An outbound call method based on artificial intelligence

Abstract

The present invention relates to the technical field of intelligent outbound calls, and discloses an outbound call method based on artificial intelligence, including obtaining a set of call times for overdue customers according to historical call records by using the best time period strategy; updating the set of opening formats according to the set of opening formats and historical call records by using the opening format language optimization strategy; obtaining the professional ranking of each business according to historical reception data by using the professional scoring strategy; forming a transfer judgment strategy according to the conversation record; executing the transfer judgment strategy to obtain a set of candidate debt collectors; and allocating the optimal debt collector for overdue customers according to the set of candidate debt collectors and the set of business to be processed by using the optimal allocation strategy. The customized call strategy can avoid communication during inefficient time periods, save resources and optimize the collection process. Optimizing the set of opening formats can improve the customer's answering quality and satisfaction, thereby possibly increasing the willingness to repay.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent outbound calls, and particularly to an outbound call method based on artificial intelligence. Background Art

[0002] Artificial intelligence outbound calls are automatic telephone call systems implemented using artificial intelligence technology. These systems can automatically make phone calls and communicate with humans through preset scripts or natural language processing technology. Artificial intelligence outbound call systems need to understand and generate natural language. Through NLP technology, the system can parse the user's language input, understand their intentions, and generate corresponding responses. ASR technology can convert voice signals into text for further processing by the system, which is a key technology for enabling the system to understand human speech. In order to communicate orally with users, the AI outbound call system also needs to convert text into speech. This process is achieved through TTS technology, enabling the machine-generated speech to be both clear and natural. Artificial intelligence outbound call systems integrate various AI technologies such as speech recognition, language understanding, and dialogue management, and are widely used in fields such as customer service, marketing, information notification, and research surveys.

[0003] Existing outbound calls need to make a large number of calls within a certain period of time, and the call times are random. When the call time happens to be when the customer is inconvenient to answer, it reduces the customer development efficiency; due to different opening remarks, it also causes different lengths of time for customers to answer the phone, thereby reducing the outbound call efficiency.

[0004] This solution proposes to determine the set of call times when customers answer based on the historical call records of customers, count the effectiveness of each opening format statement, and update the opening format set; at the same time, match the optimal debt collection specialist according to the set of pending business of the collection specialists. Summary of the Invention

[0005] The present invention provides an outbound call method based on artificial intelligence to help solve the problems mentioned in the above background art.

[0006] The present invention provides the following technical solution: An outbound call method based on artificial intelligence, including:

[0007] Obtain the historical call records of overdue customers;

[0008] According to the historical call records, adopt the best time period strategy to obtain the set of call times for overdue customers;

[0009] Obtain all the opening format statements for outbound calls to form an opening format set;

[0010] According to the opening format set and the historical call records, adopt the opening format statement optimization strategy to update the opening format set;

[0011] Obtain the historical reception data of all debt collectors;

[0012] Record the ranking of debt collectors for each business as the professional ranking;

[0013] Based on the historical reception data, adopt a professional scoring strategy to obtain the professional ranking of each business;

[0014] Obtain the importance levels of all businesses to form a business level ranking;

[0015] Record the text of overdue customers and outbound calls as the conversation record;

[0016] When an overdue customer needs to be transferred to a debt collector:

[0017] Based on the conversation record, form a transfer judgment strategy;

[0018] Execute the transfer judgment strategy to obtain a set of candidate debt collectors;

[0019] Obtain the business that the overdue customer inquires about when being transferred to a debt collector, and record it as the transferred business;

[0020] Obtain the businesses that each element in the set of candidate debt collectors has already been assigned to, and obtain the set of pending businesses for each element in the set of candidate debt collectors;

[0021] Based on the set of candidate debt collectors and the set of pending businesses, adopt an optimal allocation strategy to assign the optimal debt collector to the overdue customer.

[0022] Optionally, the adoption of the best time period strategy includes:

[0023] Set the first time, where the first time refers to Monday to Friday;

[0024] Set the second time, where the second time refers to Saturday to Sunday;

[0025] Count the total number of answer times at the first time in the historical call records of the overdue customer, and record it as the answer times at the first time;

[0026] Count the total number of answer times at the second time in the historical call records of the overdue customer, and record it as the answer times at the second time;

[0027] When the answer times at the first time are greater than the answer times at the second time, classify the overdue customer as the first type of population;

[0028] When the answer times at the first time are less than or equal to the answer times at the second time, classify the overdue customer as the second type of population;

[0029] Set the screening quantity for selecting the calling time period;

[0030] Divide 24 hours of a day into e time periods, count the number of times overdue customers answer calls in each time period in the historical call records, obtain the time periods with the number of times greater than the screening quantity, and form a set of call time periods;

[0031] When the overdue customer belongs to the first type of population, call the overdue customer at the time periods included in the set of call time periods every day at the first time;

[0032] When the overdue customer belongs to the second type of population, call the overdue customer at the time periods included in the set of call time periods every day at the second time.

[0033] Optionally, the adoption of the opening format language optimization strategy includes:

[0034] Set a threshold time for judging whether to retain the opening format language;

[0035] Obtain any element in the set of opening formats and denote it as the marked opening format language;

[0036] Denote the call duration when making a call with the marked opening format language as the marked time;

[0037] Obtain all the marked times in the historical call records and form a set of marked times;

[0038] Traverse all the elements in the set of marked times, obtain the number of marked times greater than the threshold time, and denote it as the effective number of times;

[0039] Calculate the number of elements in the set of marked times and denote it as the number of calls;

[0040] Calculate the effective number of times ÷ the number of calls, and denote the result as the effective rate;

[0041] Set a judgment ratio for judging whether to retain the opening format language;

[0042] When the effective rate is less than or equal to the judgment ratio, remove the marked opening format language from the set of opening formats;

[0043] Traverse all the elements in the set of opening formats to execute the opening format language optimization strategy and update the set of opening formats.

[0044] Optionally, the adoption of the professional scoring strategy includes:

[0045] Obtain all the services and form a set of services;

[0046] Obtain any element in the set of services and denote it as the positioned service;

[0047] Obtain the scoring data of each debt collector in the positioning business from the historical reception data, calculate the average value of multiple scoring data of each debt collector, and obtain the average scoring value of each debt collector in the positioning business;

[0048] Sort the average scoring values of all debt collectors in the positioning business from high to low to obtain the professional ranking in the positioning business;

[0049] Traverse all elements in the business set, and adopt a professional scoring strategy to obtain the professional ranking of each element in the business set.

[0050] Optionally, the adoption of the optimal allocation strategy includes:

[0051] Obtain the location of the overdue customer's phone number;

[0052] Obtain the debt collectors in the candidate debt collector set whose hometown location is the same as the location, and form a debt collector set with the same location;

[0053] Obtain all the businesses whose important level in the business level ranking is higher than that of the transfer business, and form a priority business set;

[0054] Obtain all the businesses whose important level in the business level ranking is lower than that of the transfer business, and form a lagging business set;

[0055] Record the debt collectors whose elements in the to-be-processed business set are all in the priority business set as discarded debt collectors;

[0056] Record the debt collectors whose elements in the to-be-processed business set are not in the priority business set as the first adoptive collectors;

[0057] Record the debt collectors whose elements in the to-be-processed business set are simultaneously in the priority business set and the lagging business set as the second adoptive collectors;

[0058] Traverse the to-be-processed business sets of each element in the debt collector set with the same location, obtain all the first adoptive collectors and the second adoptive collectors, and form the first adoptive set and the second adoptive set respectively;

[0059] Set a limit value for restricting the number of businesses received by debt collectors;

[0060] Count the number of elements in the to-be-processed business sets of each debt collector in the first adoptive set to obtain the first handling number set;

[0061] Compare each element in the first handling number set with the limit value, obtain the debt collectors corresponding to all elements less than the limit value, and record them as the first preferred collector set;

[0062] Count the number of elements in the to-be-processed business set of each debt collector in the second adopted set to obtain a second handling quantity set;

[0063] Compare each element in the second handling quantity set with a limit value, and obtain the debt collectors corresponding to all elements less than the limit value, denoted as a second preferred collector set.

[0064] Optionally, the adoption of the optimal allocation strategy further includes:

[0065] Set an initial order, which is the order of the business in all to-be-processed businesses when it is first assigned to a debt collector;

[0066] Set an execution order, which is the order of the business in all to-be-processed businesses of the debt collector;

[0067] Set a determination quantity for judging whether a debt collector is qualified to receive overdue customers;

[0068] Obtain historical data of the execution order of the determination quantity, denoted as a historical execution order set;

[0069] Compare all elements in the historical execution order set with the initial order:

[0070] When the initial order is less than or equal to all elements in the historical execution order set, denote the debt collector to which the business belongs as a marked collector.

[0071] Optionally, the adoption of the optimal allocation strategy further includes:

[0072] Traverse the to-be-processed business sets of each element in the first preferred collector set and the second preferred collector set, obtain all marked collectors, and form a discarded collector set;

[0073] Remove the elements in the discarded collector set from the first preferred collector set and the second preferred collector set to obtain a first target collector set and a second target collector set;

[0074] Denote the debt collector who responds to receiving overdue customers as a responding debt collector;

[0075] Obtain the debt collector with the highest professional ranking for transferring business among all responding debt collectors, denoted as a responding and answering debt collector;

[0076] When the responding and answering debt collector is an element in the first target collector set, assign the transfer business of the overdue customer to the responding and answering debt collector, and the initial order of the transfer business is the first;

[0077] When the responding and answering debt collector is an element in the second target collector set;

[0078] Obtain the elements in the set of pending operations of the response collection specialist whose importance level is higher than that of the transfer operation, and denote it as the response priority operation set;

[0079] Obtain the maximum execution order of the response priority operation set, and denote it as the connection order x;

[0080] Assign the transfer operation of the overdue customer to the response collection specialist, and the initial order of the transfer operation is x + 1.

[0081] Optionally, the execution transfer judgment strategy includes:

[0082] Extract the keywords in the conversation record to form a keyword set;

[0083] Obtain the number of elements in the keyword set, and denote the result as the number of keywords a;

[0084] Assign the conversation record to all collection specialists;

[0085] Obtain the feature words extracted from the conversation record by each collection specialist;

[0086] For the feature words of each collection specialist, adopt a feature matching strategy, which specifically includes:

[0087] Obtain the feature words of any one collection specialist, and denote it as the positioning feature word set;

[0088] Compare the elements in the positioning feature word set and the keyword set;

[0089] Obtain the equal elements in the positioning feature word set and the keyword set to form a matching feature set;

[0090] Calculate the number of elements in the matching feature set, and denote the result as the number of matches b;

[0091] Calculate the number of matches ÷ the number of keywords, and denote the result as the matching rate c, c = b ÷ a;

[0092] When the matching rate is equal to 1, the collection specialist corresponding to the positioning feature word set is denoted as the matching collection specialist;

[0093] Obtain all the matching collection specialists to form a candidate collection specialist set.

[0094] Optionally, the conversation record includes:

[0095] First, convert the voice input of the collection customer into input text, which specifically includes:

[0096] Audio collection, obtain the voice input by the collection customer, and convert the voice into an analog continuous sound wave, denoted as the voice signal;

[0097] Preprocessing: Obtain the voice signal, perform pre-emphasis, framing, and windowing operations, and record the result after preprocessing as a set of voice frames;

[0098] Feature extraction: Obtain the set of voice frames, extract the feature parameters of each element in the set of voice frames, and form a set of feature parameters;

[0099] Apply the acoustic model: Obtain the set of feature parameters, traverse each element in the set of feature parameters, apply the acoustic model to each element, and obtain the phoneme corresponding to each voice frame, forming a set of sound results;

[0100] Apply the language model: Obtain the set of voice frames, apply the language model, and obtain multiple text sequences corresponding to the set of voice frames, forming a set of text results;

[0101] Search and decoding: Obtain the set of sound results and the set of text results, apply the decoding search algorithm, and record the obtained text sequence as the input text;

[0102] Secondly, obtain the tone, expression way, and speech rate information of the collection customer, and record it as implicit information;

[0103] Obtain the input text, apply the input text and the implicit information to the natural language processing algorithm, and obtain the emotion and repayment willingness of the collection customer, which is recorded as the communication feature;

[0104] Finally, obtain the input text and the communication feature, match the reply content, and record it as the output text;

[0105] Perform language conversion operations on the output text, specifically including:

[0106] Standardize the text: Remove punctuation marks and format numbers from the output text, and record the result as the standard text;

[0107] Prosody modeling: Obtain the standard text, analyze the intonation, rhythm, and emphasis of the standard text, and record the result as the text feature;

[0108] Apply the acoustic model: Convert the standard text into phonemes, and then convert the phonemes into sound wave signals, and record the result as the output sound wave;

[0109] Synthesize language: Obtain the output sound wave and the text feature, and apply the sound synthesis technology to obtain the voice output.

[0110] The present invention has the following beneficial effects:

[0111] 1. The outbound call method based on artificial intelligence sets a first time and a second time. It counts and compares the number of answer times of overdue customers at these two times. Classify the overdue customers according to the time with more answer times. Determine the high answer time period of each day and select the call period from it. By classifying overdue customers and selecting the time period with a high answer rate, the contact success rate can be increased, thereby improving the collection efficiency. The customized call strategy can avoid communication during inefficient time periods, save resources and optimize the collection process.

[0112] 2. The outbound call method based on artificial intelligence sets a threshold time for call effectiveness, counts the call duration of various opening format languages, evaluates each opening format language according to the effectiveness of the call duration, and determines whether to retain the opening format language according to the efficiency. Evaluating the effectiveness of the opening format language helps to identify and eliminate ineffective conversation methods, ensuring the use of the opening format language that can most effectively arouse the response and communication of overdue customers. Optimizing the set of opening formats can improve the answer quality and satisfaction of customers, thereby possibly increasing the willingness to repay.

[0113] 3. The outbound call method based on artificial intelligence obtains the scoring data of collection specialists in each business. Each collection specialist calculates the average score according to the scoring data of the business. Rank the collection specialists professionally according to the average score. Through professional ranking, it can be ensured that each business is handled by the collection specialist who is best at that business, improving the collection efficiency and quality. The scoring and ranking system can also motivate collection specialists to improve their business handling capabilities.

[0114] 4. The outbound call method based on artificial intelligence selects collection specialists according to the call location of overdue customers, obtains the set of collection specialists with the same location, classifies the priority of the business, and assigns collection specialists according to the priority. Geographic matching can enhance the communication efficiency between collection specialists and overdue customers, because people in the same region may have better communication compatibility. By giving priority to handling high-priority business, the resource allocation can be optimized, and the recovery speed of high-priority debts can be accelerated. Setting a limit on the number of business processes can prevent individual collection specialists from being overloaded, ensuring the quality of the collection work and the work efficiency of the collection specialists.

[0115] 5. The outbound call method based on artificial intelligence sets the initial order and execution order of the business in the collection specialists' pending business. Set the judgment quantity for determining whether a collection specialist is suitable to receive an overdue customer. According to the comparison between the historical execution order and the initial order, mark the collection specialists suitable for handling this business, avoiding the situation that low-priority business is not processed for a long time due to the repeated occurrence of high-priority business. This matching strategy ensures that the business is correctly prioritized according to importance and urgency, while taking into account the situation that low-priority business is not processed for a long time.

[0116] 6. The outbound call method based on artificial intelligence obtains the set of discarded commissioners, removes the elements in the first preferred commissioner set and the second preferred commissioner set from the set of discarded commissioners to obtain the first target commissioner set and the second target commissioner set. The elements in the target commissioner set are qualified to receive overdue customers. According to the response speed and professional ranking of each collection commissioner, the transfer business is allocated. By accurately selecting excellent collection commissioners, customer satisfaction and collection efficiency can be improved. The transfer business is assigned to the answering collection commissioner, and the processing priority of the transfer business is higher than that of the pending business in the pending business set of the answering collection commissioner whose business level is lower than that of the transfer business. Appropriate business allocation ensures efficient processing and reduces processing delays and customer dissatisfaction.

[0117] 7. The outbound call method based on artificial intelligence extracts keywords in the conversation and performs matching. According to the matching results, the most suitable collection commissioner is selected to handle a specific conversation. Keyword matching enables collection commissioners to focus more on the types of conversations they are best at handling, improving the professionalism and efficiency of case handling. Accurate matching ensures the pertinence and effectiveness of the collection process and reduces ineffective communication.

[0118] 8. The outbound call method based on artificial intelligence realizes the full-process automated processing from voice input to text processing and then to voice output. It quickly responds to user requests, reduces waiting time, improves processing efficiency, and thus greatly enhances user satisfaction and system throughput. BRIEF DESCRIPTION OF THE DRAWINGS

[0119] Figure 1 It is a schematic diagram of the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0120] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0121] Embodiment 1. Referring to Figure 1 , an outbound call method based on artificial intelligence includes

[0122] Obtain the historical call records of overdue customers;

[0123] According to the historical call records, adopt the best time period strategy to obtain the call time period set of overdue customers;

[0124] Obtain all the opening format phrases for outbound calls and form an opening format set;

[0125] Update the opening format set according to the opening format set and historical call records, and adopt the opening format language optimization strategy.

[0126] Obtain the historical reception data of all collection specialists.

[0127] Record the ranking of collection specialists for a business as the professional ranking.

[0128] According to the historical reception data, adopt the professional scoring strategy to obtain the professional ranking of each business.

[0129] Obtain the importance levels of all businesses to form the business level ranking.

[0130] Record the text of overdue customers and outbound calls as the conversation record.

[0131] When an overdue customer needs to be transferred to a collection specialist:

[0132] Form a transfer judgment strategy based on the conversation record.

[0133] Execute the transfer judgment strategy to obtain the set of candidate collection specialists.

[0134] Obtain the business consulted by the overdue customer when being transferred to a collection specialist, and record it as the transferred business.

[0135] Obtain the businesses already assigned to each element in the set of candidate collection specialists to obtain the set of pending businesses for each element in the set of candidate collection specialists.

[0136] According to the set of candidate collection specialists and the set of pending businesses, adopt the optimal allocation strategy to assign the optimal collection specialist to the overdue customer.

[0137] The adoption of the best time period strategy includes:

[0138] Set the first time, which refers to Monday to Friday.

[0139] Set the second time, which refers to Saturday to Sunday.

[0140] Count the total number of answer times at the first time in the historical call records of the overdue customer, and record it as the answer times at the first time.

[0141] Count the total number of answer times at the second time in the historical call records of the overdue customer, and record it as the answer times at the second time.

[0142] When the answer times at the first time are greater than the answer times at the second time, classify the overdue customer into the first category of people.

[0143] When the answer times at the first time are less than or equal to the answer times at the second time, classify the overdue customer into the second category of people.

[0144] Set the screening quantity for the selected call time period;

[0145] Divide the 24 hours of a day into e time periods, count the number of times overdue customers answer calls in each time period in the historical call records, obtain the time periods with the number of times greater than the screening quantity, and form a call time period set;

[0146] When the overdue customer belongs to the first type of population, call the overdue customer's phone during the time periods included in the call time period set every day at the first time;

[0147] When the overdue customer belongs to the second type of population, call the overdue customer's phone during the time periods included in the call time period set every day at the second time.

[0148] Set the first time and the second time. Count and compare the number of times overdue customers answer calls at these two times. Classify the overdue customers according to the time with more answered calls. Determine the high-answer time period every day and select the call time period from it. By classifying the overdue customers and selecting the time period with a high answer rate, the contact success rate can be improved, thereby enhancing the collection efficiency. The customized call strategy can avoid communicating during low-efficiency time periods, save resources and optimize the collection process.

[0149] The adopted opening format language optimization strategy includes:

[0150] Set the threshold time for judging whether to retain the opening format language;

[0151] Obtain any element in the opening format set and denote it as the marked opening format language;

[0152] Denote the call duration when making a call with the marked opening format language as the marked time;

[0153] Obtain all the marked times in the historical call records of the overdue customers and form a marked time set;

[0154] Traverse all the elements in the marked time set, obtain the number of marked times greater than the threshold time, and denote it as the effective number of times;

[0155] Calculate the number of elements in the marked time set and denote it as the number of calls;

[0156] Calculate the effective number of times ÷ the number of calls, and denote the result as the effective rate;

[0157] Set the determination ratio for judging whether to retain the opening format language;

[0158] When the effective rate is less than or equal to the determination ratio, remove the marked opening format language from the opening format set;

[0159] Traverse all the elements in the opening format set to execute the opening format language optimization strategy and update the opening format set.

[0160] Set a threshold time for call validity, count the call durations of various opening phrases, evaluate each opening phrase based on the validity of the call duration, and determine whether to retain the opening phrase according to efficiency. Evaluating the effectiveness of opening phrases helps identify and eliminate ineffective conversation methods, ensuring the use of opening phrases that are most likely to elicit responses and interactions from delinquent customers. Optimizing the set of opening phrases can improve the call-answering quality and satisfaction of customers, thereby potentially increasing the willingness to repay.

[0161] The adoption of a professional scoring strategy includes:

[0162] Obtain all businesses to form a business set;

[0163] Obtain any element in the business set and denote it as the targeted business;

[0164] Obtain the scoring data of each debt collector for the targeted business in the historical reception data, calculate the average value of multiple scoring data of each debt collector, and obtain the scoring mean of each debt collector for the targeted business;

[0165] Sort the scoring means of all debt collectors for the targeted business from high to low to obtain the professional ranking of the targeted business;

[0166] Traverse all elements in the business set and adopt the professional scoring strategy to obtain the professional ranking of each element in the business set.

[0167] Obtain the scoring data of each debt collector for each business, and each debt collector calculates the scoring mean based on the scoring data of the business. Rank the debt collectors professionally for each business according to the scoring mean. Through professional ranking, it can be ensured that each business is handled by the debt collector who is best at that business, improving the debt collection efficiency and quality. The scoring and ranking system can also motivate debt collectors to improve their business handling capabilities.

[0168] The adoption of an optimal allocation strategy includes:

[0169] Obtain the location of the overdue customer's phone number;

[0170] Obtain the debt collectors whose hometown locations are the same as the location of the overdue customer's phone number in the candidate debt collector set to form a debt collector set with the same location;

[0171] Obtain all businesses whose importance levels in the business level ranking are higher than the transferred business to form a priority business set;

[0172] Obtain all businesses whose importance levels in the business level ranking are lower than the transferred business to form a lagging business set;

[0173] Record the debt collectors whose elements in the set of to-be-processed services are all in the set of priority services as discarded debt collectors;

[0174] Record the debt collectors whose elements in the set of to-be-processed services are not in the set of priority services as the first adopted debt collectors;

[0175] Record the debt collectors whose elements in the set of to-be-processed services are simultaneously in the set of priority services and the set of lagging services as the second adopted debt collectors;

[0176] Traverse the sets of to-be-processed services of each element in the set of debt collectors with the same place of customer origin, obtain all the first adopted debt collectors and the second adopted debt collectors, and form the first adopted set and the second adopted set respectively;

[0177] Set a limit value for restricting the number of services received by debt collectors;

[0178] Count the number of elements in the set of to-be-processed services of each debt collector in the first adopted set to obtain the first handling number set;

[0179] Compare each element in the first handling number set with the limit value, obtain the debt collectors corresponding to all elements less than the limit value, and record them as the first preferred debt collector set;

[0180] Count the number of elements in the set of to-be-processed services of each debt collector in the second adopted set to obtain the second handling number set;

[0181] Compare each element in the second handling number set with the limit value, obtain the debt collectors corresponding to all elements less than the limit value, and record them as the second preferred debt collector set.

[0182] Select debt collectors according to the phone number place of origin of overdue customers to obtain the set of debt collectors with the same place of customer origin, classify the priority of services, and allocate debt collectors according to the priority. Region matching can enhance the communication efficiency between debt collectors and overdue customers, because people in the same region may have better communication compatibility. By giving priority to processing high-priority services, resource allocation can be optimized and the recovery speed of high-priority debts can be accelerated. Setting a limit on the number of services processed can prevent individual debt collectors from overworking and ensure the quality of debt collection work and the work efficiency of debt collectors.

[0183] The adoption of the optimal allocation strategy further includes:

[0184] Set the initial order, which is the order of the service in all to-be-processed services when it is first assigned to a debt collector;

[0185] Set the execution order, which is the order of the service in all to-be-processed services of the debt collector;

[0186] Set the judgment quantity for determining whether a debt collection specialist is qualified to receive overdue customers;

[0187] Obtain the historical data of the execution order of the number of judgment quantities, denoted as the historical execution order set;

[0188] Compare all elements in the historical execution order set with the initial order:

[0189] When the initial order is less than or equal to all elements in the historical execution order set, denote the debt collection specialist to which the business belongs as the marked specialist.

[0190] In this embodiment, the judgment quantity is 4;

[0191] Set the initial order and execution order of the business in the debt collection specialist's pending business. Set the judgment quantity for determining whether a debt collection specialist is suitable to receive overdue customers. According to the comparison between the historical execution order and the initial order, mark the debt collection specialist suitable for handling this business, avoiding the situation where low-priority business is not processed for a long time due to the repeated occurrence of high-priority business. This matching strategy ensures that the business is correctly prioritized according to importance and urgency, while taking into account the situation where low-priority business is not processed for a long time.

[0192] The adoption of the optimal allocation strategy further includes:

[0193] Traverse the pending business sets of each element in the first preferred specialist set and the second preferred specialist set, obtain all marked specialists, and form a discarded specialist set;

[0194] In this embodiment, there are 2 elements in the first preferred specialist set, denoted as the first numbered specialist and the second numbered specialist respectively; there are 2 elements in the second preferred specialist set, denoted as the third numbered specialist and the fourth numbered specialist respectively;

[0195] Traverse the pending business sets of the first numbered specialist, the second numbered specialist, the third numbered specialist, and the fourth numbered specialist. Among them, there is 1 element in the pending business set of the second numbered specialist, denoted as the numbered business. The initial order of the numbered business is 1. During the process of the second numbered specialist handling the numbered business, the historical data of the execution order are sorted by time as 1, 2, 2, 2 respectively; obtain the historical data of 4 execution orders to form a historical execution order set, and the elements in the historical execution order set are 1, 2, 2, 2.

[0196] The initial order is less than or equal to all elements in the historical execution order set, indicating that there are continuously higher-priority businesses than the numbered business being preferentially processed by the second numbered specialist, resulting in the marked business being processed later. Then, the second numbered specialist is the marked specialist.

[0197] Remove the elements in the discarded specialist set from the first preferred specialist set and the second preferred specialist set to obtain the first target specialist set and the second target specialist set;

[0198] Denote the collection specialists who respond to overdue customers as response collection specialists;

[0199] Obtain the collection specialist with the highest professional ranking in transferring business among all response collection specialists, and denote it as the answering collection specialist;

[0200] When the answering collection specialist is an element in the first target specialist set, assign the transfer business of the overdue customer to the answering collection specialist, and the initial order of the transfer business is the first;

[0201] When the answering collection specialist is an element in the second target specialist set;

[0202] Obtain the elements in the set of pending business of the answering collection specialist whose importance level is higher than that of the transfer business, and denote it as the response priority business set;

[0203] Obtain the maximum execution order of the response priority business set, and denote it as the connection order x;

[0204] Assign the transfer business of the overdue customer to the answering collection specialist, and the initial order of the transfer business is x + 1.

[0205] Obtain the discarded specialist set, remove the elements in the discarded specialist set from the first preferred specialist set and the second preferred specialist set to obtain the first target specialist set and the second target specialist set. The elements in the target specialist set meet the conditions for receiving overdue customers. According to the response speed and professional ranking of each collection specialist, assign the transfer business. By accurately selecting excellent collection specialists, customer satisfaction and collection efficiency can be improved. Assign the transfer business to the answering collection specialist, and the processing priority of the transfer business is higher than that of the pending business in the set of pending business of the answering collection specialist whose business level is lower than that of the transfer business. Appropriate business assignment ensures efficient processing and reduces processing delays and customer dissatisfaction.

[0206] The execution of the transfer judgment strategy includes:

[0207] Extract the keywords in the conversation record to form a keyword set;

[0208] Obtain the number of elements in the keyword set, and denote the result as the number of keywords a;

[0209] Assign the conversation record to all collection specialists;

[0210] Obtain the feature words extracted from the conversation record by each collection specialist;

[0211] For the characteristic words of each debt collector, adopt a characteristic matching strategy, specifically including:

[0212] Obtain the characteristic words of any debt collector, denoted as the positioning characteristic word set;

[0213] Compare the elements in the positioning characteristic word set and the keyword set;

[0214] Obtain the equal elements in the positioning characteristic word set and the keyword set, and form a matching characteristic set;

[0215] Calculate the number of elements in the matching characteristic set, and record the result as the matching number b;

[0216] Calculate the matching number ÷ the number of keywords, and record the result as the matching rate c, c = b ÷ a;

[0217] When the matching rate is equal to 1, the debt collector corresponding to the positioning characteristic word set is denoted as the matching debt collector;

[0218] Obtain all the matching debt collectors and form a candidate debt collector set.

[0219] Extract the keywords in the conversation and perform matching. Select the most suitable debt collector to handle a specific conversation according to the matching result. Keyword matching enables debt collectors to focus more on the types of conversations they are best at handling, improving the professionalism and efficiency of case handling. Precise matching ensures the pertinence and effectiveness of the collection process and reduces ineffective communication.

[0220] The said conversation record includes:

[0221] First, convert the voice input of the debt collection customer into input text, specifically including:

[0222] Audio acquisition, obtain the voice input by the debt collection customer, and convert the voice into an analog continuous sound wave, denoted as the voice signal;

[0223] Preprocessing, obtain the voice signal, perform pre-emphasis, framing, and windowing operations, and denote the result after preprocessing as the voice frame set;

[0224] Pre-emphasis is used to enhance the high-frequency part of the voice signal and compensate for the high-frequency attenuation during the transmission of the voice signal; Framing divides the voice signal into short time periods, usually 10 - 30 milliseconds per frame, for subsequent processing; Windowing is to reduce the discontinuity at the beginning and end of the frame.

[0225] Feature extraction, obtain the voice frame set, extract the characteristic parameters of each element in the voice frame set, and form a characteristic parameter set;

[0226] Apply an acoustic model to obtain a set of feature parameters. Traverse each element in the set of feature parameters, apply the acoustic model to each element, obtain the phonemes corresponding to each speech frame, and form a sound result set.

[0227] Apply a language model to obtain a set of speech frames. Apply the language model to obtain multiple text sequences corresponding to the set of speech frames, and form a text result set.

[0228] Perform search decoding to obtain the sound result set and the text result set. Apply a decoding search algorithm, and the obtained text sequence is denoted as the input text.

[0229] Use the existing ASR algorithm to convert the voice input of the collection customer into the input text. The core of the ASR algorithm is the construction of the acoustic model and the language model. The acoustic model is mainly used to establish the mapping relationship between speech features and phonemes, while the language model is used to calculate the prior probability of the word sequence. The joint action of these two models can improve the accuracy and robustness of speech recognition. In this process, the commonly used feature extraction method is the Mel Frequency Cepstral Coefficient, which mimics the auditory characteristics of the human ear and has a certain robustness in a noisy environment.

[0230] Utilize ASR technology to receive and accurately recognize the user's voice input. No matter what environment the user is in, it can accurately capture the user's needs. ASR technology supports multiple languages and dialects, can process voice inputs from all over the world, and serves a wider user group.

[0231] Secondly, obtain the tone, expression, and speech rate information of the collection customer, denoted as implicit information.

[0232] Obtain the input text, apply the input text and the implicit information to a natural language processing algorithm, and obtain the emotion and repayment willingness of the collection customer, denoted as communication features.

[0233] The existing natural language processing algorithm NLP in the prior art is a series of computational methods and technologies for understanding and processing human natural language. Natural language is the language used in human daily communication, such as English, Chinese, French, etc. The purpose of the NLP algorithm is to enable the computer to understand, interpret, and generate natural language like humans. It mainly includes the following types of algorithms:

[0234] 1. Lexical analysis algorithm: For example, the word segmentation algorithm, which divides continuous text into meaningful units such as words and punctuation marks.

[0235] 2. Syntactic analysis algorithm: Determine the structure and grammatical relationship of a sentence.

[0236] 3. Semantic understanding algorithm: Try to understand the meaning of the text.

[0237] 4. Text Classification Algorithm: Classify texts into different categories. For example, classify news articles into categories such as sports, entertainment, finance, etc.

[0238] 5. Sentiment Analysis Algorithm: Determine whether the sentiment expressed in the text is positive, negative, or neutral. For example, analyze whether a certain comment is a praise or a complaint about a product.

[0239] 6. Machine Translation Algorithm: Achieve the conversion between different languages.

[0240] Through NLP technology, it is able to understand and process the user's language input, including text preprocessing, lexical analysis, syntactic analysis, etc., in order to accurately grasp the user's intention. Through advanced processing such as sentiment analysis and semantic understanding, it can better interact with users and provide personalized service responses.

[0241] Finally, obtain the input text and communication features, match the response content, and record it as the output text;

[0242] Perform language conversion operations on the output text, specifically including:

[0243] Text normalization: Remove punctuation marks and format numbers from the output text, and record the result as the standard text;

[0244] Prosody modeling: Obtain the standard text, analyze the intonation, rhythm, and emphasis of the standard text, and record the result as the text features;

[0245] Apply the acoustic model to convert the standard text into phonemes, and then convert the phonemes into acoustic wave signals, and record the result as the output acoustic wave;

[0246] Synthesize language: Obtain the output acoustic wave and text features, and apply voice synthesis technology to obtain the voice output.

[0247] Existing TTS technology enables intelligent outbound calls to provide feedback to users in a natural and fluent voice form, greatly improving the naturalness of interaction and user satisfaction. It can synthesize voice feedback in different styles and languages according to different users and scenarios, achieving a more personalized service experience.

[0248] The steps of the TTS algorithm include:

[0249] Text normalization: Convert the input text into a suitable format for processing, such as removing punctuation marks and number formatting.

[0250] Prosody modeling: Analyze the intonation, rhythm, and emphasis in the text, which are the natural features of speech.

[0251] Acoustic model: Convert text into phonemes, and then further convert them into acoustic wave signals.

[0252] Speech synthesis: Use sound synthesis techniques (such as concatenative synthesis or waveform editing) to generate the final speech output.

[0253] Types of TTS algorithms:

[0254] Concatenative synthesis: simulates human voice changes by changing the pitch, suitable for most text-to-speech applications.

[0255] Waveform Editing: Directly manipulate raw audio waveforms to generate speech, often used in professional sound design.

[0256] Parametric speech synthesis: uses mathematical models to simulate speech characteristics and is suitable for situations where a high degree of customization is required.

[0257] The benefits of intelligent outbound calling include:

[0258] Optimize user experience: Users do not need to face complex interfaces or perform tedious operations, and can complete all their needs through simple voice interaction. Provide barrier-free customer service channels for users with visual impairments or limited mobility, so that they can more conveniently obtain the services they need.

[0259] Enhanced security: Advanced voice recognition technology can provide secure voice services while ensuring user privacy. Integrated security measures can prevent security risks such as voice imitation and ensure stable operation of the system.

[0260] Optimization data drive: Voice data collected through ASR technology can be used to further train and optimize models to make the system more intelligent. Based on big data analysis, the system can automatically adjust and optimize service strategies to continuously improve service quality and efficiency.

[0261] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0262] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An outbound calling method based on artificial intelligence, comprising: Obtain historical call records of overdue customers; According to the historical call records, the best time period strategy is adopted to obtain the calling time period set of overdue customers; Obtain opening format phrases of all outbound calls to form an opening format set; adopt an opening format phrase optimization strategy based on the opening format set and historical call records to update the opening format set; Obtain the historical reception data of all debt collection specialists; record the ranking of debt collection specialists in the business as the professional ranking; Based on historical reception data, a professional scoring strategy is adopted to obtain the professional ranking of each business; Obtain the importance levels of all businesses and form a business ranking; Record the text of overdue customers and outbound calls as conversation records; When an overdue customer needs to be transferred to a collection specialist: a transfer decision strategy is formed based on the conversation records; Execute the transfer judgment strategy and obtain a set of candidate collection specialists; Obtain the business of overdue customer consultation when transferring to the collection specialist, which is recorded as the transferred business; Obtain the business that has been assigned to each element in the candidate collection specialist set, and obtain the pending business set of each element in the candidate collection specialist set; Based on the candidate collection specialist set and the pending business set, the optimal allocation strategy is adopted to allocate the best collection specialist to the overdue customer; The optimal allocation strategy includes obtaining the location of overdue customer phone numbers; Obtain the collection specialists whose hometown and place of origin are the same from the candidate collection specialist set, and form a collection specialist set with the same place of origin; Acquire all services that are higher in importance than the transfer service in the service level ranking to form a priority service set; Acquire all services whose importance level is lower than that of the transfer service in the service level ranking to form a lagging service set; The collection specialist whose elements in the pending business set are all in the priority business set is recorded as the discarded collection specialist; The collection specialist whose elements in the pending business set are not in the priority business set is the first collection specialist; The collection specialist whose elements in the pending business set are in both the priority business set and the delayed business set is recorded as the second collection specialist; Traverse the pending business set of each element in the collection specialist set of the same place, obtain all the first collection specialists and the second collection specialists, and form the first collection set and the second collection set respectively; Set a limit value to limit the number of businesses that collection specialists can receive; Count the number of elements in the pending business set of each collection specialist in the first set to obtain the first set of processed business; Compare each element in the first handling number set with the limit value, obtain the collection specialists corresponding to all elements whose value is less than the limit value, and record them as the first preferred specialist set; Count the number of elements in the pending business set of each collection specialist in the second set of actions to be taken, and obtain a second set of processing numbers; Compare each element in the second handling number set with the limit value, obtain the collection specialists corresponding to all elements whose value is less than the limit value, and record them as the second preferred specialist set; Set the initial order, which is the order of all pending transactions when the transaction is first assigned to a collection specialist; Set the execution order, which is the order in which the collection specialists place the business in the pending business when they are working; Set the number of judgments to determine whether the collection specialist is qualified to receive overdue customers; Obtain historical data of the execution order of the determination quantity, recorded as a historical execution order set; Compare all elements in the historical execution order set with the initial order: When the initial order is less than or equal to all elements in the historical execution order set, the collection specialist to which the business belongs is recorded as the marked specialist; Traverse the business set to be processed of each element of the first preferred specialist set and the second preferred specialist set, obtain all marked specialists, form discarded specialists and remove them, and obtain the first target specialist set and the second target specialist set; Obtain the highest professional ranking of the collection specialist who transfers the business among all the collection specialists who respond, and record him as the responding collection specialist; When the responding collection specialist is an element in the first target specialist set, the overdue customer transfer business is assigned to the responding collection specialist, and the initial order of the transfer business is first; When the responding collection specialist is an element in the second target specialist set; obtain the elements in the pending business set of the responding collection specialist that are more important than the transferred business, and record them as the response priority business set; obtain the largest execution order of the response priority business set, and record them as the connection order x; assign the transferred business of the overdue customer to the responding collection specialist, and the initial order of the transferred business is x+1.

2. The outbound calling method based on artificial intelligence according to claim 1, characterized in that: The optimal time period strategy includes: Set the first time, which refers to Monday to Friday; Set the second time, which is from Saturday to Sunday; The total number of first-time answering times in the historical call records of overdue customers is counted and recorded as the number of first-time answering times; The total number of second-time answering calls in the historical call records of overdue customers is counted and recorded as the number of second-time answering calls; When the number of first-time answers is greater than the number of second-time answers, the overdue customers are classified as the first group; When the number of calls answered at the first time is less than or equal to the number of calls answered at the second time, the overdue customers are classified as the second group of people; Set the number of filters for selecting the calling time period; Divide 24 hours a day into e time periods, count the number of times overdue customers answer calls in each time period in the historical call records, obtain the time periods with a number greater than the screening number, and form a call time period set; When the overdue customer belongs to the first group of people, the overdue customer is called at the first time in the time period included in the calling time period set every day; When the overdue customer belongs to the second group of people, a call is made to the overdue customer during a time period included in the dialing time period set every day at the second time.

3. The outbound calling method based on artificial intelligence according to claim 1, characterized in that: The opening format optimization strategy includes: Set the threshold time for judging whether the opening format words are retained; Get any element in the opening format set and record it as the marked opening format; The length of the call when the opening format is marked is recorded as the marked time; Obtain all marked times in the historical call records of overdue customers to form a marked time set; Traverse all elements in the marked time set, obtain the number of marked times greater than the threshold time, and record it as the effective number; Calculate the number of elements in the marked time set and record it as the number of dialing times; Calculate the effective number of times divided by the number of dialing times, and record the result as the effective rate; Set the judgment ratio for whether to keep the opening format; When the effective rate is less than or equal to the judgment ratio, the marked opening format words are removed from the opening format set; Traverse all elements in the opening format set to execute the opening format optimization strategy and update the opening format set.

4. The outbound calling method based on artificial intelligence according to claim 1, characterized in that: The professional scoring strategy adopted includes: Get all the businesses and form a business set; Get any element in the service set and record it as the positioning service; Obtain the scoring data of each collection specialist in the positioning business in the historical reception data, calculate the average of multiple scoring data of each collection specialist, and obtain the average score of each collection specialist in the positioning business; The average scores of all collection specialists in the location service are sorted from high to low to obtain the professional ranking of the location service; Traverse all elements in the business set, adopt professional scoring strategy, and obtain the professional ranking of each element in the business set.

5. The outbound calling method based on artificial intelligence according to claim 1 is characterized in that: The collection specialist who responds to overdue customers is called the responsive collection specialist.

6. The outbound calling method based on artificial intelligence according to claim 1, characterized in that: The execution of the transfer judgment strategy includes: Extract keywords from the conversation records to form a keyword set; Get the number of elements in the keyword set, and record the result as the number of keywords a; Distribute the conversation records to all collection agents; Obtain the characteristic words extracted from the conversation records of each debt collection specialist; A feature matching strategy is adopted for each collection specialist’s feature words, including: Obtain the characteristic words of any collection specialist and record them as the positioning characteristic word set; Compare the elements in the positioning feature word set and the keyword set; Obtaining the same elements in the positioning feature word set and the keyword set to form a matching feature set; Calculate the number of elements in the matching feature set, and record the result as the matching number b; Calculate the number of matches ÷ the number of keywords, and record the result as the match rate c, c=b÷a; When the matching rate is equal to 1, the collection specialist corresponding to the positioning feature word set is recorded as the matching collection specialist; Get all matching debt collectors to form a candidate debt collector set.

7. The outbound calling method based on artificial intelligence according to claim 1, characterized in that: The conversation record includes: First, convert the voice input of the debt collection customer into input text, including: Audio collection: obtaining the voice input by the debt collection customer, converting the voice into simulated continuous sound waves and recording them as voice signals; Preprocessing, obtaining a speech signal, performing pre-emphasis, framing and windowing operations, and recording the result after preprocessing as a speech frame set; Extract features, obtain a speech frame set, extract feature parameters of each element in the speech frame set, and form a feature parameter set; Apply the acoustic model, obtain a feature parameter set, traverse each element in the feature parameter set, apply the acoustic model to each element, obtain the phoneme corresponding to each speech frame, and form a sound result set; Applying the language model to obtain a speech frame set, applying the language model to obtain multiple text sequences corresponding to the speech frame set to form a text result set; Search and decode to obtain the sound result set and the text result set, apply the decoding search algorithm, and record the obtained text sequence as the input text; Secondly, obtain the tone, expression, and speech speed information of the debt collection customer and record it as implicit information; Obtain input text, apply natural language processing algorithms to the input text and implicit information, obtain the emotions and repayment willingness of the collection customers, and record them as communication features; Finally, the input text and communication features are obtained, and the reply content is matched and recorded as the output text; Perform language conversion operations on the output text, including: Standardize text, remove punctuation and format numbers from the output text, and record the result as standard text; Prosody modeling: obtain standard text, analyze the intonation, rhythm and emphasis of the standard text, and record the results as text features; Apply the acoustic model to convert the standard text into a phoneme set, and then convert the phoneme set into a sound wave signal, and the result is recorded as the output sound wave; Synthesize language, obtain output sound waves and text features, apply sound synthesis technology, and obtain speech output.

Images