CN112784039A - Method, device and storage medium for distributing online customer service - Google Patents
Method, device and storage medium for distributing online customer service Download PDFInfo
- Publication number
- CN112784039A CN112784039A CN201911094099.5A CN201911094099A CN112784039A CN 112784039 A CN112784039 A CN 112784039A CN 201911094099 A CN201911094099 A CN 201911094099A CN 112784039 A CN112784039 A CN 112784039A
- Authority
- CN
- China
- Prior art keywords
- user
- consultation
- customer service
- intention
- online customer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
- G06F16/353—Clustering; Classification into predefined classes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06311—Scheduling, planning or task assignment for a person or group
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/01—Customer relationship services
Abstract
The application discloses a method, a device and a storage medium for distributing online customer service, which are used for acquiring current user consultation problems, inputting the user consultation problems into a pre-trained intention prediction model to generate intention classification labels, wherein the intention prediction model is trained according to order information and operation information of a user and intention classification labels determined by the first user consultation problems of the user on a consultation page, and finally, the type of a consultation entrance is determined according to the intention classification labels, and the online customer service corresponding to the consultation entrance is distributed to the user. According to the method and the device, the intention of the user to enter the consultation page is predicted, and the online customer service is distributed according to the predicted user intention, so that the distribution strategy of the online customer service is optimized, and the service quality is improved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method, an apparatus, and a storage medium for allocating online customer services.
Background
In the e-commerce field, online manual customer service is generally adopted to serve a user, relevant problems of the user in the pre-sale field and the after-sale field are answered, the user is helped to close a loop in the whole shopping process, and the shopping experience of the user is improved. When online manual customer service is allocated to a user, the online manual customer service is generally allocated according to the type of a consultation page entered by the user. If the user enters through the order page entry, the user is considered to be likely to consult the after-sale questions, and at the moment, the user can be allocated with a customer service group responsible for solving the after-sale questions and then allocated with customer services under the customer service group.
The types of the current consultation page generally comprise a main station consultation entrance type, an order consultation entrance type, a service order consultation entrance type, a commodity page consultation entrance type and the like. Such as a user entering a consultation from an order entry, may consult issues such as delivery cycles, price guarantees, returns, and the like. The customer service groups corresponding to the several entrances are established, and obviously, the problems possibly asked by the users cannot be well subdivided. In addition, due to the fact that the granularity is too coarse and the pre-judgment on the consultation intention of the user is lacked, all customer services in each customer service group need to have multiple business skills, the training cost of new customer services is increased, the skills of the customer services are not concentrated in a certain field, and the overall answer quality is not high enough.
Disclosure of Invention
The embodiment of the application provides a method for distributing online customer service, and the problem that the online customer service is not matched with user requirements is solved.
The method comprises the following steps:
acquiring a current user consultation problem;
inputting the user consultation problems into a pre-trained intention prediction model to generate intention classification labels, wherein the intention prediction model is trained according to order information and operation information of a user and the intention classification labels determined by the first user consultation problems of a consultation page of the user;
and determining the type of a consultation entrance according to the intention classification label, and distributing online customer service corresponding to the consultation entrance for the user.
Optionally, obtaining the order information and the operation information of the user;
encoding the operation information to generate the real-time access characteristics corresponding to the operation information;
coding the order information to generate the order characteristics corresponding to the order information;
determining the intention classification label according to the first user consultation problem of the user on the consultation page;
and taking a feature group consisting of the real-time access feature, the order feature and the intention classification label corresponding to the user of each user as a sample to be input into a classifier for training, and generating the intention prediction model.
Optionally, determining a category number of each commodity according to pre-divided category information of the commodities;
and coding the operation information according to the class number to generate a real-time access characteristic.
Optionally, the first user consultation problem of each user on the consultation page is input into a pre-trained text classification model, and the intention classification label corresponding to the user consultation problem is generated.
Optionally, counting the current number of receptions of each online customer service under the consultation entrance corresponding to the intention classification label;
according to the current reception number and the total number of daily required receptions of each online customer service, respectively allocating a first weight and a second weight to the current reception number and the total number of daily required receptions, and calculating an allocation score of each online customer service;
and determining the online customer service with the lowest distribution score as the online customer service of the user in the consultation.
In another embodiment of the present invention, there is provided an apparatus for distributing online customer service, the apparatus comprising:
the acquisition module is used for acquiring the current user consultation problem;
the generation module is used for inputting the user consultation problems into a pre-trained intention prediction model and generating intention classification labels, wherein the intention prediction model is trained according to order information and operation information of a user and the intention classification labels determined by the first user consultation problems of a consultation page of the user;
and the distribution module is used for determining the type of a consultation entrance according to the intention classification label and distributing online customer service corresponding to the consultation entrance for the user.
Optionally, the generating module includes:
the obtaining subunit is used for obtaining the order information and the operation information of the user;
the first generating unit is used for encoding the operation information and generating the real-time access characteristics corresponding to the operation information;
the second generating unit is used for coding the order information and generating the order characteristics corresponding to the order information;
the determining unit is used for determining the intention classification label according to the first user consultation problem of the user on the consultation page;
and a third generating unit, configured to input, as a sample input classifier, a feature group formed by the real-time access feature, the order feature, and the intention classification label corresponding to the user of each user for training, and generate the intention prediction model.
Optionally, the first generating unit includes:
the determining subunit is used for determining the category number of each commodity according to the information of the categories of the commodities divided in advance;
and the generation subunit is used for coding the operation information according to the class number and generating a real-time access characteristic.
In another embodiment of the present invention, a non-transitory computer readable storage medium is provided, storing instructions that, when executed by a processor, cause the processor to perform the steps of a method of distributing online customer service as described above.
In another embodiment of the present invention, an electronic device is provided, which is characterized by comprising a processor and a memory, wherein the memory stores an application program executable by the processor, and the application program is used for causing the processor to execute the steps of the method for distributing online customer service.
Based on the embodiment, the current user consultation problem is firstly obtained, then the user consultation problem is input into a pre-trained intention prediction model, and an intention classification label is generated, wherein the intention prediction model is trained according to order information and operation information of the user and an intention classification label determined by the first user consultation problem of the user on a consultation page, and finally the type of a consultation entrance is determined according to the intention classification label, and online customer service corresponding to the consultation entrance is distributed to the user. According to the method and the device, the intention of the user to enter the consultation page is predicted, and the online customer service is distributed according to the predicted user intention, so that the distribution strategy of the online customer service is optimized, and the service quality is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
FIG. 1 is a flow chart illustrating a method for distributing online customer service provided by an embodiment 100 of the present application;
FIG. 2 is a schematic diagram illustrating a detailed flow of a method for distributing online customer service according to an embodiment 200 of the present application;
FIG. 3 is a schematic diagram of an apparatus for distributing online customer service according to an embodiment 300 of the present application;
fig. 4 shows a schematic diagram of a terminal device provided in embodiment 400 of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are, for example, capable of operation in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprising" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements explicitly listed, but may include other steps or elements not explicitly listed or inherent to such process, method, article, or apparatus.
Based on the problems in the prior art, the embodiment of the application provides a method for allocating online customer service, which is mainly applicable to the technical field of computers. And training an intention prediction model by constructing a sample group, and predicting according to the first consultation problem of the user on a consultation page. After the user intention is determined, appropriate online customer service is distributed to the user according to the user intention so as to improve the service quality. The technical scheme of the invention is explained in detail by specific embodiments below to realize a method for distributing online customer service. Several of the following embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Fig. 1 is a schematic flow chart of a method for allocating online customer service according to embodiment 100 of the present application. The detailed steps are as follows:
and S11, acquiring the current user consultation problem.
In this step, in the field of electronic commerce, the user can generally consult through a consultation page provided by an e-commerce. When the user inputs the user consultation problem on the consultation page, the user consultation problem currently proposed by the user is obtained. Wherein, the user consultation problem is generally a service problem related to the service. Other user consultation problems not related to the service are not in the scope of acquisition.
And S12, inputting the user consultation question into a pre-trained intention prediction model, and generating intention classification labels, wherein the intention prediction model is trained according to the order information and the operation information of the user and the intention classification label determined by the first user consultation question of the user on a consultation page.
In this step, the intention prediction is performed on the obtained user consultation problem. Wherein the intent prediction model is trained by constructing a sample set. Optionally, the user generally carries a purpose, i.e. user intention, when consulting. The data analysis shows that before the user enters the consultation page, the user generally leaves some traces on the website, and the information is collected to prejudge the intention of the user on the consultation page. For example, the user browses the white-bar related web page in the east of kyoto many times before entering the consultation for 15 minutes, and at this time, the user is likely to need to consult the services related to the "white bar". Therefore, the order information and the operation information of the user and the intention classification label determined by the first user consultation problem of the user on the consultation page are collected to form a feature group for training, and an intention prediction model is generated. The order information of the user is generally a recent order of the user, and includes information such as an order state, an order type, an order amount and the like. The operation information mainly refers to the operation behaviors of the user on the e-commerce webpage, such as shopping cart purchasing, commodity detail browsing, commodity collection clicking, commodity evaluation clicking and the like. The first user consultation question may represent the consultation intention of the user. And training an intention training model by taking the order information and the operation information of the user and a feature group formed by determining an intention classification label according to the first user consultation problem as samples.
And S13, classifying the labels according to the intentions, determining the type of the consultation entrance, and distributing the online customer service corresponding to the consultation entrance for the user.
In this step, after the user's consultation questions are input into the intention prediction model and corresponding intention classification labels are generated. The type of consulting entry is determined according to the intention classification label. The type of the consultation entrance is established and subdivided according to the skill dimension of the customer service system, and the determined intention label corresponds to the type of the consultation entrance when a sample group of the training intention prediction model is formed. Further, when the intention classification label corresponding to the user consultation question of a certain user is predicted and obtained, online customer service in the corresponding consultation entrance is distributed to the user. Optionally, there are multiple online customer services in each consultation portal, and finer granularity distribution of online customer services can be performed for the user according to the current number of receptions of each customer service and the total number of daily required receptions.
As described above, based on the above embodiment, the current user query question is obtained first, then, the user query question is input into the pre-trained intention prediction model, and the intention classification label is generated, where the intention prediction model is trained according to the order information and the operation information of the user and the intention classification label determined by the first user query question of the user on the query page, and finally, the type of the query entry is determined according to the intention classification label, and the online customer service corresponding to the query entry is assigned to the user. According to the method and the device, the intention of the user to enter the consultation page is predicted, and the online customer service is distributed according to the predicted user intention, so that the distribution strategy of the online customer service is optimized, and the service quality is improved.
Fig. 2 is a schematic diagram illustrating a specific flow of a method for allocating online customer service according to an embodiment 200 of the present application. Wherein, the detailed process of the specific flow is as follows:
s201, obtaining order information and operation information of a user.
Here, the latest order information of the user is obtained, including the order status, the order type, the order amount, and the like. If the goods return application exists, the related information of the order for applying the goods return is collected.
In addition, the method of embedding the webpage and the like is used for collecting the operation information of the user on the electronic commerce webpage. For example, the web pages browsed by the user within 24 hours are collected, the browsing dwell time is obtained, and the clicking operation is respectively carried out.
S202, determining intention classification labels according to the first user consultation problem of the user on the consultation page.
Here, through analysis of a chat log of a large number of online customer services, a general user directly indicates a core intention when making a problem consultation. Accordingly, the intention classification tag may be determined according to the first user consultation problem of the user at the consultation page. Optionally, the first service-related user query questions of the user are collected, and the text classification model can be used for performing problem classification processing on the user query questions. As shown in table 1, the intention classification label is determined for the user's consultation question, for example, the user's consultation question is "i ask for insurance", the intention classification label of the question is "inquiry for insurance", the user's consultation question is "what you buy is bad, i want to return goods", the intention classification label of the question is "return goods handling", the user's consult question is "how did my goods not yet go? ", the intent classification of the problem is labeled" delivery cycle ". Mapping each sentence of user consultation questions of the user to a preset intention classification label. And further, inputting the consultation questions and intention classification labels of the users as sample input text classification models for training, inputting the first consultation question of each user on a consultation page into the trained text classification models, and generating intention classification labels corresponding to the consultation questions of the users.
| Sample numbering | First question of user | Intention classification label |
| 1 | The bought article is bad and I want to return | Return processing |
| 2 | I ask for a price | Consultation of value insurance |
| 3 | How did my goods not get delivered? | Delivery cycle |
Table 1S203, a real-time access profile is generated.
Here, the product type number of each product is determined from the product type information of the product divided in advance, and the real-time access feature is generated by encoding the operation information based on the product type number. Alternatively, for billions of items that may be involved on an e-commerce website, each item is mapped to item information for the corresponding item. The commodity type information can be preset, for example, each commodity is mapped to a corresponding type and is divided into three levels for product type numbering, so that the sample space is greatly reduced. For each commodity, the corresponding category information is generally relatively fixed and can be collected and determined in advance. As shown in table 2, the correspondence between the item information and the item number of the partial product is shown.
TABLE 2
Further, the operation information of the user is encoded according to the item numbers of the partial commodities determined in table 2. As shown in table 3, the partial real-time access features generated for the encoded data. If the operation information is "detailed page with the browser class ID of C288", the corresponding encoded real-time access feature is "VISIT _ C288". In addition, the time range of the real-time access feature may be set in advance, for example, 30 minutes before the user enters the consultation page is set as the time range of the real-time access feature, and the operation information of the user within the 30 minutes is collected and encoded.
| Detail page with browsing category ID of C288 | VISIT_C288 |
| Click on item class ID to C288 to add to shopping cart button | CLICK_C288_add |
| Commodity evaluation tab with click item ID of C288 | CLICK_C288_comment |
| Click on item class ID C288 focus store button | CLICK_C288_followShop |
Table 3S204, generate order characteristics.
In this step, the order information is encoded to generate order characteristics corresponding to the order information. As shown in table 4, the partial order characteristics generated after encoding.
| Status of state | Encoding |
| Order type-jingdong self-operation | ORDER_TYPE_JD_SELF |
| Order type-POP | ORDER_TYPE_POP |
| Order status-out of delivery | ORDER_STATE_OVERDUE_DILIVERY |
| Order status-signed for | ORDER_STATE_RECEIVED |
| Service ticket type-return | SERVICE_TYPE_RETURN |
| Service ticket type-service | SERVICE_TYPE_REPAIR |
In table 4S205, a feature group consisting of the order characteristics, the operation characteristics, and the intention classification labels of the user is input to the classifier as a sample group and trained to generate an intention prediction model.
Here, a feature group consisting of the real-time access feature of each user, the order feature, and the intention classification label corresponding to the user is input to the classifier as a sample and trained to generate an intention prediction model. Alternatively, as shown in table 5, the real-time access characteristics and the order characteristics of each user are first combined into user characteristics, and each sample is obtained.
TABLE 5
Further, the intention classification labels corresponding to the first user consultation problems of the users on the consultation page and the determined user characteristics form characteristic groups, and each group of characteristic groups is used as a sample. As shown in table 6, the sample of number 1 is taken as an example, and the physical meaning thereof can be understood that the user has ORDER features of ORDER _ TYPE _ JD _ SELF and SERVICE _ TYPE _ RETURN, and within the last 30 minutes, real-time access features such as VISIT _ C288, client _ C288_ add, VISIT _ C374, client _ C374_ colorBlack, … … and client _ C2543_ packaging corresponding to operation information such as access or CLICK are sequentially performed, and when the user enters the consultation page, the first user consultation problem is related to "RETURN processing".
By carrying out the cleaning and the construction on the massive historical data, millions or even millions of feature groups can be obtained and used as samples for training the intention classification model.
Further, the samples can be selected to be input into a classification algorithm LibSVM for model training so as to generate an intention classification model.
The training of the intention classification model is completed based on the above steps S201 to S205.
S206, acquiring the current user consultation problem.
And S207, generating an intention classification label corresponding to the user consultation question.
In this step, the user consultation questions are input into a pre-trained intention prediction model to generate intention classification labels.
And S208, classifying the label according to the intention, and determining the type of the consultation entrance.
In the step, the online customer service corresponding to the type of the consultation entrance according to the skill dimension is constructed by rebuilding the online customer service reception system, so that the matching of the personnel to the skill is realized. And (4) the customer service system is reestablished and divided according to the skill dimension, so that the skills of a plurality of online customer services under each consultation entrance are mainly matched with the intention classification label corresponding to the consultation entrance. If the intention classification label is determined as a return process, a price guarantee consultation, a distribution cycle, etc., the type of the consultation entrance corresponds to the type of the intention classification label.
And S209, allocating online customer service to the user in the consultation entrance.
Here, the current number of receptions of each online customer service under the counseling portal corresponding to the intention classification label is counted. Secondly, according to the current reception number and the total number of the daily required receptions of each online customer service, respectively allocating a first weight and a second weight to the current reception number and the total number of the daily required receptions, and calculating the allocation score of each online customer service. And finally, determining the online customer service with the lowest distribution grade as the online customer service of the user in the consultation. Optionally, the total number of the online customer services required by the user on the day and the current number of the online customer services are counted, a weighted score of each online customer service in the plurality of online customer services corresponding to the current consultation entrance is calculated, and the online customer service is distributed to the user according to the score.
Generally, the allocation score of each online customer service can be calculated by using the formula S ═ ax + by, where x represents the total number of required receptions of the online customer service on the day, y represents the current number of receptions of the online customer service, and a and b are respectively the first weight and the second weight, which can be set in advance and adjusted in real time, and the preferred values in the embodiment of the present application are respectively 0.3 and 0.5. And when the user enters the consultation page, triggering and calculating the distribution scores of all the online customer services corresponding to the current consultation entrance, and then distributing the user to the online customer service reception with the lowest score.
The method for distributing the online customer service is realized based on the steps.
When the distribution granularity of the consultation entrance is too coarse, all online customer services under each customer service group need to have multiple service skills, and the consultation intention of the user is not pre-judged, so that the problem of low service quality can be caused. Such as a customer entering a consultation from an order entry, may be asked questions such as delivery cycles, price guarantees, returns, and the like. The online customer service groups corresponding to the several entrances are established, and obviously, the problems possibly asked by the users cannot be well subdivided. The granularity is too coarse, so that the cost of training online customer service is increased, the skill of the online customer service is not concentrated in a certain field, and the overall answer quality is not high enough.
According to the method for distributing the online customer service, the consultation intention of the user is determined by establishing the intention prediction model, and the method is matched with a system of the online customer service which is re-established and divided according to the skill dimension, so that the online customer service in a plurality of online customer services corresponding to the consultation entrance capable of solving the consultation problem is distributed to the user. Each online customer service is guaranteed to work in the field which is good at the user to the maximum extent, the optimization of reception allocation is achieved, and customers can obtain more professional services.
Based on the same inventive concept, the embodiment 300 of the present application further provides an apparatus for distributing online customer service, wherein as shown in fig. 3, the apparatus includes:
an obtaining module 31, configured to obtain a current user consultation problem;
a generating module 32, configured to input the user consultation problem into a pre-trained intention prediction model, and generate an intention classification label, where the intention prediction model is trained according to order information and operation information of the user and the intention classification label determined by a first user consultation problem of the user on a consultation page;
and the distribution module 33 is used for classifying the labels according to the intentions, determining the type of the consultation entrance and distributing the online customer service corresponding to the consultation entrance for the user.
In this embodiment, the specific functions and interaction manners of the obtaining module 31, the generating module 32 and the allocating module 33 can be referred to the record of the embodiment corresponding to fig. 1, and are not described herein again.
Optionally, the generating module 32 includes:
the acquisition subunit is used for acquiring order information and operation information of a user;
the first generating unit is used for encoding the operation information and generating real-time access characteristics corresponding to the operation information;
the second generation unit is used for coding the order information and generating order characteristics corresponding to the order information;
the determining unit is used for determining an intention classification label according to the first user consultation problem of a user on a consultation page;
and the third generating unit is used for inputting a feature group consisting of the real-time access features, the order features and the intention classification labels corresponding to the users of the users as a sample into the classifier for training to generate an intention prediction model.
Optionally, the first generating unit includes:
the determining subunit is used for determining the category number of each commodity according to the information of the categories of the commodities divided in advance;
and the generation subunit is used for coding the operation information according to the class number and generating the real-time access characteristic.
As shown in fig. 4, another embodiment 400 of the present application further provides a terminal device, which includes a processor 401, wherein the processor 401 is configured to execute the steps of the method for allocating online customer service. As can also be seen from fig. 4, the terminal device provided by the above embodiment further comprises a non-transitory computer readable storage medium 402, the non-transitory computer readable storage medium 402 having stored thereon a computer program, which when executed by the processor 401, performs the above-mentioned steps of the method of allocating online customer service. In practice, the terminal device may be one or more computers, as long as the computer-readable medium and the processor are included.
In particular, the storage medium can be a general-purpose storage medium, such as a removable disk, a hard disk, a FLASH, and the like, and when executed, the computer program on the storage medium can perform the steps of the method for distributing online customer service. In practical applications, the computer readable medium may be included in the apparatus/device/system described in the above embodiments, or may exist alone without being assembled into the apparatus/device/system. The computer readable storage medium carries one or more programs which, when executed, perform the steps of a method of distributing online customer service as described above.
According to embodiments disclosed herein, the computer-readable storage medium may be a non-volatile computer-readable storage medium, which may include, for example and without limitation: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing, without limiting the scope of the present disclosure. In the embodiments disclosed herein, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The flowchart and block diagrams in the figures of the present application illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments disclosed herein. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not explicitly recited in the present application. In particular, the features recited in the various embodiments and/or claims of the present application may be combined and/or coupled in various ways, all of which fall within the scope of the present disclosure, without departing from the spirit and teachings of the present application.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present application, and are used for illustrating the technical solutions of the present application, but not limiting the same, and the scope of the present application is not limited thereto, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can still change or easily conceive of the technical solutions described in the foregoing embodiments or equivalent replacement of some technical features thereof within the technical scope disclosed in the present application; such changes, variations and substitutions do not depart from the spirit and scope of the exemplary embodiments of the present application and are intended to be covered by the appended claims. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method of distributing online customer service, comprising:
acquiring a current user consultation problem;
inputting the user consultation problems into a pre-trained intention prediction model to generate intention classification labels, wherein the intention prediction model is trained according to order information and operation information of a user and the intention classification labels determined by the first user consultation problems of a consultation page of the user;
and determining the type of a consultation entrance according to the intention classification label, and distributing online customer service corresponding to the consultation entrance for the user.
2. The method of claim 1, wherein the training process of the intent prediction model comprises:
acquiring the order information and the operation information of the user;
encoding the operation information to generate the real-time access characteristics corresponding to the operation information;
coding the order information to generate the order characteristics corresponding to the order information;
determining the intention classification label according to the first user consultation problem of the user on the consultation page;
and taking a feature group consisting of the real-time access feature, the order feature and the intention classification label corresponding to the user of each user as a sample to be input into a classifier for training, and generating the intention prediction model.
3. The method of claim 2, wherein the step of generating the real-time access characteristic corresponding to the operational information comprises:
determining the category number of each commodity according to the information of the categories of the pre-divided commodities;
and coding the operation information according to the class number to generate a real-time access characteristic.
4. The method of claim 2, wherein the step of determining the intention classification tag according to the first of the user's consultation questions on the consultation page comprises:
inputting the first user consultation problem of each user on the consultation page into a pre-trained text classification model, and generating the intention classification label corresponding to the user consultation problem.
5. The method of claim 1, wherein said step of assigning an online customer service corresponding to said counseling portal to said user comprises:
counting the current receiving number of each online customer service under the consultation entrance corresponding to the intention classification label;
according to the current reception number and the total number of daily required receptions of each online customer service, respectively allocating a first weight and a second weight to the current reception number and the total number of daily required receptions, and calculating an allocation score of each online customer service;
and determining the online customer service with the lowest distribution score as the online customer service of the user in the consultation.
6. An apparatus for distributing online customer service, the apparatus comprising:
the acquisition module is used for acquiring the current user consultation problem;
the generation module is used for inputting the user consultation problems into a pre-trained intention prediction model and generating intention classification labels, wherein the intention prediction model is trained according to order information and operation information of a user and the intention classification labels determined by the first user consultation problems of a consultation page of the user;
and the distribution module is used for determining the type of a consultation entrance according to the intention classification label and distributing online customer service corresponding to the consultation entrance for the user.
7. The method of claim 6, wherein the generating module comprises:
the obtaining subunit is used for obtaining the order information and the operation information of the user;
the first generating unit is used for encoding the operation information and generating the real-time access characteristics corresponding to the operation information;
the second generating unit is used for coding the order information and generating the order characteristics corresponding to the order information;
the determining unit is used for determining the intention classification label according to the first user consultation problem of the user on the consultation page;
and a third generating unit, configured to input, as a sample input classifier, a feature group formed by the real-time access feature, the order feature, and the intention classification label corresponding to the user of each user for training, and generate the intention prediction model.
8. The method of claim 7, wherein the first generating unit comprises:
the determining subunit is used for determining the category number of each commodity according to the information of the categories of the commodities divided in advance;
and the generation subunit is used for coding the operation information according to the class number and generating a real-time access characteristic.
9. A non-transitory computer readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the steps of a method of distributing online customer service as recited in any of claims 1-5.
10. An electronic device, wherein the apparatus comprises: a processor and a memory; the memory has stored therein an application executable by the processor for causing the processor to perform the steps of a method of distributing online customer service as claimed in any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911094099.5A CN112784039A (en) | 2019-11-11 | 2019-11-11 | Method, device and storage medium for distributing online customer service |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911094099.5A CN112784039A (en) | 2019-11-11 | 2019-11-11 | Method, device and storage medium for distributing online customer service |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112784039A true CN112784039A (en) | 2021-05-11 |
Family
ID=75749672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911094099.5A Pending CN112784039A (en) | 2019-11-11 | 2019-11-11 | Method, device and storage medium for distributing online customer service |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112784039A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113379229A (en) * | 2021-06-08 | 2021-09-10 | 北京沃东天骏信息技术有限公司 | Resource scheduling method and device |
| CN114757186A (en) * | 2022-04-25 | 2022-07-15 | 中国电信股份有限公司 | User intention analysis method and device, computer storage medium and electronic equipment |
| CN114979723A (en) * | 2022-02-14 | 2022-08-30 | 杭州脸脸会网络技术有限公司 | Virtual intelligent customer service method, device, electronic device and storage medium |
| CN117453773A (en) * | 2023-12-21 | 2024-01-26 | 深圳市活力天汇科技股份有限公司 | Customer service matching method and system based on user intention prediction |
-
2019
- 2019-11-11 CN CN201911094099.5A patent/CN112784039A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113379229A (en) * | 2021-06-08 | 2021-09-10 | 北京沃东天骏信息技术有限公司 | Resource scheduling method and device |
| CN114979723A (en) * | 2022-02-14 | 2022-08-30 | 杭州脸脸会网络技术有限公司 | Virtual intelligent customer service method, device, electronic device and storage medium |
| CN114979723B (en) * | 2022-02-14 | 2023-08-29 | 杭州脸脸会网络技术有限公司 | Virtual intelligent customer service method, device, electronic device and storage medium |
| CN114757186A (en) * | 2022-04-25 | 2022-07-15 | 中国电信股份有限公司 | User intention analysis method and device, computer storage medium and electronic equipment |
| CN114757186B (en) * | 2022-04-25 | 2023-11-14 | 中国电信股份有限公司 | User intention analysis method and device, computer storage medium and electronic equipment |
| CN117453773A (en) * | 2023-12-21 | 2024-01-26 | 深圳市活力天汇科技股份有限公司 | Customer service matching method and system based on user intention prediction |
| CN117453773B (en) * | 2023-12-21 | 2024-03-26 | 深圳市活力天汇科技股份有限公司 | Customer service matching method and system based on user intention prediction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741146B (en) | Product recommendation method, device, equipment and storage medium based on user behaviors | |
| CN112784039A (en) | Method, device and storage medium for distributing online customer service | |
| WO2019037202A1 (en) | Method and apparatus for recognising target customer, electronic device and medium | |
| CN111784455A (en) | Article recommendation method and recommendation equipment | |
| CN109615487A (en) | Products Show method, apparatus, equipment and storage medium based on user behavior | |
| EP2416289A1 (en) | System for measuring variables from data captured from internet applications | |
| CN105868847A (en) | Shopping behavior prediction method and device | |
| CN106445905B (en) | Question and answer data processing, automatic question-answering method and device | |
| CN106503006A (en) | The sort method and device of application App neutron applications | |
| CN105573966A (en) | Adaptive Modification of Content Presented in Electronic Forms | |
| CN110322093B (en) | Information processing method, information display method, information processing device and computing equipment | |
| CN113191838B (en) | Shopping recommendation method and system based on heterogeneous graph neural network | |
| CN112001754A (en) | User portrait generation method, device, equipment and computer readable medium | |
| CN108876545A (en) | Order recognition methods, device and readable storage medium storing program for executing | |
| CN112148973B (en) | Data processing method and device for information push | |
| CN106934648A (en) | A kind of data processing method and device | |
| CN111179031A (en) | Training method, device and system for commodity recommendation model | |
| CN105335518A (en) | Method and device for generating user preference information | |
| JP7444424B2 (en) | Behavior analysis system and its program using behavior history data | |
| CN109426983A (en) | Dodge purchase activity automatic generation method and device, storage medium, electronic equipment | |
| CN111680213A (en) | Information recommendation method, data processing method and device | |
| JP5603678B2 (en) | Demand forecasting apparatus and demand forecasting method | |
| CN108510302A (en) | A kind of marketing decision-making method and trading server | |
| CN111199473A (en) | Anti-cheating method, device and system based on transaction record information | |
| Raj et al. | The Significance of Big Data for the Base of the Pyramid Segment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |