Recording data processing method and related device

Abstract

The present application relates to speech classification in the field of artificial intelligence technology, and specifically to a method for processing recording data and related devices, the method comprising: obtaining multiple first recording data, and preprocessing the multiple first recording data to obtain multiple second recording data; performing sentiment analysis on the multiple second recording data to obtain multiple sentiment scores; performing text conversion on the multiple second recording data to obtain multiple text contents; scoring the multiple second recording data according to the multiple sentiment scores and the multiple text contents to obtain multiple scoring values; among the multiple scoring values, screening out at least one scoring value that is greater than or equal to a preset scoring value, and determining the set of second recording data corresponding to the screened scoring value as the target recording data set. The use of the embodiments of the present application helps to improve the efficiency and accuracy of locating recordings including appeal information.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a recording data processing method and related devices. Background Art

[0002] At present, the process of locating recordings containing appeal information is as follows: first, randomly select a number of recordings from a large number of recordings; then, manually determine whether each of the recordings contains appeal information; finally, if the recording contains appeal information, manually filter out the recording. This method results in low efficiency and poor accuracy in locating recordings containing appeal information. Summary of the invention

[0003] The embodiments of the present application provide a recording data processing method and related devices for improving the efficiency and accuracy of locating recordings including demand information.

[0004] In a first aspect, an embodiment of the present application provides a method for processing recorded data, comprising:

[0005] Acquire a plurality of first recording data, and pre-process the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data correspond one-to-one to the plurality of first recording data;

[0006] Performing emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of second recording data;

[0007] Performing text conversion on the plurality of second recording data to obtain a plurality of text contents, wherein the plurality of text contents correspond one-to-one to the plurality of second recording data;

[0008] Scoring the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, wherein the plurality of scoring values ​​correspond one-to-one to the plurality of second recording data;

[0009] Among the multiple scoring values, at least one scoring value that is greater than or equal to a preset scoring value is screened out, and a set consisting of the second recording data corresponding to the screened out scoring value is determined as a target recording data set.

[0010] In a second aspect, an embodiment of the present application provides a recording data processing device, including:

[0011] An acquisition unit, used for acquiring a plurality of first recording data;

[0012] A preprocessing unit, configured to preprocess the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data correspond one-to-one to the plurality of first recording data;

[0013] An analysis unit, configured to perform emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of second recording data;

[0014] a conversion unit, configured to perform text conversion on the plurality of second recording data to obtain a plurality of text contents, wherein the plurality of text contents correspond one-to-one to the plurality of second recording data;

[0015] A scoring unit, configured to score the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, wherein the plurality of scoring values ​​correspond one-to-one to the plurality of second recording data;

[0016] The determination unit is used to screen out at least one score value greater than or equal to a preset score value from the multiple score values, and determine a set of second recording data corresponding to the screened score value as a target recording data set.

[0017] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for executing some or all of the steps in the method of the first aspect of the embodiment of the present application.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which is used to store a computer program. The computer program is executed by a processor to implement some or all of the steps described in the method of the first aspect of the embodiment of the present application.

[0019] It can be seen that compared to manually judging whether each recording in a number of recordings randomly selected from a large number of recordings includes appeal information, if the recording includes appeal information, then the recording is manually screened out. In the embodiment of the present application, the recording data processing device scores multiple second recording data to obtain multiple scoring values, and screens out at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines the set of second recording data corresponding to the screened scoring value as the target recording data set. In this way, at least one second recording data including appeal information is automatically selected from multiple second recording data, which helps to improve the efficiency and accuracy of locating recordings including appeal information.

[0020] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0022] Figure 1A It is a flowchart of a method for processing recorded data provided by an embodiment of the present application;

[0023] Figure 1B is a schematic diagram of pre-emphasizing first recording data provided by an embodiment of the present application;

[0024] Figure 2 It is a flowchart of another recording data processing method provided in an embodiment of the present application;

[0025] Figure 3 It is a block diagram of the functional units of a recording data processing device provided in an embodiment of the present application;

[0026] Figure 4 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application.

[0027] Specific implementation method

[0028] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0029] The following are detailed descriptions of each.

[0030] The terms "first", "second", "third" and "fourth" etc. in the specification and claims of the present application and the drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0031] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0032] Electronic devices may include various handheld devices, vehicle-mounted devices, wearable devices, computing devices or other processing devices connected to a wireless modem with wireless communication functions, as well as various forms of user equipment (UE), mobile stations (MS), terminal devices, etc.

[0033] The following is a detailed description of the embodiments of the present application.

[0034] See also Figure 1A , Figure 1A 1 is a flow chart of a recording data processing method provided in an embodiment of the present application. The recording data processing method includes steps 101-105, which are as follows:

[0035] 101: A recording data processing device obtains a plurality of first recording data, and pre-processes the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data corresponds one-to-one to the plurality of first recording data.

[0036] The first recording data may be human-computer interaction recording data or telephone recording data.

[0037] The first recording data may include appeal information or may not include appeal information, and the appeal information includes complaint information and demand information.

[0038] Preprocessing includes pre-emphasis, framing, windowing and endpoint detection. The order of pre-emphasis, framing, windowing and endpoint detection can be: pre-emphasis → framing → windowing → endpoint detection. The above is only an exemplary description of the order of pre-emphasis, framing, windowing and endpoint detection, and should not be understood as a limitation on the order of pre-emphasis, framing, windowing and endpoint detection.

[0039] The recording data processing device pre-processes the plurality of first recording data to obtain the plurality of second recording data in a serial manner or in a parallel manner.

[0040] In a possible example, the recording data processing device pre-processes the plurality of first recording data to obtain a plurality of second recording data, including:

[0041] The recording data processing device performs pre-emphasis on the plurality of first recording data to obtain the plurality of first recording data after pre-emphasis;

[0042] The recording data processing device divides the plurality of first recording data after the pre-emphasis into frames to obtain the plurality of first recording data after the frame division;

[0043] The recording data processing device performs windowing on the plurality of first recording data after the frame division to obtain the plurality of first recording data after the windowing;

[0044] The recording data processing device performs endpoint detection on the plurality of first recording data after the windowing to obtain the plurality of first recording data after the endpoint detection;

[0045] The recording data processing device determines the plurality of first recording data after the endpoint detection as a plurality of second recording data.

[0046] The recording data processing device performs pre-emphasis on the plurality of first recording data, and obtains the plurality of pre-emphasized first recording data in a serial manner or in a parallel manner.

[0047] The recording data processing device divides the plurality of first recording data after pre-emphasis into frames, and obtains the plurality of first recording data after the frame division in a serial manner or in a parallel manner.

[0048] The recording data processing device performs windowing on the plurality of first recording data after frame division, and obtains the plurality of first recording data after windowing in a serial manner or in a parallel manner.

[0049] The recording data processing device performs endpoint detection on the plurality of first recording data after windowing, and the plurality of first recording data after endpoint detection can be obtained in a serial manner or in a parallel manner.

[0050] For example, if Figure 1B As shown, Figure 1B It is a schematic diagram of pre-emphasizing first recording data provided by an embodiment of the present application. The recording data processing device pre-emphasizes the first recording data to obtain the pre-emphasized first recording data, frames the pre-emphasized first recording data to obtain the framed first recording data, windows the framed first recording data to obtain the windowed first recording data, and performs endpoint detection on the windowed first recording data to obtain the endpoint detected first recording data.

[0051] It can be seen that in this example, the recording data processing device performs pre-emphasis, frame division, windowing and endpoint detection on the first recording data to obtain the second recording data, so as to facilitate the subsequent emotion analysis and text conversion based on the second recording data.

[0052] 102: The recording data processing device performs emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, where the plurality of emotion scores correspond one-to-one to the plurality of second recording data.

[0053] The recording data processing device performs emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores in a serial manner or in a parallel manner.

[0054] In a possible example, the recording data processing device performs emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, including:

[0055] The recording data processing device performs format conversion on the plurality of second recording data to obtain the plurality of second recording data after format conversion, wherein the sound file format of each of the plurality of second recording data after format conversion is the target sound file format;

[0056] The recording data processing device slices the plurality of second recording data after the format conversion to obtain a plurality of recording segment sets, wherein the plurality of recording segment sets correspond one-to-one to the plurality of second recording data after the format conversion;

[0057] The recording data processing device inputs the multiple recording segment sets into a pre-stored emotion analysis model for processing to obtain multiple emotion scores, and the multiple emotion scores correspond to the multiple recording segment sets one by one.

[0058] The recording data processing device performs format conversion on the plurality of second recording data, and obtains the plurality of second recording data after format conversion in a serial manner or in a parallel manner.

[0059] The recording data processing device obtains the current sound file format of the second recording data, determines the target format conversion algorithm corresponding to the current sound file format according to the mapping relationship between the sound file format and the format conversion algorithm, and uses the target format conversion algorithm to convert the sound file format of the second recording data from the current sound file format to the target sound file format. The mapping relationship between the sound file format and the format conversion algorithm is pre-stored in the recording data processing device, and the target sound file format can be a waveform sound file (WAV).

[0060] The mapping relationship between the sound file format and the format conversion algorithm is pre-stored in the recording data processing device. The mapping relationship between the sound file format and the format conversion algorithm is shown in Table 1 below:

[0061] Table 1

[0062] First sound file format First format conversion algorithm Second audio file format Second format conversion algorithm Third audio file format Third format conversion algorithm …… ……

[0063] The recording data processing device may slice the plurality of second recording data after format conversion to obtain a plurality of recording segment sets in a serial manner or in a parallel manner.

[0064] The recording data processing device slices the second recording data using a slicing algorithm to obtain a set of recording segments corresponding to the second recording data, wherein the slicing algorithm is pre-stored in the recording data processing device.

[0065] The emotion analysis model is pre-stored in the recording data processing device, and the emotion analysis model is trained and targeted at the appeal information.

[0066] The recording data processing device inputs a plurality of recording segment sets into the emotion analysis model for processing, and obtains a plurality of emotion scores in a serial manner or in a parallel manner.

[0067] The recording data processing device inputs the recording segment set into the emotion analysis model for processing, and the implementation method of obtaining the emotion score can be:

[0068] The recording data processing device inputs a plurality of recording segments included in the recording segment set into the emotion analysis model;

[0069] The recording data processing device extracts emotion features from each of the plurality of recording segments to obtain a plurality of emotion feature quantities, wherein the plurality of emotion feature quantities correspond to the plurality of recording segments one by one;

[0070] The recording data processing device determines a plurality of emotion scores corresponding to a plurality of emotion feature quantities according to a mapping relationship between the emotion feature quantity range and the emotion score;

[0071] The recording data processing device determines the sum of the multiple emotion scores as the emotion score corresponding to the recording segment set.

[0072] The above is merely an exemplary description of inputting a set of recording clips into a sentiment analysis model for processing to obtain sentiment scores, and should not be understood as a limitation on inputting a set of recording clips into a sentiment analysis model for processing to obtain sentiment scores.

[0073] The mapping relationship between the range of emotion feature quantity and emotion score is pre-stored in the emotion analysis model. The mapping relationship between the range of emotion feature quantity and emotion score is shown in Table 2 below:

[0074] Table 2

[0075] Range of emotional characteristics Sentiment score ＜T1 First emotion score ≥T1 and <T2 Second emotion score ≥T2 and <T3 Third emotion score ≥T3 Fourth Emotion Score

[0076] The size of the emotion scores: the first emotion score < the second emotion score < the third emotion score < the fourth emotion score.

[0077] 103: The recording data processing device performs text conversion on the plurality of second recording data to obtain a plurality of text contents, and the plurality of text contents correspond to the plurality of second recording data in a one-to-one manner.

[0078] The recording data processing device performs text conversion on the plurality of second recording data to obtain the plurality of text contents in a serial manner or in a parallel manner.

[0079] The recording data processing device uses a speech-to-text conversion algorithm to perform text conversion on the second recording data to obtain text content corresponding to the second recording data; wherein the speech-to-text conversion algorithm is pre-stored in the recording data processing device.

[0080] 104: The recording data processing device scores the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, wherein the plurality of scoring values ​​correspond one-to-one to the plurality of second recording data.

[0081] The recording data processing device scores the multiple second recording data according to the multiple emotion scores and the multiple text contents, and the multiple score values ​​can be obtained in a serial manner or in a parallel manner.

[0082] In a possible example, the recording data processing device scores the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, including:

[0083] The recording data processing device inputs each of the multiple text contents into multiple pre-stored keyword models for processing to obtain multiple keyword matching results, wherein the multiple keyword matching results correspond to the multiple text contents one by one;

[0084] The recording data processing device determines a plurality of text scores according to the plurality of keyword matching results, wherein the plurality of text scores correspond one to one with the plurality of keyword matching results;

[0085] The recording data processing device determines a plurality of scoring values ​​according to the plurality of emotion scores, the plurality of text scores and a pre-stored scoring formula.

[0086] The recording data processing device inputs each of the multiple text contents into multiple keyword models for processing, and obtains multiple keyword matching results in a serial manner or in a parallel manner.

[0087] Multiple keyword models are pre-stored in the recording data processing device, each of the multiple keyword models is targeted at the demand information, and the multiple keyword models correspond one-to-one to multiple demand scenes, and the multiple demand scenes include angry scenes, threatening scenes, scolding scenes, etc.

[0088] The scoring formula is pre-stored in the recording data processing device, and the scoring formula is:

[0089] S=A×α+B×β,

[0090] S is the score value corresponding to the second recording data, A is the emotion score corresponding to the second recording data, α is the weight corresponding to the emotion score, B is the text score corresponding to the second recording data, β is the weight corresponding to the text score, and α+β=1.

[0091] In a possible example, the recording data processing device determines a text score according to a keyword matching result, including:

[0092] The recording data processing device obtains a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models;

[0093] The recording data processing device determines whether each of the plurality of keyword matching quantities is zero;

[0094] If so, the recording data processing device determines that the text score corresponding to the keyword matching result is the difference between the emotion score corresponding to the keyword matching result and the pre-stored penalty score.

[0095] The penalty score is pre-stored in the recording data processing device, for example, the penalty score may be 0.2.

[0096] In a possible example, the recording data processing device determines a text score according to a keyword matching result, including:

[0097] The recording data processing device obtains a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models;

[0098] The recording data processing device determines whether each of the plurality of keyword matching quantities is zero;

[0099] If not, the recording data processing device determines at least one target weight corresponding to at least one target keyword matching number according to the pre-stored weight ratio of the plurality of keyword models, and each target keyword matching number in the at least one target keyword matching number is not zero;

[0100] The recording data processing device determines the text score corresponding to the keyword matching result according to the at least one target keyword matching quantity, the at least one target weight and a pre-stored text scoring formula.

[0101] At least one target weight corresponds to at least one target keyword matching quantity.

[0102] For example, if the weight ratio of at least one keyword model corresponding to at least one target keyword matching number is 3:2:2, then at least one target weight corresponding to at least one target keyword matching number is 3 / (3+2+2):2 / (3+2+2):2 / (3+2+2).

[0103] The text scoring formula is pre-stored in the recording data processing device. If the number of at least one target weight is M, then the text scoring formula is:

[0104] P=C1×γ1+C2×γ2+……+C M ×γ M ,

[0105] P is the text score corresponding to the keyword matching result, C1 is the first target keyword matching number among M target keyword matching numbers, γ1 is the first target weight among M target weights, C2 is the second target keyword matching number among M target keyword matching numbers, γ2 is the second target weight among M target weights, C M is the number of M target keyword matches among the M target keyword matches, γ M is the Mth target weight among M target weights.

[0106] 105: The recording data processing device selects at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines a set of second recording data corresponding to the selected scoring value as a target recording data set.

[0107] It can be seen that compared to manually judging whether each recording in a number of recordings randomly selected from a large number of recordings includes appeal information, if the recording includes appeal information, then the recording is manually screened out. In the embodiment of the present application, the recording data processing device scores multiple second recording data to obtain multiple scoring values, and screens out at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines the set of second recording data corresponding to the screened scoring value as the target recording data set. In this way, at least one second recording data including appeal information is automatically selected from multiple second recording data, which helps to improve the efficiency and accuracy of locating recordings including appeal information.

[0108] In a possible example, the recording data processing device selects at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines a set of second recording data corresponding to the selected scoring value as a target recording data set, including:

[0109] The recording data processing device determines the average value or the middle value of the multiple scoring values ​​as the preset scoring value;

[0110] The recording data processing device selects at least one score value greater than or equal to the preset score value from the multiple score values;

[0111] The recording data processing device selects at least one second recording data corresponding to at least one scoring value from the plurality of second recording data, and the at least one second recording data corresponds to the at least one scoring value one-to-one;

[0112] The recording data processing device determines the set consisting of the at least one second recording data as the target recording data set.

[0113] It can be seen that in this example, the recording data processing device determines the average value or median value of multiple scoring values ​​as the preset scoring value, screens out at least one scoring value that is greater than or equal to the preset scoring value among the multiple scoring values, and determines a set consisting of at least one second recording data corresponding to at least one scoring value among the multiple second recording data as the target recording data set. This ensures that the target recording data set contains a sufficient number of recording data that include demand information, so as to facilitate the subsequent discovery and resolution of problems based on the target recording data set.

[0114] With the above Figure 1A For details on the embodiments shown in the drawings, see Figure 2 , Figure 2 2 is a flow chart of another recording data processing method provided in an embodiment of the present application. The recording data processing method includes steps 201-210, which are as follows:

[0115] 201: The recording data processing device obtains a plurality of first recording data.

[0116] 202: The recording data processing device pre-processes the plurality of first recording data to obtain a plurality of second recording data, where the plurality of second recording data correspond one-to-one to the plurality of first recording data.

[0117] 203: The recording data processing device performs format conversion on the plurality of second recording data to obtain the plurality of second recording data after format conversion, wherein the sound file format of each of the plurality of second recording data after format conversion is the target sound file format.

[0118] 204: The recording data processing device slices the plurality of second recording data after the format conversion to obtain a plurality of recording segment sets, wherein the plurality of recording segment sets correspond one-to-one to the plurality of second recording data after the format conversion.

[0119] 205: The recording data processing device inputs the plurality of recording segment sets into a pre-stored emotion analysis model for processing to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of recording segment sets.

[0120] 206: The recording data processing device performs text conversion on the plurality of second recording data to obtain a plurality of text contents, and the plurality of text contents correspond to the plurality of second recording data in a one-to-one manner.

[0121] 207: The recording data processing device inputs each of the multiple text contents into multiple pre-stored keyword models for processing to obtain multiple keyword matching results, and the multiple keyword matching results correspond to the multiple text contents one by one.

[0122] 208: The recording data processing apparatus determines a plurality of text scores according to the plurality of keyword matching results, wherein the plurality of text scores correspond one to one with the plurality of keyword matching results.

[0123] 209: The recording data processing device determines a plurality of rating values ​​according to the plurality of emotion ratings, the plurality of text ratings and a pre-stored rating formula, wherein the plurality of rating values ​​correspond one-to-one to the plurality of second recording data.

[0124] 210: The recording data processing device selects at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines a set of second recording data corresponding to the selected scoring value as a target recording data set.

[0125] It should be noted that Figure 2 The specific implementation process of each step of the method shown can refer to the specific implementation process described in the above method, which will not be described here.

[0126] The above embodiments mainly introduce the scheme of the embodiments of the present application from the perspective of the execution process on the method side. It is understandable that in order to realize the above functions, the recording data processing device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0127] The embodiment of the present application can divide the recording data processing device into functional units according to the method example. For example, each functional unit can be divided according to each function, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in the form of hardware or in the form of software functional units. It should be noted that the division of units in the embodiment of the present application is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0128] See also Figure 3 , Figure 3 : is a functional unit block diagram of a recording data processing device provided in an embodiment of the present application, and the recording data processing device 300 includes:

[0129] An acquisition unit 301 is used to acquire a plurality of first recording data;

[0130] A preprocessing unit 302 is used to preprocess the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data correspond to the plurality of first recording data one by one;

[0131] An analysis unit 303 is used to perform emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of second recording data;

[0132] A conversion unit 304, configured to perform text conversion on the plurality of second recording data to obtain a plurality of text contents, wherein the plurality of text contents correspond to the plurality of second recording data in a one-to-one manner;

[0133] A scoring unit 305, configured to score the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, wherein the plurality of scoring values ​​correspond one-to-one to the plurality of second recording data;

[0134] The determination unit 306 is configured to select at least one score value greater than or equal to a preset score value from the multiple score values, and determine a set of second recording data corresponding to the selected score value as a target recording data set.

[0135] It can be seen that compared to manually judging whether each recording in a number of recordings randomly selected from a large number of recordings includes appeal information, if the recording includes appeal information, then the recording is manually screened out. In the embodiment of the present application, the recording data processing device scores multiple second recording data to obtain multiple scoring values, and screens out at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines the set of second recording data corresponding to the screened scoring value as the target recording data set. In this way, at least one second recording data including appeal information is automatically selected from multiple second recording data, which helps to improve the efficiency and accuracy of locating recordings including appeal information.

[0136] In a possible example, in preprocessing the plurality of first recording data to obtain the plurality of second recording data, the preprocessing unit 302 is specifically configured to:

[0137] Pre-emphasize the plurality of first recording data to obtain the plurality of pre-emphasized first recording data;

[0138] Frame the plurality of first recording data after the pre-emphasis to obtain the plurality of first recording data after the frame division;

[0139] Performing windowing on the plurality of first recording data after the frame division to obtain the plurality of first recording data after the windowing;

[0140] Performing endpoint detection on the plurality of first recording data after the windowing to obtain the plurality of first recording data after the endpoint detection;

[0141] The plurality of first recording data after the endpoint detection is determined as a plurality of second recording data.

[0142] In a possible example, in performing emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, the analysis unit 303 is specifically configured to:

[0143] Performing format conversion on the plurality of second recording data to obtain the plurality of second recording data after format conversion, wherein the sound file format of each of the plurality of second recording data after format conversion is the target sound file format;

[0144] Slicing the plurality of second recording data after the format conversion to obtain a plurality of recording segment sets, wherein the plurality of recording segment sets correspond one-to-one to the plurality of second recording data after the format conversion;

[0145] The multiple recording segment sets are input into a pre-stored emotion analysis model for processing to obtain multiple emotion scores, and the multiple emotion scores correspond one-to-one to the multiple recording segment sets.

[0146] In a possible example, in terms of scoring the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, the scoring unit 305 is specifically configured to:

[0147] Input each of the plurality of text contents into a plurality of pre-stored keyword models for processing to obtain a plurality of keyword matching results, wherein the plurality of keyword matching results correspond one-to-one to the plurality of text contents;

[0148] Determine a plurality of text scores according to the plurality of keyword matching results, wherein the plurality of text scores correspond one to one to the plurality of keyword matching results;

[0149] A plurality of rating values ​​are determined according to the plurality of sentiment ratings, the plurality of text ratings, and a pre-stored rating formula.

[0150] In a possible example, in determining the text score according to the keyword matching result, the scoring unit 305 is specifically used to:

[0151] Acquire a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models;

[0152] Determining whether each of the plurality of keyword matching quantities is zero;

[0153] If so, it is determined that the text score corresponding to the keyword matching result is the difference between the sentiment score corresponding to the keyword matching result and the pre-stored penalty score.

[0154] In a possible example, in determining the text score according to the keyword matching result, the scoring unit 305 is specifically used to:

[0155] Acquire a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models;

[0156] Determining whether each of the plurality of keyword matching quantities is zero;

[0157] If not, determining at least one target weight corresponding to at least one target keyword matching quantity according to the pre-stored weight ratios of the plurality of keyword models, wherein each target keyword matching quantity in the at least one target keyword matching quantity is not zero;

[0158] The text score corresponding to the keyword matching result is determined according to the at least one target keyword matching quantity, the at least one target weight and a pre-stored text scoring formula.

[0159] In one possible example, in terms of screening out at least one score value greater than or equal to a preset score value from the multiple score values, and determining a set of second recording data corresponding to the screened score value as the target recording data set, the determination unit 306 is specifically used to:

[0160] Determine an average value or a middle value of the multiple scoring values ​​as a preset scoring value;

[0161] Filtering out at least one scoring value that is greater than or equal to the preset scoring value from the multiple scoring values;

[0162] Selecting at least one piece of second recording data corresponding to the at least one scoring value from the plurality of pieces of second recording data, wherein the at least one piece of second recording data corresponds to the at least one scoring value in a one-to-one manner;

[0163] A set consisting of the at least one second recording data is determined as a target recording data set.

[0164] With the above Figure 1A and Figure 2 For details on the embodiments shown in the drawings, see Figure 4 , Figure 4 4 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device 400 includes a processor, a memory, a communication interface, and one or more programs. The one or more programs are stored in the memory and configured to be executed by the processor. The program includes instructions for executing the following steps:

[0165] Acquire a plurality of first recording data, and pre-process the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data correspond one-to-one to the plurality of first recording data;

[0166] Performing emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of second recording data;

[0167] Performing text conversion on the plurality of second recording data to obtain a plurality of text contents, wherein the plurality of text contents correspond one-to-one to the plurality of second recording data;

[0168] Scoring the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, wherein the plurality of scoring values ​​correspond one-to-one to the plurality of second recording data;

[0169] Among the multiple scoring values, at least one scoring value that is greater than or equal to a preset scoring value is screened out, and a set consisting of the second recording data corresponding to the screened out scoring value is determined as a target recording data set.

[0170] It can be seen that compared to manually judging whether each recording in a number of recordings randomly selected from a large number of recordings includes appeal information, if the recording includes appeal information, then the recording is manually screened out. In the embodiment of the present application, the electronic device scores multiple second recording data to obtain multiple scoring values, and screens out at least one scoring value greater than or equal to a preset scoring value from the multiple scoring values, and determines the set of second recording data corresponding to the screened scoring value as the target recording data set. In this way, at least one second recording data including appeal information is automatically selected from multiple second recording data, which helps to improve the efficiency and accuracy of locating recordings including appeal information.

[0171] In a possible example, in terms of preprocessing the plurality of first recording data to obtain the plurality of second recording data, the program includes instructions specifically for executing the following steps:

[0172] Pre-emphasize the plurality of first recording data to obtain the plurality of pre-emphasized first recording data;

[0173] Framing the plurality of first recording data after the pre-emphasis to obtain the plurality of first recording data after framing;

[0174] Performing windowing on the plurality of first recording data after the frame division to obtain the plurality of first recording data after the windowing;

[0175] Performing endpoint detection on the plurality of first recording data after the windowing to obtain the plurality of first recording data after the endpoint detection;

[0176] The plurality of first recording data after the endpoint detection is determined as a plurality of second recording data.

[0177] In a possible example, in terms of performing emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, the program includes instructions specifically for executing the following steps:

[0178] Performing format conversion on the plurality of second recording data to obtain the plurality of second recording data after format conversion, wherein the sound file format of each of the plurality of second recording data after format conversion is the target sound file format;

[0179] Slicing the plurality of second recording data after the format conversion to obtain a plurality of recording segment sets, wherein the plurality of recording segment sets correspond one-to-one to the plurality of second recording data after the format conversion;

[0180] The multiple recording segment sets are input into a pre-stored emotion analysis model for processing to obtain multiple emotion scores, and the multiple emotion scores correspond one-to-one to the multiple recording segment sets.

[0181] In a possible example, in terms of scoring the plurality of second recording data according to the plurality of emotion scores and the plurality of text contents to obtain a plurality of scoring values, the program includes instructions specifically for executing the following steps:

[0182] Input each of the plurality of text contents into a plurality of pre-stored keyword models for processing to obtain a plurality of keyword matching results, wherein the plurality of keyword matching results correspond one-to-one to the plurality of text contents;

[0183] Determine a plurality of text scores according to the plurality of keyword matching results, wherein the plurality of text scores correspond one to one to the plurality of keyword matching results;

[0184] A plurality of rating values ​​are determined according to the plurality of sentiment ratings, the plurality of text ratings, and a pre-stored rating formula.

[0185] In a possible example, in terms of determining a text score according to a keyword matching result, the program includes instructions for specifically performing the following steps:

[0186] Acquire a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models;

[0187] Determining whether each of the plurality of keyword matching quantities is zero;

[0188] If so, it is determined that the text score corresponding to the keyword matching result is the difference between the sentiment score corresponding to the keyword matching result and the pre-stored penalty score.

[0189] In a possible example, in terms of determining a text score according to a keyword matching result, the program includes instructions for specifically performing the following steps:

[0190] Acquire a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models;

[0191] Determining whether each of the plurality of keyword matching quantities is zero;

[0192] If not, determining at least one target weight corresponding to at least one target keyword matching quantity according to the pre-stored weight ratios of the plurality of keyword models, wherein each target keyword matching quantity in the at least one target keyword matching quantity is not zero;

[0193] The text score corresponding to the keyword matching result is determined according to the at least one target keyword matching quantity, the at least one target weight and a pre-stored text scoring formula.

[0194] In one possible example, in terms of screening out at least one score value greater than or equal to a preset score value from the multiple score values, and determining a set of second recording data corresponding to the screened score value as a target recording data set, the program includes instructions specifically for performing the following steps:

[0195] Determine an average value or a middle value of the multiple scoring values ​​as a preset scoring value;

[0196] Filtering out at least one scoring value that is greater than or equal to the preset scoring value from the multiple scoring values;

[0197] Selecting at least one piece of second recording data corresponding to the at least one scoring value from the plurality of pieces of second recording data, wherein the at least one piece of second recording data corresponds to the at least one scoring value in a one-to-one manner;

[0198] A set consisting of the at least one second recording data is determined as a target recording data set.

[0199] An embodiment of the present application also provides a computer storage medium, which is used to store a computer program. The computer program is executed by a processor to implement part or all of the steps of any method recorded in the above method embodiment. The computer includes an electronic device.

[0200] The embodiment of the present application also provides a computer program product, the computer program product includes a non-transitory computer-readable storage medium storing a computer program, the computer program is operable to cause a computer to execute some or all of the steps of any method described in the method embodiment. The computer program product may be a software installation package, and the computer includes an electronic device.

[0201] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0202] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0203] In the several embodiments provided in the present application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the above-mentioned units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0204] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0205] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0206] If the above-mentioned integrated 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 memory. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a memory, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the above-mentioned methods of each embodiment of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or CD-ROM and other media that can store program codes.

[0207] A person skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable memory, and the memory can include: a flash drive, a read-only memory (English: Read-Only Memory, abbreviated as: ROM), a random access memory (English: Random Access Memory, abbreviated as: RAM), a magnetic disk or an optical disk, etc.

[0208] The embodiments of the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for processing recorded data, characterized in that: include: Acquire a plurality of first recording data, and pre-process the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data correspond one-to-one to the plurality of first recording data; Performing emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of second recording data; Performing text conversion on the plurality of second recording data to obtain a plurality of text contents, wherein the plurality of text contents correspond one-to-one to the plurality of second recording data; Inputting each of the plurality of text contents into a plurality of pre-stored keyword models for processing to obtain a plurality of keyword matching results, wherein the plurality of keyword matching results correspond one-to-one to the plurality of text contents; Determine a plurality of text scores according to the plurality of keyword matching results, wherein the plurality of text scores correspond one to one to the plurality of keyword matching results; The text score is determined based on the keyword matching results, including: Acquire a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models; Determining whether each of the plurality of keyword matching quantities is zero; If yes, determining that the text score corresponding to the keyword matching result is the difference between the sentiment score corresponding to the keyword matching result and the pre-stored penalty score; Determining a plurality of rating values ​​according to the plurality of sentiment ratings, the plurality of text ratings, and a pre-stored rating formula; Among the multiple scoring values, at least one scoring value that is greater than or equal to a preset scoring value is screened out, and a set consisting of the second recording data corresponding to the screened out scoring value is determined as a target recording data set.

2. The method according to claim 1, characterized in that The preprocessing of the plurality of first recording data to obtain a plurality of second recording data includes: Pre-emphasize the plurality of first recording data to obtain the plurality of pre-emphasized first recording data; Frame the plurality of first recording data after the pre-emphasis to obtain the plurality of first recording data after the frame division; Performing windowing on the plurality of first recording data after the frame division to obtain the plurality of first recording data after the windowing; Performing endpoint detection on the plurality of first recording data after the windowing to obtain the plurality of first recording data after the endpoint detection; The plurality of first recording data after the endpoint detection is determined as a plurality of second recording data.

3. The method according to claim 2, characterized in that The performing emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores includes: Performing format conversion on the plurality of second recording data to obtain the plurality of second recording data after format conversion, wherein the sound file format of each of the plurality of second recording data after format conversion is the target sound file format; Slicing the plurality of second recording data after the format conversion to obtain a plurality of recording segment sets, wherein the plurality of recording segment sets correspond one-to-one to the plurality of second recording data after the format conversion; The multiple recording segment sets are input into a pre-stored emotion analysis model for processing to obtain multiple emotion scores, and the multiple emotion scores correspond one-to-one to the multiple recording segment sets.

4. The method according to claim 1, characterized in that: Determining the text score according to the keyword matching result also includes: If each of the plurality of keyword matching quantities is not zero, determining at least one target weight corresponding to at least one target keyword matching quantity according to the pre-stored weight ratios of the plurality of keyword models, wherein each of the at least one target keyword matching quantity is not zero; The text score corresponding to the keyword matching result is determined according to the at least one target keyword matching quantity, the at least one target weight and a pre-stored text scoring formula.

5. The method according to claim 1 or 4, characterized in that: The step of selecting at least one score value greater than or equal to a preset score value from the plurality of score values, and determining a set of second recording data corresponding to the selected score value as a target recording data set includes: Determine an average value or a middle value of the multiple scoring values ​​as a preset scoring value; Filtering out at least one scoring value that is greater than or equal to the preset scoring value from the multiple scoring values; Selecting at least one piece of second recording data corresponding to the at least one scoring value from the plurality of pieces of second recording data, wherein the at least one piece of second recording data corresponds to the at least one scoring value in a one-to-one manner; A set consisting of the at least one second recording data is determined as a target recording data set.

6. A recording data processing device, characterized in that: include: An acquisition unit, used for acquiring a plurality of first recording data; A preprocessing unit, configured to preprocess the plurality of first recording data to obtain a plurality of second recording data, wherein the plurality of second recording data correspond one-to-one to the plurality of first recording data; An analysis unit, configured to perform emotion analysis on the plurality of second recording data to obtain a plurality of emotion scores, wherein the plurality of emotion scores correspond one-to-one to the plurality of second recording data; a conversion unit, configured to perform text conversion on the plurality of second recording data to obtain a plurality of text contents, wherein the plurality of text contents correspond one-to-one to the plurality of second recording data; A scoring unit, configured to input each of the plurality of text contents into a plurality of pre-stored keyword models for processing, and obtain a plurality of keyword matching results, wherein the plurality of keyword matching results correspond one-to-one to the plurality of text contents; Determine a plurality of text scores according to the plurality of keyword matching results, wherein the plurality of text scores correspond one-to-one to the plurality of keyword matching results; determine a plurality of scoring values ​​according to the plurality of emotion scores, the plurality of text scores and a pre-stored scoring formula; In determining the text score according to the keyword matching result, the scoring unit is specifically used for: Acquire a plurality of keyword matching quantities included in the keyword matching result, wherein the plurality of keyword matching quantities correspond one-to-one to the plurality of keyword models; Determining whether each of the plurality of keyword matching quantities is zero; If yes, determining that the text score corresponding to the keyword matching result is the difference between the sentiment score corresponding to the keyword matching result and the pre-stored penalty score; The determination unit is used to screen out at least one score value greater than or equal to a preset score value from the multiple score values, and determine a set of second recording data corresponding to the screened score value as a target recording data set.

7. An electronic device, characterized in that: The method comprises a processor, a memory, a communication interface and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for executing part or all of the steps in the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store a computer program, and the computer program is executed by a processor to implement the method according to any one of claims 1 to 5.

Images