Response auxiliary recommendation method based on semantic analysis

Abstract

The invention provides an auxiliary response recommendation method based on semantic analysis, and belongs to the technical field of voice recognition, and the method specifically comprises the steps: determining associated users of a user through a power distribution node where the user is located, carrying out the determination of the matching degree between different associated users and the user through the power utilization problem data of the different associated users and the power distribution node, and carrying out the recommendation of the user. Determining the processing result credibility of the prediction processing result of the user according to the matching degree and the problem processing results of different associated users; the historical processing matching degree of different customer service personnel is determined based on the power consumption problem data of the users and the historical processing data of the different customer service personnel, and the customer service personnel allocated to the users are determined in combination with queuing personnel of the different customer service personnel, so that differentiated processing of different types of users is further realized; and the processing efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of speech recognition, and in particular relates to an answer auxiliary recommendation method based on semantic analysis. Background Art

[0002] In order to solve customers' electricity problems, power companies have set up fixed service hotlines. However, since different customers are located in different grid nodes and their electricity problems also vary to a certain extent, how to allocate and process customer service according to customer needs has become a technical problem that needs to be solved urgently.

[0003] In order to solve the above technical problems, the existing technical solution in the invention patent CN202211381857.3 "A diversion system based on voice customer service" converts the received consultation content into text information, adjusts the transfer method according to the text information, and allocates the demand request to the customer service end with the highest efficiency in processing this type of demand request, and accurately classifies the demand request. However, there are the following technical problems: Different user types have different requirements for electricity reliability. For example, hospital, commercial or residential loads have different requirements for electricity reliability. Therefore, if differentiated customer allocation cannot be determined based on the identification results of different user types, the electricity reliability of users with higher electricity reliability requirements cannot be guaranteed.

[0004] The existing technical solutions ignore the evaluation of the similarity with historical users based on the user's description and the user's power supply area. Specifically, when the problems of users in a certain power supply area are similar, they are often caused by the same reasons. Therefore, the processing opinions of other users are often also applicable to the current user. Therefore, if the users cannot be screened based on the above factors, the processing efficiency will be difficult to meet the requirements.

[0005] In response to the above technical problems, the present invention provides an answer auxiliary recommendation method based on semantic analysis. Summary of the invention

[0006] To achieve the purpose of the present invention, the present invention adopts the following technical solutions: According to one aspect of the present invention, a method for auxiliary recommendation of answers based on semantic analysis is provided.

[0007] A method for assisting recommendation of answers based on semantic analysis, characterized by comprising: S1 determines the user's electricity usage type based on the user's identification result, determines the user's electricity usage reliability requirement and the core electricity users through the user's electricity usage type and historical electricity usage data, and proceeds to the next step when the user does not belong to the core electricity users; S2 determines the semantic analysis result and power usage problem data of the user's voice, determines the associated user of the user through the power distribution node where the user is located, and determines whether there is complaint data of the associated user. If yes, proceed to the next step, if not, proceed to step S4; S3 uses the electricity problem data of different associated users and the power distribution nodes to determine the matching degree between different associated users and the user, and determines the processing result credibility of the predicted processing result of the user through the matching degree and the problem processing results of different associated users. When the processing result credibility does not meet the requirements, proceed to the next step; S4 determines the historical processing matching degree of different customer service personnel based on the user's electricity problem data and the historical processing data of different customer service personnel, and determines the customer service personnel assigned to the user in combination with the queue personnel of different customer service personnel.

[0008] The beneficial effects of the present invention are: 1. In the present invention, the reliability requirements of electricity consumption of users and the determination of core electricity users are carried out through the users' electricity consumption types and historical electricity consumption data. This not only takes into account the differences in the reliability requirements of electricity consumption of different users' electricity consumption types, but also comprehensively considers the historical electricity consumption and different historical electricity idle conditions to achieve the determination of core electricity users, and lays a foundation for differentiated processing according to the differences in users' electricity consumption reliability requirements.

[0009] 2. In the present invention, the credibility of the predicted processing result of the user is determined based on the matching degree and the problem processing results of different related users, avoiding the technical problem of low answer accuracy caused by automatic answering relying solely on manual or matching of the existing problem processing database. Taking into account the impact of the differences in problem processing results and the differences in node matching degrees on the credibility of the predicted processing results, it also lays the foundation for differentiated implementation of predictive processing for users.

[0010] 3. In the present invention, the customer service personnel to be assigned to the user are determined by combining the historical processing matching degree and the queue personnel of different customer service personnel, thereby avoiding the technical problem of inaccurate allocation caused by single consideration of matching degree or queue time, which not only ensures the accuracy of the user's answer result, but also avoids the technical problem of poor user experience caused by long queues, thereby achieving a balanced processing of the accuracy of the answer result and the queue time.

[0011] A further technical solution is that the user's electricity usage type is determined based on the recognition result of the user's incoming call number. Specifically, the user's incoming call type is determined by matching the recognition result of the user's incoming call number with a preset number library.

[0012] A further technical solution is that the electricity usage type of the user is determined according to the load type of the user.

[0013] A further technical solution is that the historical electricity consumption data includes historical electricity consumption within different unit time lengths, idle time lengths, and historical electricity consumption in different time periods.

[0014] A further technical solution is that, when the user is a core electricity user, the customer service personnel assigned to the user are determined according to the user type of the user.

[0015] A further technical solution is that the electricity usage problem data includes the electricity usage problem type of the user and a matching problem time period of the electricity usage problem type.

[0016] A further technical solution is to determine the associated users of the user through the power distribution node where the user is located, specifically including: The associated transformer and distribution line of the user are determined according to the distribution node where the user is located, and the associated users of the user are determined based on the associated transformer and distribution line of the user.

[0017] A further technical solution is to determine the customer service personnel assigned to the user, specifically including: Determine the estimated queuing time of different customer service personnel based on the queue personnel of different customer service personnel, and determine the comprehensive evaluation amount of different customer service personnel in combination with the historical matching degree; The customer service personnel assigned to the user is determined based on the comprehensive evaluation amount.

[0018] On the other hand, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, the above-mentioned answer auxiliary recommendation method based on semantic parsing is executed.

[0019] Other features and advantages will be described in the following description, and partly become apparent from the description, or understood by practicing the invention. The purpose and other advantages of the invention are realized and obtained by the structures particularly pointed out in the description and the drawings.

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

[0022] Figure 1 It is a flow chart of an answer assistance recommendation method based on semantic parsing; Figure 2 It is a flow chart of a method for determining a user's power reliability requirement. DETAILED DESCRIPTION

[0023] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that the present invention will be comprehensive and complete and fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar structures, and thus their detailed description will be omitted.

[0024] The terms "a", "an", "the", and "said" are used to indicate the presence of one or more elements / components / etc.; the terms "comprising" and "having" are used to express an open-ended inclusive meaning and mean that additional elements / components / etc. may be present in addition to the listed elements / components / etc.

[0025] Example 1 To solve the above problems, according to one aspect of the present invention, Figure 1 As shown, according to one aspect of the present invention, a method for assisting recommendation of answers based on semantic parsing is provided, which is characterized by specifically comprising: S1 determines the user's electricity usage type based on the user's identification result, determines the user's electricity usage reliability requirement and the core electricity users through the user's electricity usage type and historical electricity usage data, and proceeds to the next step when the user does not belong to the core electricity users; Furthermore, the user's electricity usage type is determined based on the recognition result of the user's incoming call number. Specifically, the user's incoming call type is determined by matching the recognition result of the user's incoming call number with a preset number library.

[0026] It should be noted that the electricity consumption type of the user is determined according to the load type of the user.

[0027] Specifically, the historical electricity consumption data includes historical electricity consumption within different unit time lengths, idle time lengths, and historical electricity consumption in different time periods.

[0028] In one possible embodiment, Figure 2 As shown, the method for determining the reliability requirement of the user's electricity consumption in the above step S1 is: S11: judging whether the user's electricity usage type belongs to a preset electricity usage type based on the user's electricity usage type, and if so, determining that the user is a core electricity user, and if not, proceeding to the next step; S12 determines the historical electricity consumption of the user on different dates through the historical electricity consumption data of the user, and determines whether the number of dates on which the historical electricity consumption of the user within a preset time is greater than a preset electricity consumption threshold is greater than a preset number. If so, the user is determined to be a core electricity user. If not, proceed to the next step. S13: taking the date when the historical power consumption is greater than the preset power consumption threshold as the peak power consumption date, and determining the power consumption evaluation value of the user on the peak power consumption date according to the number of peak power consumption dates of the user within the preset time and the historical power consumption of different peak power consumption dates, and determining whether the user is a core power user based on the power consumption evaluation value, if yes, determining the user to be a core power user, if not, proceeding to the next step; S14 obtains the historical electricity consumption of the user on different dates within a preset time and the idle time on different dates, and determines the reliability requirement of the user's electricity consumption in combination with the user's electricity consumption evaluation value on peak electricity consumption dates and the user's electricity consumption type.

[0029] It should be noted that, when the user is a core electricity user, the customer service personnel assigned to the user are determined according to the user type of the user.

[0030] In another possible embodiment, the method for determining the reliability requirement of the user's electricity consumption in the above step S1 is: When the electricity usage type of the user is a preset electricity usage type, the user is determined to be a core electricity user; When the electricity consumption type of the user is not a preset electricity consumption type, the historical electricity consumption of the user within a preset time is determined by using the historical electricity consumption data of the user; when the historical electricity consumption of the user within the preset time is within a preset electricity consumption range, the user is determined to be a core electricity user; When the historical power consumption of the user within the preset time is not within the preset power consumption range, the historical power consumption of different dates is obtained, and the date on which the historical power consumption is greater than the preset power consumption threshold is used as the peak power consumption date, and the power consumption evaluation value of the user on the peak power consumption date is determined according to the number of peak power consumption dates of the user within the preset time and the historical power consumption of different peak power consumption dates, and whether the user is a core power user is determined based on the power consumption evaluation value, and if so, the user is determined to be a core power user, and if not, proceed to the next step; The busyness of electricity consumption on different dates is determined based on the user's historical electricity consumption and idle time on different dates within a preset time. The reliability requirement of the user's electricity consumption is determined based on the user's historical electricity consumption within the preset time and the busyness of electricity consumption on different dates, the user's electricity consumption evaluation value on peak electricity consumption dates, and the user's electricity consumption type.

[0031] S2 determines the semantic analysis result and power usage problem data of the user's voice, determines the associated user of the user through the power distribution node where the user is located, and determines whether there is complaint data of the associated user. If yes, proceed to the next step, if not, proceed to step S4; It should be noted that the electricity usage problem data includes the electricity usage problem type of the user and the problem time period matching the electricity usage problem type.

[0032] Specifically, determining the associated user of the user through the power distribution node where the user is located specifically includes: The associated transformer and distribution line of the user are determined according to the distribution node where the user is located, and the associated users of the user are determined based on the associated transformer and distribution line of the user.

[0033] S3 uses the electricity problem data of different associated users and the power distribution nodes to determine the matching degree between different associated users and the user, and determines the processing result credibility of the predicted processing result of the user through the matching degree and the problem processing results of different associated users. When the processing result credibility does not meet the requirements, proceed to the next step; In one possible embodiment, the method for determining the matching degree between the associated user and the user in the above step S3 is: S21 determines the matching degree between the problem type of the user and the associated user through the power usage problem type of the user and the power usage problem type of the associated user, and judges whether the problem type matching degree is less than a preset matching degree. If not, proceeds to the next step. If yes, determines the matching degree between the associated user and the user through the problem type matching degree. S22: performing problem time period matching between the user and the associated user based on the problem time period matching the power usage problem type of the user, and determining the problem matching degree in combination with the problem type matching degree, and judging whether the problem matching degree is less than a preset matching degree threshold, if not, proceeding to the next step, and if so, determining the matching degree between the associated user and the user through the problem type matching degree; S23 determines the length of the distribution line between the associated user and the distribution node of the user and the number of electrical equipment between the distribution lines based on the associated user and the distribution node of the user, and determines the matching degree between the associated user and the user through the length of the distribution line, the number of electrical equipment between the distribution lines and the matching degree of the problem type.

[0034] In another possible embodiment, the method for determining the matching degree between the associated user and the user in the above step S3 is: Determining the matching degree between the problem types of the user and the associated users based on the power usage problem type of the user and the power usage problem type of the associated users; Based on the problem time period matching the type of the user's electricity problem, the problem time period matching between the user and the associated user is determined, and the problem matching is determined in combination with the problem type matching; The length of the distribution line between the associated user and the distribution node of the user and the number of electrical equipment between the distribution lines are determined based on the associated user and the distribution node of the user, and the matching degree between the associated user and the user is determined by the length of the distribution line, the number of electrical equipment between the distribution lines and the matching degree of the problem type.

[0035] In one possible embodiment, the method for determining the credibility of the processing result of the user's prediction processing result in the above step S3 is: S31 determines the number of associated users in different matching degree intervals according to the matching degrees of different associated users, and judges whether the number of associated users in the preset matching degree interval is greater than the preset number of users. If not, proceed to step S33. If yes, proceed to the next step. S32 determines the similarity of the processing results of the associated users within the preset matching degree interval based on the processing results of the associated users within the preset matching degree interval, and determines whether the number of associated users whose similarity of the processing results meets the requirement meets the requirement, if not, proceeds to the next step, if yes, determines that the processing result credibility of the predicted processing result of the user meets the requirement; S33 determines the user association within the preset matching degree interval according to the number of associated users within the preset matching degree interval, the matching degrees between different associated users and the user, and the similarities of the processing results of different associated users, and determines whether the user association within the preset matching degree interval is greater than the preset association degree. If so, it is determined that the processing result credibility of the predicted processing result of the user meets the requirement. If not, proceed to the next step. S34 determines the user associations within different matching degree intervals according to the number of associated users in different matching degree intervals, the matching degrees between different associated users and the user, and the similarities of the processing results of different associated users, and determines the processing result credibility of the predicted processing result of the user in combination with the user associations within the preset matching degree interval and the number of associated users.

[0036] In another possible embodiment, the method for determining the credibility of the processing result of the user's prediction processing result in the above step S3 is: Determine the number of associated users within different matching degree intervals by using the matching degrees of different associated users, and determine the comprehensive credibility of different associated users according to the matching degree between the associated user and the user, the similarity between the associated user and the processing results of other associated users, and the number of associated users whose similarity meets the requirements; Determine whether the number of associated users whose comprehensive credibility meets the requirement is greater than the preset number of associated users. If so, determine that the processing result credibility of the prediction processing result of the user meets the requirement. If not, proceed to the next step. Determine the user association within the preset matching degree interval by the number of associated users within the preset matching degree interval, the comprehensive credibility of different associated users, and the similarity of the processing results of different associated users, and judge whether the user association within the preset matching degree interval is greater than the preset correlation degree. If so, determine that the processing result credibility of the predicted processing result of the user meets the requirement; if not, proceed to the next step; The user association within the preset matching degree interval is determined according to the number of associated users in different matching degree intervals, the comprehensive credibility of different associated users, and the similarity of the processing results of different associated users, and the processing result credibility of the predicted processing result of the user is determined in combination with the number of associated users and the number of associated users whose comprehensive credibility meets the requirements.

[0037] It should be noted that when the credibility of the processing result meets the requirements, the credibility of different processing results is determined by the similarity of the processing results of the associated users and the credibility of different associated users, and the processing result of the user's problem is output through the processing result with the highest credibility.

[0038] S4 determines the historical processing matching degree of different customer service personnel based on the user's electricity problem data and the historical processing data of different customer service personnel, and determines the customer service personnel assigned to the user in combination with the queue personnel of different customer service personnel.

[0039] In one possible embodiment, the method for determining the historical processing matching degree of the customer service personnel in step S4 is: Determine the similarity between the problem of the user's electricity problem type and the duration of the electricity problem type and the historical processing data of the customer service personnel and the problem of the user, and determine the similarity of the problem of the user's historical complaint customers based on the problem similarity; Determine the node similarity between the user with similar historical complaints and the user according to the distribution nodes and regions where the user with similar historical complaints is located, and determine the comprehensive similarity between the user with similar historical complaints in combination with the problem similarity; The historical processing matching degree of the customer service personnel is determined based on the number of similar historical complaint users of the customer service personnel and the comprehensive similarity of different similar historical complaint users.

[0040] Specifically, determining the customer service personnel who allocates the user includes: Determine the estimated queuing time of different customer service personnel based on the queue personnel of different customer service personnel, and determine the comprehensive evaluation amount of different customer service personnel in combination with the historical matching degree; The customer service personnel assigned to the user is determined based on the comprehensive evaluation amount.

[0041] Example 2 On the other hand, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, the above-mentioned answer auxiliary recommendation method based on semantic parsing is executed.

[0042] Through the above embodiments, the present application achieves the following technical effects: 1. In the present invention, the reliability requirements of electricity consumption of users and the determination of core electricity users are carried out through the users' electricity consumption types and historical electricity consumption data. This not only takes into account the differences in the reliability requirements of electricity consumption of different users' electricity consumption types, but also comprehensively considers the historical electricity consumption and different historical electricity idle conditions to achieve the determination of core electricity users, and lays a foundation for differentiated processing according to the differences in users' electricity consumption reliability requirements.

[0043] 2. In the present invention, the credibility of the predicted processing result of the user is determined based on the matching degree and the problem processing results of different related users, avoiding the technical problem of low answer accuracy caused by automatic answering relying solely on manual or matching of the existing problem processing database. Taking into account the impact of the differences in problem processing results and the differences in node matching degrees on the credibility of the predicted processing results, it also lays the foundation for differentiated implementation of predictive processing for users.

[0044] 3. In the present invention, the customer service personnel to be assigned to the user are determined by combining the historical processing matching degree and the queue personnel of different customer service personnel, thereby avoiding the technical problem of inaccurate allocation caused by single consideration of matching degree or queue time, which not only ensures the accuracy of the user's answer result, but also avoids the technical problem of poor user experience caused by long queues, thereby achieving a balanced processing of the accuracy of the answer result and the queue time.

[0045] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0046] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0047] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A method for answer assistance recommendation based on semantic analysis, characterized in that: Specifically include: S1 determines the user's electricity usage type based on the user's identification result, determines the user's electricity usage reliability requirement and the core electricity users through the user's electricity usage type and historical electricity usage data, and proceeds to the next step when the user does not belong to the core electricity users; S2 determines the semantic analysis result and power usage problem data of the user's voice, determines the associated user of the user through the power distribution node where the user is located, and determines whether there is complaint data of the associated user. If yes, proceed to the next step, if not, proceed to step S4; S3 uses the electricity problem data of different associated users and the power distribution nodes to determine the matching degree between different associated users and the user, and determines the processing result credibility of the predicted processing result of the user through the matching degree and the problem processing results of different associated users. When the processing result credibility does not meet the requirements, proceed to the next step; S4 determines the historical processing matching degree of different customer service personnel based on the user's electricity problem data and the historical processing data of different customer service personnel, and determines the customer service personnel assigned to the user in combination with the queue personnel of different customer service personnel.

2. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: The user's electricity usage type is determined based on the recognition result of the user's incoming call number. Specifically, the user's incoming call type is determined based on the matching result of the recognition result of the user's incoming call number and a preset number library.

3. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: The electricity usage type of the user is determined according to the load type of the user.

4. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: The historical electricity consumption data includes historical electricity consumption within different unit time lengths, idle time lengths, and historical electricity consumption in different time periods.

5. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: The method for determining the reliability requirement of the user's electricity consumption is: Based on the electricity usage type of the user, determining whether the electricity usage type of the user belongs to a preset electricity usage type, if so, determining that the user is a core electricity user, if not, proceeding to the next step; Determine the historical electricity consumption of the user on different dates through the historical electricity consumption data of the user, and judge whether the number of dates on which the historical electricity consumption of the user within a preset time is greater than a preset electricity consumption threshold is greater than a preset number, if so, determine that the user is a core electricity user, if not, proceed to the next step; The date when the historical power consumption is greater than the preset power consumption threshold is taken as the peak power consumption date, and the power consumption evaluation value of the user on the peak power consumption date is determined according to the number of peak power consumption dates of the user within the preset time and the historical power consumption of different peak power consumption dates, and whether the user is a core power user is determined based on the power consumption evaluation value, if so, the user is determined to be a core power user, if not, proceed to the next step; The historical electricity consumption of the user on different dates within a preset time and the idle time on different dates are obtained, and the reliability requirement of the user's electricity consumption is determined in combination with the user's electricity consumption evaluation value on peak electricity consumption dates and the user's electricity consumption type.

6. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: When the user is a core electricity user, the customer service personnel assigned to the user are determined according to the user type of the user.

7. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: The power usage problem data includes the power usage problem type of the user and a matching problem time period of the power usage problem type.

8. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: Determining the associated user of the user by the power distribution node where the user is located specifically includes: The associated transformer and distribution line of the user are determined according to the distribution node where the user is located, and the associated users of the user are determined based on the associated transformer and distribution line of the user.

9. The answer auxiliary recommendation method based on semantic analysis according to claim 1, characterized in that: The method for determining the historical processing matching degree of the customer service personnel is: Determine the similarity between the problem of the user's electricity problem type and the duration of the electricity problem type and the historical processing data of the customer service personnel and the problem of the user, and determine the similarity of the problem of the user's historical complaint customers based on the problem similarity; Determine the node similarity between the user with similar historical complaints and the user according to the distribution nodes and regions where the user with similar historical complaints is located, and determine the comprehensive similarity between the user with similar historical complaints in combination with the problem similarity; The historical processing matching degree of the customer service personnel is determined based on the number of similar historical complaint users of the customer service personnel and the comprehensive similarity of different similar historical complaint users.

10. A computer system comprising: A memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, it executes an answer auxiliary recommendation method based on semantic parsing as described in any one of claims 1-9.

Images