CN108228950A - A kind of information processing method and device - Google Patents

Abstract

The present invention provides a kind of information processing method and device, is related to data service technical field, to improve the accuracy of the prediction result provided to marketing personnel, so as to cut operating costs.The information processing method of the present invention, including：Obtain the information of current marketing personnel；According to the information of the current marketing personnel, the recommended models with the information matches of the current marketing personnel are obtained；Wherein, the recommended models are trained using the information of default marketing personnel and the information of the corresponding historic customer of the default marketing personnel；Obtain the information of pending client；It is that the current marketing personnel determine target customer from the pending client according to the information of the current marketing personnel, the information of the pending client and the recommended models.Present invention is mainly used in business marketing technology.

Description

An information processing method and device

technical field

The present invention relates to the technical field of data services, in particular to an information processing method and device.

Background technique

The process of modeling, mining and using the precision marketing model includes three basic processes: the historical data collection and feature selection phase for training the model; the model training phase; the prediction and recommendation phase using the model, etc.

For example, when a telecom operator uses voice customer service personnel to market specific terminals or traffic packages to customers, the customer service personnel see the recommended marketing customer group output by the system, which is the recommendation result of the precise marketing model in the background.

However, in the existing historical data collection stage for modeling training, modeling stage and subsequent model use stage, only focus on the historical data of customer groups, the characteristics of customer groups, and the marketing target customer groups recommended by different marketers It is also the result of the same recommendation model, which leads to inaccurate prediction results of the recommendation.

Contents of the invention

In view of this, the present invention provides an information processing method and device, which are used to improve the accuracy of forecast results provided to marketers, thereby reducing operating costs.

In order to solve the above technical problems, the present invention provides an information processing method, including:

Obtain information on current marketers;

According to the information of the current marketer, obtain a recommendation model that matches the information of the current marketer; wherein, the recommendation model uses the information of the preset marketer and the information of the historical customer corresponding to the preset marketer obtained by training;

Obtain information about customers to be processed;

According to the information of the current marketing personnel, the information of the customers to be processed and the recommendation model, a target customer is determined for the current marketing personnel from the customers to be processed.

Wherein, the step of obtaining a recommendation model matching the information of the current marketer according to the information of the current marketer includes:

Matching the information of the current marketer with the information of the marketer stored in each candidate recommendation model;

If the matching degree between the information of the current marketer and the information of the marketer in the first candidate recommendation model is greater than the preset threshold, then use the first candidate recommendation model as the one that matches the information of the current marketer Recommended model.

Wherein, the step of determining a target customer for the current marketer from the customers to be processed according to the information of the current marketer, the information of the customer to be processed and the recommendation model includes:

Using the information of the current marketer and the information of the customer to be processed as the input of the recommendation model, and running the recommendation model;

The target customer is determined according to the running result of the recommendation model.

Wherein, before the step of obtaining the information of current marketers, the method further includes:

The recommendation model is trained.

Wherein, the step of training the recommendation model includes:

obtain the information of said pre-set marketer;

Acquiring the information of the historical customer corresponding to the preset marketing personnel, wherein the information of the historical customer includes the identification of the historical customer, the customer characteristics of the historical customer, and the marketing results of the historical customer;

The information of the preset marketing personnel and the information of the historical customers are used as the input of the preset recommendation algorithm, and the recommendation model corresponding to the preset marketing personnel is trained.

Wherein, after the step of obtaining the information of the customer to be processed, the method further includes:

Obtain real-time emotional state information of the current marketer;

The step of determining a target customer for the current marketer from the customers to be processed according to the information of the current marketer, the information of the customer to be processed and the recommendation model is specifically:

Using the information of the current marketer, the information of the customer to be processed, and the real-time emotional state information as the input of the recommendation model, and running the recommendation model;

The target customer is determined according to the running result of the recommendation model.

In a second aspect, the present invention provides an information processing device, including:

The first information acquisition module is used to acquire the information of current marketers;

A recommendation model acquiring module, configured to acquire a recommendation model matching the information of the current marketer according to the information of the current marketer; wherein, the recommendation model utilizes the information of the preset marketer and the preset marketing It is obtained from information training of historical customers corresponding to personnel;

The second information acquisition module is used to acquire the information of the customer to be processed;

A determination module, configured to determine a target customer for the current marketer from the customers to be processed according to the information of the current marketer, the information of the customer to be processed and the recommendation model.

Wherein, the recommended model acquisition module includes:

The matching sub-module is used to match the information of the current marketer with the information of the marketer in each candidate recommendation model stored;

The acquisition sub-module is used to use the first candidate recommendation model as the first candidate recommendation model if the matching degree between the information of the current marketer and the information of the marketer in the first candidate recommendation model is greater than a preset threshold. A recommendation model for information matching for marketers.

Wherein, the determination module includes:

The first running sub-module is used to use the information of the current marketing personnel and the information of the customer to be processed as the input of the recommendation model, and run the recommendation model;

The first determination sub-module is configured to determine the target customer according to the operation result of the recommendation model.

Wherein, the device also includes:

A training module is used for training the recommendation model.

Wherein, the training module includes:

The first information acquisition sub-module is used to acquire the information of the preset marketing personnel;

The second information acquisition sub-module is used to obtain the information of the historical customers corresponding to the preset marketing personnel, wherein the information of the historical customers includes the identification of the historical customers, the customer characteristics of the historical customers, the historical customers marketing results;

The training sub-module is used to use the information of the preset marketing personnel and the information of the historical customers as the input of the preset recommendation algorithm, and train the recommendation model corresponding to the preset marketing personnel.

Wherein, the device also includes:

The third information acquisition module is used to acquire the real-time emotional state information of the current marketing personnel;

The determination module includes:

The second running sub-module is used to use the information of the current marketer, the information of the customer to be processed, and the real-time emotional state information as the input of the recommendation model, and run the recommendation model;

The second determination sub-module is used to determine the target customer according to the operation result of the recommendation model.

The beneficial effects of above-mentioned technical scheme of the present invention are as follows:

In the embodiment of the present invention, when recommending customers to the current marketer, the recommendation model corresponding to the current marketer, the information of the current marketer, and the information of the customer to be processed are used for recommendation, and the recommendation model uses the preset The information of the marketing personnel and the information of the historical customers corresponding to the preset marketing personnel are obtained through training. That is to say, in the solutions of the embodiments of the present invention, when recommending target customers to marketers, not only the information of the customers but also the information of the marketers are considered. The solution can provide marketers with more accurate forecast results, thereby reducing operating costs.

Description of drawings

FIG. 1 is a flowchart of an information processing method according to Embodiment 1 of the present invention;

FIG. 2 is a flowchart of an information processing method according to Embodiment 2 of the present invention;

FIG. 3 is a schematic diagram of a training recommendation model in Embodiment 2 of the present invention;

FIG. 4 is a schematic diagram of using a recommendation model to recommend target customers in Embodiment 2 of the present invention;

FIG. 5 is a schematic diagram of an information processing device according to Embodiment 3 of the present invention;

FIG. 6 is a structural diagram of an information processing device according to Embodiment 3 of the present invention.

Detailed ways

The specific implementation manner of the present invention will be further described in detail below with reference to the drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

In the prior art, when recommending target customers to marketers, the following processes are usually included:

First, collect historical data sets directly related to the user as a training set. These user data usually cover the basic attributes of the user (gender, age, online age, terminal model, etc.), historical consumption behavior (which voice packages, traffic packages, etc.) Which data value-added services have been ordered, whether they have participated in marketing activities) and historical bills and other feature data, and also include whether these users have accepted the marketing suggestions of customer service personnel for the products to be marketed, that is, the historical marketing results of each user; then, with The user's historical data and marketing results are used as the training set, and a specific machine learning or mining algorithm is selected for training to generate a marketing result prediction model for this marketing product; finally, for the new target user group, use the model as a voice customer service staff The recommendation basis for whether new users are worthy of marketing (whether marketing can be successful).

Since the above scheme only focuses on the historical data and characteristics of customer groups, the marketing target customer groups recommended to different marketers are also the results of the same recommendation model, which leads to inaccurate prediction results of recommendations. For this reason, the solutions of the embodiments of the present invention take the information of the marketing personnel and the information of the customers into consideration at the same time, so as to accurately recommend customers for the marketing personnel and reduce operating costs.

Embodiment one

As shown in Figure 1, the information processing method in Embodiment 1 of the present invention includes:

Step 101. Acquire information about current marketers.

Wherein, the information of the current marketer includes but is not limited to: gender, audio frequency, pitch, average speech rate, language characteristics, favorite color (personality characteristics), cultural background (educational background, cultural preference, etc.), emotional state, etc.

Step 102, according to the information of the current marketer, obtain a recommendation model matching the information of the current marketer.

Wherein, the recommendation model is trained by using the information of the preset marketing personnel and the information of the historical customers corresponding to the preset marketing personnel.

In practical applications, different recommendation models can be trained for different marketers. When training the recommendation model, the preset information of some marketers (may also include the information of the current marketer), the historical customers corresponding to the marketer, that is, the information of the customers who have been marketed by the marketer, are all taken into consideration. Through recommendation algorithms, such as decision tree, Bayesian, etc., train the recommendation model corresponding to the marketer.

Then, in this step, the information of the current marketer is matched with the stored information of the marketer in each candidate recommendation model. If the matching degree between the information of the current marketer and the information of the marketer in the first candidate recommendation model is greater than the preset threshold, then use the first candidate recommendation model as the one that matches the information of the current marketer Recommended model. Wherein, the preset threshold can be set arbitrarily.

For example, after matching, the current marketer A's information matches the marketer's information in the recommendation model B. So here, the recommendation model B can be used as the recommendation model corresponding to the current marketer A.

Step 103, acquiring the information of the customer to be processed.

Wherein, the information of the customer to be processed refers to the information of the customer to whom the marketing service is to be launched, including but not limited to: gender, age, online age, terminal model, and the like.

Step 104, according to the information of the current marketing personnel, the information of the customers to be processed and the recommendation model, determine target customers for the current marketing personnel from the customers to be processed.

Specifically, in this step, use the information of the current marketer and the information of the customer to be processed as the input of the recommendation model, run the recommendation model, and then determine the Target customers. That is, the output result of the recommendation model can be used as the target customer.

It can be seen that, in the solution of the embodiment of the present invention, when recommending target customers to marketers, not only the information of customers but also the information of marketers is considered. Therefore, compared with the prior art, using the present invention The solutions of the embodiments can provide marketing personnel with more accurate prediction results, thereby reducing operating costs.

Embodiment two

As shown in Figure 2, the information processing method in Embodiment 2 of the present invention includes:

Step 201, training a recommendation model.

In this step, the information of the preset marketing personnel is obtained, and the information of the historical customers corresponding to the preset marketing personnel is obtained, wherein the information of the historical customers includes the identification of the historical customers, the customer characteristics of the historical customers, Marketing results for said historical customer. Then, the information of the preset marketing personnel and the information of the historical customers are used as input of a preset recommendation algorithm to train a recommendation model corresponding to the preset marketing personnel.

Wherein, the preset marketing personnel may be any one or more marketing personnel, and the preset recommendation algorithm may be decision tree, Bayesian and other algorithms.

As shown in Figure 3, it is assumed that a recommendation model is trained for marketer A and marketer B here.

Among them, the information of marketer A includes: gender, audio frequency, pitch, average speech rate, language characteristics, favorite color (personality characteristics), cultural background (educational background, cultural preference, etc.), emotional state, etc. The information of marketer B includes: gender, audio frequency, pitch, average speech rate, language characteristics, favorite color (personality characteristics), cultural background (educational background, cultural preference, etc.), emotional state, etc.

For marketer A, according to his historical marketing experience, the corresponding historical customer information can be obtained as a training set for the marketing recommendation model for this marketer. Similarly, for marketer B, according to his historical marketing experience, the corresponding historical customer information can be obtained as a training set for the marketing recommendation model for this marketer.

The selected recommendation algorithm, such as decision tree, Bayesian, etc., takes the information of each marketer and the corresponding historical customer information as input to train the marketing recommendation prediction model for the marketer. As shown in Figure 3, two different recommendation models are generated for marketer A and marketer B respectively. The decision-making attributes of the recommendation model not only include the characteristic attributes of historical customers (gender, age, etc.), but also incorporate the characteristic attributes of marketers (audio, emotional state, etc.). At the same time, the characteristic attributes of marketers included in the prediction models generated for different marketers are different.

Different from the existing method, which only starts from the perspective of the marketed customers, and the collected training set only includes the characteristic data of the historical customer groups (P1...Pn), the solution of the present invention considers the characteristics of the historical customer groups and combines different marketing Characteristic data of persons (A, B, ...) are also collected as equally important training data. In this way, for each customer that has been marketed, a training sample record that needs to be collected is changed from <historical customer ID: historical characteristics: historical marketing results> to <marketer ID: historical characteristics of marketer: historical customer ID: customer Historical Features: Historical Marketing Results >.

Therefore, in the embodiment of the present invention, the training set used for modeling combines the characteristic data of the marketer, which is more in line with the requirement that the marketing interaction process is the process and behavior of both the marketer and the customer. The combination of the characteristics of the marketer and the customer datasets can lead to more accurate recommendation models. Therefore, the prediction made by this model is more targeted, more in line with the actual effect, and the prediction accuracy will be higher.

Step 202, acquiring the information of the current marketer.

Assuming that the current marketer C, the obtained information of marketer C includes: gender, audio, pitch, average speech rate, language characteristics, favorite color (personality traits), cultural background (educational background, cultural preference, etc.), emotional state, etc.

Step 203, according to the information of the current marketer, obtain a recommendation model matching the information of the current marketer.

Suppose, after matching, the recommendation model corresponding to marketer A can be used as the recommendation model of marketer C.

Step 204, acquiring the information of the customer to be processed.

It is assumed that the customers to be processed here include (P1...Pn), and the information respectively includes: gender, age, online age, terminal model and so on.

Step 205, acquiring the real-time emotional state information of the current marketing personnel.

In this step, the real-time voice segment of salesman C is collected and uploaded to the emotional state recognition device. Then the device uses an intelligent speech emotion analysis algorithm to analyze the real-time emotional state of marketing person C, and returns the analysis result. Wherein, the real-time emotional state information of the marketer C may be happy, depressed, normal, excited and so on. The real-time emotional state information is input into the recommendation model together with the real-time emotional state characteristic attribute of marketer C and the information of the customer to be processed for marketing result prediction. In this way, different recommendation results can be dynamically generated according to the real-time status of the marketers. That is, for the same batch of customer groups to be processed, different prediction and recommendation results can be generated according to the different emotional states of marketers at different times.

Step 206 , taking the information of the current marketer, the information of the customer to be processed, and the real-time emotional state information as the input of the recommendation model, and running the recommendation model.

Step 207. Determine the target customer according to the running result of the recommendation model.

As shown in Figure 4, the information of marketer C, the real-time emotional state information of marketer C, and the information of customers to be processed are used as the input of the recommendation model, and the recommendation model is run to determine marketer C from the customers to be processed target customers.

It can be seen that, in the solution of the embodiment of the present invention, when recommending target customers to marketers, not only the information of customers but also the information of marketers is considered. Therefore, compared with the prior art, using the present invention The solutions of the embodiments can provide marketing personnel with more accurate prediction results, thereby reducing operating costs and increasing the success rate of marketing.

Embodiment three

As shown in FIG. 5, the information processing device according to Embodiment 3 of the present invention includes:

The first information obtaining module 501 is used to obtain the information of the current marketing personnel; the recommendation model obtaining module 502 is used to obtain the recommendation model matching the information of the current marketing personnel according to the information of the current marketing personnel; wherein, the The recommendation model is trained by using the information of the preset marketing personnel and the information of the historical customers corresponding to the preset marketing personnel; the second information acquisition module 503 is used to obtain the information of the customers to be processed; the determination module 504 is used for According to the information of the current marketing personnel, the information of the customers to be processed and the recommendation model, a target customer is determined for the current marketing personnel from the customers to be processed.

Wherein, the recommended model acquisition module 502 includes:

The matching submodule is used to match the information of the current marketing personnel with the information of the marketing personnel in each candidate recommendation model stored; the acquisition submodule is used to match the information of the current marketing personnel with the information of the first backup If the matching degree of the marketer's information in the selected recommendation model is greater than a preset threshold, the first candidate recommendation model is used as the recommendation model that matches the current marketer's information.

Wherein, the determination module 504 includes: a first running sub-module, configured to use the information of the current marketer and the information of the customer to be processed as the input of the recommendation model, and run the recommendation model; the first determination The sub-module is used to determine the target customer according to the operation result of the recommendation model.

In order to improve marketing efficiency, as shown in Figure 6, the device also includes:

A training module 505, configured to train the recommendation model.

Wherein, the training module 505 includes: a first information acquisition submodule, used to acquire the information of the preset marketing personnel; a second information acquisition submodule, used to acquire the information of the historical customers corresponding to the preset marketing personnel , wherein the information of the historical customer includes the identification of the historical customer, the customer characteristics of the historical customer, and the marketing results of the historical customer; the training submodule is used to combine the information of the preset marketing personnel with the The historical customer information is used as the input of the preset recommendation algorithm to train the recommendation model corresponding to the preset marketing personnel.

As shown in Figure 6 again, in order to further improve the accuracy of recommendation, the device also includes:

The third information acquiring module 506 is configured to acquire the real-time emotional state information of the current marketing personnel.

At this time, the determination module 504 includes: a second running submodule, configured to use the information of the current marketer, the information of the customer to be processed, and the real-time emotional state information as the input of the recommendation model to run The recommendation model; a second determining submodule, configured to determine the target customer according to the operation result of the recommendation model.

For the working principle of the device of the present invention, reference may be made to the description of the foregoing method embodiments.

It can be seen that, in the solution of the embodiment of the present invention, when recommending target customers to marketers, not only the information of customers but also the information of marketers is considered. Therefore, compared with the prior art, using the present invention The solutions of the embodiments can provide marketing personnel with more accurate prediction results, thereby reducing operating costs and increasing the success rate of marketing.

In the several embodiments provided in this application, it should be understood that the disclosed methods and devices may be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may be physically included separately, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of hardware plus software functional units.

The above-mentioned integrated units implemented in the form of software functional units may be stored in a computer-readable storage medium. The above-mentioned software functional units are stored in a storage medium, and include several instructions to enable a computer device (which may be a personal computer, server, or network device, etc.) to execute some steps of the sending and receiving methods described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM for short), random access memory (Random Access Memory, RAM for short), magnetic disk or optical disk, etc., which can store program codes. medium.

The above description is a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications can also be made. It should be regarded as the protection scope of the present invention.

Claims (12)

1. An information processing method, characterized in that, comprising: Obtain information on current marketers; According to the information of the current marketer, obtain a recommendation model that matches the information of the current marketer; wherein, the recommendation model uses the information of the preset marketer and the information of the historical customer corresponding to the preset marketer obtained by training; Obtain information about customers to be processed; According to the information of the current marketing personnel, the information of the customers to be processed and the recommendation model, a target customer is determined for the current marketing personnel from the customers to be processed. 2. The method according to claim 1, wherein the step of obtaining a recommendation model matching the information of the current marketer according to the information of the current marketer comprises: Matching the information of the current marketer with the information of the marketer stored in each candidate recommendation model; If the matching degree between the information of the current marketer and the information of the marketer in the first candidate recommendation model is greater than the preset threshold, then use the first candidate recommendation model as the one that matches the information of the current marketer Recommended model. 3. The method according to claim 1, characterized in that, according to the information of the current marketers, the information of the customers to be processed and the recommendation model, from the customers to be processed for the current Steps for marketers to identify target customers, including: Using the information of the current marketer and the information of the customer to be processed as the input of the recommendation model, and running the recommendation model; The target customer is determined according to the running result of the recommendation model. 4. The method according to claim 1, characterized in that, before the step of obtaining the information of current marketers, the method further comprises: The recommendation model is trained. 5. The method according to claim 4, wherein the step of training the recommendation model comprises: obtain the information of said pre-set marketer; Acquiring the information of the historical customer corresponding to the preset marketing personnel, wherein the information of the historical customer includes the identification of the historical customer, the customer characteristics of the historical customer, and the marketing results of the historical customer; The information of the preset marketing personnel and the information of the historical customers are used as the input of the preset recommendation algorithm, and the recommendation model corresponding to the preset marketing personnel is trained. 6. The method according to claim 1, characterized in that, after the step of obtaining the information of the customer to be processed, the method further comprises: Obtain real-time emotional state information of the current marketer; The step of determining a target customer for the current marketer from the customers to be processed according to the information of the current marketer, the information of the customer to be processed and the recommendation model is specifically: Using the information of the current marketer, the information of the customer to be processed, and the real-time emotional state information as the input of the recommendation model, and running the recommendation model; The target customer is determined according to the running result of the recommendation model. 7. An information processing device, characterized in that it comprises: The first information acquisition module is used to acquire the information of current marketers; A recommendation model acquiring module, configured to acquire a recommendation model matching the information of the current marketer according to the information of the current marketer; wherein, the recommendation model utilizes the information of the preset marketer and the preset marketing It is obtained from information training of historical customers corresponding to personnel; The second information acquisition module is used to acquire the information of the customer to be processed; A determination module, configured to determine a target customer for the current marketer from the customers to be processed according to the information of the current marketer, the information of the customer to be processed and the recommendation model. 8. The device according to claim 7, wherein the recommendation model acquisition module comprises: The matching sub-module is used to match the information of the current marketer with the information of the marketer in each candidate recommendation model stored; The acquisition sub-module is used to use the first candidate recommendation model as the first candidate recommendation model if the matching degree between the information of the current marketer and the information of the marketer in the first candidate recommendation model is greater than a preset threshold. A recommendation model for information matching for marketers. 9. The device according to claim 7, wherein the determining module comprises: The first running sub-module is used to use the information of the current marketing personnel and the information of the customer to be processed as the input of the recommendation model, and run the recommendation model; The first determination sub-module is configured to determine the target customer according to the operation result of the recommendation model. 10. The device according to claim 7, further comprising: A training module is used for training the recommendation model. 11. The device according to claim 10, wherein the training module comprises: The first information acquisition sub-module is used to acquire the information of the preset marketing personnel; The second information acquisition sub-module is used to obtain the information of the historical customers corresponding to the preset marketing personnel, wherein the information of the historical customers includes the identification of the historical customers, the customer characteristics of the historical customers, the historical customers marketing results; The training sub-module is used to use the information of the preset marketing personnel and the information of the historical customers as the input of the preset recommendation algorithm, and train the recommendation model corresponding to the preset marketing personnel. 12. The device according to claim 7, further comprising: The third information acquisition module is used to acquire the real-time emotional state information of the current marketing personnel; The determination module includes: The second running sub-module is used to use the information of the current marketer, the information of the customer to be processed, and the real-time emotional state information as the input of the recommendation model, and run the recommendation model; The second determination sub-module is used to determine the target customer according to the operation result of the recommendation model.