CN114154727A

CN114154727A - Information processing method and device and storage medium

Info

Publication number: CN114154727A
Application number: CN202111484260.7A
Authority: CN
Inventors: 邹京甫; 钟皓明; 张海川
Original assignee: WeBank Co Ltd
Current assignee: WeBank Co Ltd
Priority date: 2021-12-07
Filing date: 2021-12-07
Publication date: 2022-03-08

Abstract

The embodiment of the application discloses an information processing method, an information processing device and a storage medium, wherein the information processing method comprises the following steps: under the condition that an electricity sales call prediction instruction is received, acquiring object identification information and time to be predicted of a target object from the electricity sales call prediction instruction; acquiring object information, historical telemarketing communication information and historical service information corresponding to the target object according to the object identification information; inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model to obtain electricity sales call connection probability corresponding to the time to be predicted; and in the case that the time to be predicted is determined to be the target electricity selling time according to the electricity selling connection probability, performing a telephone selling process between the target object and the electricity selling terminal within the target electricity selling time.

Description

Information processing method and device and storage medium

Technical Field

The present application relates to the field of machine learning technologies, and in particular, to an information processing method and apparatus, and a storage medium.

Background

With the development of internet technology, the popularization of various communication devices and the sales modes are gradually increased, and the common sales modes comprise telephone sales.

In the prior art, the telemarketing mode is that telemarketing personnel randomly determine time to call customers for telemarketing, or call customers in turn according to the priority order of the customers determined manually for telemarketing, and because the telemarketing mode depends on the experience of the telemarketing personnel, the telemarketing time determined by the telemarketing personnel with less experience is not accurate, so that the accuracy of determining the telemarketing time is reduced.

Disclosure of Invention

In order to solve the above technical problem, embodiments of the present application are expected to provide an information processing method and apparatus, and a storage medium, which can improve accuracy of determining a telemarketing time.

The technical scheme of the application is realized as follows:

an embodiment of the present application provides an information processing method, including:

under the condition that an electricity sales call prediction instruction is received, acquiring object identification information and time to be predicted of a target object from the electricity sales call prediction instruction;

acquiring object information, historical telemarketing communication information and historical service information corresponding to the target object according to the object identification information;

inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model to obtain an electricity sales call connection probability corresponding to the time to be predicted;

and in the case that the time to be predicted is determined to be the target electricity selling time according to the electricity selling connection probability, performing a telephone selling process between the target object and the electricity selling terminal within the target electricity selling time.

An embodiment of the present application provides an information processing apparatus, the apparatus including:

the device comprises an acquisition unit, a processing unit and a prediction unit, wherein the acquisition unit is used for acquiring object identification information and time to be predicted of a target object from an electricity sales call prediction instruction under the condition of receiving the electricity sales call prediction instruction; acquiring object information, historical telemarketing communication information and historical service information corresponding to the target object according to the object identification information;

the input unit is used for inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model to obtain an electricity sales call connection probability corresponding to the time to be predicted;

and the execution unit is used for executing a telephone sales process with the target object within the target electricity selling time under the condition that the time to be predicted is determined to be the target electricity selling time according to the electricity selling connection probability.

the information processing system includes a memory, a processor, and a communication bus, the memory communicating with the processor through the communication bus, the memory storing an information processing program executable by the processor, and the processor executing the information processing method when the information processing program is executed.

The embodiment of the application provides a storage medium, which stores a computer program thereon and is applied to an information processing device, wherein the computer program is used for realizing the information processing method when being executed by a processor.

The embodiment of the application provides an information processing method, an information processing device and a storage medium, wherein the information processing method comprises the following steps: under the condition that an electricity sales call prediction instruction is received, acquiring object identification information and time to be predicted of a target object from the electricity sales call prediction instruction; acquiring object information, historical telemarketing communication information and historical service information corresponding to the target object according to the object identification information; inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model to obtain electricity sales call connection probability corresponding to the time to be predicted; and in the case that the time to be predicted is determined to be the target electricity selling time according to the electricity selling connection probability, performing a telephone selling process between the target object and the electricity selling terminal within the target electricity selling time. By adopting the method, the information processing device predicts the electricity pin connection probability of successful electricity pin communication with the target object at the time to be predicted by using the electricity pin communication prediction model according to the object information, the historical electricity pin communication information and the historical service information by acquiring the object information, the historical electricity pin communication information and the historical service information corresponding to the target object, namely, the accuracy of the time to be predicted is determined by using the electricity pin connection probability, so that the target electricity pin time with high accuracy can be determined, and the accuracy of determining the target electricity pin time is improved.

Drawings

Fig. 1 is a flowchart of an information processing method according to an embodiment of the present application;

FIG. 2 is a flowchart of an exemplary method for training a plurality of electricity sales call prediction models according to an embodiment of the present disclosure;

FIG. 3 is a diagram of an exemplary series-parallel model training architecture provided in an embodiment of the present application;

fig. 4 is a processing structure diagram for processing a plurality of sets of sample object information, a plurality of sets of sample telemarketing information, and a plurality of sets of sample historical service information according to an embodiment of the present disclosure;

fig. 5 is a schematic diagram of an exemplary information processing structure provided in an embodiment of the present application;

fig. 6 is a first schematic structural diagram of an information processing apparatus according to an embodiment of the present disclosure;

fig. 7 is a schematic diagram illustrating a structure of an information processing apparatus according to an embodiment 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. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

Example one

An information processing method is provided in an embodiment of the present application, and an information processing method is applied to an information processing apparatus, and fig. 1 is a flowchart of the information processing method provided in the embodiment of the present application, and as shown in fig. 1, the information processing method may include:

s101, under the condition that the electricity sales call prediction instruction is received, object identification information and time to be predicted of the target object are obtained from the electricity sales call prediction instruction.

The information processing method is suitable for the scene of determining the electric pin connection probability corresponding to the time to be predicted.

In the embodiment of the present application, the information processing apparatus may be implemented in various forms. For example, the information processing apparatus described in the present application may include apparatuses such as a mobile phone, a camera, a tablet computer, a notebook computer, a palmtop computer, a Personal Digital Assistant (PDA), a Portable Media Player (PMP), a navigation apparatus, a wearable device, a smart band, a pedometer, and the like, and apparatuses such as a Digital TV, a desktop computer, a server, and the like.

In the embodiment of the application, the electric pin call prediction instruction carries the time to be predicted, and the electric pin call prediction instruction can be an instruction for determining the electric pin call probability corresponding to the time to be predicted.

In the embodiment of the application, the telemarketing call prediction instruction may be an instruction input into the information processing device by a user; the telemarketing call prediction instruction can also be an instruction transmitted to the information processing device by other equipment; the telemarketing call prediction instruction can also be an instruction acquired by the information processing device in other manners; the specific manner in which the information processing apparatus acquires the telephone sales call prediction instruction may be determined according to actual conditions, which is not limited in the embodiment of the present application.

In the embodiment of the present application, the target object is an electricity sales client. The object identification information of the target object may be identity information of the electricity sales client; the object identification information can also be the number information of the customer of the electronic expense; the object identification information may also be information identifying which electricity selling customer is; the specific object identification information may be determined according to actual conditions, which is not limited in this embodiment of the present application.

It should be noted that, if the object identification information is the identification information of the electricity consumer, the specific object identification information may be identification Information (ID) of the electricity consumer.

In the embodiment of the application, the time to be predicted can be in a time format of year, month, day, hour, minute and second; the time to be predicted can also be in a time format of year, month, day and time; the time to be predicted can be in other time formats; the specific time format of the time to be predicted can be determined according to actual conditions, which is not limited in the embodiment of the present application.

And S102, acquiring object information, historical telemarketing communication information and historical service information corresponding to the target object according to the object identification information.

In the embodiment of the application, after the information processing device acquires the object identification information of the target object and the time to be predicted from the electricity sales call prediction instruction, the information processing device can acquire the object information, the historical electricity sales call information and the historical service information corresponding to the target object according to the object identification information.

In the embodiment of the application, the information processing device can acquire object information, historical telemarketing communication information and historical service information corresponding to a target object in a database according to the object identification information; the information processing device can also acquire object information, historical telemarketing communication information and historical service information corresponding to the target object in other storage areas according to the object identification information; the specific manner in which the information processing device acquires the object information, the historical telemarketing information, and the historical service information may be determined according to actual conditions, which is not limited in the embodiments of the present application.

If the target object is the electricity sales client, the object information may be client basic information of the electricity sales client, and the object information includes, for example, one of the following: the age, sex, occupation, school calendar, graduation colleges, working years, working units, provinces, cities and the like of the consumer of the electricity marketing.

If the target object is the electricity sales client, the history electricity sales call information may be information obtained when the electricity sales client has made a telephone call. Illustratively, the historical telemarketing information includes one of: the time of the electricity sales staff dialing the sales call to the electricity sales client, the on-time of the electricity sales client, the hang-up time of the electricity sales client, the recording start time, the recording end time, the ringing waiting time, the recording time, the ID of the electricity sales staff and the like.

If the target object is the electricity consumer, the historical service information may be service information that the electricity consumer has transacted in history. Illustratively, the historical traffic information includes one of: the ID of the service product browsed by the telemarketing client, the ID of the service product used by the telemarketing client, and related use attributes of the service product, such as deposit, transfer, loan and other attributes, and the operation time of the corresponding service product.

S103, inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model to obtain the electricity sales call probability corresponding to the time to be predicted.

In the embodiment of the application, after the information processing device obtains the object information, the historical telepin call information and the historical service information corresponding to the target object according to the object identification information, the object information, the historical telepin call information, the historical service information and the time to be predicted can be input into the target telepin call prediction model by the processing device, and the telepin connection probability corresponding to the time to be predicted is obtained.

In the embodiment of the present application, the target electricity sales call prediction model may be a model configured in the information processing apparatus; the target electricity sales call prediction model can also be a model transmitted to the information processing device by other equipment; the specific manner in which the information processing apparatus obtains the target electricity sales call prediction model may be determined according to actual conditions, which is not limited in the embodiment of the present application.

In the embodiment of the application, before the information processing device inputs the object information, the historical electricity sales call information, the historical service information and the time to be predicted into the target electricity sales call prediction model and obtains the electricity sales call probability corresponding to the time to be predicted, the information processing device also obtains a plurality of groups of sample object information, a plurality of groups of sample electricity sales call information and a plurality of groups of sample historical service information corresponding to a plurality of sample objects; the information processing device classifies a plurality of sample objects according to the plurality of groups of sample electricity sales call information and the plurality of groups of sample historical service information to obtain a plurality of classification groups; respectively determining a plurality of first sample information, a plurality of first sample telemarketing communication information, a plurality of first sample historical service information and a plurality of first sample output targets corresponding to the plurality of classification groups; the information processing device trains an initial electricity sales call prediction model in sequence by using the first sample information, the electricity sales call information, the historical service information and the output targets of the first samples respectively to obtain a plurality of electricity sales call prediction models.

It should be noted that the plurality of electricity sales call prediction models include a target electricity sales call prediction model.

In an embodiment of the present application, the plurality of sample objects may be a plurality of historical telemarketing customers. The plurality of sets of sample object information may be customer basic information left when a plurality of historical electricity sales customers register. The plurality of historical telemarketing clients correspond to the plurality of groups of sample object information one to one, and specifically, one historical telemarketing client corresponds to one group of sample object information.

It should be noted that the plurality of sets of sample object information include ages, sexes, professions, academic calendars, graduates, working years, working units, provinces, cities, and the like of the plurality of historical electricity marketing clients.

In the embodiment of the present application, the plurality of sets of sample telemarketing communication information may be information of previous telemarketing with a plurality of historical telemarketing customers. The plurality of historical telemarketing clients correspond to the plurality of groups of sample telemarketing communication information one by one, and specifically, one historical telemarketing client corresponds to one group of sample telemarketing communication information.

It should be noted that the multi-group sample telemarketing communication information includes: the time of the electricity sales staff dialing the sales calls for the plurality of sample objects, the on-time of the plurality of sample objects, the hang-up time of the plurality of sample objects, the recording start time, the recording end time, the ringing waiting time, the recording time, the ID of the electricity sales staff and the like.

In the embodiment of the application, the plurality of sets of sample historical service information may be service information transacted by a plurality of historical telemarketing clients. The plurality of historical telemarketing clients correspond to the plurality of groups of sample historical service information one by one, and specifically, one historical telemarketing client corresponds to one group of sample historical service information.

It should be noted that the plurality of sets of sample historical service information include: the service product IDs browsed by the plurality of historical telemarketing clients, the service product IDs used by the plurality of historical telemarketing clients, and related use attributes of the service products, such as deposit, transfer, loan and other attributes, and corresponding service product operation time and the like.

In the embodiment of the present application, the process of classifying a plurality of sample objects by the information processing device according to a plurality of sets of sample telemarketing communication information and a plurality of sets of sample historical service information to obtain a plurality of classification groups may be a process of classifying a plurality of sample objects into telemarketing old customers and telemarketing new customers according to a plurality of sets of sample telemarketing communication information by the information processing device; then the information processing device divides the plurality of sample objects into service clients and non-service clients according to the plurality of groups of sample historical service information, specifically, if the number of the service types of the sample historical service information is n, the service clients can be continuously subdivided into n service clients. The information processing device firstly determines a first classification group which is a telemarketing old customer and a non-business customer; the information processing device determines a second classification group which is a new customer for electricity sale and is a non-service customer; then the information processing device determines a third classification group which is a telemarketer and is a first type service client, a fourth classification group which is the telemarketer and is a second type service client, …, and a n +2 classification group which is the telemarketer and is an nth type service client; finally, the information processing apparatus determines an n +3 th classification group which is a new customer for electricity sales and is a first-class service customer, an n +4 th classification group which is a new customer for electricity sales and is a second-class service customer, …, and a 2n +2 th classification group which is a new customer for electricity sales and is an nth-class service customer, thereby obtaining 2n +2 classification groups, namely a plurality of classification groups.

In the embodiment of the present application, after the information processing apparatus obtains the plurality of electricity sales call prediction models, the information processing apparatus may configure the correspondence between the plurality of classification groups and the plurality of electricity sales call prediction models, so that when the information processing apparatus determines the group information corresponding to the target object according to the historical electricity sales call information and the historical service information of the target object, the information processing apparatus may determine the target electricity sales call prediction model corresponding to the group according to the correspondence between the plurality of classification groups and the plurality of electricity sales call prediction models.

In the embodiment of the present application, the initial electricity sales call prediction model may be a model configured in the information processing apparatus; the initial electricity sales call prediction model can also be a model transmitted to the information processing device by other equipment; the initial electricity sales call prediction model can also be a model acquired by the information processing device in other manners; the specific manner in which the information processing apparatus obtains the initial telemarketing prediction model may be determined according to actual conditions, which is not limited in the embodiment of the present application.

In the embodiment of the application, the initial electricity sales call prediction model can be an XGboost model; the initial electricity sales call prediction model can be other models; the specific details can be determined according to actual situations, and the embodiment of the present application does not limit the details.

In an embodiment of the present application, a process in which the plurality of first sample output targets include a plurality of first tandem output targets, a plurality of second tandem output targets, and a plurality of third tandem output targets, and the information processing apparatus trains an initial telemarketing prediction model in sequence by using the plurality of first sample information, the plurality of first sample telemarketing information, the plurality of first sample historical service information, and the plurality of first sample output targets, respectively, to obtain the plurality of telemarketing prediction models includes: the information processing device trains an initial power distribution call prediction model in sequence by using the first sample information, the first sample power distribution call information, the first sample historical service information and the first series output targets to obtain a plurality of first series power distribution call prediction models; the information processing device trains an initial electricity sales call prediction model in sequence by using the first sample information, the first sample electricity sales call information, the historical service information and the second series output targets to obtain second series electricity sales call prediction models; the information processing device trains an initial electricity sales call prediction model in sequence by using the first sample information, the first sample electricity sales call information, the historical service information and the third series output targets to obtain third series electricity sales call prediction models; the information processing device uses the plurality of first serial electric sales call prediction models, the plurality of second serial electric sales call prediction models, and the plurality of third serial electric sales call prediction models as the plurality of electric sales call prediction models.

In the embodiment of the present application, the first serial output target may be a target that the client can be connected within the ring duration; the second series output target can be a target for the customer to actively click through the call; the third tandem output target may be a target that the call duration when the customer is currently on sale can satisfy the effective call duration.

For example, the initial power-off call prediction model (OBJ) may be as shown in equation (1):

it should be noted that, in the following description,

in order to be a function of the model loss,

the representative regularization term is a parameter for preventing overfitting of the model. The information processing apparatus may sort the plurality of first sample information, the plurality of first sample telemarketing communication information, and the plurality of first sample historical service information into the plurality of sets of sample characteristics

A plurality of first samples corresponding to a plurality of classification groupsThe output target is used as a plurality of target labels

Sequentially training a target function (an initial electricity sales call prediction model) by utilizing a plurality of groups of sample characteristics and a plurality of target labels, and obtaining a plurality of electricity sales call prediction models (sub-models) by using a Boosting addition training method formed by combining a plurality of CART tree models during training

)。

For example, if the number of the plurality of classification groups is 2n +2, the number of the plurality of electricity sales call prediction models is also 2n + 2.

It should be noted that any one of the plurality of electric outlets call prediction models includes 3 series electric outlets call prediction models, that is, a first series electric outlets call prediction model, a second series electric outlets call prediction model, and a third series electric outlets call prediction model. The first series electric pin call prediction model is used for outputting the probability that the electric pin customer can be connected within the ringing time; the second call connection prediction model is used for outputting the probability that the electricity sales client can actively click to connect the call; the third call distribution call prediction model is used for outputting the probability that the call duration of the current call distribution of the call distribution client can meet the effective call duration.

It should be noted that, in the process that the information processing apparatus sequentially trains the initial power-off call prediction model by using the plurality of first sample information, the plurality of first sample power-off call information, the plurality of first sample historical service information and the plurality of second tandem output targets to obtain the plurality of second tandem power-off call prediction models, the information processing apparatus may first screen out sample information that the power-off client cannot be connected within the ring-down duration from the plurality of first sample information, the plurality of first sample power-off call information and the plurality of first sample historical service information to obtain the plurality of screened-out first sample information, the plurality of screened-out first sample power-off call information and the plurality of screened-out first sample historical service information (i.e. screen out the plurality of screened-out first sample power-off call information and the plurality of screened-out first sample historical service information

Corresponding sample information is obtained

Corresponding sample information); to output targets (target labels) in series by using the first screened sample information, the first screened sample telemarketing information, the historical service information and the second series output targets

) Training the initial electric marketing communication prediction model in sequence to obtain a plurality of second serial electric marketing communication prediction models (sub-models)

)。

It should be noted that, in the process of sequentially training the initial power-off call prediction model by the information processing apparatus using the plurality of first sample information, the plurality of first sample power-off call information, the plurality of first sample historical service information and the plurality of third tandem output targets to obtain the plurality of third tandem power-off call prediction models, the information processing apparatus may first screen out sample information that the power-off client cannot be connected within the ring-down duration and sample information that the power-off client cannot actively click to connect to the call from the plurality of first sample information, the plurality of first sample power-off call information and the plurality of third sample historical service information (i.e., screen out the plurality of first sample information, the plurality of first sample power-off call information and the plurality of second sample historical service information after processing (i.e., screen out the plurality of first sample information, the plurality of first sample power-off call information and the plurality of third sample historical service information after processing)

And

corresponding sample information is obtained

Corresponding samplePresent information) to utilize the plurality of processed first sample information, the plurality of processed first sample telemarketing information, the plurality of processed first sample historical service information, and the plurality of third tandem output destinations (destination tags)

) Training the initial electric marketing communication prediction model in sequence to obtain a plurality of third series electric marketing communication prediction models (sub-models)

)。

In this embodiment, if the number of the plurality of electricity sales promotion call prediction models is also 2n +2, the information processing apparatus may obtain 2n +2 first series electricity sales promotion call prediction models, 2n +2 second series electricity sales promotion call prediction models, and 2n +2 third series electricity sales promotion call prediction models, that is, obtain the electricity sales promotion call prediction models

In this embodiment, before the information processing apparatus sequentially trains the initial power distribution call prediction model by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information, and the plurality of first sample output targets, respectively, to obtain the plurality of power distribution call prediction models, the information processing apparatus may classify the plurality of first sample information, the plurality of first sample power distribution call information, and the plurality of first sample historical service information according to a linear information feature, a nonlinear information feature, and a timing information feature, to obtain first sample linear information, first sample nonlinear information, and first sample timing information; then the information processing device carries out linear coding processing on the first sample nonlinear information to obtain first sample linear coding information; the information processing device performs time sequence derivation processing on the first sample time sequence information to obtain first sample derived time sequence information.

Accordingly, in the process of sequentially training the initial power distribution call prediction models by the information processing apparatus using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information, and the plurality of first serial output targets to obtain the plurality of first serial power distribution call prediction models, the initial power distribution call prediction models may be sequentially trained by the information processing apparatus using the first sample linear information, the first sample linear coding information, the first sample derived timing information, and the plurality of first serial output targets to obtain the plurality of first serial power distribution call prediction models.

Accordingly, the process of the information processing apparatus sequentially training the initial power distribution network prediction model by using the plurality of first sample information, the plurality of first sample power distribution network information, the plurality of first sample historical service information, and the plurality of second tandem output targets to obtain the plurality of second tandem power distribution network prediction models may be a process of sequentially training the initial power distribution network prediction model by using the first sample linear information, the first sample linear coding information, the first sample derived timing information, and the plurality of second tandem output targets for the information processing apparatus to obtain the plurality of second tandem power distribution network prediction models.

Accordingly, in the process of sequentially training the initial power distribution network call prediction model by the information processing apparatus using the plurality of first sample information, the plurality of first sample power distribution network call information, the plurality of first sample historical service information, and the plurality of third tandem output targets to obtain the plurality of third tandem power distribution network call prediction models, the initial power distribution network call prediction model may be sequentially trained by the information processing apparatus using the first sample linear information, the first sample linear coding information, the first sample derived timing information, and the plurality of third tandem output targets to obtain the plurality of third tandem power distribution network call prediction models.

In this embodiment of the application, in the process of performing linear coding processing on the first sample nonlinear information by the information processing apparatus to obtain the first sample linear coding information, the information processing apparatus may perform linear coding processing on the first sample nonlinear information by using an evidence weight coding manner to obtain the first sample linear coding information.

In this embodiment of the application, the process of performing, by the information processing apparatus, time-series derivation processing on the first sample time-series information to obtain the first sample derived time-series information may be a process of converting, by the information processing apparatus, the first sample time-series information according to a preset time conversion manner to obtain the first sample derived time-series information; and/or determining first sample derived timing information of the first sample timing information according to a time node where the first sample timing information is located; and/or counting the number of the first sample time sequence information in a preset time period to obtain first sample derived time sequence information; and/or determining a time interval between two adjacent time sequence information in the at least two first sample time sequence information under the condition that the number of the first sample time sequence information is at least two, so as to obtain the first sample derived time sequence information.

In the embodiment of the present application, the information processing apparatus may construct a plurality of sets of sample information matrices (a plurality of sets of sample characteristics) according to the plurality of first sample information, the plurality of first sample telemarketing communication information, and the plurality of first sample historical service information

The plurality of first sample output targets may be

The plurality of second sample output targets may be

The plurality of third sample output targets may be

Specifically, a feature matrix constructed by using a plurality of groups of sample features, a plurality of first sample output targets, a plurality of second sample output targets, and a plurality of third sample output targets is as shown in formula (2):

it should be noted that pi is information of a group of sample objects corresponding to the ith telephone sales call prediction model (sub-model) at each time of making a telephone call, and qi is a group of samples corresponding to the ith sub-modelThe number of the features in question,

for the sample feature corresponding to the ith sub-model,

a plurality of first series output targets, a plurality of second series output targets, and a plurality of third series output targets, respectively.

In the embodiment of the present application, the manner in which the information processing apparatus trains and obtains a plurality of electricity sales call prediction models is shown in fig. 2:

s21, the information processing device first obtains a plurality of sets of sample object information, a plurality of sets of sample telemarketing information, and a plurality of sets of sample historical service information corresponding to a plurality of sample objects.

S22, the information processing device classifies the plurality of sample objects according to the plurality of sets of sample telemarketing communication information and the plurality of sets of sample historical service information to obtain a plurality of classification groups, and determines a plurality of first sample information, a plurality of first sample telemarketing communication information, and a plurality of first sample historical service information corresponding to the plurality of classification groups, respectively.

S23, the information processing apparatus trains the initial telemarketing communication prediction model in sequence by using the plurality of first sample information, the plurality of first sample telemarketing communication information, the plurality of first sample historical service information, and the plurality of first sample output targets, respectively, to obtain a plurality of telemarketing communication prediction models.

In this embodiment, the information processing apparatus may first classify the plurality of first sample information, the plurality of first sample telemarketing communication information, and the plurality of first sample historical service information according to a linear information characteristic, a nonlinear information characteristic, and a time sequence information characteristic to obtain first sample linear information, first sample nonlinear information, and first sample time sequence information; then the information processing device carries out linear coding processing on the first sample nonlinear information by using an evidence weight coding mode to obtain first sample linear coding information (WOE value); the information processing device performs time series derivation processing on the first sample time series information to obtain first sample derived time series information (time attribute feature, period derived feature, date statistical feature, date interval feature). Finally, the information processing apparatus performs group classification on the plurality of sample objects, and thereby can obtain 2n +2 classification groups, and the 2n +2 classification groups respectively correspond to the 2n +2 first sample linear information, the 2n +2 first sample linear coding information, and the 2n +2 first sample derived time series information. Because normal once sales call of telemarketing can be split into 3 stages, stage one: ringing a mobile phone of the customer of the electronic pin; and a second stage: the electricity sales client clicks the on/off stage; and a third stage: and E, the customer voice call stage is handed off. Therefore, the three stages can respectively generate invalid calls, specifically: stage one, ringing overtime is not connected; stage two, the customer clicks to hang up; and step three, the call duration does not reach the standard. Thus, the goal of a telemarketing call prediction model is to resolve from meeting the effective call duration to meeting a series model of three steps of call completion within the ring duration, call completion by customer click, and effective call duration. The initial electric pin call prediction models are trained by setting the series output targets (a plurality of first series output targets, a plurality of second series output targets and a plurality of third series output targets) corresponding to the three stages and respectively using the plurality of first sample linear information, the plurality of first sample linear coding information and the plurality of first sample derived time sequence information and the plurality of first series output targets, the plurality of second series output targets and the plurality of third series output targets, so that a plurality of first series electric pin call prediction models, a plurality of second series electric pin call prediction models and a plurality of third series electric pin call prediction models are obtained. The information processing device uses the plurality of first serial electric sales call prediction models, the plurality of second serial electric sales call prediction models, and the plurality of third serial electric sales call prediction models as the plurality of electric sales call prediction models.

In the embodiment of the application, specifically, as shown in fig. 3, the series-parallel model training architecture diagram is a diagram that horizontally includes a plurality of electricity-selling conversation prediction models, and vertically includes a first series-connection electricity-selling conversation prediction model, a second series-connection electricity-selling conversation prediction model, and a third series-connection electricity-selling conversation prediction model, where each of the plurality of electricity-selling conversation prediction models is connected in series.For example, if the number of the plurality of electric pin call prediction models is 2n +2, the number of the series electric pin call prediction models is (2n +2) × 3 ═ 6n + 6. The information processing device firstly acquires a plurality of groups of sample object information (user information), a plurality of groups of sample telemarketing communication information (telemarketing information) and a plurality of groups of sample historical service information (service information) corresponding to a plurality of sample objects, wherein the service information comprises service 1, service 2, … and service n. The information processing device classifies a plurality of sample objects according to a plurality of groups of sample telemarketing call information and a plurality of groups of sample historical service information to obtain a plurality of classification groups (a non-service client/telemarketing old client (a client group 1), a service client/telemarketing old client (a client group 2), a service client/telemarketing new client (a client group 3) and a non-service client/telemarketing new client (a client group 4), wherein the client group 2 and the client group 3 comprise service clients, the client group 2 comprises n groups according to different services, and the client group 3 comprises n groups), and respectively determines a plurality of first sample information corresponding to the classification groups, a plurality of first sample telemarketing call information, a plurality of first sample historical service information corresponding to the classification groups; the information processing device obtains a plurality of electricity sales call prediction models (including XGB) by sequentially training an initial electricity sales call prediction model using a plurality of first sample information, a plurality of first sample electricity sales call information, a plurality of first sample historical service information, and a plurality of first sample output targets_1-1、XGB_2-1-1、…、XGB_2-n-1、XGB_3-1-1、…、XGB_3-n-1、XGB_4-1；XGB_1-2、XGB_2-1-2、…、XGB_2-n-2、XGB_3-1-2、…、XGB_3-n-2、XGB_4-2；XGB_1-3、XGB_2-1-3、…、XGB_2-n-3、XGB_3-1-3、…、XGB_3-n-3、XGB_4-3)。

In an embodiment of the present application, a process in which an information processing apparatus inputs object information, historical electricity sales call information, historical service information, and time to be predicted into a target electricity sales call prediction model to obtain an electricity sales call probability corresponding to the time to be predicted includes: the information processing device determines group information corresponding to the target object according to the historical telemarketing communication information and the historical service information; the information processing device determines a target electricity sales call prediction model corresponding to the group information according to the corresponding relation between the configured plurality of classification groups and the plurality of electricity sales call prediction models; the information processing device determines the power pin connection probability according to the object information, the historical power pin call information, the historical service information and the time to be predicted by using the target power pin call prediction model.

In the embodiment of the application, the information processing apparatus may output the power pin connection probability by inputting the object information, the historical power pin call information, the historical service information, and the time to be predicted into the target power pin call prediction model for the information processing apparatus, in a manner of determining the power pin connection probability according to the object information, the historical power pin call information, the historical service information, and the time to be predicted by using the target power pin call prediction model.

In the embodiment of the application, the information processing device can determine whether the target object is an old electricity selling customer or a new electricity selling customer according to the historical electricity selling conversation information; the information processing device can determine the service type of the target object according to the historical service information, and can determine the group information corresponding to the target object according to the service type and the old electricity sales customers or the new electricity sales customers.

In an embodiment of the present application, a process of determining a power-on probability by an information processing apparatus using a target power-off call prediction model according to object information, historical power-off call information, historical service information, and time to be predicted includes: the information processing device classifies the object information, the historical telemarketing communication information, the time to be predicted and the historical service information according to the linear information characteristic, the nonlinear information characteristic and the time sequence information characteristic to obtain linear information, nonlinear information and time sequence information; the information processing device carries out linear coding processing on the nonlinear information to obtain linear coding information; the information processing device carries out time sequence derivation processing on the time sequence information to obtain derived time sequence information; the information processing device inputs the linear information, the linear coding information and the derived time sequence information into a target power plug call prediction model to obtain power plug call connection probability.

In the embodiment of the application, the target electricity sales call prediction model comprises a first series electricity sales call prediction model, a second series electricity sales call prediction model and a third series electricity sales call prediction model; the process of inputting linear information, linear coding information and derived time sequence information into a target power pin call prediction model by an information processing device to obtain power pin call probability comprises the following steps: the information processing device inputs the linear information, the linear coding information and the derivative time sequence information into a first series telegraph-sharing call prediction model to obtain a first probability; the information processing device inputs the linear information, the linear coding information and the derived time sequence information into a second series connection electric marketing communication prediction model to obtain a second probability; the information processing device inputs the linear information, the linear coding information and the derived time sequence information into a third series connection electric marketing communication prediction model to obtain a third probability; the information processing device determines the electrical pin connection probability according to the first probability, the second probability and the third probability.

The information processing apparatus may determine the electrical pin connection probability based on the first probability, the second probability, and the third probability, and may include: the information processing apparatus determines a product between the first probability, the second probability, and the third probability, and takes the product as a pin connection probability.

In this embodiment of the present application, a process of performing timing derivation processing on timing information by an information processing apparatus to obtain derived timing information includes: the information processing device converts the time sequence information according to a preset time conversion mode to obtain derived time sequence information; and/or the information processing device determines derived time sequence information of the time sequence information according to the time node where the time sequence information is located; and/or the information processing device counts the number of the time sequence information in a preset time period to obtain derived time sequence information; and/or the information processing device determines the time interval between two adjacent time sequence information in the at least two time sequence information under the condition that the number of the time sequence information is at least two, and obtains the derived time sequence information.

In the embodiment of the present application, the preset time conversion mode may be a year-month-day time: dividing into: a second transformation mode; the preset time conversion mode may also be another time conversion mode, which may be determined specifically according to an actual situation, and this is not limited in this application embodiment.

In the embodiment of the present application, the information processing apparatus determines a time interval between two adjacent pieces of timing information in at least two pieces of timing information, and obtains the derived timing information in a manner shown in equation (3):

DateDiff＝{(TS_i-TS_j)，i≠j} (3)

note that DateDiff is derived timing information, TS_iAnd TS_jTwo adjacent pieces of timing information.

In the embodiment of the present application, the predetermined time transformation manner may be "yyyy-mm-dd hh: mm: ss' transformation mode; the preset time conversion mode may also be another time conversion mode, which may be determined specifically according to an actual situation, and this is not limited in this application embodiment. The information processing apparatus may split the derived time-series information according to attributes of year, month, day, hour, minute, second, and the like of the history tapping time.

In the embodiment of the present application, the information processing apparatus may be a derivative feature calculated based on a time period for the information processing apparatus in a manner that the information processing apparatus determines the derivative timing information of the timing information according to the time node where the timing information is located. For example, the information processing apparatus may derive the derived time series information by performing feature derivation on the seasons Q1 to Q4, the middle and upper half of the month, the week of the present month, the day of the week, whether the weekend is present, and whether the holiday is present.

In the embodiment of the present application, the information processing apparatus may obtain the derived timing information by counting the number of the timing information within a preset time period, and the derived timing information may be obtained for the information processing apparatus based on a statistical value of time. For example, the information processing apparatus may count the total number of times of dialing and turning on by year, month, day, hour and minute, for example, a statistical method of 10 times of total dialing and turning on 3 times of total in 2021 year, 2 times of total dialing and turning on 1 time of total in 10 months is similar.

In the embodiment of the present application, when the number of the time series information is at least two, the information processing apparatus determines a time interval between two adjacent time series information in the at least two time series information to obtain the derived time series information, which may be a first order difference value based on time, and mainly calculates a day interval during a day period, for example, a number of days from a first time of pin pulling to the current time of pin pulling, a number of days from a first time of pin connecting to a last time of pin connecting, and the like, where dates mainly related to which difference values need to be calculated two by two are: the time sequence data set of each telepin client can be recorded as TimeSeries ═ TS1, TS2, … and TS13, and the specific time difference calculation formula is shown as formula (3).

For example, the processing structure diagram of the information processing apparatus for processing the plurality of sets of sample object information, the plurality of sets of sample telemarketing information, and the plurality of sets of sample historical service information is shown in fig. 4: when the information processing device acquires object information (user information), historical electricity sales call information (electricity sales information) and historical service information (service information) corresponding to a target object, the information processing device classifies the object information, the historical electricity sales call information, the time to be predicted and the historical service information according to a linear information characteristic, a nonlinear information characteristic and a time sequence information characteristic to obtain linear information, the nonlinear information and the time sequence information; the information processing device encodes the nonlinear information by an evidence weight encoding method (WOE encoding) to obtain linear encoded information (WOE value); the information processing device converts (splits attributes) the time sequence information according to a preset time conversion mode to obtain derived time sequence information (time attribute information); the information processing device determines derived timing information (cycle derived information) of the timing information according to a time node (time cycle calculation) where the timing information is located; the information processing device counts the number of the time sequence information (time counting value) in a preset time period to obtain derived time sequence information (date counting information); the information processing apparatus determines a time interval between adjacent two of the at least two pieces of timing information (first order difference value calculation) to obtain derived timing information (date interval information).

In this embodiment of the present application, a process of performing linear coding processing on nonlinear information by an information processing apparatus to obtain linear coded information includes: the information processing device encodes the nonlinear information by an evidence weight encoding mode to obtain linear encoded information.

In the embodiment of the present application, the Evidence Weight encoding (WOE) manner is specifically shown in equation (4):

it should be noted that i is an attribute of the nonlinear information, Y is a power-on pin client, N is a power-off pin client, Yi is the number of power-on pin clients under the attribute of the nonlinear information, Ni is the number of power-off pin clients under the attribute of the nonlinear information, the total number of power-on pin clients under all the attributes of YT non-nonlinear information, the total number of power-off pin clients under all the attributes of NT non-nonlinear information, woe_iInformation is encoded linearly.

And S104, under the condition that the time to be predicted is determined to be the target electricity selling time according to the electricity selling connection probability, executing a telephone selling process between the target object and the electricity selling time.

In the embodiment of the application, the information processing device inputs the object information, the historical electricity sales call information, the historical service information and the time to be predicted into the target electricity sales call prediction model, obtains the electricity sales call connection probability corresponding to the time to be predicted, and executes the telephone sales process with the target object within the target electricity sales call time under the condition that the information processing device determines that the time to be predicted is the target electricity sales call time according to the electricity sales call connection probability.

In the embodiment of the application, the process that the information processing device determines the time to be predicted as the target electricity selling time according to the electricity pin connection probability can be that the information processing device determines the time to be predicted as the target electricity selling time under the condition that the electricity pin connection probability is greater than or equal to a preset probability threshold; in the case where the power pin connection probability is smaller than the preset probability threshold, the information processing apparatus determines that the time to be predicted is not the target power pin time.

It should be noted that the number of the preset probability threshold may be one; the number of the preset probability threshold values can also be two; the number of the preset probability threshold values can also be multiple; the number of the specific preset probability threshold values may be determined according to actual conditions, which is not limited in the embodiment of the present application.

It should be further noted that, if the number of the preset probability threshold values is multiple, different preset probability threshold values may be set according to different services; that is, a plurality of service information correspond to a plurality of preset probability thresholds, wherein one service information corresponds to one preset probability threshold.

An exemplary information processing architecture is shown in fig. 5: when the information processing apparatus receives the electricity sales call prediction instruction, the process of determining the electricity sales call probability corresponding to the time to be predicted in the electricity sales call prediction instruction by the information processing apparatus may be: the information processing device acquires object identification information and time to be predicted (predicted time) of a target object (target client) from the electricity sales call prediction instruction, and acquires object information (client id), historical electricity sales call information (electricity sales id) and historical business information (business id) corresponding to the target object from a database according to the object identification information (id of the target client); the information processing device determines group information corresponding to a target object according to the historical telemarketing communication information and the historical service information (the client is grouped, and the target client is determined to be a group corresponding to a non-service client/telemarketing old client, a group corresponding to a service client/telemarketing new client, a group corresponding to a non-service client/telemarketing new client or a group not containing the client); the information processing device determines a target electricity sales call prediction model corresponding to the group information (the electricity sales call prediction model corresponding to the parallel passenger group 1, or the electricity sales call prediction model corresponding to the parallel passenger group 4, or the electricity sales call prediction model corresponding to a certain service in the parallel passenger group 2, or the electricity sales call prediction model corresponding to a certain service in the parallel passenger group 3) according to the corresponding relation between the configured plurality of classification groups and the plurality of electricity sales call prediction models; the information processing apparatus can then execute the information processing process of: the information processing device classifies the object information, the historical telemarketing communication information, the time to be predicted and the historical service information according to the linear information characteristic, the nonlinear information characteristic and the time sequence information characteristic to obtain linear information, nonlinear information and time sequence information; the information processing device carries out linear coding processing on the nonlinear information to obtain linear coding information (WOE value); the information processing device carries out time sequence derivation processing on the time sequence information to obtain derived time sequence information (including time attribute information, period derivation information, date statistical information and date interval information); finally, the information processing apparatus may perform a model prediction process: the target electricity sales call prediction model comprises a first series electricity sales call prediction model corresponding to the series stage 1, a second series electricity sales call prediction model corresponding to the series stage 2 and a third series electricity sales call prediction model corresponding to the series stage 3; the information processing device inputs the linear information, the linear coding information and the derivative time sequence information into a first series telegraph-sharing call prediction model to obtain a first probability; the information processing device inputs the linear information, the linear coding information and the derived time sequence information into a second series connection electric marketing communication prediction model to obtain a second probability; the information processing device inputs the linear information, the linear coding information and the derived time sequence information into a third series connection electric marketing communication prediction model to obtain a third probability; the information processing device determines an electrical pin connection probability according to the first probability, the second probability, and the third probability and outputs the electrical pin connection probability.

It should be noted that the first serial telephone-only call prediction model corresponding to the parallel guest group 1 is XGB_1-1The second serial electric sales call prediction model corresponding to the parallel passenger group 1 is XGB_1-2And the third series connection electric pin call prediction model corresponding to the parallel connection passenger group 1 is XGB_1-3(ii) a The first series connection electric pin conversation prediction model corresponding to the parallel connection passenger group 4 is XGB_4-1The second serial electric sales call prediction model corresponding to the parallel passenger group 4 is XGB_4-2Third series pins corresponding to the parallel passenger group 4The call prediction model is XGB_4-3(ii) a The first series connection electric pin conversation prediction model corresponding to the first service in the parallel connection passenger group 2 is XGB_2-1-1The second serial electric sales call prediction model corresponding to the first service in the parallel passenger group 2 is XGB_2-1-2The third series connection electric pin conversation prediction model corresponding to the first service in the parallel connection passenger group 2 is XGB_2-1-3(ii) a The first series connection electric pin conversation prediction model corresponding to the nth service in the parallel connection passenger group 2 is XGB_2-n-1The second serial electric sales call prediction model corresponding to the nth service in the parallel passenger group 2 is XGB_2-n-2The third series connection electric sales call prediction model corresponding to the nth service in the parallel connection passenger group 2 is XGB_2-n-3(ii) a The first series connection electric marketing communication prediction model corresponding to the first service in the parallel connection passenger group 3 is XGB_3-1-1The second serial electric sales call prediction model corresponding to the first service in the parallel passenger group 3 is XGB_3-1-2The third series connection electric marketing communication prediction model corresponding to the first service in the parallel connection passenger group 3 is XGB_3-1-3(ii) a The first series connection electric pin conversation prediction model corresponding to the nth service in the parallel connection passenger group 3 is XGB_3-n-1The second serial electric sales call prediction model corresponding to the nth service in the parallel passenger group 3 is XGB_3-n-2The third series connection electric sales call prediction model corresponding to the nth service in the parallel connection passenger group 3 is XGB_3-n-3。

It should be noted that the electricity sales information in the database includes electricity sales events, electricity sales records, electricity sales seats, and the like; the service information in the database comprises service products 1, … and a service product n; the client information in the database includes client reservation information _ gender, client reservation information _ age, and the like. The information processing device can acquire the electricity sales id of the target customer from the electricity sales information in the database; acquiring a service id of a target client from service information in a database; and acquiring the client id of the target client from the client information in the database.

The information processing device can be used for predicting the electricity pin connection probability of successful communication with the electricity pin of the target object at the time to be predicted by using the electricity pin connection probability by acquiring the object information, the historical electricity pin call information and the historical service information corresponding to the target object and using the electricity pin call prediction model according to the object information, the historical electricity pin call information and the historical service information, so that the accuracy of the time to be predicted is determined by using the electricity pin connection probability, the target electricity pin time with high accuracy can be determined, and the accuracy of determining the target electricity pin time is improved.

Example two

Based on the idea of the invention together with the embodiments, the embodiments of the present application provide an information processing apparatus 1 corresponding to an information processing method; fig. 6 is a schematic diagram illustrating a first composition structure of an information processing apparatus according to an embodiment of the present application, where the information processing apparatus 1 may include:

the device comprises an acquisition unit 11, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring object identification information and time to be predicted of a target object from an electricity sales call prediction instruction under the condition of receiving the electricity sales call prediction instruction; acquiring object information, historical telemarketing communication information and historical service information corresponding to the target object according to the object identification information;

the input unit 12 is configured to input the object information, the historical electricity sales call information, the historical service information, and the time to be predicted into a target electricity sales call prediction model, so as to obtain an electricity sales call connection probability corresponding to the time to be predicted;

and the execution unit 13 is configured to execute a telemarketing process with the target object within the target telemarketing time if it is determined that the time to be predicted is the target telemarketing time according to the telemarketing connection probability.

In some embodiments of the present application, the apparatus further comprises a classification unit, the determination unit, a training unit;

the acquiring unit 11 is configured to acquire multiple sets of sample object information, multiple sets of sample telemarketing communication information, and multiple sets of sample historical service information corresponding to multiple sample objects;

the classification unit is used for classifying the plurality of sample objects according to the plurality of groups of sample electricity sales call information and the plurality of groups of sample historical service information to obtain a plurality of classification groups;

the determining unit is used for respectively determining a plurality of first sample information, a plurality of first sample telemarketing information, a plurality of first sample historical service information and a plurality of first sample output targets corresponding to the plurality of classification groups;

the training unit is configured to train an initial power distribution network call prediction model in sequence by using the plurality of first sample information, the plurality of first sample power distribution network call information, the plurality of first sample historical service information, and the plurality of first sample output targets, respectively, to obtain a plurality of power distribution network call prediction models; the plurality of electricity sales call prediction models includes the target electricity sales call prediction model.

In some embodiments of the present application, the plurality of first sample output targets comprises a plurality of first series output targets, a plurality of second series output targets, and a plurality of third series output targets;

the training unit is configured to train an initial power distribution call prediction model in sequence by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information, and the plurality of first serial output targets to obtain a plurality of first serial power distribution call prediction models; sequentially training an initial power distribution call prediction model by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information and the plurality of second series output targets to obtain a plurality of second series power distribution call prediction models; sequentially training an initial power distribution call prediction model by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information and the plurality of third series output targets to obtain a plurality of third series power distribution call prediction models; using the plurality of first series electric sales call prediction models, the plurality of second series electric sales call prediction models, and the plurality of third series electric sales call prediction models as the plurality of electric sales call prediction models.

In some embodiments of the present application, the determining unit is configured to determine group information corresponding to the target object according to the historical telemarketing communication information and the historical service information; determining the target electricity sales call prediction model corresponding to the group information according to the corresponding relation between the configured classification groups and the electricity sales call prediction models; and determining the power pin connection probability according to the object information, the historical power pin call information, the historical service information and the time to be predicted by using the target power pin call prediction model.

In some embodiments of the present application, the apparatus further comprises an encoding unit and a derivation unit;

the classification unit is used for classifying the object information, the historical telemarketing communication information, the time to be predicted and the historical service information according to linear information characteristics, nonlinear information characteristics and time sequence information characteristics to obtain linear information, nonlinear information and time sequence information;

the coding unit is used for carrying out linear coding processing on the nonlinear information to obtain linear coding information;

the derivation unit is used for performing time sequence derivation processing on the time sequence information to obtain derived time sequence information;

the input unit 12 is configured to input the linear information, the linear coding information, and the derived timing information into the target power pin call prediction model, so as to obtain the power pin call probability.

In some embodiments of the present application, the target electricity sales call prediction model comprises a first series electricity sales call prediction model, a second series electricity sales call prediction model, and a third series electricity sales call prediction model;

the input unit 12 is configured to input the linear information, the linear coding information, and the derived timing information into the first serial telemarketing prediction model to obtain a first probability; inputting the linear information, the linear coding information and the derived time sequence information into the second serial electric marketing communication prediction model to obtain a second probability; inputting the linear information, the linear coding information and the derived time sequence information into the third serial electric marketing communication prediction model to obtain a third probability;

the determining unit is used for determining the electric pin connection probability according to the first probability, the second probability and the third probability.

In some embodiments of the present application, the apparatus further comprises a conversion unit and a statistics unit;

the conversion unit is used for converting the time sequence information according to a preset time conversion mode to obtain the derived time sequence information;

and/or the determining unit is used for determining derived time sequence information of the time sequence information according to the time node where the time sequence information is located; and/or determining a time interval between two adjacent time sequence information in the at least two time sequence information under the condition that the number of the time sequence information is at least two, so as to obtain the derived time sequence information;

and/or the statistical unit is used for counting the number of the time sequence information in a preset time period to obtain the derived time sequence information.

In some embodiments of the present application, the encoding unit is configured to encode the nonlinear information by using an evidence weight encoding manner, so as to obtain the linear encoding information.

In practical applications, the obtaining Unit 11, the input Unit 12 and the execution Unit 13 may be implemented by a processor 14 on the information Processing apparatus 1, specifically implemented by a CPU (Central Processing Unit), an MPU (Microprocessor Unit), a DSP (Digital Signal processor), a Field Programmable Gate Array (FPGA), or the like; the above data storage may be realized by the memory 15 on the information processing apparatus 1.

An embodiment of the present application also provides an information processing apparatus 1, and as shown in fig. 7, the information processing apparatus 1 includes: a processor 14, a memory 15 and a communication bus 16, the memory 15 communicating with the processor 14 through the communication bus 16, the memory 15 storing a program executable by the processor 14, the program, when executed, executing the information processing method as described above through the processor 14.

In practical applications, the Memory 15 may be a volatile Memory (volatile Memory), such as a Random-Access Memory (RAM); or a non-volatile Memory (non-volatile Memory), such as a Read-Only Memory (ROM), a flash Memory (flash Memory), a Hard Disk (Hard Disk Drive, HDD) or a Solid-State Drive (SSD); or a combination of the above types of memories and provides instructions and data to processor 14.

The present embodiment provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by the processor 14, the information processing method described above is implemented.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above description is only a preferred embodiment of the present application, and is not intended to limit the scope of the present application.

Claims

1. An information processing method, characterized in that the method comprises:

2. The method according to claim 1, wherein before inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model and obtaining an electricity sales call probability corresponding to the time to be predicted, the method further comprises:

acquiring multiple groups of sample object information, multiple groups of sample telemarketing communication information and multiple groups of sample historical service information corresponding to multiple sample objects;

classifying the plurality of sample objects according to the plurality of groups of sample telemarketing communication information and the plurality of groups of sample historical service information to obtain a plurality of classification groups; respectively determining a plurality of first sample information, a plurality of first sample telemarketing information, a plurality of first sample historical service information and a plurality of first sample output targets corresponding to the plurality of classification groups;

sequentially training an initial electricity sales call prediction model by respectively utilizing the plurality of first sample information, the plurality of first sample electricity sales call information, the plurality of first sample historical service information and the plurality of first sample output targets to obtain a plurality of electricity sales call prediction models; the plurality of electricity sales call prediction models includes the target electricity sales call prediction model.

3. The method of claim 2, wherein the plurality of first sample output targets comprises a plurality of first series output targets, a plurality of second series output targets, and a plurality of third series output targets; the sequentially training an initial electricity sales call prediction model by using the plurality of first sample information, the plurality of first sample electricity sales call information, the plurality of first sample historical service information, and the plurality of first sample output targets to obtain the plurality of electricity sales call prediction models includes:

sequentially training an initial power distribution call prediction model by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information and the plurality of first serial output targets to obtain a plurality of first serial power distribution call prediction models;

sequentially training an initial power distribution call prediction model by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information and the plurality of second series output targets to obtain a plurality of second series power distribution call prediction models;

sequentially training an initial power distribution call prediction model by using the plurality of first sample information, the plurality of first sample power distribution call information, the plurality of first sample historical service information and the plurality of third series output targets to obtain a plurality of third series power distribution call prediction models;

using the plurality of first series electric sales call prediction models, the plurality of second series electric sales call prediction models, and the plurality of third series electric sales call prediction models as the plurality of electric sales call prediction models.

4. The method according to claim 1, wherein the inputting the object information, the historical electricity sales call information, the historical service information and the time to be predicted into a target electricity sales call prediction model to obtain an electricity sales call probability corresponding to the time to be predicted comprises:

determining group information corresponding to the target object according to the historical telemarketing communication information and the historical service information;

determining the target electricity sales call prediction model corresponding to the group information according to the corresponding relation between the configured classification groups and the electricity sales call prediction models;

and determining the power pin connection probability according to the object information, the historical power pin call information, the historical service information and the time to be predicted by using the target power pin call prediction model.

5. The method according to claim 4, wherein the determining the power pin connection probability according to the object information, the historical power pin call information, the historical service information and the time to be predicted by using the target power pin call prediction model comprises:

classifying the object information, the historical telemarketing communication information, the time to be predicted and the historical service information according to linear information characteristics, nonlinear information characteristics and time sequence information characteristics to obtain linear information, nonlinear information and time sequence information;

carrying out linear coding processing on the nonlinear information to obtain linear coding information;

performing time sequence derivation processing on the time sequence information to obtain derived time sequence information;

and inputting the linear information, the linear coding information and the derived time sequence information into the target power pin communication prediction model to obtain the power pin communication probability.

6. The method of claim 5, wherein the target electricity sales call prediction model comprises a first series electricity sales call prediction model, a second series electricity sales call prediction model, and a third series electricity sales call prediction model; the inputting the linear information, the linear coding information and the derived timing sequence information into the target power pin call prediction model to obtain the power pin call probability comprises:

inputting the linear information, the linear coding information and the derived time sequence information into the first serial telemarketing communication prediction model to obtain a first probability;

inputting the linear information, the linear coding information and the derived time sequence information into the second serial electric marketing communication prediction model to obtain a second probability;

inputting the linear information, the linear coding information and the derived time sequence information into the third serial electric marketing communication prediction model to obtain a third probability;

determining the pin connection probability according to the first probability, the second probability and the third probability.

7. The method of claim 5, wherein the performing timing derivation processing on the timing information to obtain derived timing information comprises:

converting the time sequence information according to a preset time conversion mode to obtain derived time sequence information;

and/or determining derived time sequence information of the time sequence information according to a time node where the time sequence information is located;

and/or counting the number of the time sequence information in a preset time period to obtain the derived time sequence information;

and/or determining a time interval between two adjacent time sequence information in the at least two time sequence information under the condition that the number of the time sequence information is at least two, so as to obtain the derived time sequence information.

8. The method according to claim 5, wherein said performing linear coding processing on the non-linear information to obtain linear coded information comprises:

and coding the nonlinear information by using an evidence weight coding mode to obtain the linear coding information.

9. An information processing apparatus characterized in that the apparatus comprises:

10. An information processing apparatus characterized in that the apparatus comprises:

a memory, a processor, and a communication bus, the memory in communication with the processor through the communication bus, the memory storing an information processing program executable by the processor, the information processing program when executed causing the processor to perform the method of any of claims 1 to 8.

11. A storage medium having stored thereon a computer program for application to an information processing apparatus, characterized in that the computer program, when executed by a processor, implements the method of any one of claims 1 to 8.