,
The step of detecting the communication service failure is
The step of detecting a communication service failure of the customer terminal by extracting a keyword that the second extended keyword ratio exceeds the threshold
Communication service failure monitoring method.
11. The method of claim 10,
receiving the second number of keywords and the second extended keyword ratio at regular intervals and updating the threshold for each time object and keyword
Communication service failure monitoring method further comprising.

Images