JP2018169843A - Information processing device, information processing method and information processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information processing device that can predict cancellation risks against those clients who come to contact for the first time and thus whose behavior history information is not obtained.SOLUTION: An information processing device 100 comprises a call storage unit 101, a voice amount feature calculation unit 102, an emotion recognition unit 103, and a cancellation prediction unit 104. The call storage unit 101 obtains and stores calls between clients and operators. The voice amount feature calculation unit 102 calculates the feature amount from the voice amount of the stored calls. The emotion recognition unit 103 recognizes the emotion of the voice activities. The cancellation prediction unit 104 predicts, based on the feature amount of the voice amount and the recognized emotion, cancellation risks in the calls.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an information processing apparatus, an information processing method, and an information processing program.

  In the above technical field, Patent Document 1 discloses that customer information, customer behavior history information, and a customer extraction rule that is regarded as a customer with a high cancellation risk are compared, and a customer that satisfies the customer extraction rule has a high cancellation risk. A technique for extracting as a customer is disclosed. Non-Patent Document 1 discloses a method for detecting an utterance section. Non-Patent Document 2 discloses an emotion recognition method. Non-Patent Document 3 discloses a machine learning method. Non-Patent Document 4 discloses a method for detecting an utterance section.

JP 2002-334200 A

Yusuke Kida and Tatsuya Kawahara, "Voice Activity Detection based on Optimally Weighted Combination of Multiple Features," Proc. INTERSPEECH 2005, pp.2621-2624, 2005. Florian Eyben, Martin Wollmer, and Bjorn Schuller, "openEAR-Introducing the Munich Open-Source Emotion and Affect Recognition Toolkit," 2009 3rd International Conference on Affective Computing and Intelligent Interaction and Workshops Pedregosa et al., "Scikit-learn: Machine Learning in Python," JMLR 12, pp. 2825-2830, 2011. JPYamron, I.Carp, L.Gillick, S.Lowe, and P.van Mulbregt, “A HIDDEN MARKOV MODEL APPROACH TO TEXT SEGMENTATION AND EVENT TRACKING,” IEEE International Conference on Acoustics, Speech and Signal Processing, pp.333- 336, 1998.

  However, since the technology described in the above document uses customer behavior history information, it has not been possible to predict churn risk even for a customer who does not have behavior history information for the first time.

  The objective of this invention is providing the technique which solves the above-mentioned subject.

In order to achieve the above object, an information processing apparatus according to the present invention provides:
Call storage means for acquiring and storing a call between a customer and an operator;
Utterance amount feature calculation means for calculating the stored utterance amount feature amount of the call;
Emotion recognition means for recognizing the emotion of the utterance interval;
A churn predicting means for predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
Equipped with.

In order to achieve the above object, an information processing method according to the present invention includes:
A call storage step of acquiring and storing a call between a customer and an operator;
An utterance amount feature calculating step of calculating an utterance amount feature amount of the stored call;
An emotion recognition step for recognizing the emotion of the utterance interval;
A churn prediction step of predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
including.

In order to achieve the above object, an information processing program according to the present invention provides:
A call storage step of acquiring and storing a call between a customer and an operator;
An utterance amount feature calculating step of calculating an utterance amount feature amount of the stored call;
An emotion recognition step for recognizing the emotion of the utterance interval;
A churn prediction step of predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
Is executed on the computer.

  According to the present invention, cancellation risk can be predicted even for a customer who does not have action history information and has contacted for the first time.

It is a block diagram which shows the structure of the information processing apparatus which concerns on 1st Embodiment of this invention. It is a block diagram which shows the structure of the information processing apparatus which concerns on 2nd Embodiment of this invention. It is a block diagram which shows the structure of the emotion characteristic calculation part of the information processing apparatus which concerns on 2nd Embodiment of this invention. It is a figure which shows the hardware constitutions of the information processing apparatus which concerns on 2nd Embodiment of this invention. It is a flowchart explaining the process sequence of the information processing apparatus which concerns on 2nd Embodiment of this invention. It is a block diagram which shows the structure of the information processing apparatus which concerns on 3rd Embodiment of this invention. It is a figure which shows the hardware constitutions of the information processing apparatus which concerns on 3rd Embodiment of this invention. It is a flowchart explaining the process sequence of the information processing apparatus which concerns on 3rd Embodiment of this invention.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be exemplarily described in detail with reference to the drawings. However, the configuration, numerical values, process flow, functional elements, and the like described in the following embodiments are merely examples, and modifications and changes are free, and the technical scope of the present invention is described in the following description. It is not intended to be limited.

[First Embodiment]
An information processing apparatus 100 as a first embodiment of the present invention will be described with reference to FIG. The information processing apparatus 100 is an apparatus that predicts a churn risk in a call based on the utterance amount characteristics and emotions in the call.

  As illustrated in FIG. 1, the information processing apparatus 100 includes a call storage unit 101, an utterance amount feature calculation unit 102, an emotion recognition unit 103, and a churn prediction unit 104. The call storage unit 101 acquires and stores a call between a customer and an operator. The utterance amount feature calculation unit 102 calculates the utterance amount feature amount of the stored call. The emotion recognition unit 103 recognizes the emotion in the utterance section. The churn prediction unit 104 predicts the churn risk in the call based on the utterance amount feature amount and the recognized emotion.

  According to the present embodiment, it is possible to predict churn risk even for a customer who does not have action history information and contacts for the first time.

[Second Embodiment]
Next, an information processing apparatus according to a second embodiment of the present invention will be described with reference to FIGS. FIG. 2 is a block diagram illustrating a configuration of the information processing apparatus according to the present embodiment.

  The information processing apparatus according to the present embodiment can extract, from a call, a customer who has a high risk of churn or a customer who can expect continuation from a call. It is a device that can be implemented.

  The information processing apparatus 200 includes a call storage unit 201, an utterance amount feature calculation unit 202, an emotion recognition unit 203, an emotion feature calculation unit 204, a customer extraction model storage unit 205, a churn prediction unit 206, and an extraction customer output unit 207.

  The call storage unit 201 acquires a call between a customer and an operator as call voice data. The call storage unit 201 stores the acquired call voice data as a file or stream data.

  When storing the call voice data, the call storage unit 201 assigns and stores a call ID (Identifier) (call ID) for specifying the call (call) to the stored call voice data. Furthermore, the call storage unit 201 detects the speech section of the call voice data with respect to the stored call voice data, and gives speaker information, that is, operator information or customer information, to each detected speech section. And remember. Furthermore, the call storage unit 201 assigns and stores the utterance time information, that is, the utterance start time and utterance end time, to each of the detected utterance sections, with respect to the stored call voice data. Therefore, when storing the call voice data, the call storage unit 201 assigns and stores the call ID, operator information, customer information, and utterance time information to the call voice data to be stored.

  Various methods are disclosed as the method for detecting an utterance section. For example, the method described in Non-Patent Document 1 may be used.

  As for the call ID, for example, in a recent call center, the call ID can be obtained by a CTI (Computer Telephony Integration) system. The customer side signal and the operator side signal can record data as separate channels. Therefore, the call ID and speaker information can be given using the call ID and channel information acquired by the CTI system.

  Further, the utterance time information can be given by acquiring the absolute time of the detected utterance section or the time information defined by the system for the start and end points.

  The utterance amount feature calculation unit 202 calculates the utterance amount feature between the operator and the customer from the time information of each utterance section stored in the call storage unit 201. That is, the total time of each utterance section of the operator and the customer in each call, the ratio of the total time of the utterance section of the operator and the customer in each call, the average value of the length of each utterance section of the operator and the customer in each call, variance At least one of a value, a median value, a mode value, or a value calculated by other statistical processing is used.

  The emotion recognition unit 203, for each utterance section stored in the call storage unit 201, valence (emotion valence indicating the quality of emotion, degree of positive and negative emotion), arousal (arousal level indicating arousal or quiet, At least one emotional value of the degree of emotion excitement), satisfaction, and expectation is calculated. Various methods for emotion recognition are disclosed. For example, the method described in Non-Patent Document 2 can be used.

  For each emotion, a numerical value in a range designated in advance or a finite number of categories is calculated as an emotion value. That is, the emotion recognition unit 203 calculates, for example, a value in the range of −3 to 3 or a category such as {−3, −2, −1, 0, 1, 2, 3} for valence. The same applies to arousal, persuasion, and expectation, but the range and category specified by each emotion need not be the same.

  The emotion feature calculation unit 204 calculates the emotion feature amount of the operator and the customer from the emotion value of each utterance section calculated by the emotion recognition unit 203. The emotion feature amount is calculated by the sum of the emotion values of the operator and the customer in each call, the emotion value ratio, the average value of the emotion values, the variance value, the median value, the mode value, or other statistical processing. Contains at least one of the values.

  FIG. 3 is a block diagram illustrating a configuration of the emotion feature calculation unit of the information processing apparatus according to the present embodiment. The emotion feature calculation unit 204 may calculate the emotion feature amounts of the operator and the customer using only the emotion value of each utterance section included in a specific section in the call voice data with the configuration shown in FIG. The emotion feature calculation unit 204 illustrated in FIG. 3 includes a topic section detection unit 241, a narrowing unit 242, and a feature calculation unit 243.

  The topic section detection unit 241 detects a topic section included in the call voice data stored in the call storage unit 201, and assigns a topic start time and a topic end time to each of the detected topic sections. Topics include, for example, opening, closing, product / service explanation, fee explanation, solicitation, hearing, personal information confirmation, and office procedures.

  Various methods for detecting a topic section are disclosed. For example, voice recognition and a method described in Non-Patent Document 4 can be used in combination. Specifically, it is possible to divide a word string obtained by speech recognition into a plurality of topic sections using a topic model in which the appearance frequency of words for various topics is learned in advance.

  Further, the topic section detection unit 241 may detect a topic section based on time information and the number of utterances. For example, the first 30 seconds section and the first five utterance sections of the call can be detected as the opening topic section. In addition, the last 30 seconds and the last five utterances of the call can be detected as closing topics. Furthermore, the latter half of the call may be detected as one topic section (for example, a section of 6 to 9 minutes for a 9-minute call).

  The narrowing-down unit 242 outputs, to the feature calculation unit 243, only the utterance section included in the specific topic detected by the topic section detection unit 208 among the utterance sections to which the emotion recognition unit 203 has given the emotion value. For example, only the speech sections included in the closing topic section and their emotion values are output to the feature calculation unit 243.

  The feature calculation unit 243 calculates the emotion feature amount of the operator and the customer from the emotion value of each utterance section output from the narrowing unit 242. Emotion feature value is calculated by total of emotion value of operator and customer, emotion value ratio, average value of emotion value, variance value, median value, mode value, or other statistical processing in specific topic section At least one of the values to be included. For example, the total or average value of the customer utterance valence and expectation emotion values included in the topic section of the opening or closing is calculated as the emotion feature amount.

  The customer extraction model storage unit 205 uses the operator and customer emotion feature amounts calculated by the emotion feature calculation unit 204 and the operator and customer utterance amounts calculated by the utterance amount feature calculation unit 202 as feature amounts, and cancels the contract. A model predicted by a customer with high risk is stored as a customer extraction model.

  In the customer extraction model, correct label information indicating whether or not the contract corresponding to each call data has been canceled is given in advance, and a model for predicting the cancellation risk is learned by a machine learning method. Various machine learning methods are disclosed. For example, Non-Patent Document 3 discloses the details and executable programs. For the cancellation prediction, a prediction method corresponding to the learning method of the model can be used.

  The churn prediction unit 206 predicts (calculates) a churn risk for each call using a customer extraction model.

  The extracted customer output unit 207 compares the predicted churn risk for each call with the churn risk threshold specified in advance and the high churn risk ratio for the number of calls, so that the customer with high churn risk, that is, the churn risk. Extract high call customers.

  In the description of the present embodiment, an example in which cancellation is predicted has been described. However, the present embodiment can be similarly applied to continuous prediction.

  FIG. 4 is a block diagram illustrating a hardware configuration of the information processing apparatus 200 according to the present embodiment. A CPU (Central Processing Unit) 410 is a processor for arithmetic control, and implements a functional component of the information processing apparatus 200 in FIG. 2 by executing a program. The CPU 410 may include a plurality of processors and execute different programs, modules, tasks, threads, and the like in parallel. A ROM (Read Only Memory) 420 stores fixed data such as initial data and programs and other programs. The network interface 430 communicates with other devices via the network. Note that the CPU 410 is not limited to one, and may be a plurality of CPUs or may include a graphics processing unit (GPU) for image processing. The network interface 430 preferably includes a CPU independent of the CPU 410 and writes or reads transmission / reception data to / from a RAM (Random Access Memory) 440 area. Further, it is desirable to provide a DMAC (Direct Memory Access Controller) that transfers data between the RAM 440 and the storage 450 (not shown). Further, the CPU 410 recognizes that the data has been received or transferred to the RAM 440 and processes the data. Further, the CPU 410 prepares the processing result in the RAM 440 and leaves the subsequent transmission or transfer to the network interface 430 or the DMAC.

  The RAM 440 is a random access memory that the CPU 410 uses as a work area for temporary storage. In the RAM 440, an area for storing data necessary for realizing the present embodiment is secured. The call ID 441 is data for identifying a call (call) between an operator and a customer, and is assigned to each call. The utterance amount feature 442 is a feature of the calculated utterance amount in a call between the operator and the customer. The emotion value 443 is a calculated emotion value for each utterance section in each call. The emotion feature 444 is a feature of emotion calculated from the emotion value of each utterance section. The customer extraction model 445 is a model for predicting a customer with a high churn risk. The churn prediction 446 is a prediction of the churn risk predicted for each call using the customer extraction model.

  The input / output data 447 is data input / output via the input / output interface 460. The transmission / reception data 448 is data transmitted / received via the network interface 430. The RAM 440 includes an application execution area 449 for executing various application modules.

  The storage 450 stores a database, various parameters, or the following data or programs necessary for realizing the present embodiment. The storage 450 stores an utterance amount feature calculation module 453, an emotion recognition module 454, an emotion feature calculation module 455, and a churn prediction module 457.

  The utterance amount feature calculation module 453 is a module that calculates the utterance amount feature of a call between an operator and a customer. The emotion recognition module 454 is a module that calculates an emotion value in an utterance section of a call between an operator and a customer. The emotion feature calculation module 455 is a module that calculates emotion features from the calculated emotion value of the utterance section. The churn prediction module 457 is a module that predicts the churn risk in a call based on the utterance amount feature quantity and the recognized emotion. These modules 453 to 457 are read by the CPU 410 into the application execution area 449 of the RAM 440 and executed. The control program 456 is a program for controlling the entire information processing apparatus 200.

  The input / output interface 460 interfaces input / output data with input / output devices. A display unit 461 and an operation unit 462 are connected to the input / output interface 460. In addition, a storage medium 464 may be connected to the input / output interface 460. Furthermore, a speaker 463 that is an audio output unit, a microphone (not shown) that is an audio input unit, or a GPS position determination unit may be connected. Note that the RAM 440 and the storage 450 shown in FIG. 4 do not show programs and data related to general-purpose functions and other realizable functions that the information processing apparatus 200 has.

  FIG. 5 is a flowchart for explaining the processing procedure of the information processing apparatus 200 according to this embodiment. This flowchart is executed by the CPU 410 in FIG. 4 using the RAM 440, and implements the functional components of the information processing apparatus 200 in FIG.

  In step S501, the information processing apparatus 200 acquires a call between an operator and a customer. In step S503, the information processing apparatus 200 detects the utterance section of the acquired call, and adds speaker information and utterance time information to each detected utterance section. In step S505, the information processing apparatus 200 calculates the feature of the amount of speech between the operator and the customer in a call. In step S507, the information processing apparatus 200 calculates the emotion value of the utterance section. In step S509, the information processing apparatus 200 calculates an emotion feature from the calculated emotion value of the utterance section. In step S513, the information processing apparatus 200 predicts the churn risk for each call using the customer extraction model. In step S515, the information processing apparatus 200 extracts a call customer having a high cancellation risk based on the predicted cancellation risk.

  According to the present embodiment, by using inquiries about customer products and services and call data at the time of application, it is possible to predict cancellation even for a customer who cannot acquire an action history that makes contact for the first time. .

  In addition, according to the present embodiment, information such as whether the operator or the customer is actively talking as an utterance feature is calculated by using statistical processing including information that cannot be ruled manually. Similarly, models that estimate churn risk with high accuracy by using statistical processing to calculate not only emotional values that can be interpreted by humans, such as satisfaction and dissatisfaction, but also information that cannot be manually ruled from emotional values. Can be learned and predicted.

  Canceling or not canceling is not determined by the customer's product or service inquiry or exchange at the time of application. Some customers tend to cancel when they purchase products that were not originally intended by up-sell or cross-sell. In such a case, the present inventor is effective in using the conversation information of the inquiry to the service and the exchange at the time of application for predicting the cancellation or continuation, and in that case, the utterance feature and the emotion feature are effective. This is what we found.

  Therefore, in the present embodiment, the predicted (calculated) risk is compared with the threshold value of the churn risk specified in advance and the ratio of high churn risk persons in the number of target calls, and the customers judged to have a high churn risk are extracted. . Thereby, in this embodiment, a customer with a high cancellation risk mainly due to telephone response can be extracted.

  Further, the present inventor has found that whether or not the customer cancels in the future depends on the feelings of the operator and the customer in a specific topic section in the call. For example, it was found that the risk of churn is low when the customer's positive feelings and expectation feeling values are large in the topic sections of the opening and closing.

  For this reason, in this embodiment, the churn risk is estimated based on the emotion feature focusing on the emotions of the operator and the customer in a specific topic section in the call. Thereby, in this embodiment, a customer with a high cancellation risk can be extracted with high accuracy.

[Third Embodiment]
Next, an information processing apparatus according to a third embodiment of the present invention will be described with reference to FIGS. FIG. 6 is a block diagram illustrating a configuration of the information processing apparatus according to the present embodiment. The information processing apparatus according to the present embodiment differs from the second embodiment in that it includes a phrase storage unit, a phrase recognition unit, and a phrase feature calculation unit. Since other configurations and operations are the same as those of the second embodiment, the same configurations and operations are denoted by the same reference numerals, and detailed description thereof is omitted.

  The information processing apparatus 600 includes a phrase storage unit 601, a phrase recognition unit 602, and a phrase feature calculation unit.

  The phrase storage unit 601 stores a phrase designated in advance. The phrase storage unit 601 stores phrases that can be regarded as those situations for at least one of satisfaction, dissatisfaction, brute force, and commitment to fees. For example, in the case of consent, “I understand”, “That ’s right”, etc. In the case of dissatisfaction, it is “for the time being”, “for the time being”, “for the moment” and so on. In the case of brute force, “Let's decide”, “This is all right”, etc. In the case of sticking to the price, there are “price”, “fee”, “free shipping”, “consumption tax”, “mottle”, and the like.

  The phrase recognition unit 602 detects and recognizes the appearance of a phrase stored in the phrase storage unit 601. The phrase recognition unit 602 performs voice recognition and word spotting, detects and recognizes the appearance of a phrase stored in the phrase storage unit 601.

  The phrase feature calculation unit 603 calculates the feature of the operator and customer phrases. The phrase feature calculation unit 603 uses the recognition result (detection result) by the phrase recognition unit 602, and performs the following calculations for at least one of satisfaction, dissatisfaction, brute force, and attention to charge. That is, the phrase feature calculation unit 603 calculates the number of phrase appearances that is the number of appearances of the phrase stored in the phrase storage unit 601. Also, the phrase feature calculation unit 603 calculates the normalized phrase appearance number obtained by normalizing the phrase appearance number by the utterance time and the phrase appearance number for each predetermined time interval. The phrase feature calculation unit 603 calculates at least one of the number of phrase appearances, the number of normalized phrase appearances, and the number of phrase appearances for each predetermined time interval.

  Similarly to the second embodiment, the customer extraction model storage unit 205 stores the operator and customer emotion feature amounts calculated by the emotion feature calculation unit 204, and the operator and customer utterances calculated by the utterance amount feature calculation unit 202. A model is generated by adding the following feature quantity to the quantity feature. In other words, the customer extraction model storage unit 205 adds the phrase feature amount of the operator and the customer calculated by the phrase feature calculation unit 603 to the emotion feature amount and the utterance amount feature, and predicts a customer with a high churn risk An extraction model is generated and stored. Note that learning (generation) of the customer extraction model and prediction using the learned model are the same as in the second embodiment except that the above-described feature amount (phrase feature amount) is used.

  FIG. 7 is a block diagram illustrating a hardware configuration of the information processing apparatus 600 according to the present embodiment. A CPU (Central Processing Unit) 410 is a processor for arithmetic control, and implements a functional component of the information processing apparatus 600 in FIG. 2 by executing a program.

  The RAM 740 is a random access memory used by the CPU 410 as a work area for temporary storage. In the RAM 740, an area for storing data necessary for realizing the present embodiment is secured. The stored phrase 741 is a phrase that is designated and stored in advance.

  The storage 750 stores a database, various parameters, or the following data or programs necessary for realizing the present embodiment. The storage 850 stores a phrase 751, a phrase recognition module 752, and a phrase feature calculation module 753. The phrase 751 is a phrase designated in advance. The phrase recognition module 752 is a module that detects and recognizes the appearance of a stored phrase by voice recognition or word spotting. Using the phrase recognition result (detection result), the phrase feature calculation module 753 calculates at least one of the number of phrases, the number of phrases normalized by the utterance time, and the number of phrases that appear at every predetermined time interval. This is a module to calculate. These modules 752 to 753 are read by the CPU 410 into the application execution area 449 of the RAM 840 and executed.

  FIG. 8 is a flowchart for explaining the processing procedure of the information processing apparatus 600 according to the present embodiment. This flowchart is executed by the CPU 410 in FIG. 7 using the RAM 840, and implements the functional components of the information processing apparatus 600 in FIG.

  In step 801, the information processing apparatus 600 detects and recognizes the appearance of a phrase specified in advance. In step S803, the information processing apparatus 600 calculates the features of the recognized phrase. The characteristics of the phrase include, for example, the number of phrases that appear, the number of phrases that are normalized by the utterance time, the number of phrases that appear at predetermined time intervals, and the like.

According to the present embodiment, the phrase related to the satisfaction, dissatisfaction, brute force, and commitment to the charge is detected, the phrase feature is calculated, and the model is generated by adding the phrase feature. Can be predicted with high accuracy in consideration of the content of the statement. [Other Embodiments]
While the present invention has been described with reference to the embodiments, the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention. In addition, a system or an apparatus in which different features included in each embodiment are combined in any way is also included in the scope of the present invention.

  In addition, the present invention may be applied to a system composed of a plurality of devices, or may be applied to a single device. Furthermore, the present invention can also be applied to a case where an information processing program that implements the functions of the embodiments is supplied directly or remotely to a system or apparatus. Therefore, in order to realize the functions of the present invention on a computer, a program installed in the computer, a medium storing the program, and a WWW (World Wide Web) server that downloads the program are also included in the scope of the present invention. . In particular, at least a non-transitory computer readable medium storing a program for causing a computer to execute the processing steps included in the above-described embodiments is included in the scope of the present invention.

[Other expressions of embodiment]
A part or all of the above-described embodiment can be described as in the following supplementary notes, but is not limited thereto.
(Appendix 1)
Call storage means for acquiring and storing a call between a customer and an operator;
Utterance amount feature calculation means for calculating the stored utterance amount feature amount of the call;
Emotion recognition means for recognizing the emotion of the utterance interval;
A churn predicting means for predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
An information processing apparatus comprising:
(Appendix 2)
The emotion recognition means further calculates an emotion value of the recognized emotion,
Emotion feature calculating means for calculating an emotion feature that is a feature of the emotion from the emotion value;
Customer extraction model storage means for storing a customer extraction model that has been learned in advance for predicting and extracting a customer with a high churn risk based on the utterance amount feature amount and the emotion feature;
Further comprising
The information processing apparatus according to appendix 1, wherein the cancellation prediction unit predicts the cancellation risk based on the customer extraction model.
(Appendix 3)
The emotion feature calculating means includes:
A topic section detecting means for detecting a specific topic section included in the call;
Narrowing means for outputting the emotion value of the utterance section included in the specific topic section;
Feature calculating means for calculating an emotion feature that is a feature of the emotion from the emotion value output by the narrowing means;
The information processing apparatus according to appendix 2, further comprising:
(Appendix 4)
Phrase storage means for storing a phrase specified in advance;
Phrase recognition means for detecting and recognizing the appearance of a phrase stored in the phrase storage means in the call;
A phrase feature calculating means for calculating a phrase feature that is a feature of the phrase that has appeared;
Further comprising
The information processing apparatus according to any one of supplementary notes 1 to 3, wherein the cancellation prediction unit further predicts the cancellation risk based on the phrase feature.
(Appendix 5)
Any one of Supplementary notes 1 to 4, wherein the call storage unit detects an utterance section of the call, and stores the utterance by adding speaker information and utterance time information to the detected utterance section. The information processing apparatus according to item 1.
(Appendix 6)
The information processing apparatus according to any one of appendices 1 to 5, wherein the emotion includes at least one of valence, arousal, satisfaction, and expectation.
(Appendix 7)
The topic section includes at least one of opening in a call, closing, product / service description, fee description, solicitation, hearing, personal information confirmation, and office procedure,
The information processing apparatus according to any one of appendices 3 to 5, wherein the emotion includes at least one of valence and expectation.
(Appendix 8)
The information processing apparatus according to any one of appendices 2 to 7, wherein the emotion feature includes at least one of a ratio of an emotion value of the emotion value, an average value, a variance value, a median value, and a mode value.
(Appendix 9)
The phrase feature includes at least one of a phrase appearance number that is the number of appearances of the phrase, a normalized phrase appearance number obtained by normalizing the phrase appearance number by an utterance time, and a phrase appearance number for each predetermined time interval. The information processing apparatus according to any one of 1 to 7.
(Appendix 10)
A call storage step of acquiring and storing a call between a customer and an operator;
An utterance amount feature calculating step of calculating an utterance amount feature amount of the stored call;
An emotion recognition step for recognizing the emotion of the utterance interval;
A churn prediction step of predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
An information processing method including:
(Appendix 11)
A call storage step of acquiring and storing a call between a customer and an operator;
An utterance amount feature calculating step of calculating an utterance amount feature amount of the stored call;
An emotion recognition step for recognizing the emotion of the utterance interval;
A churn prediction step of predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
An information processing program that causes a computer to execute.

Claims (10)

Call storage means for acquiring and storing a call between a customer and an operator;
Utterance amount feature calculation means for calculating the stored utterance amount feature amount of the call;
Emotion recognition means for recognizing the emotion of the utterance interval;
A churn predicting means for predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
An information processing apparatus comprising: The emotion recognition means further calculates an emotion value of the recognized emotion,
Emotion feature calculating means for calculating an emotion feature that is a feature of the emotion from the emotion value;
Customer extraction model storage means for storing a customer extraction model that has been learned in advance for predicting and extracting a customer with a high churn risk based on the utterance amount feature amount and the emotion feature;
Further comprising
The information processing apparatus according to claim 1, wherein the cancellation prediction unit predicts the cancellation risk based on the customer extraction model. The emotion feature calculating means includes:
A topic section detecting means for detecting a specific topic section included in the call;
Narrowing means for outputting the emotion value of the utterance section included in the specific topic section;
Feature calculating means for calculating an emotion feature that is a feature of the emotion from the emotion value output by the narrowing means;
The information processing apparatus according to claim 2, further comprising: Phrase storage means for storing a phrase specified in advance;
Phrase recognition means for detecting and recognizing the appearance of a phrase stored in the phrase storage means in the call;
A phrase feature calculating means for calculating a phrase feature that is a feature of the phrase that has appeared;
Further comprising
The information processing apparatus according to claim 1, wherein the cancellation prediction unit further predicts the cancellation risk based on the phrase feature.   The call storage means detects an utterance section of the call, and stores the utterance by adding speaker information and utterance time information to the detected utterance section. The information processing apparatus according to claim 1.   The information processing apparatus according to claim 1, wherein the emotion includes at least one of valence, arousal, satisfaction, and expectation. The topic section includes at least one of opening in a call, closing, product / service description, fee description, solicitation, hearing, personal information confirmation, and office procedure,
The information processing apparatus according to claim 3, wherein the emotion includes at least one of valence and expectation.   The information processing apparatus according to claim 2, wherein the emotion feature includes at least one of a ratio, an average value, a variance value, a median value, and a mode value of the emotion values of the emotion values. A call storage step of acquiring and storing a call between a customer and an operator;
An utterance amount feature calculating step of calculating an utterance amount feature amount of the stored call;
An emotion recognition step for recognizing the emotion of the utterance interval;
A churn prediction step of predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
An information processing method including: A call storage step of acquiring and storing a call between a customer and an operator;
An utterance amount feature calculating step of calculating an utterance amount feature amount of the stored call;
An emotion recognition step for recognizing the emotion of the utterance interval;
A churn prediction step of predicting a churn risk in the call based on the utterance amount feature and the recognized emotion;
An information processing program that causes a computer to execute.

Images