A place name recognition method and system for intelligent dispatching of service hotlines
Through voice recognition and special positioning modules combined with database matching and manual intervention, the problem of inaccurate place name recognition in intelligent order dispatch of service hotlines is solved, accurate place name recognition and database update are achieved, and the accuracy of order dispatch of service hotlines is improved and efficiency.
Patent Information
- Application Number
- CN202311420374.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-10-30
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2043-10-30
AI Technical Summary
The existing technology is inaccurate in the intelligent order dispatch of service hotlines, resulting in errors in order dispatch, affecting the efficiency and accuracy of service hotlines. Especially when it involves linkage between multiple departments and administrative levels, it is difficult to establish an effective joint disposal mechanism. .
A system including speech recognition module, special positioning module and database is adopted. The recording is converted into speech text through the speech recognition module. The special positioning module specializes in the text and corrects errors, and matches the place names in the database. The matching will be identified by manual customer service and updated the database to ensure accurate identification of place names.
It realizes accurate identification of place names and database updates, improves the accuracy and efficiency of intelligent order dispatch of service hotlines, and ensures linkage needs for multiple departments and administrative levels.
Smart Images

Figure CN117609403B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of intelligent dispatching of service hotlines, and relates to a text recognition technology, and specifically is a place name recognition method and system applied to intelligent dispatching of service hotlines. Background Art
[0002] As the level of digitalization of social public services continues to improve, government services are beginning to move towards digitalization and intelligence. As a "bridge of communication" between the government and the public and a "sensor" for the government to perceive social conditions and public opinion, the 12345 government service hotline (hereinafter referred to as the "12345 hotline") plays an increasingly important role in promoting the transformation of government functions.
[0003] After the hotline merger, the public's demands have increased sharply, and the areas involved are also rich and diverse. If the traditional three-level distribution method is followed, in the process of distribution at each level, it is possible that the large amount of manual dispatch or lack of professionalism will lead to dispatch errors and affect the accuracy. The manual dispatch method not only requires high professionalism of personnel, but also increases the number of personnel. The 12345 hotline has a variety of business types and a wide range of scope. It needs to horizontally link multiple departments, such as the Human Resources and Social Security Bureau, the Tax Bureau, the Industry and Commerce Bureau, etc., and vertically link the city, county, township, and village units. If the dispatch cannot be reasonable, efficient and accurate, and an effective joint handling mechanism cannot be established, it will be difficult to fundamentally solve the public's demands.
[0004] The invention patent with announcement number CN115935245A discloses a method for automatic classification and allocation of government hotline cases, including: step S1, establishing a sample set and training a pre-trained model; step S2, traversing the sample set according to the trained pre-trained model, generating a case category related linked list and a department related linked list; step S3, using the pre-trained model to predict the case category and the competent department with the highest probability of the sample, and combining the case category related linked list and the department related linked list to obtain the case category enhancement vector and the department enhancement vector, and obtaining the case classification model and the department allocation model through training and parameter updating; step S4, obtaining the input case content, processing it through the case classification model and the department allocation model, and outputting the predicted case classification result and allocation department result;
[0005] The invention patent with announcement number CN114841128A discloses a business interaction method, device, equipment, medium and product based on artificial intelligence, which relates to the field of data processing technology, especially to the field of natural language processing (NLP). The specific implementation scheme is: if a form filling instruction is detected, the work order text entered in the form and the fields to be filled in are determined; the model is called to parse the work order text to obtain the fields included in the work order text and the attribute values corresponding to the fields; the fields are matched with the fields to be filled in, and the attribute values corresponding to the fields matching the fields to be filled in are filled in the input box corresponding to the fields to be filled in; the completed form is displayed. After detecting the form filling instruction, the present disclosure parses the work order text through the model to obtain the fields in the work order text and the attribute values corresponding to the fields, so as to complete the automatic filling of the form and improve the efficiency of form filling.
[0006] The patent with announcement number CN112581106A discloses a method for automatically dispatching government affairs events that integrates the semantics of the handling agency grid, including the following steps: S10. Acquisition of historical data sets of government affairs events: S20. Event data preprocessing and semantic feature extraction; S30. Construction of an event handling method prediction model: using the historical data set and the geographical location corresponding to the historical data as input, and the handling agency code as the final output, a deep neural network method is used to construct and train the prediction model; S40. Constructing an event handling agency prediction model; S50. Using the trained prediction model and the prediction model to make predictions, and output the corresponding event agency code that needs to handle the event.
[0007] The patent with publication number CN111353309A discloses a method and system for processing communication quality complaint addresses based on text analysis, including: performing Chinese address segmentation on the source text through the forward longest matching algorithm; the segmentation results are analyzed by multi-level administrative division address tree reasoning to obtain accurate multi-level place name recognition results; after receiving the input transmitted by the address tree, the rule-based multi-level address recognition algorithm automatically analyzes more complex and general multi-level address fields; the reasoning results of the address tree and the recognition results of the rule matching are integrated as the final output of the algorithm system. For the specific acceptance text content of the communication quality complaint work order, after Chinese word segmentation, regular expression address extraction and address tree reasoning processing are performed, the standard address of the communication quality complaint work order is automatically extracted.
[0008] The patent with announcement number CN115062128A discloses a method for automatically filling in service hotline work orders, including the following steps: S1: obtaining a corresponding knowledge entry library, obtaining the user's voice information, converting the voice information into a work order text based on voice recognition technology, and obtaining a corresponding knowledge entry library based on the work order text information; the present invention also proposes a service hotline work order automatic filling system, including a system module assembly, the system module assembly includes an acquisition module, a calculation module, a selection module, a storage module, a display module, and the display module is connected to the selection module, S2: calculating similarity, calculating the similarity between the work order text information and each item in the work order knowledge entry library, and obtaining an item similarity set. The present invention can recommend relevant knowledge items that match the work order information to hotline service personnel, and automatically complete the work order filling, thereby improving the efficiency and effect of the service hotline.
[0009] The above patents did not take into account the problem of place name recognition when performing intelligent dispatching of service hotlines. In the existing technology, because the 12345 hotline has various types of services and a wide range of scope, it requires horizontal linkage of multiple departments. If the dispatching cannot be reasonable, efficient, and place names cannot be accurately identified, and an effective joint handling mechanism cannot be established, it will be difficult to fundamentally resolve the demands of the people.
[0010] To this end, the present invention proposes a place name recognition method and system for intelligent dispatching of service hotlines. Summary of the invention
[0011] The purpose of this application is to provide a place name recognition method and system for intelligent dispatching of service hotlines, which solves the problem of inaccurate place name recognition in service hotline dispatching in the prior art.
[0012] To achieve the above objectives, the present application provides a place name recognition system for intelligent dispatching of service hotlines, including:
[0013] A speech recognition module, which is used to recognize the service hotline intelligent dispatching recording and convert the service hotline intelligent dispatching recording into speech text;
[0014] A special positioning module, which is used to specially capture the voice text recognized by the voice recognition module; the special positioning module includes a window setting unit, a database and a positioning analysis unit:
[0015] The window setting unit is used to set the size of the capture window and traverse the voice text according to the window size;
[0016] The database is used to store Chinese place names;
[0017] The positioning and parsing unit is used to perform special positioning on Chinese place names, including the following steps:
[0018] Step 1: The special positioning module sends an acquisition signal to the speech recognition module. After receiving the acquisition signal, the speech recognition module feeds back the recognized speech text to the special positioning module;
[0019] Step 2: The window setting unit sets the size of the recognition window and traverses the speech text according to the size of the recognition window;
[0020] Step 3: When the text in the window is consistent with the Chinese place name stored in the database, the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module.
[0021] As a further solution of the present invention: the service hotline intelligent dispatch recording is a service hotline intelligent dispatch telephone recording. When a visitor dials the service hotline, the call device establishes a voice call with the electronic device corresponding to the visitor, and the call device records the call and stores the call recording in the server;
[0022] The server is used to store the call recordings sent by the call device and mark the corresponding visitors.
[0023] As a further solution of the present invention: the window setting unit is used to set the size of the recognition window according to the demander; and the size of the recognition window is set or modified according to the demand.
[0024] As a further solution of the present invention: before the speech recognition module feeds back the recognized speech text to the special positioning module, it also includes a text error correction process;
[0025] The text correction process includes:
[0026] Set the error correction word pair, and use the error correction word pair to correct the speech text to obtain the corrected text, and send the corrected text to the special positioning module.
[0027] As a further solution of the present invention: when the text in the window is consistent with the Chinese place name stored in the database, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained;
[0028] The recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database. The positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module.
[0029] As a further solution of the present invention: when the recognition windows are set in sequence from large to small, if the text in the window is still inconsistent with the Chinese place name stored in the database, the manual customer service performs manual recognition, obtains the call recording stored in the server, and performs manual recognition.
[0030] As a further solution of the present invention: when the manual customer service intervenes and finds Chinese place names that are not included in the database, new Chinese place names are added to the database by manual input to complete the update of the database.
[0031] A place name recognition method for intelligent dispatching of service hotlines, the method comprising:
[0032] Step 1: The special positioning module sends an acquisition signal to the speech recognition module. After receiving the acquisition signal, the speech recognition module feeds back the recognized speech text to the special positioning module;
[0033] Step 2: The window setting unit sets the size of the recognition window and traverses the speech text according to the size of the recognition window;
[0034] Step 3: When the text in the window is consistent with the Chinese place name stored in the database, the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module.
[0035] As a further solution of the present invention: when the text in the window is consistent with the Chinese place name stored in the database, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained;
[0036] The recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database. The positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module. When the recognition window is set from large to small, if the text in the window is still inconsistent with the Chinese place name stored in the database, the manual customer service performs manual recognition, obtains the call recording stored in the server, and performs manual recognition.
[0037] Compared with the prior art, the present invention has the following beneficial effects:
[0038] The present invention uses a special positioning module to specially capture the voice text recognized by the voice recognition module, and the capture window can be customized. When the set window size cannot capture the Chinese place name, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained; the recognition window is set in sequence from large to small until the text in the window is consistent with the Chinese place name stored in the database, and the positioning and parsing unit extracts the corresponding Chinese place name and outputs it; or manual recognition is performed with the help of manual customer service. If a Chinese place name that is not included in the database is found, the new Chinese place name is added to the database by manual input, the database is updated, and the database content is increased, so that the Chinese place name can be accurately identified and the intelligent dispatching of the service hotline can be better completed. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the prior art and the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0040] Figure 1 This is a structural block diagram of a place name recognition system applied to intelligent dispatching of service hotlines according to the present invention;
[0041] Figure 2 The present invention is a flowchart of a place name recognition method applied to intelligent dispatching of service hotlines. DETAILED DESCRIPTION
[0042] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0043] Example 1
[0044] Please refer to Figure 1 , a place name recognition system applied to intelligent dispatching of service hotlines, including a speech recognition module;
[0045] The speech recognition module is used to recognize the service hotline intelligent dispatching recording and convert the service hotline intelligent dispatching recording into speech text;
[0046] The service hotline intelligent dispatch recording is a service hotline intelligent dispatch telephone recording. When a visitor dials the service hotline, the call device establishes a voice call with the visitor's corresponding electronic device, and the call device records the call and stores the call recording in the server.
[0047] The server is used to store the call recordings sent by the call device and mark the corresponding visitors;
[0048] The server is also used to receive the extraction request sent by the voice recognition module. When the server receives the extraction request sent by the voice recognition module, it feeds back the service hotline intelligent dispatching telephone recording stored in the server to the voice recognition module.
[0049] The speech recognition module is connected to a special positioning module, and the special positioning module is used to perform special capture on the speech text recognized by the speech recognition module;
[0050] In the present application, the special positioning module includes a window setting unit, a database and a positioning analysis unit;
[0051] The window setting unit is used to set the size of the capture window and traverse the voice text according to the window size;
[0052] The database stores Chinese place names, which are classified into provinces, cities, counties, towns, villages, and towns according to the size of the regions.
[0053] Moreover, the Chinese place names include the abbreviations of certain regions;
[0054] The window setting unit is used to set the window size according to the needs of the user, which is generally 3 characters; and the window size can be set and changed according to the needs;
[0055] The positioning and parsing unit is used to perform special positioning on Chinese place names, and the specific process includes the following steps:
[0056] Step 1: The special positioning module sends an acquisition signal to the speech recognition module. After receiving the acquisition signal, the speech recognition module feeds back the recognized speech text to the special positioning module;
[0057] Step 2: The window setting unit sets the size of the recognition window and traverses the speech text according to the size of the recognition window;
[0058] It should be noted that when traversing the voice text, the Chinese place names stored in the database are combined;
[0059] Step 3: When the text in the window is consistent with the Chinese place name stored in the database, the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module;
[0060] It should be noted that before the speech recognition module feeds back the recognized speech text to the special positioning module, it also includes a text error correction process;
[0061] Specifically, the text error correction process includes:
[0062] Setting error correction word pairs, and using the error correction word pairs to correct the speech text, obtaining the corrected text, and sending the corrected text to the special positioning module;
[0063] It should be noted that when the text in the window is consistent with the Chinese place name stored in the database, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained;
[0064] The recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database, and the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module;
[0065] It should be noted that when the recognition window is set from large to small, if the text in the window is still inconsistent with the Chinese place name stored in the database, the manual customer service will perform manual recognition. The manual customer service will obtain the call recording stored in the server and perform manual recognition.
[0066] In this application, when the manual customer service intervenes and finds Chinese place names that are not included in the database, the new Chinese place names are added to the database by manual input to complete the timely update of the database.
[0067] The place name recognition system for intelligent dispatching of service hotlines proposed in this embodiment uses a special positioning module to specially capture the voice text recognized by the voice recognition module, and the capture window can be customized. When the set window size cannot capture the Chinese place name, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained; the recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database, and the positioning and parsing unit extracts the corresponding Chinese place name and outputs it; or manual recognition is performed with the help of human customer service. If Chinese place names that are not included in the database are found, new Chinese place names are added to the database by manual input, and the database is updated and the database content is increased, so that Chinese place names can be accurately identified and the intelligent dispatching of the service hotline can be better completed.
[0068] Example 2
[0069] Please refer to Figure 2 , a place name recognition method applied to intelligent dispatching of service hotlines, the method comprising:
[0070] Step 1: When a visitor dials the service hotline, the calling device establishes a voice call with the visitor's corresponding electronic device, and the calling device records the call and stores the call recording in the server;
[0071] Among them, the mark corresponds to the visitor;
[0072] Step 2: The voice recognition module sends an extraction request to the server, and the server feeds back the service hotline intelligent dispatching telephone recording stored internally to the voice recognition module; the voice recognition module recognizes the service hotline intelligent dispatching recording, and converts the service hotline intelligent dispatching recording into voice text;
[0073] Step 3: The speech recognition module is connected to the special positioning module, the window setting unit sets the size of the recognition window, and traverses the speech text according to the size of the recognition window;
[0074] Step 4: When the text in the window is consistent with the Chinese place name stored in the database, the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module;
[0075] When the text in the window is consistent with the Chinese place name stored in the database, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained;
[0076] The recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database. The positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module. When the recognition window is set from large to small, if the text in the window is still inconsistent with the Chinese place name stored in the database, the manual customer service performs manual recognition, obtains the call recording stored in the server, and performs manual recognition.
[0077] The place name recognition method applied to the intelligent dispatching of service hotlines proposed in this embodiment uses a special positioning module to specially capture the voice text recognized by the voice recognition module, and the capture window can be customized. When the set window size cannot capture the Chinese place name, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained; the recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database, and the positioning and parsing unit extracts the corresponding Chinese place name and outputs it; or manual recognition is performed with the help of human customer service. If Chinese place names that are not included in the database are found, new Chinese place names are added to the database by manual input, and the database is updated and the content included in the database is increased, so that Chinese place names can be accurately identified and the intelligent dispatching of the service hotline can be better completed.
[0078] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A place name recognition system for intelligent dispatching of service hotlines, characterized in that: include: A speech recognition module, which is used to recognize the service hotline intelligent dispatching recording and convert the service hotline intelligent dispatching recording into speech text; A special positioning module, which is used to specially capture the voice text recognized by the voice recognition module; the special positioning module includes a window setting unit, a database and a positioning analysis unit: The window setting unit is used to set the size of the capture window and traverse the voice text according to the window size; The database is used to store Chinese place names; The positioning and parsing unit is used to perform special positioning on Chinese place names, including the following steps: Step 1: The special positioning module sends an acquisition signal to the speech recognition module. After receiving the acquisition signal, the speech recognition module feeds back the recognized speech text to the special positioning module; Step 2: The window setting unit sets the size of the recognition window and traverses the speech text according to the size of the recognition window; Step 3: When the text in the window is consistent with the Chinese place name stored in the database, the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module; The database stores Chinese place names, wherein the Chinese place names included are provinces, cities, counties, towns, villages and towns according to the size of the regions; the Chinese place names also include the abbreviations of the regions; The service hotline intelligent dispatch recording is a service hotline intelligent dispatch telephone recording. When a visitor dials the service hotline, the call device establishes a voice call with the visitor's corresponding electronic device, and the call device records the call and stores the call recording in the server; The server is used to store the call recordings sent by the call device and mark the corresponding visitors; The window setting unit is used to set the size of the recognition window according to the demander; and the size of the recognition window is set or changed according to the demand; Before the speech recognition module feeds back the recognized speech text to the special positioning module, it also includes a text error correction process; The text correction process includes: Setting error correction word pairs, and using the error correction word pairs to correct the speech text, obtaining the corrected text, and sending the corrected text to the special positioning module; When the text in the window is consistent with the Chinese place name stored in the database, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained; The recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database, and the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module; When the recognition window is set from large to small, if the text in the window is still inconsistent with the Chinese place name stored in the database, the manual customer service will perform manual recognition. The manual customer service will obtain the call recording stored in the server and perform manual recognition; When the manual customer service intervenes and finds Chinese place names that are not included in the database, the new Chinese place names are added to the database by manual input to complete the update of the database.
2. A place name recognition method for intelligent dispatching of service hotlines, implemented based on a place name recognition system for intelligent dispatching of service hotlines as claimed in claim 1, characterized in that: Methods include: Step 1: The special positioning module sends an acquisition signal to the speech recognition module. After receiving the acquisition signal, the speech recognition module feeds back the recognized speech text to the special positioning module; Step 2: The window setting unit sets the size of the recognition window and traverses the speech text according to the size of the recognition window; Step 3: When the text in the window is consistent with the Chinese place name stored in the database, the positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module.
3. A place name recognition method for intelligent dispatching of service hotlines as claimed in claim 2, characterized in that: When the text in the window is consistent with the Chinese place name stored in the database, the longest character and the shortest character corresponding to the Chinese place name included in the database are obtained; The recognition window is set from large to small until the text in the window is consistent with the Chinese place name stored in the database. The positioning and parsing unit extracts the corresponding Chinese place name and feeds it back to the place name output module. When the recognition window is set from large to small, if the text in the window is still inconsistent with the Chinese place name stored in the database, the manual customer service performs manual recognition, obtains the call recording stored in the server, and performs manual recognition.
