WO2018045915A1 - Method and apparatus for adjusting service voice playback sequence - Google Patents

Abstract

Provided by the present application are a method and apparatus for adjusting a service voice playback sequence, the specific solution being: acquiring a user-related historical call record according to identity information of a user; determining a service selection priority of the user according to the user-related historical call record; and adjusting a service voice playback sequence of the user according to the service selection priority. Compared to existing technology, the present application may acquire a user-related historical call record according to identity information of a user, and further predict a service selection priority of the user during a current call on the basis of a historical call so as to rearrange playback sequence of a service voice to be currently played back to the user. Playback of a service voice of a high service selection priority is prioritized, so that an adjustment mode is flexible and the time spent by the user listening to the service voice may be shortened, thereby reducing the occupation rate of a call circuit, improving the utilization efficiency of the circuit and saving time cost of the user and a service provider.

Description

Method and device for adjusting business voice playing sequence Technical field

The present application relates to the field of information technology, and in particular, to a method and device for adjusting a service voice playing sequence.

Background technique

因此，对于一个需要咨询财务问题的用户，需要先听前面8种业务与其无关的业务语音之后，才能听到符合自己需要的业务语音，等待时间较长。At present, when the user calls the service call request service, the voice customer service system will play the preset voice to guide the user to perform subsequent operations. During this process, the order in which the user hears the service voice is fixed. For example, when a user dials the customer service phone of the cloud computing service provider, the call center customer service system automatically plays the service voice to inform the user of the service classification option. : Please press 1 for the cloud server problem, 2 for the database problem, 3 for the domain name problem, and press 9 for the financial and account questions. The order of service options that each user hears when making a call is the same. As shown in Figure 1, after the customer calls the customer service phone, the customer service system determines the account information of the user, and then plays the service voice in a fixed playback order for any user. For example, the time of playing the voice message of the cloud server problem is 1 For t ₁ , the time spent playing the “database problem 2” business voice is t ₂ , and the time t _{n of} playing the nth business voice is analogized, so that the waiting time for the user to listen to the nth business voice is

Therefore, for a user who needs to consult financial problems, it is necessary to listen to the business voices of the previous eight kinds of services and their unrelated business voices before they can hear the business voices that meet their needs, and the waiting time is longer.

The existing solution is: pre-customize a fixed service voice playing sequence for each user, for example, the bank's voice customer service system, and customize different services for each user according to the information filled in when the user opens the account and the opened service. voice. However, in this solution, when the user dials the customer service call at any time, the voice voice played by the voice customer service system is still fixed, and the customized service voice can only be modified when the account is opened or the customer service call is specifically made. Therefore, such a scheme is not flexible enough to adjust the order in which the business voices are played in real time according to the user's recent situation.

Application content

An object of the present invention is to provide a method and a device for adjusting a service voice play sequence, which are used to solve the problem that the service voice play sequence cannot be flexibly changed in the prior art.

To achieve the above objective, the present application provides a method for adjusting a service voice play sequence, where the method includes:

Obtaining a historical call record related to the user according to the identity information of the user;

Determining a service selection priority of the user according to a historical call record related to the user;

Adjusting the service voice play sequence of the user according to the service selection priority.

Further, before acquiring the historical call record related to the user according to the identity information of the user, the method further includes:

The identity information of the user is obtained based on the voice input of the user.

Further, acquiring the identity information of the user based on the voice input of the user includes:

Acquiring the feature speech in the voice input based on the voice input of the user;

Acquiring the identity information of the user according to the audio content and/or the spectral feature of the feature speech.

Further, the method further includes:

If the identity information of the user cannot be obtained, new identity information is generated for the user.

Further, the historical call record related to the user is obtained according to the identity information of the user, including:

Querying the historical call record of the user according to the identity information of the user;

If there is a historical call record of the user, the historical call record of the user is obtained as a historical call record related to the user.

Further, the historical call record related to the user is obtained according to the identity information of the user, including:

Querying the historical call record of the user according to the identity information of the user;

If the history call record of the user does not exist, obtaining a historical call record of the associated user as a historical call record related to the user, where the associated user is a user associated with any attribute of the user .

Further, determining, according to the historical call record related to the user, the service selection priority of the user, including:

Obtaining, according to the historical call record related to the user, the number or frequency of service selection in the historical call record in the preset time period;

The service selection priority of the user is determined in descending order according to the number or frequency of service selections in the preset time period.

Further, the method further includes:

The business voice is played to the user according to the adjusted business voice playing sequence.

Further, the method further includes:

Adjusting the user's business voice play content according to the historical call record associated with the user.

Further, adjusting the voice play content of the user according to the historical call record related to the user, including:

Obtaining, according to a historical call record associated with the user, a number or frequency of service selections in the historical call record within a preset time period;

The number of times of service selection or the service voice whose frequency is higher than the preset value in the preset time period is determined as the service voice play content of the user.

Further, determining, according to the historical call record related to the user, the service selection priority of the user, including:

Obtaining, according to a historical call record associated with the user, a number or frequency of service selections in the historical call record within a preset time period;

Determining, according to the number or frequency of service selections in the preset time period, the user selection priority of the user in descending order;

At least one service voice with a higher priority of the service preference is determined as the service voice play content of the user.

Further, the method further includes:

The service voice is played to the user according to the adjusted service voice play sequence and the service voice play content.

In addition, the present application also provides a method for adjusting an output order of candidate information, where the method includes:

Obtaining, according to the identity information of the user, a historical operation record of the associated user, where the associated user is a user associated with any attribute of the user;

The historical operation record of the root associated user determines the selection priority of the candidate information of the user;

Adjusting the candidate information output order of the user according to the selection priority of the candidate information of the user.

According to another aspect of the present application, there is also provided a device for adjusting a service voice playing sequence, the device comprising:

Querying means, configured to acquire a historical call record related to the user according to the identity information of the user;

a sorting device, configured to determine a service selection priority of the user according to a historical call record related to the user;

The sequence adjustment device is configured to adjust a service voice play sequence of the user according to the service selection priority.

Further, the device further includes:

The identity obtaining means is configured to acquire the identity information of the user based on the voice input of the user before acquiring the historical call record related to the user according to the identity information of the user.

Further, the identity obtaining means is configured to acquire feature voices in the voice input based on voice input of the user; and acquire identity information of the user according to audio content and/or spectrum features of the feature voice.

Further, the identity obtaining apparatus is further configured to generate new identity information for the user when the identity information of the user cannot be obtained.

Further, the querying device is configured to query the historical call record of the user according to the identity information of the user; and acquire the historical call record of the user when the historical call record of the user exists, as A historical call record related to the user.

Further, the querying apparatus is configured to query the historical call record of the user according to the identity information of the user; and acquire an association when there is no historical call record of the user A historical call record of the user as a historical call record associated with the user, wherein the associated user is a user associated with any of the attributes of the user.

Further, the sorting device is configured to acquire, according to a historical call record related to the user, a number or frequency of service selections in the historical call record in a preset time period; and according to the preset time period The number or frequency of selections determines the priority of the user's business selection in descending order.

Further, the device further includes:

The playing device is configured to play the business voice to the user according to the adjusted service voice playing sequence.

Further, the device further includes:

And content adjustment means, configured to adjust a service voice play content of the user according to a historical call record related to the user.

Further, the content adjustment apparatus is configured to acquire, according to a historical call record related to the user, a number or frequency of service selections in the historical call record in a preset time period; and the preset time period The number of times the service is selected or the frequency of the service voice whose frequency is higher than the preset value is determined to be the content of the user's business voice.

Further, the content adjustment apparatus is configured to acquire, according to a historical call record associated with the user, a number or frequency of service selections in the historical call record in a preset time period; Determining the number of times or frequency, determining the service selection priority of the user in descending order; and determining at least one service voice with a higher service preference priority as the service voice play content of the user.

Further, the device further includes:

The playing device is configured to play the service voice to the user according to the adjusted service voice playing sequence and the service voice playing content.

The application also provides a device for adjusting the output order of the candidate information, wherein the device includes:

a querying device, configured to acquire a historical operation record of the associated user according to the identity information of the user, where the associated user is a user associated with any attribute of the user;

a sorting device, configured to determine a priority of selection of the candidate information of the user by a historical operation record of the root associated user;

And a sequence adjustment device, configured to adjust an output order of the candidate information of the user according to a selection priority of the candidate information of the user.

Compared with the prior art, the solution provided by the present application can acquire a historical call record related to the user according to the identity information of the user, and further predict the priority of the service selection of the user at the current call based on the historical call. Therefore, the order of playing the service voices played back to the user is reordered, and the service voice with a higher priority is preferentially played, and the adjustment mode is more flexible, and the time for the user to listen to the service voice can be shortened, and the occupation of the telephone line can be reduced. The rate increases the utilization efficiency of the line and saves the time cost of users and service providers.

DRAWINGS

Other features, objects, and advantages of the present application will become more apparent from the detailed description of the accompanying drawings.

1 is a schematic diagram of interaction between a voice customer service system and a user in the prior art;

2 is a schematic diagram of interaction between a voice customer service system and a user after the service voice play sequence is adjusted by using the method provided in this embodiment;

FIG. 3 is a schematic diagram of interaction between the voice customer service system and the user after adjusting the service voice play content and the play sequence by using the method provided in this embodiment;

FIG. 4 is a schematic diagram of interaction between the voice customer service system and the user after adjusting the service voice play content and the play sequence by using another method provided by this embodiment;

FIG. 5 is a flow chart of interaction between a voice customer service system and a user in a voice customer service scenario in the cloud computing field;

6 is a flow chart of another interaction between a voice customer service system and a user in a voice customer service scenario in the cloud computing field;

7 is a flow chart of interaction between a voice customer service system and a user in a voice customer service scenario in the B2C e-commerce field;

The same or similar reference numerals in the drawings denote the same or similar components.

detailed description

The present application is further described in detail below with reference to the accompanying drawings.

In a typical configuration of the present application, the terminal, the device of the service network, and the trusted party each include one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include non-persistent memory, random access memory (RAM), and/or non-volatile memory in a computer readable medium, such as read only memory (ROM) or flash memory. Memory is an example of a computer readable medium.

Computer readable media includes both permanent and non-persistent, removable and non-removable media. Information storage can be implemented by any method or technology. The information can be computer readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, A magnetic tape cartridge, magnetic tape storage or other magnetic storage device or any other non-transportable medium that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory computer readable media, such as modulated data signals and carrier waves.

The embodiment of the present application provides a method for adjusting a service voice play sequence, where the method specifically includes the following steps:

Step S101: Acquire a historical call record related to the user according to the identity information of the user;

Step S102: Determine a service selection priority of the user according to a historical call record related to the user.

Step S103: Adjust a service voice play sequence of the user according to the service selection priority.

The solution is particularly applicable to a voice call customer service system of a call center, capable of acquiring a historical call record related to the user according to the identity information of the user, and further based on the historical call to the user Predicting the priority of the service selection at the time of the current call, thereby reordering the playing order of the service voices played to the user, and preferentially playing the service voice with a high priority, and the adjustment mode is flexible and capable of The time for the user to listen to the service voice is shortened, the occupation rate of the telephone line is reduced, the utilization efficiency of the line is improved, and the time cost of the user and the service provider is saved.

Here, the identity information refers to an identifier capable of determining the identity of the user, such as a name, an account ID (identification), an account alias, a number, or a combination thereof. In different usage scenarios, different forms of identity information may be used. For example, for a scenario of an individual user, the identity information may be a number that is pre-assigned to the user, and the identity of the user may be determined by the number.

For another application scenario, such as a customer service scenario in the cloud computing field, the user's account is usually a corporate account, corresponding to a team of the enterprise. Within the team, people usually have a division of labor, so when someone on the team calls the customer service to request service, the services they need are different. For example, the person in charge of the database in the team and the person in charge of finance use the same account number. If the voice sequence is adjusted only by the voice content identification account ID, the waiting time process of some personnel may be caused. Therefore, for such a scenario, the identity information of the user may be combined by multiple information, for example, the enterprise account ID may be used as the primary identity information, and the employee name may be used as the secondary identity information, and the primary identity information is used. The way identity information is combined is as complete identity information. The essence is still the same as the foregoing individual user. In the process of determining the identity information, the various information contained in the identity information may be separately processed, and only when all kinds of information in the identity information are clear, the identity The information was finalized.

In the scenario of a voice call, the identity information of the user can be obtained based on the voice input of the user. Therefore, the method for adjusting the service voice playing sequence may be further improved, that is, before the historical call record related to the user is obtained according to the user's identity information, the method further includes: acquiring the User identity information. Because in a voice call scenario, users often make calls by telephone, mobile phone, and the like. The user's voice input is used to obtain the user's identity information, which can prevent the user from operating the phone, the phone's button or the touch screen to manually input the identity information during the call process, thereby effectively improving the user's convenience.

As a preferred embodiment, the user's identity letter is obtained based on the user's voice input. Specifically, the following processing steps may be adopted: First, the feature voice in the voice input is acquired based on the voice input of the user. Then, the identity information of the user is obtained according to the voice content and/or the spectrum feature of the feature voice.

For example, in the scenario where the user dials the customer service phone, not all the voices sent by the user have a specific meaning or are easy to identify, such as "hello" and "hello" which are often used when making a call. In order to improve the accuracy of determining the identity information of the user, a part of the voice input of the user may be extracted as a feature voice to perform a subsequent recognition process. For example, when the user dials the customer service phone, the voice customer service system can prompt the user to issue a specific voice through the preset voice, thereby identifying the voice that the user sends at this time as the feature voice. For example, the voice customer service system can prompt the user to "please report your account ID". At this time, the user will report his/her account ID according to the requirements of the preset voice, and then the voice customer service system can use the voice input of the account ID as the feature voice. , performing identification processing to determine identity information of the user.

Specifically, based on the feature speech, at least two aspects of the content, that is, the speech content of the feature speech and the spectral feature of the feature speech, the speech content represents the meaning expressed by the feature speech, and the spectral feature represents the characteristics of the sound itself. In the scenario of an enterprise user, the preferred method is to determine the complete identity information including the enterprise account ID and the employee name according to the voice content and the spectrum feature of the feature voice. In this embodiment, a feasible manner is provided. For example, when an employee of the enterprise dials a voice service call, the voice customer service system prompts “please report your account ID”, and the employee reports the enterprise account ID by voice. ZH123456", after the voice customer service system obtains the feature voice, the voice content is obtained through voice recognition, thereby obtaining the enterprise account ID of the incoming user.

At the same time, the spectrum features are obtained through spectrum analysis and compared with the spectrum feature records of all employees under the enterprise account ID stored in the database. For a specific spectral feature matching method, different algorithms may be implemented, for example, extracting multiple feature values by Fourier transform for comparison, or performing vectorization processing and vector feature matching, and the like. If the spectrum feature record whose similarity exceeds the threshold can be matched, it can be determined which incoming employee is the specific employee in the enterprise, thereby obtaining complete identity information, that is, the enterprise account ID + the employee name. And when matching When recording to multiple spectral features, the one with the highest similarity can be selected as the matching result.

As another feasible implementation manner, the user can report the name of the corresponding person corresponding to his or her identity while reporting the enterprise account ID, thereby realizing multi-level in the customer service scene in the cloud computing field only by recognizing the voice content. Identification of identity tags. However, in this solution, the user needs to provide more information (including the enterprise account ID and the name of the person). In order to improve the convenience of the user, the user may use the foregoing method to report the enterprise account ID only, and determine the enterprise by means of voice content identification. The account ID is further determined by the spectrum feature identification method.

In the scenario of the aforementioned individual user, both the voice content and the spectrum feature have unique identification, and the identity information of a certain individual user can be determined separately. Therefore, the processing of the individual user scene can refer to the process, and only the processing process can be simplified accordingly.

In the above solution, if the identity information of the user cannot be obtained, new identity information is generated for the user. For example, in the foregoing scenario, the voice content is obtained by voice recognition, thereby obtaining the enterprise account ID of the incoming user, but when the obtained spectrum feature is matched with the spectrum feature record in the database, the similarity is not matched. The spectrum characteristic record of the threshold indicates that the incoming user has not dialed the customer service phone before, so the user cannot obtain the complete identity information of the user, and the incoming user can be provided with an employee name or the voice customer service system automatically according to the preset rule. Generate an employee name for the user, thereby generating identity information for the enterprise account ID + employee name. For example, under a certain enterprise account ID, the person in charge of the database A, the person in charge of the cloud server B, etc. have repeatedly dialed the customer service telephone to consult the relevant field, and the financial staff C dialed the customer service phone for the first time. Since the second database does not contain the spectrum feature record of the person C, the new identity information is generated for the person C under the enterprise account ID, and the spectrum feature record of the person C is saved in the database, so that the person C can dial again. Customer service can be successfully matched.

In step S101, the historical call record associated with the user includes two types, one is the user's own historical call record, and the other is the historical call record of the associated user. The associated user refers to a user that is associated with any attribute of the user. The attribute of the user is related information for arbitrarily classifying the user, for example, any item for constructing a user portrait. For example, for an individual user, it can be the user’s age, occupation, Interest categories, etc., and for enterprise users, may be products purchased by the enterprise user, type of business, main business of the enterprise, and the like.

In order to ensure that the service selection priority determined according to the historical call record has high accuracy, when the historical call record related to the user is acquired according to the identity information of the user, the user is first queried according to the identity information of the user. Historical call record. According to the results of the query, different processing is performed. In the first case, if the result of the query is that there is a historical call record of the user, the historical call record of the user is obtained and used as a historical call record related to the user. In the second case, if the result of the query is: there is no historical call record of the user, the historical call record of the associated user is obtained as a historical call record related to the user.

Since the historical call record includes the service selected by a certain user in the call, the historical call record of the user itself or the related user may select the service providing prediction basis when the user calls this time. Therefore, regardless of whether the current incoming user has a historical call record, the user can accurately predict the service selection priority according to the corresponding historical data, thereby providing an appropriate service voice playing sequence and reducing the waiting time of the user. .

After the historical call record related to the user is obtained, the service selection priority of the user may be determined according to the following manner. First, the number or frequency of service selections in the historical call record within a preset time period is obtained according to a historical call record associated with the user. For example, for a certain user, a total of 100 historical call records of the user in the most recent week can be queried, among which 50 financial call records are selected for consulting financial problems, and 30 historical call records are selected for consulting cloud server problems, 20 The consulting database problem is selected in the historical call record, so that the number of times of various business selections in the historical call record can be directly obtained. The frequency refers to the number of times per unit time, which can be calculated mathematically according to the number of times obtained. For example, if the unit time is set to day, the frequency of selecting a financial question is 7.143 times/day. When the number or frequency is obtained, the service selection priority of the user may be determined in descending order based on the number or frequency of service selections in the preset time period. That is, in the historical call record, the higher the number or frequency of selection of a certain service, the higher the priority of the user selecting the service. For the preceding scenario, the service selection priority is: financial related services > cloud server related services > Database related business > Other.

For the foregoing second case, it is found that there is no historical call record for an incoming user, and other users who purchase the same product may be used as an associated user according to a preset association rule. Thus, 100 related users are found. In all historical call records of the last week of these associated users, the domain name problem is selected among 6000 historical call records, and the cloud server problem is selected among 3000 historical call records, 1000 history. A content distribution network issue was selected in the call log. Based on the number of times or the frequency, it can be determined that the priority of the current inbound user's service selection for each type of service is from high to low: domain name related service > cloud server related service > content distribution network related service > other.

如果采用本申请提供的方案，通过声音匹配到其历史呼叫记录中80％是咨询财务问题，则将原先第9项播放的财务与账号的业务语音向动态调整至第一位播放，使该财务人员的等待时间能够缩短到t₉，如图2所示。Further, after the service voice playing sequence is adjusted according to the service selection probability, the solution further includes: playing a service voice to the user according to the adjusted service voice playing sequence. Using this method can greatly shorten the waiting time for the user's incoming call. For a financial caller's incoming call, if the prior art has a fixed play order, the financial person needs to listen to the first 8 business voices before hearing the match. Demand business voice, waiting time is

If the solution provided by the present application is used, and 80% of the historical call records are matched to the financial question by the voice, the financial voice of the original 9th item and the business voice of the account are dynamically adjusted to the first place to play, so that the financial The waiting time for personnel can be shortened to t ₉ , as shown in Figure 2.

As another preferred implementation manner, the solution may adjust the service voice play content of the user according to the historical call record related to the user while adjusting the service voice play sequence of the user. For example, in the scene of the customer service call, the content of the service voice is generally preset play content, for example, the preset voice corresponding to the nine services of the buttons 1 to 9, respectively, when the content of the service voice is adjusted, The number of preset voices is reduced. For example, based on the history call record, the possibility that the user selects some services is very small, and the preset voices corresponding to the services may not be played; in addition, the conventional services may not be played. Voice, while playing other forms of voice, or accessing the manual service after playing the prompt voice, the specific processing strategy can be set according to different application scenarios.

The embodiments of the present application provide the following specific processing manners when the number of preset voices to be played is reduced.

In the first mode, first, according to the historical call record related to the user, the number or frequency of service selection in the historical call record in the preset time period is obtained. For the specific processing procedure of this step, refer to the foregoing process for determining the priority of the service selection, and the description is not repeated here. After obtaining the number or frequency of service selections, the number of times of service selection or the service voice whose frequency is higher than the preset value in the preset time period may be determined as the service voice play content of the user.

Still taking the foregoing scenario as an example, in a history call record of a user in the last week, 50 financial call records are selected for consulting financial problems, 30 historical call records are selected for consulting cloud server problem, and 20 historical call records are selected. I chose the consulting database question. By setting a preset value, you can filter out services with a higher priority. For example, you can set 1/20 of the total number of historical call records as the preset value, that is, the service that does not exceed 5 times in the most recent week. The business voice will not be played. In combination with the sorting result of the business voice playback order, only three of the business voices will be played, that is, "Please press 9 for financial problems, 1 for cloud server problems, 2 for database problems, and no corresponding button 1 will be played." All nine business voices of ~9, as shown in Figure 3.

In the second mode, the content of the service voice playback content of the user may be determined based on the result of the service selection priority that has been acquired. That is, first, according to the historical call record related to the user, the number or frequency of service selection in the historical call record in the preset time period is obtained. Then, according to the number or frequency of service selections in the preset time period, the service selection priority of the user is determined in descending order. Finally, at least one service voice with a higher priority of the service preference is determined as the service voice play content of the user.

In an actual scenario, the specific number of the at least one service voice may be set according to actual requirements, and may be less than the total number of service voices and greater than or equal to one. Taking a user who does not have a history call record as an example, the 1000 associated users of the user select a domain name problem and 3000 history call records in 4000 historical call records in all historical call records of the last week of the associated users. The cloud server problem was selected, the content distribution network problem was selected among 100 historical call records, the record problem was selected in 2000 historical call records, the database problem was selected in 50 historical call records, and 2100 historical call records were selected. Financial question question. If the number of business voices played last time is set to 4, the adjusted result is: "Please press 3 for the domain name problem, 1 for the cloud server problem, 9 for the financial problem, and 5 for the record problem." 4 is shown.

Similarly, after adjusting the service voice play sequence and the service voice play content according to the service selection probability, the solution further includes: playing the service voice to the user according to the adjusted service voice play sequence and the service voice play content. Thereby, the playing time of the business voice can be reduced, and the occupation time of the communication line can be reduced.

The following provides a complete solution of the present application in combination with an example of a voice customer service scenario in the field of cloud computing. The users in this field are mainly enterprise users, that is, the user identity information is composed of at least two levels, such as enterprise account ID + employee name, company name + employee number. The user identity information in this example is the financial staff F1 of the enterprise C1. When the financial personnel F1 calls the customer service telephone of the service provider's voice customer service system, the specific information processing flow is as shown in FIG. 5 .

In step S501, the voice customer service system first informs the user to provide the identity information by voice in a preset voice manner. Since the specific personnel can be identified by the spectral characteristics of the voice, the voice customer service system can only request the user to provide the enterprise account ID. Of course, if the user has other input methods, the user may be prompted to provide his or her identity information in other ways, such as using a touch screen of the mobile phone, a button of the phone to input the enterprise account ID, and a personnel number, etc., in order to determine the identity of the user.

In step S502, the user issues a voice according to the requirement, and the content of the voice is a corporate account ID.

Step S503, after the voice customer service system obtains the voice, by identifying the voice content, the enterprise account ID is “ZH123456”, and the spectrum feature of the voice is recognized.

Step S504, the voice customer service system performs a query in the database based on the spectrum feature, searches for the spectrum feature record matching the spectrum feature, and queries the database for the spectrum feature record of the financial person F1, thereby determining that the voice is from the voice. The financial personnel F1 of the enterprise C1 determines the identity information of the user. During the search process, the query range of the spectrum feature matching can be narrowed based on the enterprise account ID, and the query is only performed in the spectrum feature record of the relevant personnel of the enterprise to improve the query efficiency.

Step S505, after the voice customer service system determines the identity information, the historical call record of the user may be searched according to the identity information. The historical call record shows that the user has dialed the customer service The preference of the business question selected during the call, the preference can indicate the possibility of selecting a certain type of service problem when the user calls this call, and the higher the probability, the higher the priority of the service selection of the service. For example, the financial staff F1 has called 70 customer service calls in the past month, that is, there are 70 historical call records in the most recent month, of which 60 are consulting financial questions, 6 are consulting cloud server problems, and 4 are consulting database problems. .

Step S506, the voice customer service system performs sorting according to the number of times the user selects in the historical call record, thereby determining the service selection priority of the user. That is, financial related business > cloud server related business > database related business > other.

Step S507: The voice customer service system reorders the order of the service voice according to the financial related service > cloud server related service > database related service > other priority order, "Please press 9 for financial problems, 1 for cloud server problems, database Please press 2,...". Since there are no historical call records for other business issues, the order between other business voices may not be adjusted.

Step S508, the voice customer service system plays the business voice to the user for the user to select. If the financial officer F1 still needs to ask financial related questions, he can make a choice after hearing the first voice.

In another example, the identity information of the user is the technical staff T1 of the enterprise C2. When the technician T1 calls the customer service phone of the voice service system of the service provider, the specific information processing flow is as shown in FIG. 6.

Step S601, after the phone is connected, the user actively reports the enterprise account ID in the form of voice.

Step S602, after the voice customer service system obtains the voice, by identifying the voice content, the enterprise account ID is “ZH654321”, and the spectrum feature of the voice is recognized.

Step S603, the voice customer service system performs a query in the database based on the spectrum features, and searches for a spectrum feature record that matches the spectrum feature. However, if the matching spectrum feature record is not found, it is impossible to determine which person the user is from the enterprise C2, that is, the identity information of the user cannot be obtained. Thus, the person can be assigned a person number according to a preset rule, thereby generating a complete identity information for the person.

Step S604, the voice customer service system determines the identity information, and can find the identity information according to the identity information. User's history call record. Since the identity information is newly generated this time, the history call record for the user cannot be found. At this time, the related user may be searched based on an attribute of the user, for example, according to the cloud computing product used by the enterprise, and the enterprise C3, C4, C5, etc. that purchased the same cloud computing product are found. The historical call record of the user is replaced with a historical call record for all of the personnel of these enterprises to determine the priority of the business selection.

Step S605, the voice customer service system performs sorting according to the number of selections of the associated users in the historical call record, thereby determining the service selection priority of the user.

In step S606, the voice customer service system adjusts the service voice play content according to the number of times the associated user is selected in the historical call record, and only selects the service voice whose selection frequency is higher than a certain preset value as the service voice play content. For example, in this example, the service voice play content ultimately includes: financial related services, cloud server related services, database related services, and business voices of the related services.

Step S607, adjusting the service voice playing sequence. Since the service voice play content includes only four service voices, when the play sequence is adjusted, only the four service voices can be adjusted. At this time, the content that the voice customer service system will play is: "Please press 9 for financial problems, 1 for cloud server problems, 2 for database problems, and 5 for filing problems."

The following provides a complete solution provided by the present application in conjunction with an example of a voice service scenario in the B2C (BusinesstoCustomer) e-commerce field. The users in this field are mainly individual users, that is, the user identity information is composed only of one type of information, such as user ID, user name, user binding mailbox, and the like. In this example, the user U1 dials the logistics problem of the voice customer service consulting order of the e-commerce platform, and the specific information processing flow is shown in FIG. 7 .

Step S701, after the phone is connected, the user reports his account ID in the form of voice.

Step S702: After the voice customer service system obtains the voice, the voice content can be identified, thereby obtaining the account ID of the user as “USER123”, thereby determining the identity information of the user. In addition, the identity information of the user can also be determined by identifying the spectral features of the speech and querying the matching spectral feature records in the database.

Step S703, after the voice customer service system determines the identity information, the historical call record of the user may be searched according to the identity information. According to the specific query result (query to the history of the user) Calling the record or failing to query the historical call record of the user), the order of the business voice can be adjusted according to the historical call record of the user or the historical call record of the associated user. In this scenario, the associated user may be a user who has similar spending habits with the user, such as purchasing a similar product, having similar spending power, the user being of the same age, and the like.

Step S704, if there is no historical call record of the user, the voice customer service system obtains 50 related users to dial 240 customer service calls in the latest month, 170 of which are logistics problems, 35 times are consulting account questions, 25 times It is a question of asking for a refund, and 10 times is a question of consulting a transaction.

Step S705: The voice customer service system performs sorting according to the number of times the associated user selects in the historical call record, thereby determining the service selection priority of the user. That is, logistics related business > account related business > refund related business > transaction related business.

Step S706, the voice customer service system adjusts the service voice according to the foregoing sequence, and obtains the current service voice play sequence.

Step S707, the voice customer service system broadcasts the business voice to the user, specifically, "Please press 3 for the logistics problem, 1 for the account number, 2 for the refund question, and 4 for the transaction problem."

In addition, the embodiment of the present application further provides a method for adjusting an output order of an alternative information, which is applicable not only to a voice playing scene, but also to other forms of outputting a scene, such as display of alternative information such as text and graphics. Output. Specifically, the method includes:

Step S801: Acquire a historical operation record of the associated user according to the identity information of the user, where the associated user is a user that is associated with any attribute of the user.

Step S802, the historical operation record of the root associated user determines the selection priority of the candidate information of the user. The candidate information may be in different forms according to different scenarios, for example, in a voice customer service scenario, the candidate information may be a service voice played to the user, and the customer service is provided through an instant online chat tool (eg In the scenario where the Taobao seller provides customer service to the Taobao seller through Ali Wangwang, the alternative information may also be text information.

Step S803, adjusting an order of outputting the candidate information of the user according to a selection priority of the candidate information of the user.

Take Taobao sellers as an example of how to provide customer service to Taobao sellers through Ali Wangwang. The method of adjusting the output order of the alternative information will be described in detail. At present, in the process of shopping, if the buyer needs to provide customer service, in addition to calling the seller's phone, the buyer can also send a message to the seller through an instant online chat tool, such as Aliwangwang. In order to improve processing efficiency, sellers often set up automatic reply or automatically send chat text, for example, when the buyer opens a chat window with the seller's customer service staff, automatically send the corresponding text options, such as "Refund, please reply 1; return, Please reply 2; exchange, please reply 3; order modification, please reply 4; pre-sale consultation, please reply 5." When the text of the alternative information is longer, or the information is more complicated, the buyer needs to view the alternative information one by one. The time will also be extended. If the problem that the buyer actually wants to consult is in the back position, it will take a long time to find the corresponding reply option. For example, if a buyer wants to make a pre-sales consultation, they need to go through the fifth article one by one to know that they need to reply 5 in the chat window.

In this scenario, since the buyer's identity information can be determined when the buyer establishes a chat window with the buyer, the identity information of the buyer can be directly obtained. Of course, for other scenarios, the buyer's identity information can also be obtained based on the buyer's input, for example, prompting the buyer to input his or her identity information through text, or sending a QR code image indicating his identity.

After adopting the method for adjusting the output order of the alternative information provided by the embodiment of the present application, the candidate information automatically sent by the seller to the buyer may be recorded according to historical operations of the associated user (such as other buyers who have purchased the same product), for example, recently. Of the 150 historical operation records of 100 related users in a month, 100 selected pre-sales consultation, 30 selected order modification, 15 selected replacement, 4 selected refund, and 1 selected return . Therefore, it can be determined that the selection priority of the candidate information of the user is: pre-sale consultation > order modification > exchange > refund > return, and then adjust the display of the automatically issued alternative information output order: "pre-sales consultation Please reply 5; order modification, please reply 4; exchange, please reply 3; refund, please reply 1; return, please reply 2.) This makes the buyer who has consulted have a higher probability in the first few The required information is obtained in an alternative message, thereby shortening the consultation time of the buyer.

In this scenario, the output content of the candidate information may be further adjusted. For the specific manner, reference may be made to the foregoing adjustment manner of the service voice playback content.

Based on the same inventive concept, the voice playback of the service is also provided in the embodiment of the present application. The device of the sequence, the device for adjusting the output order of the candidate information, the method corresponding to the device is the corresponding method in the foregoing embodiment, and is similar to the principle of solving the problem by the method, so the implementation of the device can refer to the corresponding method. Implementation, repetition will not be repeated.

The specific structure of the device for adjusting the service voice playing sequence includes a querying device, a sorting device, and a sequence adjusting device. Specifically, the querying device is configured to acquire a historical call record related to the user according to the identity information of the user; and the sorting device is configured to determine the service selection of the user according to the historical call record related to the user The priority adjustment device is configured to adjust a service voice play sequence of the user according to the service selection priority.

Further, the identity obtaining means is configured to acquire the identity information of the user based on the voice input of the user before acquiring the historical call record related to the user according to the identity information of the user.

The specific processing procedure of the identity obtaining apparatus includes acquiring feature voices in the voice input based on voice input of the user; and acquiring identity information of the user according to audio content and/or spectrum features of the feature voice.

Further, the identity obtaining apparatus is further configured to generate new identity information for the user when the identity information of the user cannot be obtained.

Further, the querying device is configured to query the historical call record of the user according to the identity information of the user; and acquire the historical call record of the user when the historical call record of the user exists, as A historical call record related to the user.

In another case, the querying apparatus is configured to query the historical call record of the user according to the identity information of the user; and acquire the historical call record of the associated user when the historical call record of the user does not exist. And as a historical call record associated with the user, wherein the associated user is a user associated with any attribute of the user.

The sorting device is configured to acquire, according to a historical call record associated with the user, a number or frequency of service selections in the historical call record in a preset time period; and a number of times of service selection according to the preset time period Or frequency, determining the priority of the user's service selection in descending order.

In addition, the device also includes:

The playing device is configured to play the business voice to the user according to the adjusted service voice playing sequence.

As another preferred embodiment, the device further includes:

And content adjustment means, configured to adjust a service voice play content of the user according to a historical call record related to the user.

Specifically, the content adjusting apparatus is configured to acquire, according to a historical call record related to the user, a number or frequency of service selections in the historical call record in a preset time period; and the preset time period The number of times the service is selected or the frequency of the service voice whose frequency is higher than the preset value is determined to be the content of the user's business voice.

In another implementation manner of the content adjustment apparatus, the method is configured to acquire, according to a historical call record related to the user, a number or frequency of service selections in the historical call record in a preset time period; The number or frequency of service selections in the time period, determining the service selection priority of the user in descending order; and determining at least one service voice with a higher service preference priority as the service voice of the user Play the content.

Further, the playing device is configured to play a business voice to the user according to the adjusted service voice playing sequence and the service voice playing content.

The embodiment of the present application further provides an apparatus for adjusting an output order of an alternative information, where the apparatus includes a querying device, a sorting device, and a sequence adjusting device. The querying device is configured to acquire a historical operation record of the associated user according to the identity information of the user, where the associated user is a user associated with any attribute of the user; and the sorting device is used for the root associated user The historical operation record determines a selection priority of the candidate information of the user; the sequence adjustment means is configured to adjust an output order of the candidate information of the user according to a selection priority of the candidate information of the user.

In summary, the solution provided by the embodiment of the present application can acquire a historical call record related to the user according to the identity information of the user, and further predict the priority of the service selection of the user at the current call based on the historical call. Therefore, the order of playing the service voices played back to the user is reordered, and the service voice with a higher priority is preferentially played, and the adjustment mode is more flexible, and the time for the user to listen to the service voice can be shortened, and the occupation of the telephone line can be reduced. The rate increases the utilization efficiency of the line and saves the time cost of users and service providers. At the same time, after the solution is improved, it can also be applied to other application scenarios other than the voice customer service scene.

It should be noted that the present application can be implemented in software and/or a combination of software and hardware, for example, using an application specific integrated circuit (ASIC), a general purpose computer, or any other similar hardware device. In one embodiment, the software program of the present application can be executed by a processor to implement the steps or functions described above. Likewise, the software programs (including related data structures) of the present application can be stored in a computer readable recording medium such as a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some of the steps or functions of the present application may be implemented in hardware, for example, as a circuit that cooperates with a processor to perform various steps or functions.

In addition, a portion of the present application can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide a method and/or technical solution in accordance with the present application. The program instructions for invoking the method of the present application may be stored in a fixed or removable recording medium, and/or transmitted by a data stream in a broadcast or other signal bearing medium, and/or stored in a The working memory of the computer device in which the program instructions are run. Herein, an embodiment in accordance with the present application includes a device including a memory for storing computer program instructions and a processor for executing program instructions, wherein when the computer program instructions are executed by the processor, triggering The apparatus operates based on the aforementioned methods and/or technical solutions in accordance with various embodiments of the present application.

It is obvious to those skilled in the art that the present application is not limited to the details of the above-described exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application. Therefore, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the invention is defined by the appended claims instead All changes in the meaning and scope of equivalent elements are included in this application. Any reference signs in the claims should not be construed as limiting the claim. In addition, it is to be understood that the word "comprising" does not exclude other elements or steps. A plurality of units or devices recited in the device claims may also be implemented by a unit or device by software or hardware.

Claims (26)

A method for adjusting a service voice play sequence, wherein the method includes: Obtaining a historical call record related to the user according to the identity information of the user; Determining a service selection priority of the user according to a historical call record related to the user; Adjusting the service voice play sequence of the user according to the service selection priority. The method according to claim 1, wherein before the historical call record related to the user is obtained according to the identity information of the user, the method further includes: The identity information of the user is obtained based on the voice input of the user. The method according to claim 2, wherein the obtaining the identity information of the user based on the voice input of the user comprises: Acquiring the feature speech in the voice input based on the voice input of the user; Acquiring the identity information of the user according to the audio content and/or the spectral feature of the feature speech. The method of claim 2, wherein the method further comprises: If the identity information of the user cannot be obtained, new identity information is generated for the user. The method according to any one of claims 1 to 4, wherein the historical call record related to the user is obtained according to the identity information of the user, including: Querying the historical call record of the user according to the identity information of the user; If there is a historical call record of the user, the historical call record of the user is obtained as a historical call record related to the user. The method according to any one of claims 1 to 4, wherein the historical call record related to the user is obtained according to the identity information of the user, including: Querying the historical call record of the user according to the identity information of the user; If the history call record of the user does not exist, obtaining a historical call record of the associated user as a historical call record related to the user, where the associated user is a user associated with any attribute of the user . The method according to any one of claims 1 to 4, wherein The user-related historical call record determines the priority of the user's service selection, including: Obtaining, according to the historical call record related to the user, the number or frequency of service selection in the historical call record in the preset time period; The service selection priority of the user is determined in descending order according to the number or frequency of service selections in the preset time period. The method of the method of any of claims 1 to 4, wherein the method further comprises: The business voice is played to the user according to the adjusted business voice playing sequence. The method of claim 1 wherein the method further comprises: Adjusting the user's business voice play content according to the historical call record associated with the user. The method of claim 9, wherein adjusting the user's business voice play content according to the historical call record associated with the user comprises: Obtaining, according to a historical call record associated with the user, a number or frequency of service selections in the historical call record within a preset time period; The number of times of service selection or the service voice whose frequency is higher than the preset value in the preset time period is determined as the service voice play content of the user. The method of claim 9, wherein determining a priority of the user's service selection based on a historical call record associated with the user comprises: Obtaining, according to a historical call record associated with the user, a number or frequency of service selections in the historical call record within a preset time period; Determining, according to the number or frequency of service selections in the preset time period, the user selection priority of the user in descending order; At least one service voice with a higher priority of the service preference is determined as the service voice play content of the user. The method according to the method of any one of claims 9 to 11, wherein the method further comprises: According to the adjusted business voice playing sequence and the business voice playing content to the user Play business voice. A method of adjusting an order of outputting an alternative information, wherein the method comprises: Obtaining, according to the identity information of the user, a historical operation record of the associated user, where the associated user is a user associated with any attribute of the user; The historical operation record of the root associated user determines the selection priority of the candidate information of the user; Adjusting the candidate information output order of the user according to the selection priority of the candidate information of the user. A device for adjusting a service voice playing sequence, wherein the device includes: Querying means, configured to acquire a historical call record related to the user according to the identity information of the user; a sorting device, configured to determine a service selection priority of the user according to a historical call record related to the user; The sequence adjustment device is configured to adjust a service voice play sequence of the user according to the service selection priority. The device of claim 14, wherein the device further comprises: The identity obtaining means is configured to acquire the identity information of the user based on the voice input of the user before acquiring the historical call record related to the user according to the identity information of the user. The device according to claim 15, wherein the identity obtaining means is configured to acquire feature speech in the voice input based on a voice input of the user; and acquire the audio content and/or the spectrum feature according to the feature voice The identity information of the user. The device according to claim 15, wherein the identity obtaining means is further configured to generate new identity information for the user when the identity information of the user cannot be obtained. The device according to any one of claims 14 to 17, wherein the querying means is configured to query a historical call record of the user according to the identity information of the user; and in the presence of the historical call record of the user At the time, the historical call record of the user is obtained as a historical call record associated with the user. Apparatus according to any one of claims 14 to 17, wherein said inquiry is loaded Locating, for querying the historical call record of the user according to the identity information of the user; and acquiring a historical call record of the associated user when there is no historical call record of the user, as a historical call related to the user A record, wherein the associated user is a user associated with any of the attributes of the user. The device according to any one of claims 14 to 17, wherein the sorting means is configured to acquire, according to a historical call record associated with the user, a service selection in the historical call record within a preset time period The number of times or frequency; and determining the priority of the user's service selection in descending order according to the number or frequency of service selections within the preset time period. The device of the device of any one of claims 14 to 17, wherein the device further comprises: The playing device is configured to play the business voice to the user according to the adjusted service voice playing sequence. The device of claim 14, wherein the device further comprises: And content adjustment means, configured to adjust a service voice play content of the user according to a historical call record related to the user. The device according to claim 22, wherein the content adjusting means is configured to acquire the number or frequency of service selections in the historical call record within a preset time period according to a historical call record associated with the user; The number of times of service selection or the service voice whose frequency is higher than the preset value in the preset time period is determined as the service voice play content of the user. The device according to claim 22, wherein the content adjustment means is configured to acquire, according to a historical call record associated with the user, the number or frequency of service selections in the historical call record within a preset time period; Determining the service selection priority of the user in a high-to-low order, and determining at least one service voice with a higher service preference priority as the said number or frequency of service selection in the preset time period User's business voice playback content. The device of the device of any one of claims 22 to 24, wherein the device further comprises: The playing device is configured to play the service voice to the user according to the adjusted service voice playing sequence and the service voice playing content. A device for adjusting an order of outputting an alternative information, wherein the device comprises: a querying device, configured to acquire a historical operation record of the associated user according to the identity information of the user, where the associated user is a user associated with any attribute of the user; a sorting device, configured to determine a priority of selection of the candidate information of the user by a historical operation record of the root associated user; And a sequence adjustment device, configured to adjust an output order of the candidate information of the user according to a selection priority of the candidate information of the user.

Images