Data evaluation method and system based on combination of double dictionary trees and AC automaton

Through the combination of the double dictionary tree and AC automaton, the disable and negative words in the call are quickly identified, which solves the problem of low recognition efficiency in the existing technology, realizes a comprehensive evaluation of customer service quality, supports flexible rules and high efficiency quality inspection improves customer satisfaction.

CN119089899APending Publication Date: 2024-12-06CHINA CITIC BANK CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411023525.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing technology cannot quickly and accurately identify disable words and negative words in calls, resulting in the inability to achieve timely and flexible evaluation of customer service quality and the inability to adapt to changing customer needs and service scenarios.

Method used

Using the data evaluation method based on the double dictionary tree combined with the AC automaton, the rapid recognition of disable words and negative words is achieved by establishing a predefined word set, configuring the double dictionary tree, and constructing the state transfer relationship and output table of the AC automaton. Flexible evaluation of service quality.

Benefits of technology

It improves quality inspection efficiency, realizes comprehensive quality inspection of customer service call content, supports flexible service quality evaluation rules, can be customized and adjusted according to actual needs, significantly improving customer satisfaction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089899A_ABST
    Figure CN119089899A_ABST
Patent Text Reader

Abstract

The invention relates to a data evaluation method and system based on double dictionary trees combined with an AC automaton, and the method comprises the steps: building a first word set, configuring a first double dictionary tree, recognizing a node association path in the first double dictionary tree, and determining a mismatch pointer through the combination with the judgment of a transfer condition, thereby carrying out the targeted matching evaluation operation of to-be-evaluated data, and achieving the evaluation of the to-be-evaluated data. Through combination of the double-dictionary tree and the AC automaton, rapid identification of forbidden words and negative words is realized, and quality inspection efficiency is improved, so that comprehensive quality inspection of customer service call content is realized, limitation of sampling inspection is avoided, flexible service quality evaluation rules can be supported, customization and adjustment can be performed according to actual demands, and user experience is improved. And great help is provided for improving customer service and improving customer satisfaction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of automated text data analysis and data processing, and in particular to a data evaluation method and system based on a dual dictionary tree combined with an AC automaton. Background Art

[0002] With the development of call center customer service business, manual service data can be obtained in real time. However, in the face of ever-increasing service scenarios and ever-changing customer needs, the quality inspection dimensions have increased and the workload is heavy. In the traditional manual processing method, only certain calls can be sampled, and all calls cannot be quality inspected. In particular, the large number of banned words and negative words in the calls cannot be quickly identified, and the quality of the service cannot be effectively evaluated.

[0003] In order to improve the shortcomings of traditional manual information quality inspection methods, the existing technology has made various attempts to improve through technical means. For example, firstly, voice recognition technology is used to convert the customer service call content into text data, and then automatic semantic recognition is used to assist in customer service data quality inspection, and the text data obtained by recognition can be automatically matched and tested using preset matching indicators to evaluate customer service quality.

[0004] However, when processing large amounts of customer service data, existing technologies are unable to adapt to ever-changing customer needs and service scenarios. In particular, they are unable to quickly and accurately identify banned words, negative words, etc. involved in customer service, and are unable to flexibly respond to current service scenarios and timely evaluate the quality of each service. Summary of the invention

[0005] In order to address the deficiencies of the prior art, the present invention proposes a data evaluation method and system based on a dual dictionary tree combined with an AC automaton. Through the combination of the dual dictionary tree and the AC automaton, rapid identification of banned words and negative words is achieved, and quality inspection efficiency is improved, thereby achieving comprehensive quality inspection of customer service call content, avoiding the limitations of random inspections, and supporting flexible service quality evaluation rules, which can be customized and adjusted according to actual needs, and are of great help in helping to improve customer service and enhance customer satisfaction.

[0006] To achieve the above objectives, the technical solutions adopted by the present invention include:

[0007] A data evaluation method based on dual dictionary trees combined with AC automata, characterized by comprising:

[0008] S1. Establish a first word set, wherein the first word set includes predefined banned words and / or negative words;

[0009] S2, configuring a first dual dictionary tree, and saving the first word set to the first dual dictionary tree in a dual dictionary tree data structure;

[0010] S3, identifying the node association path in the first dual dictionary tree, and adding association pointers to each node having the same prefix node according to the prefix node corresponding to the node;

[0011] S4, constructing a state transition relationship corresponding to the AC automaton according to the first double dictionary tree;

[0012] S5. judging whether the associated pointer satisfies the transfer condition according to the state transfer relationship, and marking the associated pointer as a mismatched pointer when it is judged that the associated pointer satisfies the transfer condition;

[0013] S6, establishing an output table according to the mismatch pointer, wherein the output table at least includes output information corresponding to the first word set;

[0014] S7. Use the AC automaton containing the output table to perform data evaluation operations on the data to be evaluated and generate evaluation results.

[0015] Furthermore, the data to be evaluated includes text data generated by performing a speech recognition operation on the collected customer service call voice data.

[0016] Further, the using of the AC automaton with the output table to perform the data evaluation operation on the data to be evaluated includes:

[0017] Input the data to be evaluated into the AC automaton and split it into characters to be recognized;

[0018] According to the mismatch pointers contained in the output table, the characters to be recognized are matched and correlated to determine whether the data to be evaluated contains the output information defined in the output table;

[0019] When it is determined that the data to be evaluated contains the output information defined in the output table, the occurrence position of the output information in the data to be evaluated is marked and the weight of the output information is calculated;

[0020] The output information and the occurrence position and weight of the corresponding output information are integrated to generate the evaluation result.

[0021] Furthermore, the method further comprises:

[0022] Based on the evaluation results, a service quality score is generated for the evaluation information using the preset scoring rules.

[0023] Furthermore, the method further comprises:

[0024] A preprocessing operation is performed on the text data to generate data to be evaluated, wherein the preprocessing operation includes any one or more combinations of a text segmentation operation, a text deduplication operation, and a missing text association supplement operation.

[0025] Further, the identifying the node association path in the first dual dictionary tree includes:

[0026] The conductive path from the root node to the leaf node is labeled according to the first word set.

[0027] The present invention also relates to a data evaluation system based on a dual dictionary tree combined with an AC automaton, characterized in that it comprises:

[0028] A word set management module, used for establishing and modifying a first word set;

[0029] A dual dictionary tree management module, used for configuring a first dual dictionary tree, and saving the first word set to the first dual dictionary tree in a dual dictionary tree data structure;

[0030] An association pointer identification module is used to identify the node association path in the first dual dictionary tree, and add association pointers to each node having the same prefix node according to the prefix node corresponding to the node;

[0031] A mismatch pointer marking module is used to construct a state transition relationship of the corresponding AC automaton according to the first double dictionary tree, and judge whether the associated pointer meets the transition condition according to the state transition relationship, and mark the associated pointer as a mismatch pointer when it is judged that the associated pointer meets the transition condition;

[0032] An output table generation module, used for establishing an output table according to the mismatch pointer;

[0033] The evaluation execution module is used to use an AC automaton containing an output table to perform data evaluation operations on the data to be evaluated and generate evaluation results.

[0034] The present invention also relates to a computer-readable storage medium, characterized in that a computer program is stored on the storage medium, and the computer program implements the above method when executed by a processor.

[0035] The present invention also relates to an electronic device, characterized in that it comprises a processor and a memory;

[0036] The memory is used to store the first word set, the first dual dictionary tree and the output table;

[0037] The processor is used to execute the above method by calling the first word set, the first dual dictionary tree and the output table.

[0038] The present invention also relates to a computer program product, comprising a computer program and / or instructions, characterized in that the computer program and / or instructions implement the steps of the above method when executed by a processor.

[0039] The beneficial effects of the present invention are:

[0040] The data evaluation method and system based on the double dictionary tree combined with the AC automaton described in the present invention reduces the complexity required for detection and matching through the combination of the double dictionary tree and the AC automaton, improves the efficiency of the matching algorithm, is suitable for processing large-scale data, realizes rapid identification of banned words and negative words, and improves quality inspection efficiency, thereby realizing comprehensive quality inspection of customer service call content, avoiding the limitations of random inspections, and can support flexible service quality evaluation rules, which can be customized and adjusted according to actual needs, and is of great help in helping to improve customer service and enhance customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The figure is a flow chart of the data evaluation method based on the dual dictionary tree combined with the AC automaton of the present invention.

[0042] Figure 2 It is a structural schematic diagram of the data evaluation system based on the dual dictionary tree combined with the AC automaton of the present invention.

[0043] Figure 3 Schematic diagram of the dictionary tree of the preferred embodiment of the present invention.

[0044] Figure 4 Schematic diagram of state transition relationship of a preferred embodiment of the present invention.

[0045] Figure 5 Output a representation diagram for a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to more clearly understand the content of the present invention, it will be described in detail with reference to the accompanying drawings and embodiments. It should be noted that in the embodiments of the present application, some existing solutions in the industry such as certain software, components, models, etc. may be mentioned, which should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0047] The acquisition, transmission, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0048] In order to achieve the purpose of full-scale rapid quality inspection of customer service speech, the present invention introduces dual dictionary trees and AC automata into the call center service quality evaluation system. At the same time, because Chinese is profound and extensive, any word composed of a certain character is likely to be a banned word or a negative word. Therefore, when matching these characters during pattern matching, the weight is directly identified according to the probability, without further determining which banned words the character matches. In this way, after adding weights to the banned characters, the score items are directly output in the quality inspection link, saving matching time and improving efficiency.

[0049] The AC automaton algorithm was first invented by Bell Labs in 1975. The main idea is to preprocess the pattern string into a deterministic finite state automaton, which will perform pattern string matching based on the input. 2 ), reduced to O(n), which is independent of the length and number of pattern strings. In the prior art, it is generally believed that AC automata are implemented in combination with non-compressed dictionary trees.

[0050] The first aspect of the present invention relates to a process of steps as follows Figure 1 The data evaluation method based on the dual dictionary tree combined with the AC automaton shown includes:

[0051] S1. Establish a first word set, wherein the first word set includes predefined banned words and / or negative words.

[0052] In the specific implementation, the banned word and negative word corpus corresponding to the customer service call process is used as input.

[0053] S2. Configure a first dual dictionary tree, and save the first word set in the first dual dictionary tree in a dual dictionary tree data structure.

[0054] In the specific implementation, the typical dual dictionary tree preferred embodiment structure is as follows Figure 3 As shown, the constructed double dictionary tree can start searching from the root node, select the corresponding child node after obtaining the keyword, and finally find the required information, which corresponds to the success state transfer in the AC automaton algorithm.

[0055] S3, identifying the node association path in the first dual dictionary tree, and adding association pointers to each node with the same prefix node according to the prefix node corresponding to the node. Wherein, identifying the node association path in the first dual dictionary tree includes: marking the conduction path from the root node to the leaf node according to the first word set.

[0056] S4. Construct a state transition relationship corresponding to the AC automaton based on the first double dictionary tree.

[0057] S5. Determine whether the associated pointer satisfies the transfer condition according to the state transfer relationship. When it is determined that the associated pointer satisfies the transfer condition, mark the associated pointer as a mismatch pointer.

[0058] In the specific implementation, the mismatch pointer corresponds to the failure state in the AC automaton algorithm. At this time, if the string matching fails, it jumps to a specific node, and the node from the root node is exactly the part of the failed text. Figure 4 Shown is the corresponding Figure 3 A schematic diagram of the state transition relationship of a preferred embodiment of a dual dictionary tree is shown.

[0059] S6. Establish an output table according to the mismatch pointer, wherein the output table at least includes output information corresponding to the first word set.

[0060] In the specific implementation, corresponding to the above Figure 3 The preferred embodiment of the dual dictionary tree shown and Figure 4 The state transition relationship diagram shown in the figure can be obtained as follows Figure 5 In practical applications, it is preferred to use a depth-first algorithm to construct the output table.

[0061] S7, using an AC automaton containing an output table to perform a data evaluation operation on the data to be evaluated, and generate an evaluation result. Preferably, performing the data evaluation operation includes: inputting the data to be evaluated into the AC automaton and splitting it into characters to be recognized; performing association matching on the characters to be recognized according to the mismatch pointers contained in the output table, and judging whether the data to be evaluated contains the output information defined in the output table; when judging that the data to be evaluated contains the output information defined in the output table, marking the occurrence position of the output information in the data to be evaluated and calculating the weight of the output information; integrating the output information and the occurrence position and weight of the corresponding output information, and generating an evaluation result.

[0062] Preferably, the data to be evaluated includes text data generated by performing a speech recognition operation on the collected customer service call voice data, and in particular also includes performing a preprocessing operation on the text data to generate the data to be evaluated, wherein the preprocessing operation includes any one or more combinations of text segmentation operations, text deduplication operations, and missing text association and supplement operations.

[0063] Further preferably, the method of the present invention may also include generating a service quality score for the information to be evaluated using a preset scoring rule according to the evaluation result.

[0064] In the preferred embodiment, when trying to match ashehis, we start from 0, 0->7->8->9 to match the first she, then we go back to 0, and then from 0->1->2->3 to his. Here we have successfully reduced the complexity from O(n 2 ), reduced to O(n), and we can see that after the dictionary tree and mismatch pointer are created, no matter how long the text length and pattern matching string are, we can always get the matching result in O(n) complexity. Then assume that h is a banned word, and the words with h are marked as banned words. When h is matched, the match is returned directly.

[0065] In specific implementation, when the call recording files of the agents are translated into text through ASR, during semantic analysis, they are identified through AC automata, and all banned words and negative words can be marked to evaluate the service quality. Compared with other matching algorithms, the system that introduces dual dictionary trees and AC automata can match all words in linear time without increasing the time complexity due to the increase in the number of pattern strings.

[0066] According to actual test results, this method is 5 to 10 times faster than the traditional sequential matching method, making it possible to match large-scale text. In call centers, this method can be used to match banned words, negative words, modal particles, business content, etc. on a large scale in the corpus of customer service calls, so as to conduct a refined assessment of customer service status, without relying on traditional sampling methods.

[0067] Another aspect of the present invention also relates to a data evaluation system based on a dual dictionary tree combined with an AC automaton, the structure of which is as follows: Figure 2 As shown, including:

[0068] A word set management module, used for establishing and modifying a first word set;

[0069] A dual dictionary tree management module, used for configuring a first dual dictionary tree, and saving the first word set to the first dual dictionary tree in a dual dictionary tree data structure;

[0070] An association pointer identification module is used to identify the node association path in the first dual dictionary tree, and add association pointers to each node having the same prefix node according to the prefix node corresponding to the node;

[0071] A mismatch pointer marking module is used to construct a state transition relationship of the corresponding AC automaton according to the first double dictionary tree, and judge whether the associated pointer meets the transition condition according to the state transition relationship, and mark the associated pointer as a mismatch pointer when it is judged that the associated pointer meets the transition condition;

[0072] An output table generation module, used for establishing an output table according to the mismatch pointer;

[0073] The evaluation execution module is used to use an AC automaton containing an output table to perform data evaluation operations on the data to be evaluated and generate evaluation results.

[0074] By using this system, the above-mentioned operation processing method can be executed and the corresponding technical effects can be achieved.

[0075] An embodiment of the present invention also provides a computer-readable storage medium capable of implementing all the steps in the method in the above embodiment, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, all the steps in the method in the above embodiment are implemented.

[0076] An embodiment of the present invention also provides an electronic device for executing the above method. As an implementation device of the method, the electronic device at least has a processor and a memory, and in particular, the memory stores data and related computer programs required for executing the method, such as a first word set, a first dual dictionary tree and an output table, etc., and the processor calls the data and programs in the memory to execute all the steps of the implementation method and obtains the corresponding technical effect.

[0077] Preferably, the electronic device may include a bus architecture, the bus may include any number of interconnected buses and bridges, and the bus will include various circuits linked together by one or more processors and memories. The bus may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are all well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the receiver and transmitter. The receiver and transmitter may be the same component, i.e., a transceiver, providing a unit for communicating with various other systems over a transmission medium. The processor is responsible for managing the bus and general processing, while the memory may be used to store data used by the processor when performing operations.

[0078] In addition, the electronic device may further include components such as a communication module, an input unit, an audio processor, a display, and a power supply. The processor (or controller, operation control) used may include a microprocessor or other processor device and / or a logic device, which receives input and controls the operation of various components of the electronic device; the memory may be one or more of a cache, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory, or other suitable devices, which may store the above-mentioned related data information, and may also store a program for executing related information, and the processor may execute the program stored in the memory to realize information storage or processing, etc.; the input unit is used to provide input to the processor, for example, it may be a key or a touch input device; the power supply is used to provide power to the electronic device; the display is used to display display objects such as images and text, for example, it may be an LCD display. The communication module is a transmitter / receiver that sends and receives signals via an antenna. The communication module (transmitter / receiver) is coupled to the processor to provide input signals and receive output signals, which may be the same as the case of a conventional mobile communication terminal. Based on different communication technologies, multiple communication modules may be provided in the same electronic device, such as a cellular network module, a Bluetooth module and / or a wireless local area network module, etc. The communication module (transmitter / receiver) is also coupled to a speaker and a microphone via an audio processor to provide an audio output via the speaker and receive an audio input from the microphone, thereby realizing a common telecommunication function. The audio processor may include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor is also coupled to a central processor, so that recording can be performed on the machine through the microphone, and the sound stored on the machine can be played through the speaker.

[0079] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0080] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or multiple boxes.

[0081] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0082] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps of the functions specified in a box or multiple boxes. Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0083] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A data evaluation method based on dual dictionary trees combined with AC automata, characterized in that: include: S1. Establish a first word set, wherein the first word set includes predefined banned words and / or negative words; S2, configuring a first dual dictionary tree, and saving the first word set to the first dual dictionary tree in a dual dictionary tree data structure; S3, identifying the node association path in the first dual dictionary tree, and adding association pointers to each node having the same prefix node according to the prefix node corresponding to the node; S4, constructing a state transition relationship corresponding to the AC automaton according to the first double dictionary tree; S5. judging whether the associated pointer satisfies the transfer condition according to the state transfer relationship, and marking the associated pointer as a mismatched pointer when it is judged that the associated pointer satisfies the transfer condition; S6, establishing an output table according to the mismatch pointer, wherein the output table at least includes output information corresponding to the first word set; S7. Use the AC automaton containing the output table to perform data evaluation operations on the data to be evaluated and generate evaluation results.

2. The method according to claim 1, characterized in that The data to be evaluated includes text data generated by performing a speech recognition operation on the collected customer service call voice data.

3. The method according to claim 1, characterized in that The use of an AC automaton with an output table to perform a data evaluation operation on the data to be evaluated comprises: Input the data to be evaluated into the AC automaton and split it into characters to be recognized; According to the mismatch pointers contained in the output table, the characters to be recognized are matched and correlated to determine whether the data to be evaluated contains the output information defined in the output table; When it is determined that the data to be evaluated contains the output information defined in the output table, the occurrence position of the output information in the data to be evaluated is marked and the weight of the output information is calculated; The output information and the occurrence position and weight of the corresponding output information are integrated to generate the evaluation result.

4. The method according to claim 3, characterized in that The method further comprises: Based on the evaluation results, a service quality score is generated for the evaluation information using the preset scoring rules.

5. The method according to claim 2, characterized in that The method further comprises: A preprocessing operation is performed on the text data to generate data to be evaluated, wherein the preprocessing operation includes any one or more combinations of a text segmentation operation, a text deduplication operation, and a missing text association supplement operation.

6. The method according to claim 1, characterized in that The identifying of the node association path in the first dual dictionary tree comprises: The conductive path from the root node to the leaf node is labeled according to the first word set.

7. A data evaluation system based on dual dictionary trees combined with AC automata, characterized in that: include: A word set management module, used for establishing and modifying a first word set; A dual dictionary tree management module, used for configuring a first dual dictionary tree, and saving the first word set to the first dual dictionary tree in a dual dictionary tree data structure; An association pointer identification module is used to identify the node association path in the first dual dictionary tree, and add association pointers to each node having the same prefix node according to the prefix node corresponding to the node; A mismatch pointer marking module is used to construct a state transition relationship of the corresponding AC automaton according to the first double dictionary tree, and judge whether the associated pointer meets the transition condition according to the state transition relationship, and mark the associated pointer as a mismatch pointer when it is judged that the associated pointer meets the transition condition; An output table generation module, used for establishing an output table according to the mismatch pointer; The evaluation execution module is used to use an AC automaton containing an output table to perform data evaluation operations on the data to be evaluated and generate evaluation results.

8. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

9. An electronic device, characterized in that: including a processor and a memory; The memory is used to store the first word set, the first dual dictionary tree and the output table; The processor is used to execute the method according to any one of claims 1 to 6 by calling the first word set, the first dual dictionary tree and the output table.

10. A computer program product comprising a computer program and / or instructions, characterized in that: When the computer program and / or the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Citation Information

Cited By

  • Standard term speech recognition system and method based on double AC automaton

    CN120148482A