Interactive data processing method and processing system

Abstract

The invention discloses an interactive data processing method and system, and relates to the technical field of intelligent voice recognition. The method comprises the following steps: collecting user language information, and converting the voice information into text information; performing semantic analysis and understanding on the converted text information by using a natural language processing technology; according to the historical interaction record and the current context of the user, making a personalized interaction strategy, and generating questions and answers; and synthesizing questions and answers into voice and outputting to the user. By means of the interactive processing method, more natural, smoother and more intelligent interactive experience can be achieved, and the satisfaction degree and the use efficiency of users are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent speech recognition technology, and in particular to an interactive data processing method and processing system. Background Art

[0002] With the increasing development of voice recognition technology and artificial intelligence, intelligent voice assistants are widely used in people's lives and work. Intelligent voice assistants are applications that help users solve problems through intelligent interactions with intelligent dialogue and instant questions and answers. Intelligent voice assistants enable users to operate devices or obtain information through voice commands through voice control programs, thereby providing convenience and efficiency.

[0003] However, existing intelligent voice recognition assistants still have certain limitations in interactivity and cannot meet the increasingly complex and diverse needs of users. Summary of the invention

[0004] The present invention provides an interactive data processing method, comprising:

[0005] Collect user language information and convert voice information into text information;

[0006] Use natural language processing technology to perform semantic analysis and understanding on the converted text information;

[0007] Develop personalized interaction strategies and generate questions and answers based on the user's historical interaction records and current context;

[0008] The questions and answers are synthesized into speech output to the user.

[0009] The interactive data processing method as described above, wherein the voice information is converted into text information, specifically comprises the following sub-steps:

[0010] Build a speech recognition model and use a large amount of annotated speech data to input the speech recognition model for model training so that it can learn the correspondence between speech and text;

[0011] Acquire user voice input through a microphone device and perform audio preprocessing on the acquired voice information;

[0012] Extracting representative acoustic features from the preprocessed speech;

[0013] The extracted representative acoustic features are input into the speech recognition model, and the predicted text sequence is output. The output text sequence is corrected, punctuation is added, and grammar is optimized to improve the readability and accuracy of the text.

[0014] In the interactive data processing method as described above, the method for constructing the speech recognition model specifically includes the following sub-steps:

[0015] Collect a large amount of speech data, including audio data of different speakers, accents, speaking speeds, and background noise, and pre-process the audio data, including audio cropping, normalization, and framing operations;

[0016] Extract meaningful features from speech data, including Mel-frequency cepstral coefficients (MFCC) and perceptual linear prediction (PLP);

[0017] Select a suitable deep learning model and input the extracted features into the deep learning model for model training;

[0018] Separate the test set from the extracted features, use the test set to evaluate the trained model, and adjust the model hyperparameters, including learning rate, number of layers, and number of neurons, based on the evaluation results to optimize model performance.

[0019] In the above interactive data processing method, text sequence optimization is performed, which specifically includes the following sub-steps:

[0020] Use the natural language processing library to perform lexical analysis on the input text sequence, including word segmentation and part-of-speech tagging;

[0021] Formulate grammar rules and use syntax tree analysis algorithm to detect grammatical errors in text sequences based on the defined grammar rules;

[0022] Establish a grammar library of common words and replace the marked complex words with common words according to the context and semantics to improve the accuracy and richness of expression;

[0023] Analyze sentence structure and split or reorganize long sentences to make them clearer and easier to read;

[0024] Check whether the punctuation is correct based on the sentence grammar and logic, and correct incorrect punctuation.

[0025] An interactive data processing method as described above, wherein, in order to improve the accuracy of the speech recognition model, it also includes training the model using a large amount of text with correct standard grammar, and continuously updating and improving grammar rules and optimization strategies based on new error types and optimization requirements.

[0026] The present invention also provides an interactive data processing system, comprising:

[0027] An interactive module, used to collect user language information and convert voice information into text information;

[0028] Semantic analysis module, used to use natural language processing technology to perform semantic analysis and understanding on the converted text information;

[0029] The interaction strategy generation module is used to formulate personalized interaction strategies and generate questions and answers based on the user's historical interaction records and current context;

[0030] The output module is used to synthesize questions and answers into speech and output it to the user.

[0031] An interactive data processing system as described above, wherein the interactive module is specifically used to: construct a speech recognition model, use a large amount of annotated speech data to input the speech recognition model for model training, so that it can learn the correspondence between speech and text; obtain user voice input through a microphone device, and perform audio preprocessing on the obtained voice information; extract representative acoustic features from the preprocessed speech; input the extracted representative acoustic features into the speech recognition model, output a predicted text sequence, and perform error correction, punctuation addition, and grammar optimization on the output text sequence to improve the readability and accuracy of the text.

[0032] An interactive data processing system as described above, wherein the semantic analysis module is specifically used to: collect a large amount of speech data, including audio data of different speakers, accents, speaking speeds, and background noise, and preprocess the audio data, including audio cropping, normalization, and framing operations; extract meaningful features from the speech data, including Mel-frequency cepstral coefficients MFCC and perceptual linear prediction PLP; select a suitable deep learning model, and input the extracted features into the deep learning model for model training; separate a test set from the extracted features, use the test set to evaluate the trained model, and adjust the model hyperparameters according to the evaluation results, including the learning rate, number of layers, and number of neurons, to optimize the model performance.

[0033] An interactive data processing system as described above, wherein the semantic analysis module is specifically used to use a natural language processing library to perform lexical analysis on an input text sequence, including word segmentation and part-of-speech tagging; formulate grammatical rules, and use a syntax tree analysis algorithm to detect grammatical errors in the text sequence based on the defined grammatical rules; establish a grammar library for common words, and replace the marked complex words with common words according to the context and semantics to improve the accuracy and richness of expression; analyze sentence structure, split or reorganize long sentences to make them clearer and easier to read; detect whether punctuation is correct based on sentence grammar and logic, and correct incorrect punctuation.

[0034] An interactive data processing system as described above, wherein, in order to improve the accuracy of the speech recognition model, it also includes training the model using a large amount of text with correct standard grammar, and continuously updating and improving grammar rules and optimization strategies based on new error types and optimization requirements.

[0035] The beneficial effects achieved by the present invention are as follows: Through the interactive processing method of the present invention, a more natural, smooth and intelligent interactive experience can be achieved, thereby improving user satisfaction and usage efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0037] Figure 1 This is a flow chart of an interactive data processing method provided in Example 1 of the present application;

[0038] Figure 2 It is a flow chart of a method for converting voice information into text information;

[0039] Figure 3 It is a flow chart of the method of using natural language processing technology to perform semantic analysis and understanding on the converted text information;

[0040] Figure 4 It is a flow chart of a method for developing personalized interaction strategies based on the user's historical interaction records and current context. DETAILED DESCRIPTION

[0041] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0042] Embodiment 1

[0043] like Figure 1 As shown, the first embodiment of the present application provides an interactive data processing method, including:

[0044] Step 110: Collect user language information and convert the voice information into text information;

[0045] Step 120: Using natural language processing technology to perform semantic analysis and understanding on the converted text information;

[0046] Step 130: formulate a personalized interaction strategy and generate questions and answers based on the user's historical interaction records and current context;

[0047] Step 140: synthesize the question and answer into speech and output it to the user.

[0048] like Figure 2 As shown, converting voice information into text information specifically includes the following sub-steps:

[0049] Step 210: construct a speech recognition model, and use a large amount of annotated speech data to input the speech recognition model for model training so that it can learn the correspondence between speech and text;

[0050] The method for constructing the speech recognition model specifically includes:

[0051] Step S11, collecting a large amount of speech data, including audio data of different speakers, accents, speaking speeds, background noises, etc., and preprocessing the audio data, such as audio cropping, normalization, framing, etc.;

[0052] Step S12, extracting meaningful features from the speech data, such as Mel-frequency cepstral coefficients MFCC, perceptual linear prediction PLP, etc.;

[0053] Step S13: Select a suitable deep learning model, such as a recurrent neural network (RNN), a long short-term memory (LSTM), a gated recurrent unit (GRU), or a combination of a convolutional neural network (CNN) and RNN, and input the extracted features into the deep learning model for model training;

[0054] Step S14: Separate a test set from the extracted features, use the test set to evaluate the trained model, and adjust the model hyperparameters such as learning rate, number of layers, number of neurons, etc. according to the evaluation results to optimize the model performance.

[0055] Step 220: obtaining user voice input through a microphone device, and performing audio preprocessing on the obtained voice information;

[0056] Step 230: extracting representative acoustic features from the preprocessed speech;

[0057] Representative acoustic features include Mel-frequency cepstral coefficients (MFCC), perceptual linear prediction (PLP), etc.

[0058] Step 240: Input the extracted representative acoustic features into a speech recognition model, output a predicted text sequence, and perform error correction, punctuation addition, grammar optimization, etc. on the output text sequence to improve the readability and accuracy of the text;

[0059] The text sequence optimization specifically includes the following sub-steps:

[0060] Step S21, using a natural language processing library to perform lexical analysis on the input text sequence, including word segmentation, part-of-speech tagging, etc.;

[0061] Step S22: Formulate grammar rules, and use a grammar tree analysis algorithm to detect possible grammatical errors in the text sequence according to the defined grammar rules;

[0062] Grammatical rules include but are not limited to the completeness of the subject-predicate-object structure, the importance of word-part collocation, and the principles of correct use of clauses.

[0063] Step S23: Establish a common word grammar library, and replace the marked complex words with common words according to the context and semantics to improve the accuracy and richness of expression;

[0064] Step S24, analyzing the sentence structure, splitting or reorganizing long sentences to make them clearer and easier to read;

[0065] Step S25: Check whether the punctuation is correct according to the sentence grammar and logic, and correct incorrect punctuation;

[0066] In addition, in order to improve the accuracy of the speech recognition model, the model is trained using a large amount of text with correct standard grammar, and the grammar rules and optimization strategies are continuously updated and improved based on new error types and optimization needs.

[0067] like Figure 3 As shown, natural language processing technology is used to perform semantic analysis and understanding on the converted text information, which specifically includes the following sub-steps:

[0068] Step 310: segment the text information into basic units such as sentences and words;

[0069] Step 320: Analyze the sentence structure, determine the subject, predicate, object and other components in the sentence, and mark the part of speech for each word, such as verb, noun, adjective, etc., identify specific entities such as names of people, places, and organization names in the text, build a text tree, and use word vectors to convert elements in the text tree into vector form;

[0070] Step 330: perform text classification and sentiment analysis based on the text content, determine the role of each entity in the action, and extract key information in the text, such as events, relationships, etc.;

[0071] Step 340: Combine the knowledge graph to associate entities and concepts in the text with information in the knowledge base for deeper semantic understanding.

[0072] like Figure 4 As shown in the figure, a personalized interaction strategy is formulated based on the user's historical interaction records and current context, including the following sub-steps:

[0073] Step 410: Comprehensively collect historical user interaction information, including user input text, system response, interaction time, scenario, etc.

[0074] Step 420: extracting key elements of interaction in the current context, such as topics, user emotions, scenes, etc., to determine the user's preferences and behavior patterns in the current context;

[0075] Step 430: Create a user profile based on historical interaction information, including user interests, habits, language style, knowledge level, etc.;

[0076] Step 440: Analyze historical interaction information to identify common interaction patterns and rules of users;

[0077] Step 450: formulate a personalized interaction strategy based on user portrait, context analysis and interaction rules;

[0078] For example, if the user has shown a good response to humorous styles in historical interactions, then the current interaction can choose a humorous style of expression;

[0079] Step 460: During the interaction process, the interaction strategy is adjusted in a timely manner according to the real-time feedback of the user, and the new interaction strategy is re-incorporated into the historical data to continuously optimize and improve the model.

[0080] Embodiment 2

[0081] Embodiment 2 of the present invention provides an interactive data processing system, including:

[0082] An interactive module, used to collect user language information and convert voice information into text information;

[0083] Semantic analysis module, used to use natural language processing technology to perform semantic analysis and understanding on the converted text information;

[0084] The interaction strategy generation module is used to formulate personalized interaction strategies and generate questions and answers based on the user's historical interaction records and current context;

[0085] The output module is used to synthesize questions and answers into speech and output it to the user.

[0086] The interactive module is specifically used to: build a speech recognition model, use a large amount of annotated speech data to input the speech recognition model for model training, so that it can learn the correspondence between speech and text; obtain user voice input through a microphone device, and perform audio preprocessing on the obtained voice information; extract representative acoustic features from the preprocessed speech; input the extracted representative acoustic features into the speech recognition model, output the predicted text sequence, and perform error correction, punctuation addition, and grammar optimization on the output text sequence to improve the readability and accuracy of the text.

[0087] The semantic analysis module is specifically used to: collect a large amount of speech data, including audio data from different speakers, accents, speaking speeds, and background noise, and preprocess these audio data, including audio cropping, normalization, and framing operations; extract meaningful features from the speech data, including Mel-frequency cepstral coefficients MFCC and perceptual linear prediction PLP; select a suitable deep learning model and input the extracted features into the deep learning model for model training; separate a test set from the extracted features, use the test set to evaluate the trained model, and adjust the model hyperparameters based on the evaluation results, including the learning rate, number of layers, and number of neurons, to optimize model performance.

[0088] The semantic analysis module is specifically used to use the natural language processing library to perform lexical analysis on the input text sequence, including word segmentation and part-of-speech tagging; formulate grammar rules, and use the syntax tree analysis algorithm to detect grammatical errors in the text sequence based on the defined grammar rules; establish a grammar library for common words, and replace the marked complex words with common words according to the context and semantics to improve the accuracy and richness of expression; analyze the sentence structure, split or reorganize long sentences to make them clearer and easier to read; detect whether the punctuation is correct based on the sentence grammar and logic, and correct incorrect punctuation.

[0089] In order to improve the accuracy of the speech recognition model, it also includes training the model with a large amount of text with correct standard grammar, and constantly updating and improving grammar rules and optimization strategies based on new error types and optimization needs.

[0090] Corresponding to the above embodiment, an embodiment of the present invention provides a computer storage medium, including: at least one memory and at least one processor;

[0091] The memory is used to store one or more program instructions;

[0092] The processor is used to run one or more program instructions to execute an interactive data processing method.

[0093] Corresponding to the above-mentioned embodiment, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer storage medium contains one or more program instructions, and the one or more program instructions are used for a processor to execute an interactive data processing method.

[0094] The embodiment disclosed in the present invention provides a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are executed on a computer, the computer executes the above-mentioned interactive data processing method.

[0095] In the embodiment of the present invention, the processor may be an integrated circuit chip having the ability to process signals. The processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0096] The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present invention can be directly embodied as a hardware decoding processor for execution, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The processor reads the information in the storage medium and completes the steps of the above method in combination with its hardware.

[0097] The storage medium may be a memory, which may be, for example, a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memory.

[0098] Among them, the non-volatile memory can be a read-only memory (ROM), a programmable ROM (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory.

[0099] The volatile memory may be a random access memory (RAM) which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDRSDRAM), enhanced synchronous DRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus DRAM (DRRAM).

[0100] The storage media described in the embodiments of the present invention are intended to include, but are not limited to, these and any other suitable types of memory.

[0101] Those skilled in the art will appreciate that in one or more of the above examples, the functions described in the present invention can be implemented using a combination of hardware and software. When software is used, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. Storage media can be any available media that can be accessed by general or special-purpose computers.

[0102] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. An interactive data processing method, characterized in that: include: Collect user language information and convert voice information into text information; Use natural language processing technology to perform semantic analysis and understanding on the converted text information; Develop personalized interaction strategies and generate questions and answers based on the user's historical interaction records and current context; The questions and answers are synthesized into speech output to the user.

2. An interactive data processing method as claimed in claim 1, characterized in that: Converting voice information into text information specifically includes the following sub-steps: Build a speech recognition model and use a large amount of annotated speech data to input the speech recognition model for model training so that it can learn the correspondence between speech and text; Acquire user voice input through a microphone device and perform audio preprocessing on the acquired voice information; Extracting representative acoustic features from the preprocessed speech; The extracted representative acoustic features are input into the speech recognition model, and the predicted text sequence is output. The output text sequence is corrected, punctuation is added, and grammar is optimized to improve the readability and accuracy of the text.

3. An interactive data processing method as claimed in claim 2, characterized in that: The method for constructing a speech recognition model specifically includes the following sub-steps: Collect a large amount of speech data, including audio data of different speakers, accents, speaking speeds, and background noise, and pre-process the audio data, including audio cropping, normalization, and framing operations; Extract meaningful features from speech data, including Mel-frequency cepstral coefficients (MFCC) and perceptual linear prediction (PLP); Select a suitable deep learning model and input the extracted features into the deep learning model for model training; Separate the test set from the extracted features, use the test set to evaluate the trained model, and adjust the model hyperparameters, including learning rate, number of layers, and number of neurons, based on the evaluation results to optimize model performance.

4. An interactive data processing method as claimed in claim 2, characterized in that: Optimizing the text sequence includes the following sub-steps: Use the natural language processing library to perform lexical analysis on the input text sequence, including word segmentation and part-of-speech tagging; Formulate grammar rules and use syntax tree analysis algorithm to detect grammatical errors in text sequences based on the defined grammar rules; Establish a grammar library of common words and replace the marked complex words with common words according to the context and semantics to improve the accuracy and richness of expression; Analyze sentence structure and split or reorganize long sentences to make them clearer and easier to read; Check whether the punctuation is correct based on the sentence grammar and logic, and correct incorrect punctuation.

5. An interactive data processing method as claimed in claim 2, characterized in that: In order to improve the accuracy of the speech recognition model, it also includes training the model with a large amount of text with correct standard grammar, and constantly updating and improving grammar rules and optimization strategies based on new error types and optimization needs.

6. An interactive data processing system, characterized in that: include: An interactive module, used to collect user language information and convert voice information into text information; Semantic analysis module, used to use natural language processing technology to perform semantic analysis and understanding on the converted text information; The interaction strategy generation module is used to formulate personalized interaction strategies and generate questions and answers based on the user's historical interaction records and current context; The output module is used to synthesize questions and answers into speech and output it to the user.

7. An interactive data processing system as claimed in claim 6, characterized in that: The interactive module is specifically used to: build a speech recognition model, use a large amount of annotated speech data to input the speech recognition model for model training, so that it can learn the correspondence between speech and text; obtain user voice input through a microphone device, and perform audio preprocessing on the obtained voice information; extract representative acoustic features from the preprocessed speech; input the extracted representative acoustic features into the speech recognition model, output the predicted text sequence, and perform error correction, punctuation addition, and grammar optimization on the output text sequence to improve the readability and accuracy of the text.

8. An interactive data processing system as claimed in claim 7, characterized in that: The semantic analysis module is specifically used to: collect a large amount of speech data, including audio data from different speakers, accents, speaking speeds, and background noise, and preprocess these audio data, including audio cropping, normalization, and framing operations; extract meaningful features from the speech data, including Mel-frequency cepstral coefficients MFCC and perceptual linear prediction PLP; select a suitable deep learning model and input the extracted features into the deep learning model for model training; separate a test set from the extracted features, use the test set to evaluate the trained model, and adjust the model hyperparameters based on the evaluation results, including the learning rate, number of layers, and number of neurons, to optimize model performance.

9. An interactive data processing system as claimed in claim 7, characterized in that: The semantic analysis module is specifically used to use the natural language processing library to perform lexical analysis on the input text sequence, including word segmentation and part-of-speech tagging; formulate grammar rules, and use the syntax tree analysis algorithm to detect grammatical errors in the text sequence based on the defined grammar rules; establish a common word grammar library, and replace the marked complex words with common words according to the context and semantics to improve the accuracy and richness of expression; Analyze sentence structure and split or reorganize long sentences to make them clearer and easier to read; check whether punctuation is correct based on sentence grammar and logic, and correct incorrect punctuation.

10. An interactive data processing system as claimed in claim 7, characterized in that: In order to improve the accuracy of the speech recognition model, it also includes training the model with a large amount of text with correct standard grammar, and constantly updating and improving grammar rules and optimization strategies based on new error types and optimization needs.

Images