WO2019037391A1 - 客户购买意向的预测方法、装置、电子设备及介质 - Google Patents

客户购买意向的预测方法、装置、电子设备及介质 Download PDF

Info

Publication number
WO2019037391A1
WO2019037391A1 PCT/CN2018/074872 CN2018074872W WO2019037391A1 WO 2019037391 A1 WO2019037391 A1 WO 2019037391A1 CN 2018074872 W CN2018074872 W CN 2018074872W WO 2019037391 A1 WO2019037391 A1 WO 2019037391A1
Authority
WO
WIPO (PCT)
Prior art keywords
customer
purchase
task
follow
sales
Prior art date
Application number
PCT/CN2018/074872
Other languages
English (en)
French (fr)
Inventor
李芳�
王建明
肖京
Original Assignee
平安科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安科技（深圳）有限公司 filed Critical 平安科技（深圳）有限公司
Priority to US16/099,425 priority Critical patent/US20210224832A1/en
Priority to SG11201809952XA priority patent/SG11201809952XA/en
Publication of WO2019037391A1 publication Critical patent/WO2019037391A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0202—Market predictions or forecasting for commercial activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G06Q30/0255—Targeted advertisements based on user history
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/20—Ensemble learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/04—Inference or reasoning models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06311—Scheduling, planning or task assignment for a person or group
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/26—Speech to text systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M3/00—Automatic or semi-automatic exchanges
    • H04M3/42—Systems providing special services or facilities to subscribers
    • H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
    • H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
    • H04M3/523—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing with call distribution or queueing
    • H04M3/5231—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing with call distribution or queueing with call back arrangements

Definitions

  • the present application belongs to the field of information processing, and in particular, to a method, a device, an electronic device and a medium for predicting a customer's purchase intention.
  • product marketing methods include telemarketing, email marketing, and SMS marketing.
  • Telemarketing is the use of the phone to achieve a planned, organized, and efficient way to expand the customer base.
  • all major enterprises have begun to implement personalized and precise marketing. Specifically, through in-depth analysis of the collected personal characteristics data of each user, the different consumption characteristics of different customers are determined, so that the customer is confirmed as a potential customer when the sales product and the customer's consumption characteristics are more consistent.
  • the agent In order to enable the agent to conduct telemarketing to the potential customer, it can ensure that after each telemarketing, there is a greater probability that the customer will be converted into the actual customer who purchases the product, thereby improving marketing efficiency.
  • the prior art only directly evaluates whether the customer is a potential customer based on the customer's personal characteristic data, and the consideration factor is single, and the method cannot quantify the customer's product purchase intention, and thus it is difficult to find a customer who truly has the intention to purchase the electric product. .
  • the embodiment of the present application provides a method, a device, an electronic device, and a medium for predicting a customer's purchase intention, so as to solve the problem that the prior art considers a single factor and cannot quantify the customer's product purchase intention when determining a potential customer in the prior art. .
  • a first aspect of the embodiments of the present application provides a method for predicting a customer's purchase intention, including:
  • the customer whose actual purchase propensity is greater than a preset threshold is determined as a potential customer, so that the sales agent makes a telephone call back to the potential customer and promotes the electric product.
  • a second aspect of the embodiments of the present application provides a prediction apparatus for a customer purchase intention, including:
  • a first obtaining unit configured to acquire personal characteristic data of the customer
  • a first output unit configured to input the personal characteristic data into a pre-established random forest model related to the electric product, to output an objective purchase propensity value of the customer to the electric product;
  • a second obtaining unit configured to acquire a subjective purchase propensity value of the customer for the electric product according to the emotional tendency of the customer in the historical electric sales process
  • a weighting unit configured to perform weighting processing on the objective purchase propensity value and the subjective purchase propensity value, and output the weighted result as an actual purchase propensity degree of the customer;
  • the determining unit is configured to determine the customer whose actual purchase propensity is greater than a preset threshold as a potential customer, so that the sales agent makes a telephone call back to the potential customer and promotes the sales product.
  • a third aspect of the embodiments of the present application provides an electronic device including a memory and a processor, wherein the memory stores computer readable instructions executable on the processor, the processor executing the computer The step of implementing the prediction method of the customer purchase intention as described in the first aspect when the instruction is read.
  • a fourth aspect of the embodiments of the present application provides a computer readable storage medium storing computer readable instructions, the computer readable instructions being executed by a processor to implement the first aspect as described in the first aspect The steps of the customer's method of predicting the purchase intention.
  • the purchasing tendency value of the customer on the electronic product can be calculated at an objective level; and the emotional tendency of the customer in the process of obtaining historical electric sales can be obtained.
  • the potential customers identified in the end are potential customers who have combined the multi-faceted considerations, thus improving the forecast accuracy of potential customers.
  • by allowing the sales agents to call back and sell the products to potential customers Avoid neglecting historical sales customers, which further reduces customer churn.
  • FIG. 1 is a flowchart of implementing a method for predicting a customer purchase intention according to an embodiment of the present application
  • FIG. 2 is a specific implementation flowchart of a method for predicting customer purchase intention S103 provided by an embodiment of the present application
  • FIG. 3 is a specific implementation flowchart of a method for predicting customer purchase intention S104 provided by an embodiment of the present application
  • FIG. 4 is a flowchart of implementing a method for predicting a customer purchase intention according to another embodiment of the present application
  • FIG. 5 is a flowchart of implementing a method for predicting a customer purchase intention according to another embodiment of the present application.
  • FIG. 6 is a structural block diagram of a device for predicting a customer's purchase intention according to an embodiment of the present application
  • FIG. 7 is a structural block diagram of a device for predicting a customer purchase intention according to an embodiment of the present application.
  • FIG. 8 is a structural block diagram of a device for predicting a customer purchase intention according to another embodiment of the present application.
  • FIG. 9 is a structural block diagram of a device for predicting a customer purchase intention according to another embodiment of the present application.
  • FIG. 10 is a schematic diagram of an electronic device according to an embodiment of the present application.
  • FIG. 1 shows an implementation flow of a method for predicting a customer purchase intention provided by an embodiment of the present application, and the method flow includes steps S101 to S105.
  • the specific implementation principles of each step are as follows:
  • S101 Acquire personal characteristic data of the customer.
  • a customer is a historical sales customer who has the possibility to purchase a product for sale. It is used to dig out customers who have a high degree of product purchase and need to conduct telemarketing again. In the case that the sales products and the sales service meet the needs of the customers, the customers can be transformed into actual customers who purchase the products. Among them, the products sold by the sales agents to the customers by means of telephone communication, including but not limited to insurance products and credit products and other financial products.
  • the customers that have been promoted by each of the sales agents and the sales information related to the customers are recorded in the database.
  • the sales promotion information related to the customer includes the personal characteristic data of the customer. Therefore, for one of the customers, the customer's personal characteristic data can be read from the database.
  • Personal characteristics data includes, but is not limited to, age, income, hobbies, education, historical consumption of financial products, and life insurance premiums.
  • S102 Input the personal characteristic data into a pre-established random forest model related to the electric product, to output an objective purchase propensity value of the customer to the electric product.
  • a pre-trained random forest model is obtained.
  • the random forest model includes a plurality of decision trees, each of which is used for classification selection based on input parameters. After the classification and selection results of each decision tree are statistically summarized, the final output parameters of the random forest model are obtained.
  • the input parameter is the personal characteristic data of the current customer.
  • the output parameter is the objective purchase propensity value of the customer.
  • the size of the objective purchase propensity value characterizes the purchase possibility of the customer's product for the objective product, and also the degree of matching between the customer's personal characteristic data and the characteristics of the electric product.
  • a plurality of training sample data are input to a pre-built random forest model.
  • Each training sample data includes a personal sales profile of the customer and a customer type.
  • the historical sales customers in each training sample data are customers that the sales agents sell for the same electric products, and the random forest model obtained by training is also related to the electric products.
  • the above customer types are actual customers or non-real customers. That is, whether the customer that the sales agent once sold has finally purchased the electric product. If so, the historical sales customer is the actual customer, and if not, the historical sales customer is a non-real customer.
  • the model parameters in the random forest model are adjusted based on the obtained training sample data. Specifically, in the received N training sample data, the random extraction with the reversal is repeatedly performed to obtain the extracted M (0 ⁇ M ⁇ N, and M is an integer) training sample data as a new training sample set. Based on the new training sample set, K (K is an integer greater than 1) decision trees for classification are generated.
  • the decision tree includes a binary tree and a non-binary tree.
  • S103 Obtain a subjective purchase propensity value of the customer for the electric product according to the emotional tendency of the customer in the historical electric sales process.
  • the customer's emotional tendency in the historical sales process can be obtained by the following methods: the sales agent can judge the customer's emotional tendency in the process of each sales, that is, in the process of making contact with the customer every time. Which type is included, and the judgment result is recorded as the sales information related to the customer, and then stored in the database. Therefore, when predicting the purchase intention of the customer's product, the latest data record corresponding to the customer can be read from the database, and the sentiment tendency of the customer stored therein can be read.
  • the foregoing S103 specifically includes:
  • S1031 Audio recording of the historical electric sales process to obtain audio data.
  • the sales agent communicates with the customer through the smart terminal installed with the communication software to perform product promotion.
  • the smart terminal detects that the phone number dialed by the current sales agent is turned on, the audio recording function carried by the communication software is triggered, and the audio recording is started, so that the time is t 1 .
  • the recording of the audio is stopped, so that the moment is t 2 .
  • the time t 1 to time t 2 between the saved recorded audio data obtained as an audio file, audio file and the time t 1 to time t 2 between the client agent pin electrical contact.
  • S1033 Perform recognition processing on the text data based on a preset positive emotion dictionary and a negative emotion dictionary to determine an emotional tendency corresponding to the text data.
  • the positive sentiment dictionary contains pre-collected words for expressing positive emotions, such as “very good”, “satisfied” and "good”.
  • the negative sentiment dictionary contains pre-collected words for expressing negative emotions, such as "very bad”, “harassment”, and "very annoying”.
  • each participle exists in a positive emotion dictionary or a negative emotion dictionary. If there is a participle in the text data that exists in the positive sentiment dictionary, the cumulative value used to express the degree of sentiment orientation is incremented by one; if a participle exists in the negative sentiment dictionary, the cumulative value used to express the degree of sentiment tendency is reduced. One. According to the correspondence between the cumulative value and the emotional tendency, the emotional tendency corresponding to the finally obtained cumulative value is determined.
  • the sentiment classification model may also be trained based on a plurality of text training data marked with an emotional tendency. At this time, if the word segmentation does not appear in the positive sentiment dictionary and the negative sentiment dictionary in the text data obtained by the voice data conversion, the text data is input into the pre-trained sentiment classification model to output the emotional tendency corresponding to the text data. .
  • S1034 Obtain a subjective purchase propensity value that matches the sentiment tendency.
  • different sentimental tendencies correspond to different subjective purchasing propensity values. Based on the correspondence between the subjective purchase tendency value and the sentiment tendency, the subjective purchase tendency value corresponding to the current time emotional tendency is determined.
  • the corresponding subjective purchasing tendency value is 100%; if the emotional tendency is positive, the corresponding subjective purchasing tendency value is 90%; if the emotional tendency is aversive, the corresponding subjective purchasing tendency value It is 0%.
  • S104 Perform weighting processing on the objective purchase propensity value and the subjective purchase propensity value, and output the weighted result as the actual purchase propensity degree of the customer.
  • the weight ratio corresponding to the subjective purchase propensity value is greater than the weight ratio corresponding to the objective purchase propensity value.
  • the weight ratio corresponding to the objective purchase propensity value is preferably 35%, and the weight ratio corresponding to the subjective purchase propensity value is preferably 65%, and the actual purchase propensity C of the customer can be calculated by the following formula. inferred:
  • the above A is an objective purchase tendency value
  • the above B is a subjective purchase tendency value
  • the subjective purchase tendency value of the customer can more accurately reflect the subjective attitude tendency of the customer for the electric product, and whether the customer performs the purchase operation, it is usually directly related to the subjective attitude tendency.
  • the weight ratio corresponding to the subjective purchase propensity value is preferably 65% and the weight ratio of the objective purchase propensity value is 35%.
  • the calculated actual purchase propensity of the customer will have a high reference value, which can further improve the recognition accuracy of the potential customer.
  • S105 Determine the customer whose actual purchase tendency is greater than a preset threshold as a potential customer, so that the sales agent makes a telephone call back to the potential customer and promotes the sales product.
  • the sales agent makes a telemarketing for the customer, the customer is difficult to convert into an actual customer. Therefore, in order to improve the marketing efficiency of the sales agent, only A customer whose actual purchase propensity is greater than a preset threshold is determined to be a potential customer. By recommending the identified potential customers to the sales agents, the sales agents can make a telephone call back for the limited time to purchase the historical sales customers with higher probability of product purchase, so as to maximize the promotion. Customer conversion rate.
  • the purchasing tendency value of the customer on the electronic product can be calculated at an objective level; and the emotional tendency of the customer in the process of obtaining historical electric sales can be obtained.
  • the potential customers identified in the end are potential customers who have combined the multi-faceted considerations, thus improving the forecast accuracy of potential customers.
  • by allowing the sales agents to call back and sell the products to potential customers Avoid neglecting historical sales customers, which further reduces customer churn.
  • the weighting manner of the actual purchase tendency of the customer is further defined.
  • the above S104 includes:
  • S1041 Acquire a satisfaction score of the customer feedback at the end of the historical sales process.
  • customers can receive satisfaction rating reminders.
  • the satisfaction rating reminder information is used to prompt the customer to score the sales level of the current sales service or the sales agent.
  • the customer can receive the satisfaction score of the customer feedback by pressing the rating value in the dialing keypad of the communication terminal or replying to the rating value by means of the short message. The higher the score, the higher the customer's satisfaction.
  • the satisfaction scores returned by each customer at the end of the sales process are also stored in the database.
  • the satisfaction score of the customer's most recent feedback is read from the database before calculating the actual purchase propensity of the customer.
  • S1042 Perform weighting processing on the satisfaction score, the objective purchase propensity value, and the subjective purchase propensity value, and output the weighted result as the actual purchase propensity degree of the customer.
  • the satisfaction score shows the satisfaction degree of the customer to the electric product or the electric service, it can accurately reflect the true subjective emotional tendency of the customer to a certain extent, and therefore, based on the satisfaction score and objective purchase
  • the three factors of the propensity value and the subjective purchase propensity value jointly calculate the actual purchase propensity of the customer, which can reduce the prediction error of the actual purchase propensity due to the theoretically analyzed value of the subjective purchase propensity value, thus improving the prediction error.
  • the accuracy of the customer's actual purchase propensity since the satisfaction score shows the satisfaction degree of the customer to the electric product or the electric service, it can accurately reflect the true subjective emotional tendency of the customer to a certain extent, and therefore, based on the satisfaction score and objective purchase
  • the three factors of the propensity value and the subjective purchase propensity value jointly calculate the actual purchase propensity of the customer, which can reduce the prediction error of the actual purchase propensity due to the theoretically analyzed value of the subjective purchase propensity value, thus improving the prediction error.
  • the method further includes:
  • S106 In the sales task management interface, in accordance with the order of the actual purchase tendency of each of the customers, sequentially displaying the sales follow-up tasks based on each of the customers.
  • a power-up follow-up task based on the customer is generated.
  • any of the sales agents can log in to the sales task management system through their own sales account to view the sales follow-up tasks displayed in the sales task management interface.
  • the sales follow-up tasks corresponding to each customer are sorted, so that the electricity corresponding to the customer with a large purchase tendency is relatively large.
  • the sales follow-up task is arranged before the sales follow-up task corresponding to the customer with a small purchase tendency.
  • S107 When receiving the scheduling instruction of the power-up follow-up task, changing an implementation state of the power-up follow-up task from a first state to a second state to shield the other salesperson from the salesperson Follow up on the task.
  • the sales agent can click to select a sales follow-up task that needs to be followed up.
  • the dispatching instruction based on the sales follow-up task is received.
  • the sales follow-up task is bound to the sales agent account that issues the dispatch instruction, and the customer information related to the sales follow-up task is sent to the sales account account bound with the sales follow-up task.
  • the customer information related to the sales follow-up task includes the customer's personal characteristic data, contact information, actual purchase tendency and satisfaction score, etc., thereby enabling the sales agent who obtained the customer information to return to the customer in time. And carry out electric sales operations.
  • the implementation status of the power-up follow-up task is used to indicate the real-time processing progress of the power-up follow-up task
  • the implementation state includes the first state and the second state.
  • the first state is an unprocessed state and the second state is an allocated state.
  • the implementation status of the sales follow-up task can be represented by the color presented by the sales follow-up task in the sales task management interface. For example, the power-up follow-up task is marked in red to indicate that its implementation state is the first state; the power-up follow-up task is marked in yellow to indicate that its implementation state is the second state.
  • the implementation status of the sales follow-up task is changed to the second state. Since the electric power follow-up task in the second state cannot be clicked again, the dispatching instruction based on the electric power follow-up task is not received again, and the shielding of other electric sales agents is realized.
  • the sales follow-up task based on each customer is sequentially displayed on the sales task management interface, so that the sales agent can know in real time which customer's real-time purchase tendency according to the order of the sales follow-up tasks. The highest degree and which sales follow-up task can achieve better return visit results.
  • the implementation state of the power-up follow-up task is changed from the first state to the second state, so that other sales agents cannot repeatedly schedule the same power-selling task to avoid A number of sales agents followed up with the same customer, which improved the follow-up efficiency of the sales task and the efficiency of the sales staff, thereby avoiding excessive telephone call harassment for the customer.
  • the foregoing method for predicting the purchase intention of the customer further includes:
  • S108 Acquire, according to the creation time point of the power-up follow-up task, the created duration of the power-up follow-up task at each moment.
  • the sales task management interface when the actual purchase propensity degree of the first calculated customer is greater than a preset threshold, in the sales task management interface, a power sales follow-up task based on the customer is generated and displayed, and the sales follow-up task is performed.
  • the generation time is the time point of creation of the sales follow-up task.
  • the difference between the system real time at that time and the creation time point of the sales follow-up task is determined as the created duration of the sales follow-up task.
  • S109 Calculate a purchase tendency degree decrease value corresponding to the created time length, and the purchase tendency degree decrease value is proportional to the created time length.
  • S110 Output, by the difference between the actual purchase tendency degree corresponding to the electric power follow-up task and the purchase tendency degree decrease value, an actual purchase tendency degree corresponding to the current sales follow-up task.
  • the purchase tendency degree decrease value ⁇ s corresponding to the created duration is output via the preset purchase tendency decrease value calculation formula.
  • the actual purchase propensity corresponding to the real-time correspondence of the sales follow-up task is adjusted to S- ⁇ s.
  • S represents the actual purchase tendency corresponding to the sales follow-up task at the time of creation.
  • the above formula for calculating the purchase tendency reduction value is a proportional function, and the formula may be, for example,
  • a ⁇ x.
  • a is the preset constant coefficient
  • x is the created duration of the sales follow-up task
  • y is the decrease in the purchasing tendency corresponding to the created duration x. It can be seen that the greater the created duration x of the sales follow-up task, the greater the decrease in the purchasing tendency.
  • S111 Adjust an arrangement order of the sales follow-up tasks in the sales task management interface based on an actual purchase tendency corresponding to the current sales follow-up task at the current time.
  • the order of the various sales follow-up tasks in the sales task management interface indicates the actual purchase tendency corresponding to the sales follow-up task. Therefore, if the electric power follow-up task is not scheduled, according to the above S108 to S110, the actual purchase tendency corresponding to the electric power follow-up task will become smaller and smaller, so the electric power follow-up task is in the sales task management interface. The order in which they are arranged will also be adjusted in real time. When the customer's actual purchase propensity is less than the preset threshold, the sales follow-up task is deleted.
  • the identified potential customer is a customer who is more likely to purchase the electric product selected from the historical electric sales customer, if the potential customer is not returned to the customer for a long time, the customer The purchase intention of the sales products will also become smaller and smaller as time goes by.
  • the agents can understand which customers are losing more and more based on the subsequent sales follow-up tasks. This has played a role in urging follow-up.
  • FIG. 6 is a structural block diagram of the prediction device of the customer purchase intention provided by the embodiment of the present application. For the convenience of explanation, only the parts related to the present embodiment are shown.
  • the apparatus includes:
  • the first obtaining unit 601 is configured to acquire personal characteristic data of the client.
  • the first output unit 602 is configured to input the personal characteristic data into a pre-established random forest model related to the electric product to output an objective purchase propensity value of the customer for the electric product.
  • the second obtaining unit 603 is configured to obtain a subjective purchase propensity value of the customer for the electric product according to the emotional tendency of the customer in the historical electric sales process.
  • the weighting unit 604 is configured to perform weighting processing on the objective purchase propensity value and the subjective purchase propensity value, and output the weighted result as the actual purchase propensity degree of the customer.
  • the determining unit 605 is configured to determine the customer whose actual purchase propensity is greater than a preset threshold as a potential customer, so that the sales agent makes a telephone call back to the potential customer and promotes the sales product.
  • the second obtaining unit 603 includes:
  • Recording subunit for audio recording of the historical lighting process to obtain audio data.
  • a conversion subunit for converting the audio data into text data.
  • the identification subunit is configured to perform recognition processing on the text data based on the preset positive emotion dictionary and the negative emotion dictionary to determine an emotional tendency corresponding to the text data.
  • the first obtaining subunit is configured to obtain a subjective purchasing propensity value that matches the sentiment tendency.
  • the weighting unit 604 includes:
  • a second obtaining subunit configured to obtain a satisfaction score of the customer feedback at the end of the historical sales process.
  • a weighting subunit configured to perform weighting processing on the satisfaction score, the objective purchase propensity value, and the subjective purchase propensity value, and output the weighted result as the actual purchase propensity degree of the customer.
  • the predicting device of the customer purchase intention further includes:
  • the display unit 606 is configured to sequentially display the power-up follow-up tasks based on each of the customers in the order of the actual purchase tendency of each of the customers in the sales task management interface.
  • the changing unit 607 is configured to change the implementation state of the power-up follow-up task from the first state to the second state when receiving the scheduling instruction of the power-up follow-up task, so as to block the other power-seat seats The electric sales follow up task.
  • the predicting device of the customer purchase intention further includes:
  • the third obtaining unit 608 is configured to obtain, according to the creation time point of the power-up follow-up task, the created duration of the power-up follow-up task at each moment.
  • the calculating unit 609 is configured to calculate a purchase tendency degree decrease value corresponding to the created time length, and the purchase tendency degree decrease value is proportional to the created time length.
  • the second output unit 610 is configured to output a difference between the actual purchase propensity degree corresponding to the electric power follow-up task and the purchase propensity decrease value as an actual purchase corresponding to the current sales follow-up task at the current time. Degree of inclination.
  • the adjusting unit 611 is configured to adjust an arrangement order of the electric vehicle follow-up tasks in the sales task management interface based on an actual purchase tendency degree corresponding to the electric power follow-up task at the current time.
  • FIG. 10 is a schematic diagram of an electronic device according to an embodiment of the present application.
  • the electronic device 10 of this embodiment includes a processor 1000 and a memory 1001 in which computer readable instructions 1002 executable on the processor 1000 are stored, such as a customer purchase intention. Forecasting program.
  • the processor 1000 executes the computer readable instructions 1002 to implement the steps in the foregoing method for predicting a method of invoking each customer purchase, such as steps 101 to 105 shown in FIG.
  • the processor 1000 implements the functions of the modules/units in the various apparatus embodiments described above when the computer readable instructions 1002 are executed, such as the functions of the units 601 to 605 shown in FIG.
  • the computer readable instructions 1002 can be partitioned into one or more modules/units that are stored in the memory 1001 and executed by the processor 1000, To complete this application.
  • the one or more modules/units may be a series of computer readable instruction segments capable of performing a particular function for describing the execution of the computer readable instructions 1002 in the electronic device 10.
  • the so-called processor 1000 can be a central processing unit (Central Processing Unit, CPU), can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc.
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the memory 1001 may be an internal storage unit of the electronic device 10, such as a hard disk or a memory of the electronic device 10.
  • the memory 1001 may also be an external storage device of the electronic device 10, such as a plug-in hard disk equipped on the electronic device 10, a smart memory card (SMC), and a secure digital (SD). Card, flash card (Flash Card) and so on.
  • the memory 1001 may also include both an internal storage unit of the electronic device 10 and an external storage device.
  • the memory 1001 is configured to store the computer readable instructions and other programs and data required by the electronic device.
  • the memory 1001 can also be used to temporarily store data that has been output or is about to be output.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.
  • the integrated unit if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium.
  • a computer readable storage medium A number of instructions are included to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Data Mining & Analysis (AREA)
  • General Business, Economics & Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Educational Administration (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Signal Processing (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本方案提供了一种客户购买意向的预测方法、装置、电子设备及介质，适用于信息处理领域，该方法包括：获取客户的个人特征数据；将个人特征数据输入预先建立的随机森林模型，以输出客户的客观购买倾向值；根据历史电销过程中客户的情感倾向，获取客户的主观购买倾向值；对客观购买倾向值以及主观购买倾向值进行加权处理，并将加权结果输出为客户的实际购买倾向度；将实际购买倾向度大于预设阈值的客户确定为潜在客户，以使电销坐席对潜在客户进行电话回访并推销电销产品。本方案综合了多方面的考量因子来确定出潜在客户，因而提高了潜在客户的预测准确率；通过对客观购买倾向值以及主观购买倾向值进行加权处理，实现了对于客户购买意向的量化计算。

Description

客户购买意向的预测方法、装置、电子设备及介质
本申请要求于2017年08月24日提交中国专利局、申请号为201710736859.2、发明名称为“客户购买意向的预测方法及电子设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请属于信息处理领域，尤其涉及一种客户购买意向的预测方法、装置、电子设备及介质。
 背景技术
目前，产品的营销手段包括电话营销、邮件营销以及短信营销等。电话营销是通过使用电话来实现有计划、有组织并且高效率地扩大顾客群的手法。为了避免电销的坐席人员只能随机地打出大量电话，凭借运气地去给各个电话接听者推销产品，目前，各大企业都开始着手于实现个性化的精准营销。具体地，通过对收集得到的各个用户的个人特征数据进行深入分析，确定出不同客户的不同消费特点，从而在推销产品与客户的消费特点较为吻合的情况下，才将该客户确认为潜在客户，以令坐席人员向该潜在客户进行电话推销，由此可保证每一次电话推销后，能够有更大的概率令客户转化为购买产品的实际客户，从而提高营销效率。
然而，现有技术仅根据客户的个人特征数据来直接评估该客户是否为潜在客户，考虑因素单一，且该方式无法量化客户的产品购买意向，因而难以找出真正具有电销产品购买意向的客户。
 技术问题
有鉴于此，本申请实施例提供了一种客户购买意向的预测方法、装置、电子设备及介质，以解决现有技术在确定潜在客户时，考虑因素单一以及无法量化客户的产品购买意向的问题。
 技术解决方案
本申请实施例的第一方面提供了一种客户购买意向的预测方法，包括：
获取客户的个人特征数据；
将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值；
根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值；
对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度；
将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
本申请实施例的第二方面提供了一种客户购买意向的预测装置，包括：
第一获取单元，用于获取客户的个人特征数据；
第一输出单元，用于将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值；
第二获取单元，用于根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值；
加权单元，用于对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度；
确定单元，用于将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
本申请实施例的第三方面提供了一种电子设备，包括存储器以及处理器，所述存储器中存储有可在所述处理器上运行的计算机可读指令，所述处理器执行所述计算机可读指令时实现如第一方面所述的客户购买意向的预测方法的步骤。
本申请实施例的第四方面提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机可读指令，所述计算机可读指令被处理器执行时实现如第一方面所述的客户购买意向的预测方法的步骤。
 有益效果
本申请实施例中，通过将客户的个人特征数据输入预设的随机森林模型，可计算出客观层面上客户对电销产品的购买倾向值；通过获取历史电销过程中客户的情感倾向，可计算出主观层面上客户对电销产品的购买倾向值；由于最终输出的客户实际购买倾向度为客观购买倾向值以及主观购买倾向值的加权结果，因而实现了对于客户购买意向的量化计算，使得最终所确定出的潜在客户为综合了多方面考量因子所得出的潜在客户，因而提高了潜在客户的预测准确率；同时，通过令电销坐席对潜在客户进行电话回访并推销电销产品，能够避免对历史电销客户的忽视，由此也进一步地降低了客户的流失率。
 附图说明
图1是本申请实施例提供的客户购买意向的预测方法的实现流程图；
图2是本申请实施例提供的客户购买意向的预测方法S103的具体实现流程图；
图3是本申请实施例提供的客户购买意向的预测方法S104的具体实现流程图；
图4是本申请另一实施例提供的客户购买意向的预测方法的实现流程图；
图5是本申请又一实施例提供的客户购买意向的预测方法的实现流程图；
图6是本申请实施例提供的客户购买意向的预测装置的结构框图；
图7是本申请实施例提供的客户购买意向的预测装置的结构框图；
图8是本申请另一实施例提供的客户购买意向的预测装置的结构框图；
图9是本申请又一实施例提供的客户购买意向的预测装置的结构框图；
图10是本申请实施例提供的电子设备的示意图。
 本发明的实施方式
为了说明本申请所述的技术方案，下面通过具体实施例来进行说明。
图1示出了本申请实施例提供的客户购买意向的预测方法的实现流程，该方法流程包括步骤S101至S105。各步骤的具体实现原理如下：
S101：获取客户的个人特征数据。
客户是指具有购买电销产品可能性的历史电销客户，其用于从中挖掘出产品购买倾向度较高且需要再次进行电话推销的客户。在电销产品以及电销服务满足客户的需求的情况下，客户能够转化为购买产品的实际客户。其中，电销产品为电销坐席通过电话沟通的方式向客户推荐的产品，包括但不限于保险产品以及信贷产品等各类金融产品。
每一电销坐席所曾经推销的客户以及与客户相关的推销信息均记录于数据库中。其中，与客户相关的推销信息包括客户的个人特征数据。因此，对于其中的一个客户而言，可从数据库中读取出该客户的个人特征数据。个人特征数据包括但不限于年龄、收入、兴趣爱好、学历、金融产品历史消费金额以及寿险交付保费等。
S102：将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值。
本申请实施例中，获取预先训练完毕的随机森林模型。随机森林模型包括多个决策树，每一决策树用于根据输入参数来进行分类选择。将各个决策树的分类选择结果进行统计汇总后，获取随机森林模型的最终输出参数。其中，输入参数为当前客户的个人特征数据。输出参数为该客户的客观购买倾向值。客观购买倾向值的大小表征了在客观条件上，客户对于电销产品的购买可能性大小，也表征了客户的个人特征数据与电销产品的特征匹配程度。
具体地，将多个训练样本数据输入预先构建的随机森林模型。每一训练样本数据包括一历史推销客户的各项个人特征数据以及客户类型。其中，每一训练样本数据中的历史推销客户均为电销坐席对于同一电销产品所推销的客户，且训练得到的随机森林模型也与该电销产品相关。上述客户类型为实际客户或非实际客户。即，电销坐席所曾经推销的客户是否最终购买了该电销产品。若是，则该历史推销客户为实际客户，若否，则该历史推销客户为非实际客户。
基于获取得到的各个训练样本数据，对随机森林模型中的模型参数进行调整。具体地，在接收到的N个训练样本数据中，重复执行多次有放回地随机抽取，以将抽取到的M（0< M< N，且M为整数）个训练样本数据作为一个新的训练样本集。根据新的训练样本集，生成K（K为大于1的整数）个用于分类的决策树。其中，决策树包括二叉树以及非二叉树。
由于随机森林模型的模型参数调整方法为本领域的现有技术，因此不再详细论述。
S103：根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值。
对于每一客户而言，其在历史电销过程中的情感倾向基于客户的主观反应态度来体现，其包括积极、中立、厌恶等多种类型的情感倾向。
示例性地，客户在历史电销过程中的情感倾向可通过以下方式获取：电销坐席在每一次的电销过程中，即每一次与客户电话接触的过程中，可自行判断客户的情感倾向属于哪一类型，并将判断结果记录为与客户相关的推销信息后，存储于数据库中。因此，在对客户的产品购买意向进行预测时，可从数据库中读取该客户所对应的最新一条数据记录，并读取其中存储的客户的情感倾向。
作为本申请的一个实施例，如图2所示，上述S103具体包括：
S1031：对历史电销过程进行音频录制，得到音频数据。
本申请实施例中，电销坐席通过安装有通信软件的智能终端来与客户进行电话通信，以进行产品推销。当智能终端检测到当前电销坐席所拨打的电话号码接通时，触发通信软件所携带的音频录制功能，开始执行音频录制，令该时刻为t 1。当检测到当前所拨打的电话中断时，停止录制音频，令该时刻为t 2。将时刻t 1至时刻t 2之间所录制得到的音频数据保存为一音频文件，且该音频文件与时刻t 1至时刻t 2之间电销坐席所接触的客户对应。
S1032：将所述音频数据转换为文本数据。
S1033：基于预设的积极情感词典以及消极情感词典，对所述文本数据进行识别处理，以确定所述文本数据对应的情感倾向。
积极情感词典中包含有预先收集得到的用于表达积极情感的各个词语，如“很好”、“满意”以及“不错”等。消极情感词典中包含有预先收集得到的用于表达消极情感的各个词语，如“很差”、“骚扰”以及“很烦”等。对文本数据进行识别处理时，先对该文本数据进行分词处理，以得到文本数据对应的多个分词。
判定各个分词是否存在于积极情感词典或消极情感词典。若文本数据中，出现有一分词存在于积极情感词典，则将用于表示情感倾向程度的累积数值加一；若出现有一分词存在于消极情感词典，则将用于表示情感倾向程度的累积数值减一。根据累积数值与情感倾向的对应关系，确定出与最终得到的累积数值对应的情感倾向。
优选地，本申请实施例中，还可基于标记有情感倾向的多个文本训练数据，训练情感分类模型。此时，若上述语音数据转换得到的文本数据中，各个分词均未出现于积极情感词典以及消极情感词典，则将该文本数据输入预先训练完毕的情感分类模型，以输出文本数据对应的情感倾向。
S1034：获取与所述情感倾向匹配的主观购买倾向值。
本申请实施例中，不同的情感倾向对应于不同的主观购买倾向值。根据主观购买倾向值与情感倾向的对应关系，确定出与当前时刻情感倾向对应的主观购买倾向值。
例如，若情感倾向为十分满意，则对应的主观购买倾向值为100%；若情感倾向为积极，则对应的主观购买倾向值为90%；若情感倾向为厌恶，则对应的主观购买倾向值为0%。
S104：对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
客观购买倾向值以及主观购买倾向值分别具有对应的一个权重比值。若通过上述步骤所得到的客观购买倾向值为a，其对应预设的权重比值为A，主观购买倾向值为b，其对应预设的权重比值为B，则计算得到的客户的实际购买倾向度C=A*a+B*b。
优选地，主观购买倾向值所对应的权重比值大于客观购买倾向值所对应的权重比值。
作为本申请的一个具体实施示例，客观购买倾向值所对应的权重比值优选为35%，主观购买倾向值所对应的权重比值优选为65%，则客户的实际购买倾向度C可通过以下公式计算得出：
C=A*35%+B*65%
其中，上述A为客观购买倾向值，上述B为主观购买倾向值。
本申请实施例中，由于客户的主观购买倾向值能够较为准确地反映出客户对于电销产品的主观态度倾向，而客户是否会执行购买操作，通常与其主观态度倾向具有较为直接的关联，因此，通过令主观购买倾向值所对应的权重比值大于客观购买倾向值所对应的权重比值，在主观购买倾向值所对应的权重比值优为65%以及客观购买倾向值所对应的权重比值为35%的情况下，计算得到的客户的实际购买倾向度将具有较高的参考价值，能够进一步提高潜在客户的识别准确率。
S105：将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
若客户在当前时刻的实际购买倾向度低于预设阈值，则表示即使电销坐席为其进行电话推销，该客户也难以转化为实际客户，因此，为了提高电销坐席的营销效率，仅将实际购买倾向度大于预设阈值的客户确定为潜在客户。通过将确定出的潜在客户推荐至电销坐席，使得电销坐席能够在有限的时间内，争取为产品购买概率更高的历史电销客户进行电话回访，以进行再次推销，从而最大限度地提高客户转化率。
本申请实施例中，通过将客户的个人特征数据输入预设的随机森林模型，可计算出客观层面上客户对电销产品的购买倾向值；通过获取历史电销过程中客户的情感倾向，可计算出主观层面上客户对电销产品的购买倾向值；由于最终输出的客户实际购买倾向度为客观购买倾向值以及主观购买倾向值的加权结果，因而实现了对于客户购买意向的量化计算，使得最终所确定出的潜在客户为综合了多方面考量因子所得出的潜在客户，因而提高了潜在客户的预测准确率；同时，通过令电销坐席对潜在客户进行电话回访并推销电销产品，能够避免对历史电销客户的忽视，由此也进一步地降低了客户的流失率。
作为本申请的一个实施例，在上述实施例的基础之上，对客户的实际购买倾向度的加权方式做进一步地限定。如图3所示，上述S104包括：
S1041：获取所述历史电销过程结束时所述客户反馈的满意度评分。
在每一次的电销过程结束后，客户均可以接收到满意度评分提示信息。满意度评分提示信息用于提示客户对本次的电销服务或电销坐席的推销水平进行评分。客户在通信终端的拨号键盘中按下评分数值或者通过短信方式回复评分数值后，便可接收到客户反馈的满意度评分。评分数值越高，客户的满意度程度越高。
每一客户在电销过程结束时所反馈的满意度评分同样存储于数据库中。在计算客户的实际购买倾向度之前，从数据库中读取客户最近一次所反馈的满意度评分。
S1042：对所述满意度评分、所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
本申请实施例中，由于满意度评分表现了客户对电销产品或者电销服务的满意程度，其在一定程度上能够准确反映客户最真实的主观情感倾向，因此，基于满意度评分、客观购买倾向值以及主观购买倾向值这三个因子来共同计算客户的实际购买倾向度，能够降低了因主观购买倾向值为理论上分析得到的数值而造成实际购买倾向度的预测误差，因此，提高了客户实际购买倾向度的预测准确性。
作为本申请的一个实施例，如图4所示，在上述S105之后，还包括：
S106：在电销任务管理界面中，依照各个所述客户的所述实际购买倾向度的高低顺序，依次展示基于各个所述客户的电销跟进任务。
本申请实施例中，当客户的实际购买倾向度大于预设阈值时，在电销任务管理界面中，生成基于该客户的电销跟进任务。其中，任一电销坐席均可以通过自己的电销坐席账号来登录电销任务管理系统，以查看电销任务管理界面中所展示的电销跟进任务。
对于实际购买倾向度大于预设阈值的多个客户，按照实际购买倾向度的数值大小，对各个客户所对应的电销跟进任务进行排序，使得实际购买倾向度较大的客户所对应的电销跟进任务排在实际购买倾向度较小的客户所对应的电销跟进任务之前。
S107：当接收到所述电销跟进任务的调度指令时，将所述电销跟进任务的实施状态由第一状态变变更为第二状态，以对其他电销坐席屏蔽所述电销跟进任务。
电销坐席可在电销任务管理界面中，点击选取自己所需跟进的一项电销跟进任务，此时，即接收到基于该电销跟进任务的调度指令。将该电销跟进任务与发出调度指令的电销坐席账号进行绑定，并将与该电销跟进任务相关的客户信息发送至与电销跟进任务绑定的电销坐席账号。其中，与该电销跟进任务相关的客户信息包括客户的个人特征数据、联系方式、实际购买倾向度以及满意度评分等，由此使得获取到客户信息的电销坐席能够及时对客户进行回访以及进行电销操作。
本申请实施例中，电销跟进任务的实施状态用于表示电销跟进任务的实时处理进度，实施状态包括第一状态以及第二状态。示例性地，第一状态为未处理状态，第二状态为已分配状态。电销跟进任务的实施状态可通过电销跟进任务在电销任务管理界面中所呈现的色彩来表示。例如，将电销跟进任务标记为红色，以表示其实施状态为第一状态；将电销跟进任务标记为黄色，以表示其实施状态为第二状态。
当电销跟进任务被任一电销坐席选取时，该电销跟进任务的实施状态变更为第二状态。由于第二状态下的电销跟进任务无法被再次点击选取，因而不会再次接收到基于该电销跟进任务的调度指令，实现了对其他电销坐席的屏蔽。
本申请实施例中，通过在电销任务管理界面依次展示基于各个客户的电销跟进任务，使得电销坐席能够根据电销跟进任务的排列顺序，实时了解到哪一客户的实时购买倾向度最高以及哪一电销跟进任务能够达到较好的回访效果。当接收到电销跟进任务的调度指令时，通过将电销跟进任务的实施状态由第一状态变变更为第二状态，使得其他电销坐席无法重复调度同一电销跟进任务，避免了多个电销坐席跟进同一客户的情况发生，提高了电销任务的跟进效率以及电销坐席人员的工作效率，由此也避免了对客户造成过多的电话回访骚扰。
作为本申请的一个实施例，在上一实施例的基础之上，当电销跟进任务的实施状态为第一状态时，对与该电销跟进任务对应的客户的实际购买倾向度的预测方式做进一步地限定。如图5所示，上述客户购买意向的预测方法还包括：
S108：根据所述电销跟进任务的创建时间点，获取所述电销跟进任务在各个时刻的已创建时长。
本申请实施例中，当首次计算得到的客户的实际购买倾向度大于预设阈值时，在电销任务管理界面中，生成并展示基于该客户的电销跟进任务，则电销跟进任务的生成时间即为电销跟进任务的创建时间点。
当电销跟进任务未被任一电销坐席调度时，其实施状态保持为第一状态。随着时间的推移，电销跟进任务在电销任务管理界面中所存在的时间越长，即，电销跟进任务的已创建时长越长。
在任一时刻，将该时刻的系统实时时间与电销跟进任务的创建时间点的差值确定为电销跟进任务的已创建时长。
S109：计算所述已创建时长对应的购买倾向度下降值，所述购买倾向度下降值与所述已创建时长成正比。
S110：将所述电销跟进任务对应的所述实际购买倾向度与所述购买倾向度下降值的差值输出为当前时刻所述电销跟进任务对应的实际购买倾向度。
根据电销跟进任务在当前时刻的已创建时长，经由预设的购买倾向度下降值计算公式，输出已创建时长对应的购买倾向度下降值Δs。将电销跟进任务所实时对应的实际购买倾向度调整为S-Δs。其中，S表示电销跟进任务在创建时间点所对应的实际购买倾向度。上述购买倾向度下降值计算公式为正比例函数，该公式例如可以是：|y|=a×x。其中，a为预设的常系数，x为电销跟进任务的已创建时长，y为已创建时长x所对应的购买倾向度下降值。由此可知，电销跟进任务的已创建时长x越大，购买倾向度下降值越大。
S111：基于当前时刻所述电销跟进任务对应的实际购买倾向度，对所述电销跟进任务在所述电销任务管理界面中的排列顺序进行调整。
由于电销任务管理界面中各个电销跟进任务的排列顺序表示了电销跟进任务所对应的实际购买倾向度的大小。因此，若电销跟进任务未被调度，则根据上述S108至S110可知，电销跟进任务所对应的实际购买倾向度将越来越小，故电销跟进任务在电销任务管理界面中的排列顺序也将实时发生调整。当客户的实际购买倾向度小于预设阈值时，将电销跟进任务进行删除。
本申请实施例中，由于确定出的潜在客户为从历史电销客户中所筛选出的电销产品购买可能性较高的客户，因此，若较长时间未对潜在客户进行回访，则该客户对电销产品的购买意向也会随着时间的流逝而变得越来越小。通过对电销跟进任务在电销任务管理界面中的排列顺序进行实时调整，使得坐席人员基于排列顺序较后的电销跟进任务，能够了解到哪些客户的流失可能性将越来越大，由此起到了督促跟进的作用。
应理解，上述实施例中各步骤的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本申请实施例的实施过程构成任何限定。
对应于上文实施例所述的客户购买意向的预测方法，图6示出了本申请实施例提供的客户购买意向的预测装置的结构框图。为了便于说明，仅示出了与本实施例相关的部分。
参照图6，该装置包括：
第一获取单元601，用于获取客户的个人特征数据。
第一输出单元602，用于将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值。
第二获取单元603，用于根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值。
加权单元604，用于对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
确定单元605，用于将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
可选地，所述第二获取单元603包括：
录制子单元，用于对历史电销过程进行音频录制，得到音频数据。
转换子单元，用于将所述音频数据转换为文本数据。
识别子单元，用于基于预设的积极情感词典以及消极情感词典，对所述文本数据进行识别处理，以确定所述文本数据对应的情感倾向。
第一获取子单元，用于获取与所述情感倾向匹配的主观购买倾向值。
可选地，所述加权单元604包括：
第二获取子单元，用于获取所述历史电销过程结束时所述客户反馈的满意度评分。
加权子单元，用于对所述满意度评分、所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
可选地，如图7所示，所述客户购买意向的预测装置还包括：
展示单元606，用于在电销任务管理界面中，依照各个所述客户的所述实际购买倾向度的高低顺序，依次展示基于各个所述客户的电销跟进任务。
变更单元607，用于当接收到所述电销跟进任务的调度指令时，将所述电销跟进任务的实施状态由第一状态变变更为第二状态，以对其他电销坐席屏蔽所述电销跟进任务。
可选地，当所述电销跟进任务的所述实施状态为第一状态时，如图8所示，所述客户购买意向的预测装置还包括：
第三获取单元608，用于根据所述电销跟进任务的创建时间点，获取所述电销跟进任务在各个时刻的已创建时长。
计算单元609，用于计算所述已创建时长对应的购买倾向度下降值，所述购买倾向度下降值与所述已创建时长成正比。
第二输出单元610，用于将所述电销跟进任务对应的所述实际购买倾向度与所述购买倾向度下降值的差值输出为当前时刻所述电销跟进任务对应的实际购买倾向度。
调整单元611，用于基于当前时刻所述电销跟进任务对应的实际购买倾向度，对所述电销跟进任务在所述电销任务管理界面中的排列顺序进行调整。
图10是本申请一实施例提供的电子设备的示意图。如图10所示，该实施例的电子设备10包括：处理器1000以及存储器1001，所述存储器1001中存储有可在所述处理器1000上运行的计算机可读指令1002，例如客户购买意向的预测程序。所述处理器1000执行所述计算机可读指令1002时实现上述各个客户购买意向的预测方法实施例中的步骤，例如图1所示的步骤101至105。或者，所述处理器1000执行所述计算机可读指令1002时实现上述各装置实施例中各模块/单元的功能，例如图6所示单元601至605的功能。
示例性的，所述计算机可读指令1002可以被分割成一个或多个模块/单元，所述一个或者多个模块/单元被存储在所述存储器1001中，并由所述处理器1000执行，以完成本申请。所述一个或多个模块/单元可以是能够完成特定功能的一系列计算机可读指令段，该指令段用于描述所述计算机可读指令1002在所述电子设备10中的执行过程。
所述电子设备10可以是桌上型计算机、笔记本、掌上电脑及云端服务器等计算设备。所述电子设备可包括，但不仅限于，处理器1000、存储器1001。本领域技术人员可以理解，图10仅仅是电子设备10的示例，并不构成对电子设备10的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件，例如所述电子设备还可以包括输入输出设备、网络接入设备、总线等。
所称处理器1000可以是中央处理单元(Central Processing Unit，CPU)，还可以是其他通用处理器、数字信号处理器 (Digital Signal Processor，DSP)、专用集成电路 (Application Specific Integrated Circuit，ASIC)、现成可编程门阵列 (Field-Programmable Gate Array，FPGA) 或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
所述存储器1001可以是所述电子设备10的内部存储单元，例如电子设备10的硬盘或内存。所述存储器1001也可以是所述电子设备10的外部存储设备，例如所述电子设备10上配备的插接式硬盘，智能存储卡（Smart Media Card, SMC），安全数字（Secure Digital, SD）卡，闪存卡（Flash Card）等。进一步地，所述存储器1001还可以既包括所述电子设备10的内部存储单元也包括外部存储设备。所述存储器1001用于存储所述计算机可读指令以及所述电子设备所需的其他程序和数据。所述存储器1001还可以用于暂时地存储已经输出或者将要输出的数据。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备（可以是个人计算机，服务器，或者网络设备等）执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器（Read-Only Memory，ROM）、随机存取存储器（Random Access Memory，RAM）、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，以上实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。

Claims (20)

  1. 一种客户购买意向的预测方法，其特征在于，包括：
    获取客户的个人特征数据；
    将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值；
    根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值；
    对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度；
    将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
  2. 如权利要求1所述的客户购买意向的预测方法，其特征在于，所述根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值，包括：
    对历史电销过程进行音频录制，得到音频数据；
    将所述音频数据转换为文本数据；
    基于预设的积极情感词典以及消极情感词典，对所述文本数据进行识别处理，以确定所述文本数据对应的情感倾向；
    获取与所述情感倾向匹配的主观购买倾向值。
  3. 如权利要求1所述的客户购买意向的预测方法，其特征在于，所述对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度，包括：
    获取所述历史电销过程结束时所述客户反馈的满意度评分；
    对所述满意度评分、所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
  4. 如权利要求1所述的客户购买意向的预测方法，其特征在于，还包括：
    在电销任务管理界面中，依照各个所述客户的所述实际购买倾向度的高低顺序，依次展示基于各个所述客户的电销跟进任务；
    当接收到所述电销跟进任务的调度指令时，将所述电销跟进任务的实施状态由第一状态变变更为第二状态，以对其他电销坐席屏蔽所述电销跟进任务。
  5. 如权利要求4所述的客户购买意向的预测方法，其特征在于，当所述电销跟进任务的所述实施状态为第一状态时，还包括：
    根据所述电销跟进任务的创建时间点，获取所述电销跟进任务在各个时刻的已创建时长；
    计算所述已创建时长对应的购买倾向度下降值，所述购买倾向度下降值与所述已创建时长成正比；
    将所述电销跟进任务对应的所述实际购买倾向度与所述购买倾向度下降值的差值输出为当前时刻所述电销跟进任务对应的实际购买倾向度；
    基于当前时刻所述电销跟进任务对应的实际购买倾向度，对所述电销跟进任务在所述电销任务管理界面中的排列顺序进行调整。
  6. 一种客户购买意向的预测装置，其特征在于，包括：
    第一获取单元，用于获取客户的个人特征数据；
    第一输出单元，用于将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值；
    第二获取单元，用于根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值；
    加权单元，用于对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度；
    确定单元，用于将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
  7. 根据权利要求6所述的客户购买意向的预测装置，其特征在于，所述第二获取单元包括：
    录制子单元，用于对历史电销过程进行音频录制，得到音频数据；
    转换子单元，用于将所述音频数据转换为文本数据；
    识别子单元，用于基于预设的积极情感词典以及消极情感词典，对所述文本数据进行识别处理，以确定所述文本数据对应的情感倾向；
    第一获取子单元，用于获取与所述情感倾向匹配的主观购买倾向值。
  8. 根据权利要求6所述的客户购买意向的预测装置，其特征在于，所述加权单元包括：
    第二获取子单元，用于获取所述历史电销过程结束时所述客户反馈的满意度评分；
    加权子单元，用于对所述满意度评分、所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
  9. 根据权利要求6所述的客户购买意向的预测装置，其特征在于，还包括：
    展示单元，用于在电销任务管理界面中，依照各个所述客户的所述实际购买倾向度的高低顺序，依次展示基于各个所述客户的电销跟进任务；
    变更单元，用于当接收到所述电销跟进任务的调度指令时，将所述电销跟进任务的实施状态由第一状态变变更为第二状态，以对其他电销坐席屏蔽所述电销跟进任务。
  10. 根据权利要求9所述的客户购买意向的预测装置，其特征在于，当所述电销跟进任务的所述实施状态为第一状态时，还包括：
    第三获取单元，用于根据所述电销跟进任务的创建时间点，获取所述电销跟进任务在各个时刻的已创建时长；
    计算单元，用于计算所述已创建时长对应的购买倾向度下降值，所述购买倾向度下降值与所述已创建时长成正比；
    第二输出单元，用于将所述电销跟进任务对应的所述实际购买倾向度与所述购买倾向度下降值的差值输出为当前时刻所述电销跟进任务对应的实际购买倾向度；
    调整单元，用于基于当前时刻所述电销跟进任务对应的实际购买倾向度，对所述电销跟进任务在所述电销任务管理界面中的排列顺序进行调整。
  11. 一种电子设备，其特征在于，包括存储器以及处理器，所述存储器中存储有可在所述处理器上运行的计算机可读指令，所述处理器执行所述计算机可读指令时实现如下步骤：
    获取客户的个人特征数据；
    将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值；
    根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值；
    对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度；
    将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
  12. 根据权利要求11所述的电子设备，其特征在于，所述根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值，包括：
    对历史电销过程进行音频录制，得到音频数据；
    将所述音频数据转换为文本数据；
    基于预设的积极情感词典以及消极情感词典，对所述文本数据进行识别处理，以确定所述文本数据对应的情感倾向；
    获取与所述情感倾向匹配的主观购买倾向值。
  13. 根据权利要求11所述的电子设备，其特征在于，所述对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度，包括：
    获取所述历史电销过程结束时所述客户反馈的满意度评分；
    对所述满意度评分、所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
  14. 根据权利要求11所述的电子设备，其特征在于，所述处理器执行所述计算机可读指令时，还实现如下步骤：
    在电销任务管理界面中，依照各个所述客户的所述实际购买倾向度的高低顺序，依次展示基于各个所述客户的电销跟进任务；
    当接收到所述电销跟进任务的调度指令时，将所述电销跟进任务的实施状态由第一状态变变更为第二状态，以对其他电销坐席屏蔽所述电销跟进任务。
  15. 根据权利要求14所述的电子设备，其特征在于，若所述电销跟进任务的所述实施状态为第一状态时，则所述处理器执行所述计算机可读指令时，还实现如下步骤：
    根据所述电销跟进任务的创建时间点，获取所述电销跟进任务在各个时刻的已创建时长；
    计算所述已创建时长对应的购买倾向度下降值，所述购买倾向度下降值与所述已创建时长成正比；
    将所述电销跟进任务对应的所述实际购买倾向度与所述购买倾向度下降值的差值输出为当前时刻所述电销跟进任务对应的实际购买倾向度；
    基于当前时刻所述电销跟进任务对应的实际购买倾向度，对所述电销跟进任务在所述电销任务管理界面中的排列顺序进行调整。
  16. 一种计算机可读存储介质，所述计算机可读存储介质存储有计算机可读指令，其特征在于，所述计算机可读指令被至少一个处理器执行时实现如下步骤：
    获取客户的个人特征数据；
    将所述个人特征数据输入预先建立的与电销产品相关的随机森林模型，以输出所述客户对所述电销产品的客观购买倾向值；
    根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值；
    对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度；
    将所述实际购买倾向度大于预设阈值的所述客户确定为潜在客户，以使电销坐席对所述潜在客户进行电话回访并推销所述电销产品。
  17. 根据权利要求16所述的计算机可读存储介质，其特征在于，所述根据历史电销过程中所述客户的情感倾向，获取所述客户对所述电销产品的主观购买倾向值，包括：
    对历史电销过程进行音频录制，得到音频数据；
    将所述音频数据转换为文本数据；
    基于预设的积极情感词典以及消极情感词典，对所述文本数据进行识别处理，以确定所述文本数据对应的情感倾向；
    获取与所述情感倾向匹配的主观购买倾向值。
  18. 根据权利要求16所述的计算机可读存储介质，其特征在于，所述对所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度，包括：
    获取所述历史电销过程结束时所述客户反馈的满意度评分；
    对所述满意度评分、所述客观购买倾向值以及所述主观购买倾向值进行加权处理，并将加权结果输出为所述客户的实际购买倾向度。
  19. 根据权利要求16所述的计算机可读存储介质，其特征在于，所述计算机可读指令被至少一个处理器执行时，还实现如下步骤：
    在电销任务管理界面中，依照各个所述客户的所述实际购买倾向度的高低顺序，依次展示基于各个所述客户的电销跟进任务；
    当接收到所述电销跟进任务的调度指令时，将所述电销跟进任务的实施状态由第一状态变变更为第二状态，以对其他电销坐席屏蔽所述电销跟进任务。
  20. 根据权利要求19所述的计算机可读存储介质，其特征在于，若所述电销跟进任务的所述实施状态为第一状态，则所述计算机可读指令被至少一个处理器执行时，还实现如下步骤：
    根据所述电销跟进任务的创建时间点，获取所述电销跟进任务在各个时刻的已创建时长；
    计算所述已创建时长对应的购买倾向度下降值，所述购买倾向度下降值与所述已创建时长成正比；
    将所述电销跟进任务对应的所述实际购买倾向度与所述购买倾向度下降值的差值输出为当前时刻所述电销跟进任务对应的实际购买倾向度；
    基于当前时刻所述电销跟进任务对应的实际购买倾向度，对所述电销跟进任务在所述电销任务管理界面中的排列顺序进行调整。
PCT/CN2018/074872 2017-08-24 2018-01-31 客户购买意向的预测方法、装置、电子设备及介质 WO2019037391A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US16/099,425 US20210224832A1 (en) 2017-08-24 2018-01-31 Method and apparatus for predicting customer purchase intention, electronic device and medium
SG11201809952XA SG11201809952XA (en) 2017-08-24 2018-01-31 Method and apparatus for predicting customer purchase intention, electronic device and medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710736859.2A CN107688967A (zh) 2017-08-24 2017-08-24 客户购买意向的预测方法及终端设备
CN201710736859.2 2017-08-24

Publications (1)

Publication Number Publication Date
WO2019037391A1 true WO2019037391A1 (zh) 2019-02-28

Family

ID=61154934

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/074872 WO2019037391A1 (zh) 2017-08-24 2018-01-31 客户购买意向的预测方法、装置、电子设备及介质

Country Status (4)

Country Link
US (1) US20210224832A1 (zh)
CN (1) CN107688967A (zh)
SG (1) SG11201809952XA (zh)
WO (1) WO2019037391A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111144108A (zh) * 2019-12-26 2020-05-12 北京百度网讯科技有限公司 情感倾向性分析模型的建模方法、装置和电子设备
CN112070519A (zh) * 2019-06-11 2020-12-11 中国科学院沈阳自动化研究所 一种基于数据全局搜索和特征分类的预测方法
CN112825175A (zh) * 2019-11-20 2021-05-21 顺丰科技有限公司 一种客户异常预警方法、装置和设备
CN115660722A (zh) * 2022-12-26 2023-01-31 平安银行股份有限公司 银寿客户转化的预测方法、装置和电子设备
CN117422071A (zh) * 2023-12-19 2024-01-19 中南大学 一种文本词项多重分割标注转换方法及装置

Families Citing this family (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108510313A (zh) * 2018-03-07 2018-09-07 阿里巴巴集团控股有限公司 一种信息转化率的预测、信息推荐方法和装置
CN108711067A (zh) * 2018-04-09 2018-10-26 平安科技（深圳）有限公司 电销时间段的选取方法、终端设备及介质
CN108648011B (zh) * 2018-05-11 2022-01-25 上海赢科信息技术有限公司 模型生成、识别客户购买车险意向的方法及系统
CN108898429B (zh) * 2018-06-19 2023-04-18 平安科技（深圳）有限公司 电子装置、偏好倾向预测方法和计算机可读存储介质
CN109300050A (zh) * 2018-08-31 2019-02-01 平安科技（深圳）有限公司 基于用户画像的保险推送方法、装置及存储介质
CN109285011A (zh) * 2018-09-12 2019-01-29 田雪娇 一种智能客户回访方法、装置及设备
CN110956479A (zh) * 2018-09-26 2020-04-03 北京高科数聚技术有限公司 基于销售线索交互记录的产品推荐方法
CN111160936A (zh) * 2018-11-07 2020-05-15 广汽菲亚特克莱斯勒汽车销售有限公司 一种汽车销售系统
CN109685560A (zh) * 2018-12-17 2019-04-26 泰康保险集团股份有限公司 大数据处理方法、装置、介质及电子设备
CN109961786B (zh) * 2019-01-31 2023-04-14 平安科技（深圳）有限公司 基于语音分析的产品推荐方法、装置、设备和存储介质
CN110166642A (zh) * 2019-05-31 2019-08-23 上海上湖信息技术有限公司 一种预测式外呼方法及设备
CN111091832B (zh) * 2019-11-28 2022-12-30 秒针信息技术有限公司 一种基于语音识别的意向评估方法和系统
CN111210281B (zh) * 2020-01-21 2023-07-25 上海克而瑞信息技术有限公司 一种房产交易中客户兴趣度统计方法
US11423432B2 (en) * 2020-04-21 2022-08-23 Capital One Services, Llc System, method and computer-accessible medium for providing financial account recommendation based on use of benefits
CN111784411A (zh) * 2020-07-14 2020-10-16 中国工商银行股份有限公司 一种客户预测方法、装置、电子设备及存储介质
CN111914169A (zh) * 2020-07-16 2020-11-10 中信银行股份有限公司 产品的推荐方法、装置、电子设备及计算机可读存储介质
CN112163154B (zh) * 2020-09-30 2024-05-03 深圳前海微众银行股份有限公司 数据处理方法、装置、设备及存储介质
WO2022086994A1 (en) * 2020-10-20 2022-04-28 ServiceTitan, Inc. Automated customer review matching
US20220129921A1 (en) * 2020-10-23 2022-04-28 Sony Group Corporation User intent identification from social media post and text data
CN112488750A (zh) * 2020-11-27 2021-03-12 上海容大数字技术有限公司 一种保险场景智能推荐续保系统
CN112651569A (zh) * 2020-12-31 2021-04-13 中国平安财产保险股份有限公司 基于状态节点的产品推荐方法、装置、设备及存储介质
CN112950261A (zh) * 2021-02-20 2021-06-11 北京读我网络技术有限公司 一种用户价值的确定方法及系统
US20220270117A1 (en) * 2021-02-23 2022-08-25 Christopher Copeland Value return index system and method
CN112950268A (zh) * 2021-03-02 2021-06-11 深圳市前海房极客网络科技有限公司 一种实时计算客户购买房产意愿程度的算法
CN114943225A (zh) * 2022-04-20 2022-08-26 南京优数信息技术有限公司 一种电话营销回访处理方法及系统
CN114757595A (zh) * 2022-06-15 2022-07-15 中邮消费金融有限公司 一种基于自定义调度策略的电话销售任务调度方法及系统
CN115952354B (zh) * 2022-12-28 2024-04-12 海南星捷安科技集团股份有限公司 一种药械销售代表工作台管理系统
CN117236984A (zh) * 2023-09-21 2023-12-15 元保科创(北京)科技有限公司 用户分层的管理方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6363346B1 (en) * 1999-12-22 2002-03-26 Ncr Corporation Call distribution system inferring mental or physiological state
CN101454771A (zh) * 2006-03-31 2009-06-10 依玛奇灵控股有限公司 基于使用多媒体调查特征匹配以划分和标记个体的系统和方法
CN105354183A (zh) * 2015-10-19 2016-02-24 Tcl集团股份有限公司 一种家电产品互联网评论的分析方法、装置及系统
CN106022800A (zh) * 2016-05-16 2016-10-12 北京百分点信息科技有限公司 一种用户特征数据的处理方法和装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6363346B1 (en) * 1999-12-22 2002-03-26 Ncr Corporation Call distribution system inferring mental or physiological state
CN101454771A (zh) * 2006-03-31 2009-06-10 依玛奇灵控股有限公司 基于使用多媒体调查特征匹配以划分和标记个体的系统和方法
CN105354183A (zh) * 2015-10-19 2016-02-24 Tcl集团股份有限公司 一种家电产品互联网评论的分析方法、装置及系统
CN106022800A (zh) * 2016-05-16 2016-10-12 北京百分点信息科技有限公司 一种用户特征数据的处理方法和装置

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112070519A (zh) * 2019-06-11 2020-12-11 中国科学院沈阳自动化研究所 一种基于数据全局搜索和特征分类的预测方法
CN112070519B (zh) * 2019-06-11 2024-03-05 中国科学院沈阳自动化研究所 一种基于数据全局搜索和特征分类的预测方法
CN112825175A (zh) * 2019-11-20 2021-05-21 顺丰科技有限公司 一种客户异常预警方法、装置和设备
CN111144108A (zh) * 2019-12-26 2020-05-12 北京百度网讯科技有限公司 情感倾向性分析模型的建模方法、装置和电子设备
CN111144108B (zh) * 2019-12-26 2023-06-27 北京百度网讯科技有限公司 情感倾向性分析模型的建模方法、装置和电子设备
CN115660722A (zh) * 2022-12-26 2023-01-31 平安银行股份有限公司 银寿客户转化的预测方法、装置和电子设备
CN115660722B (zh) * 2022-12-26 2023-03-31 平安银行股份有限公司 银寿客户转化的预测方法、装置和电子设备
CN117422071A (zh) * 2023-12-19 2024-01-19 中南大学 一种文本词项多重分割标注转换方法及装置
CN117422071B (zh) * 2023-12-19 2024-03-15 中南大学 一种文本词项多重分割标注转换方法及装置

Also Published As

Publication number Publication date
CN107688967A (zh) 2018-02-13
SG11201809952XA (en) 2019-03-28
US20210224832A1 (en) 2021-07-22

Similar Documents

Publication Publication Date Title
WO2019037391A1 (zh) 客户购买意向的预测方法、装置、电子设备及介质
US11004013B2 (en) Training of chatbots from corpus of human-to-human chats
WO2019037202A1 (zh) 目标客户的识别方法、装置、电子设备及介质
US20230139628A1 (en) Supporting automation of customer service
US20150071418A1 (en) Techniques for topical customer service menu reconfiguration based on social media
US20180253734A1 (en) Automated upsells in customer conversations
EP4036823A1 (en) Customization platform and method for service quality evaluation product
CN112231584B (zh) 基于小样本迁移学习的数据推送方法、装置及计算机设备
US20170308903A1 (en) Satisfaction metric for customer tickets
US12002454B2 (en) Method and apparatus for intent recognition and intent prediction based upon user interaction and behavior
US20210118432A1 (en) Generating training datasets for a supervised learning topel model from outputs of a discovery topic model
CN114547475B (zh) 一种资源推荐方法、装置及系统
CN113297365B (zh) 一种用户意向判定方法、装置、设备及存储介质
CN111192082A (zh) 产品卖点的分析方法、终端设备及计算机可读存储介质
CN109426983A (zh) 闪购活动自动生成方法及装置、存储介质、电子设备
CN114186024A (zh) 推荐方法及装置
CN114202203A (zh) 投诉工单处理方法、装置、存储介质及电子设备
CN111445263A (zh) 信息处理方法、装置和计算机系统及介质
US20210319457A1 (en) Utilizing models to aggregate data and to identify insights from the aggregated data
CN116433324A (zh) 一种产品推荐方法、装置、介质和计算机设备
CN116151840A (zh) 一种基于大数据的用户服务数据智能管理系统及方法
WO2021073258A1 (zh) 基于情绪分析的任务跟催方法、装置、设备及存储介质
CN115186179A (zh) 保险产品推送方法和装置
CN113688222A (zh) 基于上下文语义理解的保险销售任务话术推荐方法、系统以及设备
CN113283979A (zh) 一种贷款申请人的贷款信用评估方法、装置及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18848918

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 24/09/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18848918

Country of ref document: EP

Kind code of ref document: A1