Data processing method and device

Abstract

The present disclosure relates to the field of data processing technology and provides a data processing method. In this method, based on the text information and time information corresponding to a number of audios to be processed, the text information can be processed to obtain the character string corresponding to the text information and the time information corresponding to the character string, and based on the character string and the time information corresponding to the character string, the abnormal character string can be determined from the character information dimension of the character string and the time dimension of the character string appearance. In this way, when the audio to be processed is obtained, the processed audio can be processed to obtain the abnormal character string, which can avoid the problem of abnormal character string recognition errors or omissions caused by operational errors in the process of manual inspection and sorting of abnormal character strings, thereby improving the accuracy and efficiency of determining abnormal character strings and improving user experience.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, and in particular to a data processing method, apparatus, computer equipment, and computer-readable storage medium. Background Art

[0002] With the in-depth development of digital and intelligent technologies in various fields, more intelligent devices are playing a role in our lives. Various intelligent voice devices, such as smart speakers, smartphones, and smart robots, already support voice conversations.

[0003] In an actual conversation environment, due to the influence of the hardware of the voice device or the surrounding environment of the voice package device, there may be abnormal input in the voice recognition results of the voice device in the robot. For example, sometimes although there is no user talking around the robot, but there is only ambient sound (such as wind, footsteps, etc.) around the robot, the voice device in the robot may mistakenly recognize the ambient sound as a wake-up word (such as text such as "play") during voice recognition, thereby affecting the conversation interaction recognition of the voice device in the robot. In the prior art, for abnormal character strings that the voice device in the robot often mistakenly recognizes as wake-up words during voice recognition, the only way is to rely on manual inspection and sorting. It is found that the voice device mistakenly recognizes abnormal character strings as wake-up words during voice recognition. However, since the process of manual inspection and sorting of abnormal character strings is repetitive and tedious, prone to errors, and has low accuracy, the manual inspection and sorting of abnormal character strings leads to low data processing efficiency. Summary of the Invention

[0004] In view of this, the embodiments of the present disclosure provide a data processing method, apparatus, computer device, and computer-readable storage medium to address the problem in the prior art of manually checking and sorting abnormal character strings, which is prone to incorrect recognition or omission of abnormal character strings due to operational errors, thereby resulting in low data processing efficiency.

[0005] A first aspect of the present disclosure provides a data processing method, the method comprising:

[0006] Get some audio to be processed;

[0007] Determining text information and time information corresponding to the plurality of audios to be processed;

[0008] Determining, based on the text information and time information corresponding to the plurality of audios to be processed, a character string corresponding to the text information and time information corresponding to the character string;

[0009] An abnormal character string is determined according to the character string and time information corresponding to the character string.

[0010] According to a second aspect of the present disclosure, a data processing device is provided, the device comprising:

[0011] An audio acquisition unit, used to acquire a number of audios to be processed;

[0012] A first determining unit, configured to determine text information and time information corresponding to the plurality of audios to be processed;

[0013] A second determining unit is configured to determine, based on the text information and time information corresponding to the plurality of audios to be processed, a character string corresponding to the text information and time information corresponding to the character string;

[0014] The abnormality determination unit is used to determine an abnormal character string according to the character string and the time information corresponding to the character string.

[0015] According to a third aspect of an embodiment of the present disclosure, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0016] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps of the above method are implemented.

[0017] Compared with the prior art, the beneficial effects of the embodiments of the present disclosure are as follows: the embodiments of the present disclosure can first obtain a number of audios to be processed. Then, the text information and time information corresponding to the audios to be processed can be determined. Then, based on the text information and time information corresponding to the audios to be processed, the character string corresponding to the text information and the time information corresponding to the character string can be determined. Finally, the abnormal character string can be determined based on the character string and the time information corresponding to the character string. It can be seen that in this embodiment, based on the text information and time information corresponding to the audios to be processed, the text information can be processed to obtain the character string corresponding to the text information and the time information corresponding to the character string, and the abnormal character string can be determined from the character information dimension of the character string and the time dimension of the character string appearance based on the character string and the time information corresponding to the character string. In this way, when the audios to be processed are obtained, the data of the audios to be processed can be processed to obtain the abnormal character string, which can avoid the problem of abnormal character string recognition errors or omissions caused by operational errors in the process of manual inspection and sorting of abnormal character strings, thereby improving the accuracy and efficiency of determining abnormal character strings and improving user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0019] Figure 1 is a schematic diagram of an application scenario of an embodiment of the present disclosure;

[0020] Figure 2 is a flow chart of a data processing method provided by an embodiment of the present disclosure;

[0021] Figure 3 is a block diagram of a data processing device provided by an embodiment of the present disclosure;

[0022] Figure 4 Schematic diagram of a computer device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present disclosure. However, it will be apparent to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present disclosure with unnecessary detail.

[0024] A data processing method and apparatus according to an embodiment of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0025] In the existing technology, since the voice equipment in the robot often misidentifies abnormal character strings as wake-up words during the voice recognition process, it can only rely on manual inspection and sorting. It is found that the voice equipment often misidentifies abnormal character strings as wake-up words during the voice recognition process. However, since the process of manual inspection and sorting of abnormal character strings is repetitive and cumbersome, prone to errors, and low in accuracy, the manual inspection and sorting of abnormal character strings leads to low data processing efficiency.

[0026] In order to solve the above problems. The present invention provides a data processing method, in which a number of audios to be processed can be obtained first. Then, the text information and time information corresponding to the audios to be processed can be determined. Then, based on the text information and time information corresponding to the audios to be processed, the character string corresponding to the text information and the time information corresponding to the character string can be determined. Finally, the abnormal character string can be determined based on the character string and the time information corresponding to the character string. It can be seen that in this embodiment, based on the text information and time information corresponding to the audios to be processed, the text information can be processed to obtain the character string corresponding to the text information and the time information corresponding to the character string, and based on the character string and the time information corresponding to the character string, the abnormal character string can be determined from the character information dimension of the character string and the time dimension of the character string appearance. In this way, when the audio to be processed is obtained, the audio to be processed can be processed to obtain the abnormal character string, which can avoid the problem of abnormal character string recognition errors or omissions caused by operational errors in the process of manual inspection and sorting of abnormal character strings, thereby improving the accuracy and efficiency of determining abnormal character strings and improving user experience.

[0027] For example, the embodiment of the present invention can be applied to Figure 1 The application scenario shown in FIG. In this scenario, a terminal device 1 and a server 2 may be included.

[0028] The terminal device 1 may be an intelligent voice device, such as a smart speaker, a smart phone, an intelligent robot, etc. As an example, the terminal device 1 may include a microphone and a computing device. The microphone may be used to collect audio from the terminal device 1 itself or its surroundings, and the computing device may be used to process the collected audio to obtain an abnormal character string, or may be used to send the collected audio to the server 2.

[0029] Server 2 can be a server that provides various services, such as a backend server that receives requests sent by terminal devices that establish communication connections with it. This backend server can receive and analyze audio sent by the terminal devices, and process the collected audio to obtain abnormal character strings based on the collected audio. Server 2 can be a single server, a server cluster consisting of multiple servers, or a cloud computing service center, all of which are not limited in the present disclosure.

[0030] It should be noted that the server 2 can be either hardware or software. When the server 2 is hardware, it can be various electronic devices that provide various services to the terminal device 1. When the server 2 is software, it can be multiple software programs or software modules that provide various services to the terminal device 1, or it can be a single software program or software module that provides various services to the terminal device 1, and this is not limited in the present embodiment.

[0031] The terminal device 1 and the server 2 can be connected to each other through a network. The network can be a wired network connected by coaxial cable, twisted pair, or optical fiber, or a wireless network that can interconnect various communication devices without wiring, such as Bluetooth, Near Field Communication (NFC), infrared, etc., which is not limited in the embodiments of the present disclosure.

[0032] Specifically, the audio to be processed can be collected by the terminal device 1, and the terminal device 1 sends the audio to be processed to the server 2. The server 2 can first obtain the audio to be processed. Then, the server 2 can determine the text information and time information corresponding to the several audios to be processed. Then, the server 2 can determine the character string corresponding to the text information and the time information corresponding to the character string based on the text information and time information corresponding to the several audios to be processed. Then, the server 2 can determine the abnormal character string based on the character string and the time information corresponding to the character string. Finally, the server 2 can send the abnormal character string to the terminal device 1 so that the terminal device 1 does not need to respond when collecting the audio corresponding to the abnormal character string. It can be seen that in this embodiment, based on the text information and time information corresponding to the several audios to be processed, the text information can be processed to obtain the character string corresponding to the text information and the time information corresponding to the character string, and based on the character string and the time information corresponding to the character string, the abnormal character string can be determined from the character information dimension of the character string and the time dimension of the occurrence of the character string. In this way, when the audio to be processed is obtained, data processing can be performed on the audio to be processed to obtain abnormal strings, which can avoid the problem of abnormal string recognition errors or omissions caused by operational errors during the manual inspection and sorting of abnormal strings, thereby improving the accuracy and efficiency of determining abnormal strings and improving user experience.

[0033] It should be noted that the specific types, quantities and combinations of the terminal devices 1, the server 2 and the network can be adjusted according to the actual needs of the application scenario, and the embodiments of the present disclosure do not limit this.

[0034] It should be noted that the above application scenarios are only shown to facilitate understanding of the present disclosure, and the embodiments of the present disclosure are not limited in this respect. On the contrary, the embodiments of the present disclosure can be applied to any applicable scenario.

[0035] Figure 2 This is a flow chart of a data processing method provided by an embodiment of the present disclosure. Figure 2 A data processing method can be Figure 1 The terminal device or server executes. Figure 2As shown, the data processing method includes:

[0036] S201: Acquire some audio to be processed.

[0037] In this embodiment, the audio to be processed can be understood as the audio from which the abnormal character string needs to be extracted. As an example, the audio to be processed can be the audio from which the abnormal character string needs to be extracted using the target device (such as Figure 1 The audio collected by the terminal device 1) in the embodiment of the present invention can also be read from a storage device that has pre-stored audio data. The target device can be understood as a voice device that can collect audio, for example, a smart speaker, smart phone, smart robot, etc. that has the function of collecting audio.

[0038] Next, we will take the target device as an example to illustrate how to obtain a number of audios to be processed. In one implementation, the target device can be used to collect a number of audios to be processed at a preset time. For example, assuming that the target device can pre-set a timer to collect audio to be processed, it can be a preset time period for audio collection (for example, collecting from 8 am to 10 pm), or it can be a preset time period for audio collection (for example, collecting from 8 am to 10 pm), or it can be a preset time period for voice collection (for example, five minutes). After collecting a sentence, if the preset interval time (for example, one minute) is exceeded or no sound is detected after silence detection, the audio collection is stopped.

[0039] In another implementation, if the target device detects a preset trigger operation, the target device is used to collect a number of audios to be processed. The preset trigger operation includes at least one of the following: collecting preset facial information, triggering a preset button, and collecting a preset wake-up word. As an example, the target device can be a physical robot placed in a fixed position, for example, a robot placed in a shopping mall, exhibition hall, hospital entrance, etc.; when the target device collects preset facial information, such as the target device recognizes a face or a specific user face (such as an employee of the unit) through a face recognition function, the target device can start to enter a conversation mode, thereby collecting a number of audios to be processed; when the target device detects that a preset button is triggered, such as the target device detects that the user presses a preset button or clicks a preset button on the screen, the target device can turn on the microphone to receive sound, thereby collecting a number of audios to be processed.

[0040] It should be noted that the above two methods can be implemented either one or both at the same time.

[0041] S202: Determine text information and time information corresponding to the plurality of audios to be processed.

[0042] When the audio to be processed is acquired, the time at which the audio to be processed was collected can be recorded. Furthermore, the text information corresponding to the audio to be processed needs to be determined. The text information corresponding to the audio to be processed can be understood as the conversation content in the audio to be processed; and the time information corresponding to the audio to be processed can be understood as the time at which the audio to be processed was collected. It should be noted that the time information corresponding to the audio to be processed can be a specific time point (e.g., 3:07 PM) or a time period (e.g., 8:00 AM to 10:00 AM).

[0043] As an example, for each audio to be processed, speech recognition processing can be performed on the audio to be processed (for example, using a trained neural network model to perform speech recognition) to obtain a text recognition result corresponding to the audio to be processed. The text recognition result can then be used as the text information corresponding to the audio to be processed, and the acquisition time of the audio to be processed can be used as the time information corresponding to the audio to be processed.

[0044] For example, within a set period of time (e.g., one week), the target device collects n audios to be processed. Then, speech recognition is performed on the n audios to be processed, and the text recognition results corresponding to each audio to be processed are obtained, that is, n text recognition results are obtained. The text information and time information corresponding to the n audios to be processed can be Data ASR For {s1,s2,…,s n}, where s i Represents the text information and time information corresponding to the i-th audio to be processed.

[0045] S203: Based on the text information and time information corresponding to the plurality of audios to be processed, determine the character string corresponding to the text information and the time information corresponding to the character string.

[0046] Since an audio to be processed may be audio content of a period of time, the character strings in the text information corresponding to the audio to be processed are relatively long. In order to better identify abnormal character strings, the text information corresponding to the audio to be processed can be split into several character strings for processing. Specifically, for each text information corresponding to the audio to be processed, the text information corresponding to the audio to be processed can be divided and processed to obtain several character strings corresponding to the text information. For example, each sentence in the text information can be divided into a character string, or each word in the text can be divided into a character string; and the time information corresponding to the audio to be processed is used as the time information corresponding to the several character strings.

[0047] For example, the string corresponding to the text information is a sentence. After obtaining the text information and time information corresponding to the audio to be processed, each sentence of the text information corresponding to the i-th audio to be processed is divided into several strings, that is, one sentence in the text information corresponding to the i-th audio to be processed is divided into one string. Specifically, the text information and time information s of the i-th audio to be processed are i The corresponding string and the time information corresponding to the string can be expressed as (time i ,text i ), where text i It can be expressed as the i-th sentence in the text information corresponding to the i-th audio to be processed, time i It can represent the collection time of the i-th sentence. For example, s i ("2022-02-02-00-00", "Play"), a character string representing the text information corresponding to the i-th audio to be processed is "Play", and the time information corresponding to the character string "Play" is 00:00 on February 2, 2022, that is, after the target device turns on voice recognition at 00:00 on February 2, 2022, the text information of the recognized audio is "Play".

[0048] S204: Determine an abnormal character string based on the character string and the time information corresponding to the character string.

[0049] It should be noted that the source of abnormal speech recognition results may be due to problems with the target device itself (for example, the hardware microphone of the target device itself may have noise), or there may be problems with the surrounding environment of the target device (for example, there is always wind noise or frequent promotional sounds around, etc.), which will cause certain words to always appear in the text recognition results, and this problem may also be closely related to the time period, for example, the wind is stronger in the early morning.

[0050] Therefore, in this embodiment, after determining the character string corresponding to the text information and the time information corresponding to each character string, the frequency of occurrence of each character string and the time period in which each character string appears can be counted first. Then, the abnormal character strings can be filtered out from the character strings corresponding to the text information based on the frequency dimension and the time period dimension of the character string. It can be understood that abnormal character strings are character strings that are easily misidentified, such as character strings that are easily misidentified as wake-up words. For example, assuming that the surrounding ambient sounds (such as wind, footsteps, etc.) will be recognized as strings such as "play" through voice recognition, the character strings corresponding to the ambient sounds can be used as abnormal character strings. In this way, when the target device collects the audio corresponding to the abnormal character string, it does not need to respond, for example, there is no need for dialogue interaction, thereby improving the user experience.

[0051] As an example, we can first determine several candidate strings and the frequency information corresponding to each candidate string based on all the strings and the time information corresponding to the strings. Each candidate string is different, and the frequency information corresponding to each candidate string can include the total number of times the candidate string appears and the number of times the candidate string appears in each time period. In other words, we can count the total number of times all the strings appear, as well as the time period in which they appear and the number of times they appear in that time period. The number of times the candidate string appears in each time period can be expressed as (text i ,timeRange i , count i ), where text i Indicates the i-th candidate string, timeRange i Indicates the time period when the i-th candidate string appears (for example, it can be a certain time period, not a specific day), count i Indicates the i-th candidate string in the time range i The number of times it appears in the string; for example, ("play", "00-00", 124), which means that the candidate string "play" appears 124 times at zero o'clock.

[0052] For example, assuming that the character strings corresponding to the text information include {("2022-02-02-00-00", "play"), ("2022-02-02-01-00", "play"), ("2022-02-02-02-01", "hee hee")}, it can be determined that there are two candidate character strings, namely "play" and "hee hee"; the frequency information freq1 corresponding to the candidate character string "play" may include the total number of occurrences of the candidate character string "play" is 2 and the candidate character string The number of times in each time period ("play", "01-00:03-00", 2) indicates that the candidate character string "candidate" appears 2 times in the time period from 1 o'clock to 3 o'clock; the frequency information freq2 corresponding to the candidate character string "hehe" may include the total number of times the candidate character string "hehe" appears as 1 and the number of times the candidate character string appears in each time period ("hehe", "02-00:03-00", 1), which indicates that the candidate character string "candidate" appears once in the time period from 1 o'clock to 3 o'clock.

[0053] Then, for each candidate string, an anomaly weight value can be determined based on the number of characters and frequency information in the candidate string. It should be noted that a higher anomaly weight value indicates a higher probability that the candidate string is an anomaly, while a lower anomaly weight value indicates a lower probability that the candidate string is an anomaly.

[0054] In one implementation of this embodiment, the method for determining the abnormal weight value of a candidate string may include the following method: if the number of characters in the candidate string is less than a first preset threshold, the abnormal weight value of the candidate string may be determined based on the total number of occurrences of the candidate string and the sum of the total number of occurrences of all candidate strings; if the number of characters in the candidate string is greater than or equal to the first preset threshold, the abnormal weight value of the candidate string may be determined based on the number of characters in the candidate string, the total number of occurrences of the candidate string, the number of times the candidate string appears in each time period, the total number of occurrences of candidate strings contained in the candidate string, and the sum of the total number of occurrences of all candidate strings. Specifically, the abnormal weight value of the candidate string may be calculated using the following formula:

[0055]

[0056] Among them, weight(atext) represents the abnormal weight value of the candidate string; atext is the candidate string; allcount represents the total number of occurrences of all candidate strings, that is, atext i Indicates the i-th candidate string, f(atext i ) represents the total number of occurrences of the i-th candidate string; |atext| represents the length of the candidate string, that is, the number of characters contained in the candidate string (that is, the number of characters); f(atext) represents the total number of occurrences of the candidate string atext. It should be noted that if a candidate string appears in another longer candidate string, for example, "play" appears in "play the wind", then the total number of occurrences of "play" can include the total number of occurrences of "play the wind"; g(atext) represents the total number of occurrences of the candidate string in each time period timeRange i The number of times (i.e. the number of times it appears); c i ∈atext represents c i Contained by atext, where c i Can represent candidate strings different from atext, for example, "play" is included in "play the wind sound".

[0057] It should be noted that in this embodiment, different anomaly weight calculation methods are used based on the number of characters in the candidate string in order to allow for the recognition of even longer strings. For example, the string "play" is found to be relatively high frequency, making it easy to identify as an anomaly. However, the occurrence of the string "play the wind sound" can also be considered a possible problem identification result, i.e., an anomaly.

[0058] Then, based on the abnormal weight values ​​of each candidate string, the candidate string that meets the preset abnormal condition can be determined, and the candidate string that meets the preset abnormal condition can be used as an abnormal string. The preset abnormal condition can be that the abnormal weight value of the candidate string is greater than the preset weight threshold, or that the abnormal weight value of the candidate string ranks high among the abnormal weight values ​​of all candidate strings (that is, the ranking of the abnormal weight value of the candidate string is within the range of the preset ranking threshold, for example, ranking in the top three). In other words, when the abnormal weight value of the candidate string is greater than the preset weight threshold, the candidate string can be used as an abnormal string; alternatively, the abnormal weight values ​​of all candidate strings are first ranked from large to small, and then the candidate strings ranked in the top N positions can be used as abnormal strings.

[0059] It can be seen that in the embodiment of the present disclosure, a number of audios to be processed can be obtained first. Then, the text information and time information corresponding to the audios to be processed can be determined. Then, based on the text information and time information corresponding to the audios to be processed, the character string corresponding to the text information and the time information corresponding to the character string can be determined. Finally, the abnormal character string can be determined based on the character string and the time information corresponding to the character string. That is to say, in this embodiment, based on the text information and time information corresponding to the audios to be processed, the text information can be processed to obtain the character string corresponding to the text information and the time information corresponding to the character string, and the abnormal character string can be determined from the character information dimension of the character string and the time dimension of the character string appearance based on the character string and the time information corresponding to the character string. In this way, when the audio to be processed is obtained, the data processing of the audio to be processed can be performed to obtain the abnormal character string, which can avoid the problem of abnormal character string recognition errors or omissions caused by operational errors in the process of manual inspection and sorting of abnormal character strings, thereby improving the accuracy and efficiency of determining abnormal character strings and improving user experience.

[0060] To further improve the efficiency of identifying abnormal character strings, in one implementation of this embodiment, before the step of determining the abnormal character string based on the character string and the time information corresponding to the character string, the method may further include: performing a cleaning process on the character string to obtain a cleaned character string.

[0061] It should be noted that the cleaning process includes removing character strings whose number of characters is greater than a second preset threshold value and character strings whose characters are null values. That is to say, after obtaining the character string corresponding to the text information, the character strings whose number of characters is greater than a certain threshold value (such as 40) and the character strings whose characters are null values ​​can be deleted from the character string corresponding to the text information to obtain the cleaned character string corresponding to the text information. Then, the abnormal character string can be determined based on the cleaned character string and the time information corresponding to the cleaned character string. In this way, before determining the candidate character string, inappropriate character strings can be removed, reducing the amount of data that needs to be processed, thereby improving the efficiency of data processing, and further improving the efficiency of identifying abnormal character strings.

[0062] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present disclosure, and will not be described in detail here.

[0063] The following are embodiments of the apparatus disclosed herein, which can be used to implement the method embodiments disclosed herein. For details not disclosed in the apparatus embodiments disclosed herein, please refer to the method embodiments disclosed herein.

[0064] Figure 3 Schematic diagram of a data processing device provided by an embodiment of the present disclosure. Figure 3 As shown, the data processing device includes:

[0065] The audio acquisition unit 301 is used to acquire a number of audios to be processed;

[0066] A first determining unit 302 is configured to determine text information and time information corresponding to the plurality of audios to be processed;

[0067] The second determining unit 303 is configured to determine, based on the text information and time information corresponding to the plurality of audio files to be processed, a character string corresponding to the text information and time information corresponding to the character string;

[0068] The abnormality determination unit 304 is configured to determine an abnormal character string based on the character string and the time information corresponding to the character string.

[0069] In some embodiments, the audio acquisition unit 301 is configured to:

[0070] Using the target device to collect a number of audios to be processed at a preset time; and / or,

[0071] If the target device detects a preset trigger operation, the target device is used to collect a number of audios to be processed; wherein the preset trigger operation includes at least one of the following: collecting preset facial information, triggering a preset button, and collecting a preset wake-up word.

[0072] In some embodiments, the first determining unit 302 is configured to:

[0073] For each audio to be processed, speech recognition processing is performed on the audio to be processed to obtain a text recognition result corresponding to the audio to be processed, and the text recognition result is used as the text information corresponding to the audio to be processed; and the collection time of the audio to be processed is used as the time information corresponding to the audio to be processed.

[0074] In some embodiments, the second determining unit 303 is configured to:

[0075] For each text information corresponding to the audio to be processed, the text information corresponding to the audio to be processed is divided and processed to obtain a plurality of character strings corresponding to the text information, and the time information corresponding to the audio to be processed is used as the time information corresponding to the plurality of character strings.

[0076] In some embodiments, the abnormality determination unit 304 is configured to:

[0077] Determining, based on all the character strings and the time information corresponding to the character strings, a number of candidate character strings and frequency information corresponding to each candidate character string; wherein each candidate character string is different, and the frequency information corresponding to the candidate character string includes the total number of occurrences of the candidate character string and the number of occurrences of the candidate character string in each time period;

[0078] For each candidate string, determine the abnormality weight value of the candidate string based on the number of characters and frequency information of the candidate string;

[0079] According to the abnormality weight values ​​of the candidate character strings, a candidate character string that meets a preset abnormality condition is determined, and the candidate character string that meets the preset abnormality condition is used as an abnormal character string.

[0080] In some embodiments, the abnormality determination unit 304 is specifically configured to:

[0081] If the number of characters in the candidate character string is less than a first preset threshold, determining an abnormality weight value of the candidate character string based on the total number of occurrences of the candidate character string and the sum of the total number of occurrences of all candidate character strings;

[0082] If the number of characters in the candidate string is greater than or equal to the first preset threshold, the abnormality weight value of the candidate string is determined based on the number of characters in the candidate string, the total number of occurrences of the candidate string, the number of times the candidate string appears in each time period, the total number of occurrences of candidate strings contained in the candidate string, and the sum of the total number of occurrences of all candidate strings.

[0083] In some embodiments, the apparatus further includes a data cleaning unit configured to:

[0084] Cleaning the character string to obtain a cleaned character string;

[0085] The cleaning process includes removing character strings with a number of characters greater than a second preset threshold and character strings with null values;

[0086] Accordingly, the abnormality determination unit 304 is specifically configured to:

[0087] An abnormal character string is determined according to the cleaned character string and time information corresponding to the cleaned character string.

[0088] According to an embodiment of the present disclosure, a data processing device is provided, comprising: an audio acquisition unit for acquiring a plurality of audios to be processed; a first determination unit for determining text information and time information corresponding to the plurality of audios to be processed; a second determination unit for determining, based on the text information and time information corresponding to the plurality of audios to be processed, a character string corresponding to the text information and the time information corresponding to the character string; and an anomaly determination unit for determining an abnormal character string based on the character string and the time information corresponding to the character string. It can be seen that in this embodiment, based on the text information and time information corresponding to the plurality of audios to be processed, the text information can be processed to obtain a character string corresponding to the text information and the time information corresponding to the character string, and an abnormal character string can be determined based on the character information dimension of the character string and the time dimension of the character string appearance. In this way, when the audio to be processed is acquired, the data of the audio to be processed can be processed to obtain an abnormal character string, which can avoid the problem of abnormal character string recognition errors or omissions caused by operational errors during manual inspection and sorting of abnormal character strings, thereby improving the accuracy and efficiency of determining abnormal character strings and improving the user experience.

[0089] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0090] Figure 4 Schematic diagram of the computer device 4 provided in the embodiment of the present disclosure. Figure 4 As shown, the computer device 4 of this embodiment includes: a processor 401, a memory 402, and a computer program 403 stored in the memory 402 and executable on the processor 401. When the processor 401 executes the computer program 403, the steps of the above-mentioned method embodiments are implemented. Alternatively, when the processor 401 executes the computer program 403, the functions of the modules / units in the above-mentioned apparatus embodiments are implemented.

[0091] For example, computer program 403 may be divided into one or more modules / units, which are stored in memory 402 and executed by processor 401 to implement the present disclosure. One or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of computer program 403 in computer device 4.

[0092] The computer device 4 can be a desktop computer, a notebook computer, a palmtop computer, a cloud server, etc. The computer device 4 can include but is not limited to a processor 401 and a memory 402. It will be understood by those skilled in the art that Figure 4 This is merely an example of the computer device 4 and does not constitute a limitation on the computer device 4. The computer device 4 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0093] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0094] Memory 402 can be an internal storage unit of computer device 4, such as a hard disk or memory of computer device 4. Memory 402 can also be an external storage device of computer device 4, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on computer device 4. Furthermore, memory 402 can include both an internal storage unit of computer device 4 and an external storage device. Memory 402 is used to store computer programs and other programs and data required by the computer device. Memory 402 can also be used to temporarily store data that has been output or is about to be output.

[0095] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this disclosure. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0096] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0097] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0098] In the embodiments provided in the present disclosure, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of modules or units is merely a logical function division. In actual implementation, there may be other division methods. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which may be electrical, mechanical or other forms.

[0099] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0100] In addition, the functional units in the various embodiments of the present disclosure may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0101] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of the above-mentioned various method embodiments. The computer program may include computer program code, which may be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0102] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present disclosure, and should all be included in the scope of protection of the present disclosure.

Claims

1. A data processing method, characterized in that: The method comprises: Get some audio to be processed; Determining text information and time information corresponding to the plurality of audios to be processed; Determining, based on the text information and time information corresponding to the plurality of audios to be processed, a character string corresponding to the text information and time information corresponding to the character string; determining an abnormal character string based on the character string and the time information corresponding to the character string, wherein the abnormal character string is a character string that is easily misidentified as a wake-up word during the speech recognition process; The determining of an abnormal character string according to the character string and the time information corresponding to the character string includes: Determining, based on all the character strings and the time information corresponding to the character strings, a number of candidate character strings and frequency information corresponding to each candidate character string; wherein each candidate character string is different, and the frequency information corresponding to the candidate character string includes the total number of occurrences of the candidate character string and the number of occurrences of the candidate character string in each time period; For each candidate string, determine the abnormality weight value of the candidate string based on the number of characters and frequency information of the candidate string; According to the abnormality weight values ​​of the candidate character strings, a candidate character string that meets a preset abnormality condition is determined, and the candidate character string that meets the preset abnormality condition is used as an abnormal character string.

2. The method according to claim 1, characterized in that The step of obtaining a plurality of audio files to be processed includes: Using the target device to collect a number of audios to be processed at a preset time; and / or, If the target device detects a preset trigger operation, the target device is used to collect a number of audios to be processed; wherein the preset trigger operation includes at least one of the following: collecting preset facial information, triggering a preset button, and collecting a preset wake-up word.

3. The method according to claim 1, characterized in that The determining of text information and time information corresponding to the plurality of audios to be processed includes: For each audio to be processed, speech recognition processing is performed on the audio to be processed to obtain a text recognition result corresponding to the audio to be processed, and the text recognition result is used as the text information corresponding to the audio to be processed; and the collection time of the audio to be processed is used as the time information corresponding to the audio to be processed.

4. The method according to claim 1, wherein The determining, based on the text information and time information corresponding to the plurality of audios to be processed, a character string corresponding to the text information and time information corresponding to the character string includes: For each text information corresponding to the audio to be processed, the text information corresponding to the audio to be processed is divided and processed to obtain a plurality of character strings corresponding to the text information, and the time information corresponding to the audio to be processed is used as the time information corresponding to the plurality of character strings.

5. The method according to claim 1, wherein Determining the abnormality weight value of the candidate character string based on the number of characters and frequency information of the candidate character string includes: If the number of characters in the candidate character string is less than a first preset threshold, determining an abnormality weight value of the candidate character string based on the total number of occurrences of the candidate character string and the sum of the total number of occurrences of all candidate character strings; If the number of characters in the candidate string is greater than or equal to the first preset threshold, the abnormality weight value of the candidate string is determined based on the number of characters in the candidate string, the total number of occurrences of the candidate string, the number of times the candidate string appears in each time period, the total number of occurrences of candidate strings contained in the candidate string, and the sum of the total number of occurrences of all candidate strings.

6. The method according to any one of claims 1 to 5, characterized in that: Before the step of determining an abnormal character string based on the character string and the time information corresponding to the character string, the method further includes: Cleaning the character string to obtain a cleaned character string; The cleaning process includes removing character strings with a number of characters greater than a second preset threshold and character strings with null values; Accordingly, determining an abnormal character string based on the character string and the time information corresponding to the character string includes: An abnormal character string is determined according to the cleaned character string and time information corresponding to the cleaned character string.

7. A data processing device, characterized in that: The device comprises: An audio acquisition unit, used to acquire a number of audios to be processed; A first determining unit, configured to determine text information and time information corresponding to the plurality of audios to be processed; A second determining unit is configured to determine, based on the text information and time information corresponding to the plurality of audios to be processed, a character string corresponding to the text information and time information corresponding to the character string; an abnormality determination unit, configured to determine an abnormal character string based on the character string and time information corresponding to the character string, wherein the abnormal character string is a character string that is easily misidentified as a wake-up word during speech recognition; The anomaly determination unit is specifically configured to: determine, based on all character strings and time information corresponding to the character strings, a plurality of candidate character strings and frequency information corresponding to each candidate character string; wherein each candidate character string is different, and the frequency information corresponding to the candidate character string includes the total number of occurrences of the candidate character string and the number of occurrences of the candidate character string in each time period; for each candidate character string, determine, based on the number of characters in the candidate character string and the frequency information, an anomaly weight value of the candidate character string; determine, based on the anomaly weight value of each candidate character string, a candidate character string that satisfies a preset anomaly condition, and define the candidate character string that satisfies the preset anomaly condition as an anomaly character string.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images