Call function management system and method based on big data
By building a dialogue recognition model and encrypted storage technology, the problems of low efficiency and insufficient security of call recording data are solved, and the extraction and acceleration of dialogue segments are realized, which significantly reduces memory usage and improves data security.
Patent Information
- Application Number
- CN202411161453.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-19
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, the call recording data storage method directly occupies the memory of the user's mobile terminal and cannot effectively remove non-dialogue segments, resulting in low storage efficiency and insufficient security.
Design a call function management system based on big data, including data acquisition, processing, optimization and storage modules. By building a dialogue recognition model, extracting dialogue fragments, accelerating processing and encrypting storage, reducing memory footprints and improving data security.
Effectively remove non-dialogue clips, reduce the memory usage of call recording data, improve the utilization rate of user mobile terminals, and significantly improve data security through encryption processing and protect user privacy.
Smart Images

Figure CN118984349A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of call management, and in particular to a call function management system and method based on big data. Background Art
[0002] Call recording refers to the function of recording and saving the audio content of a telephone call or an Internet call. The call recording function can be used for a variety of purposes, such as recording the content of the conversation, backing up or retaining important calls, as evidence or supervision, etc. The call recording function is usually provided by communication equipment or call software and can be enabled through corresponding settings or operations. Once the call recording function is enabled, the device or software will automatically record the audio signal during the call and save it in a specific audio file format; In the prior art, call recordings are often stored directly in audio format. This storage method not only occupies the memory of the user's mobile terminal, but also cannot guarantee the security of the recorded data. A considerable portion of call recordings contain a large number of non-conversation fragments. If these can be removed, conversation fragments with storage value can be obtained, and they can also be reorganized according to the randomness of their cutting to achieve encryption. In view of the shortcomings of the prior art, the present invention provides a call function management system and method based on big data. Summary of the invention
[0003] The purpose of the present invention is to provide a call function management system and method based on big data.
[0004] The object of the present invention can be achieved by the following technical solutions: A call function management system based on big data, comprising a main control center, wherein the main control center is communicatively connected with a data acquisition module, a data processing module, a data optimization module, and a data storage module; The data collection module is used to collect the user's call recording data; The data processing module is used to pre-process the collected call recording data to obtain corresponding usable recording data; The data optimization module is used to construct a dialogue recognition model, and extract dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data; and is also used to accelerate the obtained dialogue recording data to obtain corresponding accelerated recording data; The data storage module is used to reorganize the acquired accelerated recording data to obtain corresponding primary encrypted data and generate a primary encrypted identifier; and is also used to perform secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and store the secondary encrypted data.
[0005] Furthermore, the process of the data collection module collecting the user's call recording data includes: Setting a registration unit and a user database, inputting the user's personal information into the registration unit to complete the registration, and uploading the obtained personal information to the user database for storage; A login unit and an input unit are set up, and a user who has completed registration completes the login through the login unit, and inputs his own call recording data through the input unit, so that the user's personal information and the call recording data are bound.
[0006] Furthermore, the process of the data processing module preprocessing the collected call recording data to obtain corresponding usable recording data includes: The pre-processing includes noise removal processing, gain adjustment processing, and format conversion processing. The noise removal processing, gain adjustment processing, and format conversion processing are only for the audio content in the collected call recording data, and no other operations are performed on other data therein, and the subsequent processing is the same; A noise removal unit is provided, and the noise removal unit is used to perform noise removal on the collected call recording data to reduce or remove environmental noise and interference therein; a gain adjustment unit is provided, and the gain adjustment is performed on the collected call recording data to improve voice clarity and ensure audio consistency; a format conversion unit is provided, and the format conversion unit is used to perform format conversion on the collected call recording data to convert the audio content into a unified format; and the call recording data obtained after the above-mentioned noise removal processing, gain adjustment processing, and format conversion processing are marked as available recording data.
[0007] Furthermore, the data optimization module constructs a dialogue recognition model, and the process of extracting dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data includes: Marking the conversation segments and non-conversation segments in the audio content respectively, wherein the conversation segments refer to the audio content when the two parties of the call are talking, and the non-conversation segments refer to the other audio content except the conversation segments, and using the same method to obtain several segments of marked audio content, and marking them as an available audio set; Perform feature extraction on the obtained available audio set to obtain corresponding audio features, select a deep learning model as an initial dialogue recognition model, divide the available audio set into a training set and a test set, use the training set to train the initial dialogue recognition model, obtain a trained dialogue recognition model by learning the audio features of dialogue segments and non-dialogue segments in the available audio set, use the test set to evaluate the trained dialogue recognition model, tune the trained dialogue recognition model according to the evaluation result to obtain a tuned dialogue recognition model, and use the most recently tuned dialogue recognition model as the current dialogue recognition model; The constructed conversation recognition model is used to identify conversation segments in the available recording data, and the identified conversation segments are extracted. The extracted conversation segments are combined in their original playback order to obtain corresponding conversation recording data. The obtained conversation recording data only contains conversation segments, but does not contain non-conversation segments.
[0008] Furthermore, the process of the data optimization module accelerating the acquired conversation recording data to obtain corresponding accelerated recording data includes: A recording acceleration unit is provided, and the obtained conversation recording data is accelerated by the recording acceleration unit. Different acceleration multiples are preset in the recording acceleration unit, and the memory size occupied by the conversation recording data before acceleration is obtained, and the memory size occupied by the conversation recording data after different acceleration multiples is calculated, and the corresponding calculation results are obtained and fed back to the user for viewing, and the user selects the acceleration multiple that he wants to adopt according to the calculation results; The recording acceleration unit accelerates the acquired conversation recording data according to the selected acceleration ratio to obtain corresponding accelerated recording data, and a recording deceleration unit is set, which is used to decelerate the acquired accelerated recording data to restore it to conversation recording data at a normal rate.
[0009] Furthermore, the data storage module reorganizes the acquired accelerated recording data to obtain corresponding primary encrypted data, and generates a primary encrypted identifier, including: Since the conversation recording data is composed of several extracted conversation segments, the accelerated recording data is also composed of several accelerated conversation segments. The conversation segments are numbered according to their playback order to obtain the segment number of each conversation segment, and the obtained conversation segments are randomly reorganized to obtain corresponding one-time encrypted data. The segment numbers of each conversation segment constituting the one-time encrypted data are combined in sequence to obtain a corresponding numbering sequence, and marked as the one-time encrypted identifier of the one-time encrypted data.
[0010] Furthermore, the data storage module performs secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data, and stores the secondary encrypted data, including: Setting a binary conversion unit, through which the obtained primary encrypted data and primary encrypted identifier are converted into binary data, wherein the binary conversion unit only changes the form of the data but does not change the content of the data; An encryption unit is provided, wherein a plurality of encryption methods are provided in the encryption unit, and the obtained binary data is encrypted by the encryption unit, and the encryption method of the binary data is randomly selected, and the encrypted binary data is marked as secondary encrypted data, and a corresponding encryption key is generated, and the secondary encrypted data is bound to its corresponding encryption key; A recording database is set up, and all the secondary encrypted data obtained are uploaded to the recording database for storage, and the encryption key of the secondary encrypted data obtained is uploaded to the corresponding resident database for storage.
[0011] A call function management method based on big data comprises the following steps: Step S1: collecting the user's call recording data, and preprocessing the collected call recording data to obtain corresponding usable recording data; Step S2: constructing a dialogue recognition model, extracting dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data; and accelerating the obtained dialogue recording data to obtain corresponding accelerated recording data; Step S3: reorganize the acquired accelerated recording data to obtain corresponding primary encrypted data and generate a primary encrypted identifier; and perform secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and store the secondary encrypted data.
[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. By building a conversation recognition model to identify the conversation segments in each audio content, and extracting them to obtain the corresponding conversation recording data, it can effectively remove useless audio such as non-conversation segments, and accelerate the processing of the original conversation recording data to obtain the corresponding accelerated recording data. The above two methods can effectively reduce the memory size occupied by the call recording data, which is conducive to improving the utilization rate of the user's mobile terminal; 2. After extracting the conversation fragments, the obtained conversation fragments are numbered and rearranged, which is conducive to encrypting the call recording data once according to its random segmentation characteristics. On the basis of the first encryption, it is converted into binary data and then encrypted again, which can significantly improve the security of the call recording data and improve the protection of user privacy. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] Figure 1 It is the principle diagram of the present invention; Figure 2 It is a flow chart of the present invention. DETAILED DESCRIPTION
[0014] like Figure 1 As shown, a call function management system based on big data includes a main control center, and the main control center is communicatively connected with a data acquisition module, a data processing module, a data optimization module, and a data storage module; The data collection module is used to collect the user's call recording data; The data processing module is used to pre-process the collected call recording data to obtain corresponding usable recording data; The data optimization module is used to construct a dialogue recognition model, and extract dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data; and is also used to accelerate the obtained dialogue recording data to obtain corresponding accelerated recording data; The data storage module is used to reorganize the acquired accelerated recording data to obtain corresponding primary encrypted data and generate a primary encrypted identifier; and is also used to perform secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and store the secondary encrypted data.
[0015] It should be further explained that, in the specific implementation process, the process of the data collection module collecting the user's call recording data includes: A registration unit and a user database are set up, and the user's personal information is input into the registration unit, and the personal information includes but is not limited to name, gender, age, occupation, mobile phone number, ID number, and ID photo. The input personal information is reviewed by the main control center, and a corresponding registration verification code is generated after the review is passed. The generated registration verification code is sent to the mobile terminal corresponding to the user's mobile phone number, and the obtained registration verification code is input into the registration unit for verification. After the verification is passed, the registration is completed, and the obtained personal information is uploaded to the user database for storage; A login unit and an entry unit are set up. During the subsequent login process, the user who has completed the registration inputs the corresponding mobile phone number into the login unit. The input mobile phone number is reviewed by the main control center, and a corresponding login verification code is generated after the review is passed. The generated login verification code is sent to the mobile terminal corresponding to the mobile phone number, and the obtained login verification code is input into the login unit for verification. After the verification is passed, the login is completed. After completing the registration and login, the user enters his or her own call recording data through the entry unit. The call recording data includes but is not limited to audio content, call start and end timestamps, call duration, phone numbers of participants in the call, call type, call device information, and related notes or tags. The entered call recording data is reviewed by the main control center, and the user's personal information and call recording data are bound after the review is passed.
[0016] It should be further explained that, in a specific implementation process, the process in which the data processing module pre-processes the collected call recording data to obtain corresponding usable recording data includes: The pre-processing includes noise removal processing, gain adjustment processing, and format conversion processing. The noise removal processing, gain adjustment processing, and format conversion processing are only for the audio content in the collected call recording data, and no other operations are performed on other data therein, and the subsequent processing is the same; A noise removal unit is provided, and the noise removal unit is used to perform noise removal on the collected call recording data to reduce or remove environmental noise and interference therein; a gain adjustment unit is provided, and the gain adjustment is performed on the collected call recording data to improve voice clarity and ensure audio consistency; a format conversion unit is provided, and the format conversion unit is used to perform format conversion on the collected call recording data to convert the audio content into a unified format; and the call recording data obtained after the above-mentioned noise removal processing, gain adjustment processing, and format conversion processing are marked as available recording data.
[0017] It should be further explained that, in a specific implementation process, the data optimization module constructs a dialogue recognition model, and the process of extracting dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data includes: Taking a certain audio content as an example, the conversation segment and the non-conversation segment in the audio content are marked respectively, wherein the conversation segment refers to the audio content when the two parties are talking, and the non-conversation segment refers to the other audio content except the conversation segment. The same method is adopted to obtain several segments of marked audio content, and they are marked as an available audio set; Perform feature extraction on the obtained available audio set to obtain corresponding audio features, select a deep learning model as an initial dialogue recognition model, divide the available audio set into a training set and a test set, use the training set to train the initial dialogue recognition model, obtain a trained dialogue recognition model by learning the audio features of dialogue segments and non-dialogue segments in the available audio set, use the test set to evaluate the trained dialogue recognition model, tune the trained dialogue recognition model according to the evaluation result to obtain a tuned dialogue recognition model, and use the most recently tuned dialogue recognition model as the current dialogue recognition model; The constructed conversation recognition model is used to identify conversation segments in the available recording data, and the identified conversation segments are extracted. The extracted conversation segments are combined in their original playback order to obtain corresponding conversation recording data. The obtained conversation recording data only contains conversation segments, but does not contain non-conversation segments.
[0018] It should be further explained that, in a specific implementation process, the process in which the data optimization module accelerates the acquired conversation recording data to obtain corresponding accelerated recording data includes: A recording acceleration unit is provided, and the obtained conversation recording data is accelerated by the recording acceleration unit. Different acceleration magnifications are preset in the recording acceleration unit, namely, 2x speed, 3x speed, 5x speed, and 10x speed. The memory size occupied by the conversation recording data before acceleration is obtained, and the memory size occupied by the conversation recording data after different acceleration magnifications is calculated. The corresponding calculation results are obtained and fed back to the user for viewing, and the user selects the acceleration magnification that he wants to adopt according to the calculation results; The recording acceleration unit accelerates the acquired conversation recording data according to the selected acceleration ratio to obtain corresponding accelerated recording data, and a recording deceleration unit is set, which is used to decelerate the acquired accelerated recording data to restore it to conversation recording data at a normal rate.
[0019] It should be further explained that, in a specific implementation process, the data storage module reorganizes the acquired accelerated recording data to obtain corresponding primary encrypted data, and generates a primary encrypted identifier, including: Since the conversation recording data is composed of several extracted conversation segments, the accelerated recording data is also composed of several accelerated conversation segments. Taking any accelerated recording data as an example, the conversation segments therein are numbered according to their playback order, denoted as i, where i=1, 2, ..., the segment number of each conversation segment is obtained, the obtained conversation segments are randomly reorganized to obtain the corresponding first-time encrypted data, the segment numbers of the various conversation segments constituting the first-time encrypted data are combined in sequence to obtain the corresponding numbering sequence, and it is marked as the first-time encrypted identifier of the first-time encrypted data.
[0020] It should be further explained that, in a specific implementation process, the data storage module performs secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data, and stores the data in a process including: Setting a binary conversion unit, through which the obtained primary encrypted data and primary encrypted identifier are converted into binary data, wherein the binary conversion unit only changes the form of the data but does not change the content of the data; An encryption unit is provided, wherein a plurality of encryption methods are provided in the encryption unit, and the obtained binary data is encrypted by the encryption unit, and the encryption method of the binary data is randomly selected, and the encrypted binary data is marked as secondary encrypted data, and a corresponding encryption key is generated, and the secondary encrypted data is bound to its corresponding encryption key; A recording database is set up, and all the secondary encrypted data obtained are uploaded to the recording database for storage, and the encryption key of the secondary encrypted data obtained is uploaded to the corresponding resident database for storage.
[0021] like Figure 2 As shown, a call function management method based on big data includes the following steps: Step S1: collecting the user's call recording data, and preprocessing the collected call recording data to obtain corresponding usable recording data; Step S2: constructing a dialogue recognition model, extracting dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data; and accelerating the obtained dialogue recording data to obtain corresponding accelerated recording data; Step S3: reorganize the acquired accelerated recording data to obtain corresponding primary encrypted data and generate a primary encrypted identifier; and perform secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and store the secondary encrypted data.
[0022] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.
Claims
1. A call function management system based on big data, including a main control center, characterized in that: The main control center is communicatively connected with a data acquisition module, a data processing module, a data optimization module, and a data storage module; The data collection module is used to collect the user's call recording data; The data processing module is used to pre-process the collected call recording data to obtain corresponding usable recording data; The data optimization module is used to construct a dialogue recognition model, and extract dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data; and is also used to accelerate the obtained dialogue recording data to obtain corresponding accelerated recording data; The data storage module is used to reorganize the acquired accelerated recording data to obtain corresponding primary encrypted data and generate a primary encrypted identifier; and is also used to perform secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and store the secondary encrypted data.
2. A call function management system based on big data according to claim 1, characterized in that: The process of the data collection module collecting the user's call recording data includes: Setting a registration unit and a user database, inputting the user's personal information into the registration unit to complete the registration, and uploading the obtained personal information to the user database for storage; A login unit and an input unit are set up, and a user who has completed registration completes the login through the login unit, and inputs his own call recording data through the input unit, so that the user's personal information and the call recording data are bound.
3. A call function management system based on big data according to claim 2, characterized in that: The process of the data processing module preprocessing the collected call recording data to obtain corresponding usable recording data includes: The preprocessing includes noise removal processing, gain adjustment processing, and format conversion processing. A noise removal unit is set up, and noise is removed from the collected call recording data through the noise removal unit. A gain adjustment unit is set up, and gain adjustment is performed on the collected call recording data through the gain adjustment unit. A format conversion unit is set up, and format conversion is performed on the collected call recording data through the format conversion unit. The call recording data obtained after the above-mentioned noise removal processing, gain adjustment processing, and format conversion processing is marked as usable recording data.
4. A call function management system based on big data according to claim 3, characterized in that: The data optimization module constructs a dialogue recognition model, and the process of extracting dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data includes: The dialogue segments and non-dialogue segments in the audio content are respectively marked to obtain the corresponding available audio sets, a deep learning model is selected as the initial dialogue recognition model, the available audio set is divided into a training set and a test set, and the initial dialogue recognition model is trained and tested using the available audio set to obtain the latest dialogue recognition model; The constructed dialogue recognition model is used to identify dialogue segments in the available recording data, and the identified dialogue segments are extracted. The extracted dialogue segments are combined according to their original playback order to obtain corresponding dialogue recording data.
5. A call function management system based on big data according to claim 4, characterized in that: The process of the data optimization module accelerating the acquired conversation recording data to obtain corresponding accelerated recording data includes: A recording acceleration unit is provided, and the obtained conversation recording data is accelerated by the recording acceleration unit. Different acceleration rates are preset in the recording acceleration unit, and the user selects the acceleration rate he wants to use; The recording acceleration unit accelerates the acquired conversation recording data according to the selected acceleration ratio to obtain corresponding accelerated recording data, and a recording deceleration unit is set, which is used to decelerate the acquired accelerated recording data to restore it to conversation recording data at a normal rate.
6. A call function management system based on big data according to claim 5, characterized in that: The data storage module reorganizes the acquired accelerated recording data to obtain corresponding primary encrypted data, and generates a primary encrypted identifier, which includes: The conversation segments in the accelerated recording data are numbered to obtain segment numbers of the conversation segments, the obtained conversation segments are reorganized to obtain corresponding primary encrypted data, and the primary encrypted identifiers of the primary encrypted data are obtained according to the segment numbers of the conversation segments.
7. A call function management system based on big data according to claim 6, characterized in that: The process of the data storage module performing secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and storing the secondary encrypted data comprises: A binary conversion unit is provided, through which the obtained primary encrypted data and the primary encrypted identifier are converted into binary data; an encryption unit is provided, through which the obtained binary data is encrypted, the encrypted binary data is marked as secondary encrypted data, and a corresponding encryption key is generated; A recording database is set up, and all the secondary encrypted data obtained are uploaded to the recording database for storage, and the encryption key of the secondary encrypted data obtained is uploaded to the corresponding resident database for storage.
8. A call function management method based on a call function management system based on big data according to any one of claims 1 to 7, characterized in that: The following steps are involved: Step S1: collecting the user's call recording data, and preprocessing the collected call recording data to obtain corresponding usable recording data; Step S2: constructing a dialogue recognition model, extracting dialogue segments from available recording data through the dialogue recognition model to obtain corresponding dialogue recording data; and accelerating the obtained dialogue recording data to obtain corresponding accelerated recording data; Step S3: reorganize the acquired accelerated recording data to obtain corresponding primary encrypted data and generate a primary encrypted identifier; and perform secondary encryption on the primary encrypted data and the primary encrypted identifier to obtain corresponding secondary encrypted data and store the secondary encrypted data.
Citation Information
Patent Citations
Method and mobile terminal for processing calling record
CN101394634A
Call recording method and system
CN110536000A
Voice communication method, device, system and storage medium
CN116682452A
