JP5189354B2 - Operator reception monitoring and switching system - Google Patents

Description

  The present invention monitors the reception work of the primary operator who responds first and reduces the stress received from customer utterances in a call center where the system of the operator who responds to the customer's telephone is composed of multiple stages of primary reception and secondary reception. In addition, the present invention relates to an operator reception monitoring / switching system that switches a call reception task whose stress state is above a certain level to a secondary operator.

  Operators who accept calls at a call center are required to respond so that customers can achieve their objectives and get satisfaction. However, there are various situations that occur during answering a call, and there may be situations where the operator has difficulty in answering calls.

  In particular, the operator may be greatly stressed depending on the utterance content and the way of speaking of the customer who has been accepted. For this reason, it is necessary to assist the reception work by reducing the stress received by the operator as much as possible.

Patent Document 1 describes a system that assists an operator who is hesitant to answer a telephone call with a customer in order to reduce stress when the operator receives the customer. The system of Patent Document 1 evaluates the operator's confusion by applying emotion analysis technology from the transmitted voice when the operator talks to the customer, and when the evaluation result satisfies a predetermined condition, The call voice is used for listening to the operator supporter.
JP 2007-4000

  The present invention monitors the reception of the primary operator in a call center that operates the operator reception system in a plurality of stages, reduces the stress received from the utterance content of the customer and how to speak, and further from the stressed primary operator. By taking over the response to the secondary operator, it supports the smooth operation of the operator reception work.

  When the reception by the operator is configured in multiple stages, the timing of switching the call reception from the primary operator to the secondary operator even if the level of stress received increases due to customer utterances and the reception work cannot be performed as usual. In many cases, it is left to the subjective judgment of the primary operator itself or the judgment based on the rule of thumb of the secondary operator or the manager.

  For example, when the primary operator felt that it was difficult to respond, he / she signaled himself / herself and requested switching of reception. Alternatively, the secondary operator or operator manager monitors the response and elapsed time of the primary operator, and prompts to switch the reception when it is determined that it is difficult to continue the response.

  For this reason, there was no relation between the strength of stress received from customers and the timing of reception switching, and there was variation in ensuring objectivity regarding the effect of reducing the amount of stress received. In particular, if the timing of reception switching is too late, the following problems have occurred.

  ・ Primary operators feel strong stress, making it difficult to respond with normal feelings. Eventually, the operator's operating rate declines and the employee turnover rate rises, increasing call center operating costs.

  -As the situation where it is difficult for the primary operator to respond normally is prolonged, the customer becomes dissatisfied with the response of the primary operator, and the customer's satisfaction is likely to be lowered.

  • If the stress of the primary operator is not noticed, it will hinder smooth business operation in the entire call center, and the quality and efficiency of work will be reduced.

  The present invention has been made in order to solve the above-described problems. The purpose of the present invention is to monitor the reception of the primary operator and visualize the stress state of the primary operator received from the customer's utterance. It is intended to provide a system that can respond to customer utterances so as not to be stressed as much as possible, and can promptly switch the response to a secondary operator when the stress state of the primary operator exceeds a certain level.

  In the present invention, a call center control system including a front terminal used by a primary operator and a back terminal used by a secondary operator is set, the primary operator accepts and responds to a customer call, and the secondary operator performs the call. In the call center that takes over and responds, an operator reception monitoring / switching system that monitors a reception state of the primary operator and switches reception of the call from the front terminal to the back terminal based on a switching request.

  This system includes 1) a dangerous phrase storage unit that stores dangerous phrases that may cause stress to the operator, and 2) conditions for determining the output of a switching request for switching the reception of a customer's call from the primary operator to the secondary operator. A determination condition storage unit that stores a determination condition based on the number of times of detection of the dangerous phrase, 3) an utterance voice information acquisition unit that acquires utterance voice data of a customer of a call accepted by the primary operator, and 4) the utterance. A dangerous word / phrase detection unit for detecting the dangerous word / phrase from the voice data; and 5) extracting partial voice data corresponding to the dangerous word / phrase from the utterance voice data and performing a predetermined conversion process on the partial voice data. , A dangerous phrase corresponding unit that outputs the converted partial voice data, and 6) when the dangerous phrase is detected a predetermined number of times or more within a predetermined time. , A status display processing unit for displaying a monitoring classification according to the stress passive state, and 7) switching the call accepted at the front terminal to the back terminal when the number of times the dangerous phrase is detected satisfies the determination condition. An acceptance switching determination unit for outputting a request; 8) an acceptance switching unit for accepting the switching request and performing an acceptance switching for switching the acceptance of the call to the back terminal; and 9) an answer to the call generated at the front terminal. And a reception information takeover unit that transmits the reception information regarding to the back terminal.

  The present invention operates as follows.

  In this system, the utterance voice information acquisition unit acquires the utterance voice data of the customer of the call accepted by the primary operator, and the dangerous phrase detection unit detects the dangerous phrase from the utterance voice information data.

  The dangerous phrase correspondence unit extracts partial voice data corresponding to the dangerous phrase from the utterance voice information data, performs a predetermined conversion process on the partial voice data, and outputs the converted partial voice data at the front terminal. To do.

  Here, the dangerous word / phrase handling unit converts, as the conversion process, a process of reducing the volume of the partial audio data, a process of modulating the partial audio data to a predetermined tone, a process of synthesizing the partial audio data with a predetermined sound source, or a partial A function of processing for converting voice data into text data is provided, and any one or a plurality of the conversion processing is performed.

  Further, the status display processing unit displays a monitoring classification according to the stress passive state of the primary operator when a dangerous word or phrase is detected a predetermined number of times or more within a predetermined time.

  In addition, the reception switching determination unit outputs a switching request for switching the call being received at the front terminal to the back terminal when the number of times of detection of the dangerous phrase satisfies the determination condition. The reception switching unit receives the switching request, performs reception switching to switch the reception of the call to the back terminal, and the reception information transfer unit transmits the reception information regarding the reception of the call generated at the front terminal to the back terminal.

  In this way, in this system, when the customer's remark includes an expression (dangerous phrase) that may cause the primary operator to be stressed, this dangerous phrase part is detected, and the voice data of the detected part is included. A predetermined conversion process can be performed and output. As a result, it is possible to reduce the stress experienced by the operator by not placing the primary operator in a situation where the customer directly hears comments such as “sue” and “dead”.

  Further, the dangerous word / phrase handling unit can set a conversion processing function to be executed based on an instruction input from the primary operator from a plurality of conversion processing functions provided in advance. As a result, it is possible to cause the primary operator to select a method for responding to the dangerous phrase, and to perform a response suitable for each operator.

  In addition, in this system, the monitoring classification according to the stress passive state that the primary operator is receiving can be displayed, and the state of the stress that the operator is receiving and the necessary monitoring can be visualized.

  “Normal continuity monitoring”, “warning monitoring”, “switching”, and “desired continuation monitoring” are set as examples of primary operator monitoring according to the stress passive state.

  “Normal continuous monitoring” is a monitoring for the situation where the operator is responding as usual without stress, and “warning monitoring” is considered that the operator is stressed by the customer's utterance. This is a monitoring for a state that needs to be monitored.

  “Switching” is a monitoring for a state where the operator is considered to have received a considerable stress from the customer's utterance and the reception switching to the secondary operator is necessary.

  “Wanted continuation monitoring” is a monitoring for a state in which a response is continued in response to a primary operator's request, although it is a stress passive state corresponding to reception switching. These categories are displayed on the display screen in a predetermined manner such as “blue display”, “yellow display”, “red display”, and “red flashing display”.

  As a result, the operator manager and the secondary operator can visually grasp the classification of the monitoring performed on the primary operator.

  Also, in this system, when the stress received by the operator becomes a condition that satisfies the predetermined judgment condition, it is judged that it is difficult to continue the reception and the acceptance switching is automatically executed, causing a problem. It is possible to promptly take over the response of the primary operator considered to be the secondary operator. As a result, it is possible to objectively determine whether or not reception switching is necessary for the reception of the primary operator, and it is possible to perform reception switching without delay.

  Furthermore, the system according to the present invention includes: 10) a standard voice pitch acquisition unit that acquires a voice pitch in a customer's standard utterance from the utterance voice data and sets it as a standard voice pitch; and 11) a customer's voice pitch from the utterance voice data. And a voice pitch fluctuation amount detection unit for calculating a fluctuation amount of the detected voice pitch with respect to the standard voice pitch.

  And in the said structure of this system, a standard audio | voice pitch acquisition part acquires the audio | voice pitch in a customer's standard utterance from utterance audio | voice data, and makes it a standard audio | voice pitch. The voice pitch fluctuation amount detection unit detects the voice pitch of the customer from the speech voice data, and calculates the fluctuation amount of the detected voice pitch with respect to the standard voice pitch.

  Then, the state display processing unit displays a predetermined section indicating the stress passive state of the primary operator when the fluctuation amount fluctuates beyond a predetermined value. In addition, the reception switching determination unit can output a switching request for switching the call being accepted at the front terminal to the back terminal when the variation amount satisfies the determination condition.

  In general, when the voice pitch of a customer's utterance changes more than a certain amount compared to normal times, it can be considered that the customer is in an emotional state, and the customer's emotional utterance is given to the operator. There is a risk of stress.

  Therefore, the fluctuation amount of the voice pitch is detected from the utterance voice data of the customer, and the state of the stress received by the operator is determined based on the number of times the fluctuation amount is detected. By using a combination of detection of the customer's dangerous phrase and voice pitch fluctuation amount, it is possible to perform detailed monitoring of the stress passive state and switching of reception.

  Furthermore, in the system according to the present invention, the acceptance switching processing unit can inquire about the consent of acceptance switching with the primary operator and can perform acceptance switching to the back terminal when acceptance of acceptance switching is accepted.

  In general, it is desirable for a single operator to complete the response. In addition, the operator may want to challenge the continuation of response by his / her will and improve his / her skills. Therefore, even when it is determined that the state corresponds to the reception switching, the primary operator is requested to agree to the reception switching, and the primary operator can select to continue the reception.

  Furthermore, the system according to the present invention can store determination conditions used by the acceptance switching determination unit for each primary operator. This makes it possible to flexibly perform the determination process of the stress passive state and switching according to the skill level of the primary operator.

  In addition, the system according to the present invention, when a switching request is output, detects the number of detections of the amount of change in the dangerous phrase or voice pitch detected from the customer utterance voice data received by the primary operator related to the output of the switching request. The determination condition update part which updates determination conditions using is provided.

  In addition, the determination condition update unit can update the determination condition of the primary operator with a predetermined weight when the primary operator completes the reception of the call without agreeing to the reception switching.

  Thereby, the determination process of the stress passive state or the reception switching can be performed using the determination condition reflecting the skill level of each primary operator.

  Further, in the system according to the present invention, the reception state display processing unit associates the stress passive states of the primary operator in time series with each other, or associates the stress passive states of a plurality of primary operators with the arrangement of the front terminal. Any one of the display screens shown can be displayed.

  Thereby, focusing on a specific primary operator, it is possible to confirm the transition of the category associated with the stress passive state, or to list the category at a certain point in time for the entire primary operator or the group unit.

  The system according to the present invention further includes an extracted dangerous phrase storage unit that accumulates dangerous phrases detected from customer utterance voice data, and a dangerous phrase statistical processing unit that performs predetermined statistical processing on the accumulated dangerous phrases.

  As a result, statistical information on the dangerous phrases that the customer actually uttered can be obtained, and it can be used as a material for the operator's educational plan planning in accordance with the actual situation.

  This system can be implemented by computer hardware resources and processing means implemented as software programs. Further, the software program can be stored in an appropriate recording medium such as a portable medium readable by a computer, a semiconductor memory, and a hard disk, and is provided by being recorded on these recording media, or has a communication interface. It is provided by transmission and reception using various communication networks.

  In a call center, a situation in which an operator is stressed by a customer's emotional or violent remarks not only directly harms the operator's mental health, but also worsens the work environment by increasing the turnover rate. In addition, the quality and efficiency of call center reception work will be reduced, and the impression on customers will also deteriorate.

  According to the present invention, when a dangerous word / phrase is included in a customer's statement, the dangerous word / phrase is converted so that the operator can cope with the stress without being stressed, so that the operator's stress can be reduced. .

  In addition, a stressed primary operator can be detected at an early stage, and the primary operator's stress can be reduced in order to cope with the changeover (escalation) from the primary operator to the secondary operator.

  Therefore, the work efficiency and turnover rate of the primary operator can be improved.

  In addition, the stress passive state of the operator and necessary monitoring can be visually confirmed, and further, the timing of escalation can be determined objectively. Therefore, escalation automation and efficiency can be achieved.

  In addition, since the escalation to the secondary operator can be performed at an early stage, it is possible to prevent a decrease in customer satisfaction.

  Furthermore, according to the present invention, it is possible to escalate with the consent of the primary operator and leave an opportunity for the operator to deal with himself / herself. As a result, the operator can be improved in skill.

  Further, according to the present invention, the operation can be started at the individual setting level of the individual primary operator, and the operation corresponding to the skill of the operator can be performed.

  Further, according to the present invention, it is possible to obtain statistical information of dangerous phrases that a customer actually speaks, and it can be used for making an operator's education plan according to the actual situation.

  Hereinafter, the best mode for carrying out the present invention will be described.

  The system and the apparatus constituting the system according to the present invention are incorporated in a call center control system installed in a call center or separately configured.

  In the call center in which the present invention is implemented, it is assumed that an inquiry about a method of operating a personal computer, a defect, a repair or the like is received from a customer and responded, and a call is received in a primary system and a secondary system.

  The primary operator that first receives a customer call and the secondary operator that takes over the response of the primary operator are configured at a ratio of 10 to 1, for example. The secondary operator has a longer work experience than the primary operator and covers the situation where the primary operator feels difficult to respond.

  FIG. 1 shows a configuration example of the present invention and a device configuration example of a call center in which the present invention is implemented.

  A call center control system 1 installed in a call center includes a front terminal 2, a front reception control device 3, a back terminal 4, a back reception control device 5, and a CTI (Computer Telephony Integration) device 6.

  Each of the front terminal 2 and the back terminal 4 includes a headphone, a microphone, and a display unit, is connected to the CTI device 6, and accepts a customer call via the telephone network 7. The front terminal 2 is used by the primary operator, and the back terminal 4 is used by the secondary operator.

  An operator reception monitoring / switching system 10 according to the present invention includes a determination condition storage device 11 and a display device 12.

  FIG. 2 shows a configuration example of the operator reception monitoring / switching system 10 according to the present invention.

  The operator reception monitoring / switching system 10 is a computer composed of a CPU and a memory, and is composed of a software program or the like, and includes an utterance voice information acquisition unit 151, a voice pitch variation detection unit 152, a dangerous phrase detection unit 153, and a dangerous phrase correspondence. Unit 154, reception switching determination unit 155, reception continuation notification unit 156, status display processing unit 171, reception switching unit 172, reception information takeover unit 173, determination condition update unit 174, dangerous phrase statistical processing unit 175, extracted dangerous phrase storage unit 176.

  The configuration example shown in FIG. 2 shows an example in which the operator reception monitoring / switching system 10 is configured as two devices, that is, a front device 15 and a back device 17. In this case, each of the front side device 15 and the back side device 17 is provided with communication units 157 and 177 to communicate necessary data.

  The determination condition storage device 11 of the operator reception monitoring / switching system 10 stores a determination condition for reception switching determination processing, a threshold value for detecting a stress passive state (the number of detections of a dangerous phrase or voice pitch variation), and the like.

  The determination condition storage device 11 includes a dangerous phrase storage unit 111 that stores dangerous phrases.

  The dangerous phrase in the dangerous phrase storage unit 111 is an emotional or violent expression (phrase) that may cause stress to the operator.

  In the configuration shown in FIG. 2, the front-side processing (front-side device 15) and the back-side processing (back-side device 17) are separately configured. However, these two devices may be configured integrally.

  FIG. 3 shows an example of dangerous phrase levels. The dangerous words and phrases shown in FIG. 3 are examples in the case of being divided into two levels according to the intensity of stress applied to the operator.

  The dangerous phrase level 1 is an expression that feels anxiety when it is spoken and causes stress, but it is an expression that can be accepted by an operator if it is spoken several times.

  The dangerous phrase level 2 is a utterance that feels a lot of anxiety when it is uttered, and even if it is a single utterance, it is expressed so as to cause a stress that exceeds the level that the operator can tolerate.

  FIG. 4 shows examples of determination conditions.

  The determination condition is a determination condition used in a process for specifying a monitoring category based on an operator's stress passive state or a process for determining acceptance switching from a primary operator to a secondary operator. Here, it is assumed that determination conditions for each operator associated with each primary operator are stored.

For example, the following conditions are set as determination conditions.
“Dangerous phrase level 1 = 3 times / 10 minutes;
Dangerous phrase level 2 = 1 or more times / 10 minutes;
Voice pitch threshold th1 = 2 times or more / 10 minutes;
Voice pitch threshold th2 = 1 or more / 10 minutes;
Timeout = 10 minutes or more per customer "
For example, “dangerous phrase level 1 = 3 times / 10 minutes” means that a dangerous phrase classified into dangerous phrase level 1 is detected from customer's speech data within a predetermined time (10 minutes), and the number of detections is Indicates that a request for switching acceptance is output when the number of times is three or more.

  “Voice pitch threshold th1 = 2 times / 10 minutes” means that the fluctuation amount from the standard voice pitch of the voice pitch extracted from the customer's utterance voice data within a predetermined time (10 minutes) is equal to or greater than the threshold th1. It is detected that the request has been changed, and when the number of detections is three or more, a request for switching acceptance is output.

  The voice pitch tends to change when the speaker is speaking emotionally and tends to be shorter than the normal voice pitch. The voice pitch example shown in FIG. 5 indicates that the voice pitch changes in the second half of the utterance. This voice pitch example shows that the speaker is emotional in the second half.

  In the present invention, this is utilized to compare the voice pitch acquired at a predetermined time point from the utterance voice data of the customer who is being handled with the standard voice pitch, and the amount of variation based on the standard voice pitch is examined. If it is shorter than the normal time, the speaker is assumed to be emotional rather than calm.

  The standard voice pitch is collected from the customer's normal utterance voice data acquired at the initial stage of accepting the customer's call. For example, it is acquired using voice data when the customer responds with a name.

  The voice pitch threshold th1 is a value obtained by reducing the standard voice pitch by 10%, and the voice pitch threshold th2 is a value obtained by reducing the standard voice pitch by 15%.

  “Timeout = 10 minutes or more / 1 customer” indicates that a switching request for acceptance switching is output when 10 minutes have elapsed since the start of customer response.

  The display device 12 is a device that displays information processed by the operator reception monitoring / switching system 10 and a classification corresponding to the stress passive state of the primary operator.

  The front side device 15 of the operator reception monitoring / switching system 10 includes an utterance voice information acquisition unit 151, a voice pitch fluctuation amount detection unit 152, a dangerous word / phrase detection unit 153, a dangerous word / phrase correspondence unit 154, a reception switching determination unit 155, and a continuous response. A notification unit 156 is provided.

  In addition, the back side device 17 of the operator reception monitoring / switching system 10 includes a status display processing unit 171, a reception switching unit 172, a reception information transfer unit 173, a judgment condition update unit 174, a dangerous phrase statistical processing unit 175, and an extracted dangerous phrase storage. A portion 176 is provided.

  The utterance voice information acquisition unit 151 is processing means for acquiring, in real time, the utterance voice data of the customer who is answering the call by the primary operator recorded through the front terminal 2 via the front reception control device 3.

  The voice pitch variation detection unit 152 extracts the voice pitch from the utterance voice data of the customer, holds the voice pitch extracted at the initial stage as the standard voice pitch, and further extracts the voice pitch (monitoring voice pitch). This is a processing means for detecting a fluctuation amount of the monitoring voice pitch with respect to the standard voice pitch, and measuring the number of times of detecting a fluctuation amount that is equal to or greater than a predetermined fluctuation amount threshold among the fluctuation amounts detected in a predetermined time.

  The dangerous word / phrase detection unit 153 is a processing unit that detects a dangerous word / phrase from customer utterance voice data, identifies the level of the detected dangerous word / phrase, and measures the number of times the dangerous word / phrase is detected within a predetermined time.

  The dangerous phrase corresponding unit 154 extracts partial voice data corresponding to the dangerous phrase from the utterance voice data, performs a predetermined conversion process on the partial voice data, and outputs the converted partial voice data. It is.

  This is because there is no emotion in the mechanical sound compared to the raw words in which customer emotion is injected, so the effect of reducing the acceptance stress is used.

  The dangerous phrase correspondence unit 154 performs conversion processing such as processing for reducing the volume of partial audio data, processing for modulating partial audio data to a predetermined tone, processing for synthesizing partial audio data with a predetermined sound source, or partial audio data Is converted into text data, and one or more of these conversion processes are performed.

As shown in FIG. 6, the conversion process and the output of the conversion result are set in association with the following corresponding modes.
“Correspondence mode 1: Performs a process of outputting the corresponding partial audio data by reducing the volume (output volume);
Corresponding mode 2: Partial audio data obtained by cutting the corresponding partial audio data is output as audio, the partial audio data is converted into text data, and displayed on the display unit of the front terminal 2;
Corresponding mode 3: The corresponding partial voice data is cut (replaced with no voice), the partial voice data is synthesized using a predetermined sound source (machine voice), and synthesized to the location of the partial voice data of the utterance voice data. Incorporate audio data (machine audio) and output audio;
Correspondence mode 4: The corresponding partial voice data is cut, the partial voice data is voice-synthesized using a predetermined sound source (machine voice), and the voice data (machine voice) synthesized at the location of the partial voice data of the utterance voice data Is output by voice, and the partial voice data is converted into text data and displayed on the display unit of the front terminal 2. Note that not only the partial voice data but also the entire speech voice data may be converted into text data and displayed. "
In the corresponding modes 3 and 4, the speech synthesis process using a mechanical sound source uses a known processing method. The voice synthesis processing means may be configured to be provided in the front terminal 2 or the front side device 15.

  Further, the response mode may be set by a selection based on the primary operator, a predetermined group, or the whole.

  FIG. 7 is used to show a specific example of the handling process of the dangerous phrase included in the customer's utterance.

  As shown in FIG. 7 (A), it is assumed that the customer who the primary operator is responding remarks, “This system is strange. The utterance voice information acquisition unit 151 acquires the utterance voice data.

  Then, as shown in FIG. 7B, the dangerous phrase detection unit 153 detects the dangerous phrase “write to the net” from the acquired utterance voice data.

  Subsequently, the dangerous phrase correspondence unit 154 performs conversion processing according to the set corresponding mode on the voice data (partial voice data) corresponding to “write to the net” in the speech voice data.

  For example, when the correspondence mode 1 is set, the volume of the portion of the partial voice data “write to the net” is lowered when the speech voice data is output.

  Further, when the corresponding mode 2 is set, as shown in FIG. 7C, the partial voice data “write to the net” is cut (replaced with no voice), and “This system is strange. If it is this kind of response, please answer me more ... " Further, the character string “write to the net” converted into the text data is displayed on the display unit of the front terminal 2.

  If the corresponding mode 3 is set, when outputting speech voice data, the partial voice data “write to the net” is synthesized with the machine voice and incorporated into the corresponding part of the original speech voice data. , "This system is strange. If this is the case, <I will write on the net> Please respond to my relatives more (the part that is synthesized with speech is indicated by <>)."

  Also, when the correspondence mode 4 is set, the partial voice data “write to the net” is synthesized with the machine voice, and the voice is output in the same manner as in the correspondence mode 3, and the converted voice data is written to the “net”. Is displayed on the display unit of the front terminal 2.

  The reception switching determination unit 155 is a processing unit that outputs a switching request for switching reception of a call from the front terminal 2 to the back terminal 4 when the number of detections of the fluctuation amount of the dangerous phrase or voice pitch satisfies the determination condition.

  The reception continuation notification unit 156 is a processing unit that notifies the continuation of the response when the primary operator does not agree to the reception switching in response to the consent to the reception switching notified from the reception switching unit 172.

  The state display processing unit 171 specifies the monitoring category corresponding to the stress passive state of the primary operator based on the detection of the fluctuation amount exceeding the predetermined threshold value of the dangerous phrase and the voice pitch extracted from the utterance voice data of the customer. , A processing means for displaying the identified section with a mark.

  The status display processing unit 171 displays a display screen showing the sections at regular intervals in time series for each primary operator. Alternatively, for a plurality of primary operators, a display screen is displayed that indicates the classification of each primary operator acquired at a certain time point in correspondence with the arrangement of the front terminal 2.

  FIG. 8 shows a classification example and a display example thereof.

As shown in FIG. 8, the predetermined categories are classified into four categories of “normal continuity monitoring”, “warning monitoring”, “switching”, and “desired continuation monitoring”.
“Normal continuous monitoring” means monitoring the state where the primary operator is responding as usual without being stressed by the customer's utterance, and is displayed with a blue, double-circle mark, etc. on the display screen.
“Warning monitoring” means monitoring for a situation where the operator is stressed by the customer's remarks and needs to be monitored carefully, and is displayed on the display screen with yellow, circle marks, and the like.
“Switching” is a monitoring for the situation where the operator is receiving a considerable amount of stress from the customer's remarks, and the reception switching to the secondary operator is necessary. The state of the primary operator is the condition for outputting the switching request. Means that It is displayed on the display screen in red and circles with diagonal lines.
“Wanted continuation monitoring” is a stress passive state in which reception switching is necessary, and the original classification is “switching”, but indicates a state in which a response is continued according to the primary operator's request. Flashes red on the display screen, and is displayed with a thick circle and a hatched mark.

The classification conditions for these categories are shown below.
“Dangerous phrase detected once in speech data = warning monitoring (yellow);
The voice pitch of the utterance voice data exceeds / greater than the voice pitch threshold th1 = warning monitoring (yellow);
Dangerous phrases more than a predetermined number of times have been detected from utterance voice data within a certain period of time = switching (red);
The voice pitch of the speech voice data exceeds / greater than the voice pitch threshold th2 = switching (red);
Conditions other than the above = normal continuous monitoring (blue);
Although the status falls under “Switching”, the person wants to continue = Desired continuation monitoring (flashing red)
FIG. 9 and FIG. 10 show examples of display screens for the primary operator classification.

  FIG. 9 is a diagram showing a display screen 200 showing the progress of monitoring for each primary operator, and FIG. 10 is a diagram showing a display screen 300 showing a plurality of primary operators at a certain point in time.

  In the display screen 200 of FIG. 9, for each primary operator, the categories for the stress passive state acquired at intervals of 5 minutes are displayed in time series. When the primary operator is taking a break, a predetermined “break” mark is displayed.

  When the “display entire state” button 210 on the display screen 200 is selected by a mouse click operation or the like, the state display processing unit 171 switches the display to the display screen 300 shown in FIG.

  In the display screen 300 of FIG. 10, the classification of each primary operator with respect to the stress passive state acquired at a certain time is displayed in correspondence with the arrangement of the front terminal 2 indicated by “row” and “seat”.

  When the “display state over time” button 310 on the display screen 300 is selected by a mouse click operation or the like, the state display processing unit 171 switches the display to the display screen 200 shown in FIG.

  The reception switching unit 172 is a processing unit that receives a switching request and switches reception of a call from the front terminal 2 to the back terminal 4. Specifically, the front reception control device 3 and the back reception control device 5 are requested to switch the corresponding call.

  When the reception switching is executed by the reception switching unit 172, the reception information takeover unit 173 transmits the reception information regarding the reception of the call generated by the front terminal 2 to the back terminal 4 via the back reception control device 5. Processing means.

  The reception information is information input at the front terminal 2 by the operation of the primary operator during reception, information about reception searched and acquired via the CTI device 6, and the like.

  When the switching request is output, the determination condition update unit 174 displays the determination condition of the corresponding primary operator as the amount of fluctuation of the dangerous phrase or voice pitch detected from the utterance voice data of the customer who the primary operator has responded to. This is a processing means for updating using the number of detection times.

  The determination condition update unit 174 updates the determination condition of the primary operator with a predetermined weight when the primary operator completes the response without agreeing to the reception switching.

  An example of determination condition updating will be described with reference to FIG.

  FIG. 4A shows an example of determination conditions set as initial values. Here, the determination condition is a condition for determining the number of detections of dangerous words / phrases at each level and the number of detections within a predetermined time of a fluctuation amount exceeding a predetermined fluctuation amount threshold value (threshold th1, threshold th2) of the voice pitch. It is. The determination condition of FIG. 4A is used for the primary operator before the determination condition is updated.

  FIG. 4B is an example of determination conditions for the primary operator E.

  In the determination process of the primary operator E, the determination condition (initial value) shown in FIG. Then, it is assumed that a dangerous word / phrase level 1 dangerous word / phrase is detected from the utterance voice data being handled and the number of detections exceeds this condition (3 times / 10 minutes or less). It is assumed that the primary operator Mr. E has continued to respond without completing the acceptance switching.

  In this case, the acceptance switching determination unit 155 stores the record information of the state of the primary operator E when the switching request is output. Based on the stored record information, the determination condition update unit 174 performs predetermined weighting (adding one time) for the dangerous phrase level 1 condition item of the determination condition that triggered the output of the switching request. To update. In the updated determination condition, the condition for level 1 of the dangerous phrase is “4 times / 10 minutes or less”.

  FIG. 4C is an example of the determination condition of the primary operator Mr. R.

  In the determination process by the primary operator R, the determination condition (initial value) shown in FIG. 4A is used first. After that, the amount of fluctuation exceeding the voice pitch threshold th1 was detected from the speech data being handled exceeding the predetermined number of detections (within 3 times / 10 minutes). Then, although the switching request is output, it is assumed that the primary operator Mr. R does not agree to the reception switching and continues to respond.

  The record information of the state of the primary operator R is saved when the switching request is output, and the determination condition update unit 174 sets a predetermined weight (1) to the condition item of the voice pitch threshold th1 of the determination condition that triggered the output of the switching request. Update). In the determination condition after the update, the condition of the voice pitch value threshold th1 is “3 times / 10 minutes or less”.

  The dangerous word / phrase statistics processing unit 175 is a processing unit that accumulates the dangerous words / phrases detected from the customer's speech data by the dangerous word / phrase detection unit 153 in the extracted dangerous word / phrase storage unit 176 and performs predetermined statistical processing on the accumulated dangerous words / phrases. is there.

  As predetermined statistical processing, for example, the number of detected dangerous words and the frequency are calculated in units of words or operators. Statistical materials such as the usage tendency of dangerous phrases are used for education planning such as improvement of operator manners and response training for newly introduced products and services.

  11 to 14 show the processing flow of the present invention. FIG. 11 is a process flow diagram of the main process of the operator reception monitoring / switching system 10.

  It is assumed that, before the processing of FIG. 11 is started, identification information of the primary operator, dangerous phrases, determination conditions, and the like are recorded in the determination condition storage device 11.

  Moreover, the process flow of FIG. 11 is a process with respect to each primary operator, and the process of step S1-S10 is performed for every primary operator who is using the front terminal 2. FIG.

  The operator reception monitoring / switching system 10 identifies a temporary operator who uses the front terminal 2 and sets a response mode to be executed by the dangerous phrase corresponding unit 154 (step S1).

  The identification process of the primary operator is a known process using an identification number given in advance for each primary operator and input at the front terminal 2 by a processing means for managing the reception work of the primary operator provided in the operator reception monitoring / switching system 10. The identification method is used. Further, information such as a determination condition of the identified primary operator is acquired from the determination condition storage device 11. Further, the dangerous phrase correspondence unit 154 sets a response mode to be executed based on the response mode setting instruction input at the front terminal 2.

  Next, the utterance voice information acquisition unit 151 performs timer monitoring for a predetermined time (step S2), and executes monitoring processing for every elapse of the predetermined time (step S3).

  And the display screen 200 or the display screen 300 which displays the division | segmentation corresponding to the stress passive state of a primary operator obtained by the monitoring process of step S3 is displayed on the display apparatus 12 (step S4).

  Here, the acceptance switching determination unit 155 checks the reception start time with the customer to whom the primary operator responds, and checks whether the reception time with the same customer exceeds the timeout time of the determination condition (step S5).

  Only when the response time with the same customer exceeds the timeout time (YES in step S5), the primary operator classification is changed to “switching”, and the status display processing unit 171 changes the corresponding primary operator classification to “red”. The display is changed to “display (switching required)” (step S6).

  The reception switching unit 172 checks whether the primary operator category is “red display (switching)” (step S7). If it is “red display (switching)” (YES in step S7), the reception switching unit 172 , An acceptance switching process is performed (step S8).

  In the acceptance switching process in step S8, it is checked whether the primary operator has agreed to acceptance switching (step S9). Only when the primary operator has agreed to the acceptance switching (YES in step S9), the primary operator is instructed to take a break (step S10).

  Thereafter, the determination condition update unit 174 performs management processing (step S11), and returns to timer monitoring for a predetermined time (step S2).

  If the primary operator classification is not “red display (switch)” in the process of step S7, the process returns to the process of step S2.

  FIG. 12 is a process flowchart of the monitoring process in step S3.

  The utterance voice information acquisition unit 151 acquires the customer utterance voice data received by the primary operator in real time (step S301).

  Subsequently, the voice pitch fluctuation amount detection unit 152 extracts a portion responding to the name confirmation of the primary operator from the utterance voice data, and acquires and holds the standard voice pitch of the customer from the extracted utterance voice data (step S302). .

  The dangerous phrase detection unit 153 monitors whether or not a dangerous phrase appears in the utterance voice data of the customer while the primary operator responds to the customer (step S303).

  Here, if no dangerous word or phrase has appeared (NO in step S304), the voice pitch variation detection unit 152 identifies the voice pitch from the utterance voice data, obtains the monitoring voice pitch, and determines the monitoring voice pitch. The amount of variation from the standard voice pitch is calculated by comparing with the standard voice pitch (step S305).

  If the fluctuation amount of the monitoring voice pitch with respect to the standard voice pitch is not equal to or greater than the voice pitch threshold th1 (NO in step S306), the acceptance switching determination unit 155 determines that the primary operator is classified as “normal continuation monitoring” and displays the status display. The processing unit 171 sets the corresponding division mark on the display screen to “blue display” (step S307).

  If a dangerous phrase is detected from the utterance voice data in the process of step S304 (YES in step S304), the dangerous phrase correspondence unit 154 selects a portion (partial voice data) corresponding to the dangerous phrase of the utterance voice data. The extracted partial voice data is processed according to the set corresponding mode (step S308).

  Subsequently, the dangerous phrase detection unit 153 identifies the level of the detected dangerous phrase, and the acceptance switching determination unit 155 checks whether the number of detections of the prohibited determination phrase within a predetermined time satisfies the determination condition (step S309). . The number of times of dangerous phrase detection is compared with the determination condition, and if the number of detection does not meet the determination condition (NO in step S309), the voice pitch variation detection unit 152 calculates the monitoring voice pitch and the standard voice pitch. Compare. If the fluctuation amount of the monitoring voice pitch with respect to the standard voice pitch is not equal to or greater than the voice pitch threshold th2 (NO in step S310), the reception switching determination unit 155 determines that the primary operator is classified as “warning monitoring”, and the status display process The unit 171 changes the mark of the corresponding section on the display screen to “yellow display” (step S311).

  If it is determined in step S309 that the number of detected dangerous words corresponds to the determination condition (YES in step S309), the reception switching determination unit 155 determines that the primary operator classification is “switching”, and the status display processing unit In 171, the mark of the corresponding section on the display screen is set to “red display” (step S 312).

  If it is determined in step S310 that the fluctuation amount of the monitoring voice pitch with respect to the standard voice pitch is greater than or equal to the voice pitch threshold th2 (YES in step S310), the process in step S312 is performed.

  FIG. 13 is a process flow diagram of the acceptance switching process in step S8.

  The reception switching unit 172 displays “switching consent” on the display device of the front terminal 2 of the primary operator that is the target of the reception switching process (step 801). When the primary operator selects and agrees with “switch agreement” (YES in step S802), the response state and determination conditions when the primary operator switch request is output (switching) are saved (step S803), and the front reception is performed. The control device 3 transfers / stores the reception data supported by the front terminal 2 to the back reception control device 5 (step S804).

  The reception switching unit 172 notifies the customer of the switching of the operator, and sends a message such as “Please wait for a while since switching to the secondary support person” by voice (step S805). Then, a free secondary operator of the back terminal 4 is searched (step S806), and if there is a free operator (YES in step S807), the customer call reception switching is executed (step S808). The status display processing unit 171 deletes “red display” in the primary operator category of the display screen (step S809).

When the primary operator does not agree with “switch agreement” in the process of step S802 (NO in step S802), the reception switching unit 172 displays the response state when the primary operator switch request is output (during non-switching) and The determination condition is saved (step S810). Then, the reception switching determination unit 155 determines that the primary operator's response status is “desired continuous monitoring”, and the status display processing unit 171 displays the primary operator “red flashing” on the display screen and sounds the buzzer ( Step S811)
FIG. 14 is a process flowchart of the management process in step S11.

  The determination condition update unit 174 reads the stored response state and determination condition at the time of switching the primary operator (step S111), and when the primary operator completes the reception alone (YES in step S112), Update the judgment condition.

  Although the present invention has been described above with reference to the embodiments, it is obvious that the present invention can be variously modified within the scope of the gist thereof.

  For example, in the processing flow of the monitoring process shown in FIG. 11, the primary operator identification process is performed using the identification number information. However, when the front terminal 2 used by the primary operator is fixed, the primary operator may be identified using information specifying the front terminal 2. Regardless of the identification process described, any known identification method may be used.

It is a figure which shows the structural example of this invention, and the apparatus structural example of the call center by which this invention is implemented. It is a figure which shows the structural example of the operator reception monitoring and switching system concerning this invention. It is a figure which shows the example of a level of a dangerous phrase. It is a figure for demonstrating the example of a determination condition, and the update process of a determination condition. It is a figure which shows the example of a change of an audio | voice pitch. It is a figure which shows the example of response mode. It is a figure for demonstrating the corresponding process of a dangerous word / phrase. It is a figure which shows the example of a division | segmentation and its display example. It is a figure which shows the example of the display screen of the division | segmentation of a primary operator. It is a figure which shows the example of the display screen of the division | segmentation of a primary operator. It is a processing flow figure of the main process of an operator reception monitoring and switching system. It is a processing flowchart of a monitoring process. It is a processing flow figure of reception switching processing. It is a processing flow figure of management processing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Call center control system 2 Front terminal 3 Front reception control apparatus 4 Back terminal 5 Back reception control apparatus 6 CTI apparatus 7 Telephone network 10 Operator reception monitoring and switching system 11 Judgment condition storage apparatus 111 Hazard phrase storage section 12 Display apparatus 151 Utterance voice information Acquiring unit 152 Voice pitch variation detection unit 153 Dangerous phrase detection unit 154 Dangerous phrase correspondence unit 155 Reception switching determination unit 156 Reception continuation notification unit 157 Communication unit 171 Status display processing unit 172 Reception switching unit 173 Response information takeover unit 174 Update of determination condition Part 175 dangerous word / phrase statistics processing part 176 extracted dangerous word / phrase storage part 177 communication part

Claims (12)

A call center control system including a front terminal used by a primary operator and a back terminal used by a secondary operator is set, the primary operator accepts a customer call and responds, and the secondary operator takes over the call and responds In a call center, an operator reception monitoring / switching system for monitoring a response state of the primary operator and switching reception of the call from the front terminal to the back terminal based on a switching request,
A dangerous phrase storage unit for storing dangerous phrases that may stress the operator;
A determination condition storage unit for storing a determination condition based on the number of detections of the dangerous phrase, which is a condition for determining an output of a switching request for switching a customer call reception from a primary operator to a secondary operator;
An utterance voice information acquisition unit for acquiring utterance voice data of a customer of a call accepted by the primary operator;
A dangerous phrase detection unit that detects the dangerous phrase from the utterance voice data;
Extracting a partial voice data corresponding to the dangerous phrase from the utterance voice data, performing a predetermined conversion process on the partial voice data, and outputting the converted partial voice data;
A state display processing unit for displaying a monitoring classification according to a stress passive state when the dangerous phrase is detected a predetermined number of times or more within a predetermined time;
An acceptance switching determination unit for outputting a switching request for switching a call being accepted at the front terminal to a back terminal when the number of detections of the dangerous phrase satisfies the determination condition;
An acceptance switching unit for accepting the switching request and performing acceptance switching for switching the acceptance of the call to the back terminal;
An operator reception monitoring / switching system, comprising: a reception information takeover unit that transmits reception information about reception of the call generated at the front terminal to the back terminal. The dangerous word / phrase correspondence unit includes, as the conversion process, a process of reducing the volume of the partial voice data, a process of modulating the partial voice data to a predetermined tone, a process of synthesizing the partial voice data with a predetermined sound source, The operator reception monitoring / switching system according to claim 1, further comprising a process function of converting the partial voice data into text data, and performing any one or more of the conversion processes. Obtaining a voice pitch in a customer's standard utterance from the utterance voice data and setting it as a standard voice pitch;
A voice pitch fluctuation amount detection unit for detecting a voice pitch of a customer from the utterance voice data and calculating a fluctuation amount of the detected voice pitch with respect to the standard voice pitch;
When the determination condition storage unit stores a determination condition based on the amount of change in the customer's voice pitch as the determination condition,
The state display processing unit displays a predetermined section indicating a stress passive state of the primary operator when the amount of change fluctuates to a predetermined value or more,
3. The acceptance switching determination unit outputs a switching request for switching a call being accepted at the front terminal to a back terminal when the fluctuation amount satisfies the determination condition. The operator reception monitoring / switching system described. The said acceptance switching determination part outputs the said switching request | requirement, when the response time with respect to the said customer of the said primary operator exceeds fixed time. The switching request | requirement as described in any one of Claims 1-3 characterized by the above-mentioned. Operator reception monitoring and switching system. The determination condition storage unit stores a determination condition in which a plurality of threshold values are set for each of the number of detection times of the dangerous phrase or the number of detection times of the variation amount of the voice pitch. The operator reception monitoring and switching system according to any one of the above. The said acceptance switching part inquires about the consent of the said acceptance switching to the said primary operator, and when the acceptance consent of the said acceptance switching is received, the said acceptance switching is performed. The operator reception monitoring / switching system described in one item. When the determination condition for each primary operator is stored in the determination condition storage unit, the reception switch determination unit extracts the determination condition of the primary operator from the determination condition storage unit and uses the determination condition for the reception switch The operator reception monitoring / switching system according to any one of claims 1 to 6, wherein When the switching request is output, the determination condition of the primary operator related to the switching request is set to the number of detections of the fluctuation amount of the dangerous phrase or the voice pitch detected from the utterance voice data of the customer accepted by the primary operator. An operator reception monitoring / switching system according to any one of claims 1 to 7, further comprising: a determination condition updating unit that performs weighting and updating. The state display processing unit displays a display screen showing the sections indicating the stress passive state of the primary operator in time series, or a display indicating the sections indicating the stress passive state of a plurality of primary operators corresponding to the arrangement of the front terminal The operator reception monitoring / switching system according to any one of claims 1 to 8, wherein any one of the display screens is displayed. An extracted dangerous phrase storage unit that accumulates dangerous phrases detected from the utterance voice data;
The operator reception monitoring / switching system according to any one of claims 1 to 9, further comprising: a dangerous word / phrase statistical processing unit that performs predetermined statistical processing on the accumulated dangerous words / phrases. 3. The operator according to claim 2, wherein the dangerous word / phrase handling unit sets a conversion processing function to be executed based on an instruction input of the primary operator from a plurality of conversion processing functions provided in advance. Reception monitoring / switching system. The determination condition update unit updates the determination condition of the primary operator with a predetermined weight when the primary operator completes the reception of the received call without agreeing to the reception switching. The operator reception monitoring / switching system according to claim 8.

Images