A method, system, device and storage medium for gathering alarm information

Abstract

The present invention discloses an alarm information aggregation method, system, device and storage medium, the alarm information aggregation method comprising: obtaining voice data of multiple original alarm information; converting the voice data into text data and performing natural language processing optimization, extracting keywords from the optimized text data; for alarm content with dialect or heavy accent, probabilistically estimating the voice features through a Gaussian mixture model, and combining the dialect data in a language database, using the Bayesian inference method to calculate the occurrence probability of keywords, and selecting the keywords with the highest occurrence probability according to the maximum a posteriori probability rule; based on the extracted keywords, performing dependency grammar analysis on the text data optimized by natural language processing, and extracting feature information; based on the feature information, aggregating multiple original alarm information of the same time period and the same location to obtain a complete aggregation alarm information. The present invention can solve the problem that the dialect alarm content is difficult to identify.

Description

Technical Field

[0001] The present invention relates to alarm information processing, and in particular to an alarm information aggregation method, system, device and storage medium. Background Art

[0002] With the development of computer and Internet technologies, more and more online services (such as Internet applications) can be provided to users, and the number of visits to online services is increasing, which brings challenges to operation and maintenance. In order to promptly detect abnormal situations during the operation of online services, operation and maintenance personnel will monitor online services through the alarm system, and after receiving the alarm information sent by the online service, they will alarm the operation and maintenance personnel in a specified way, such as sending an alarm by sending a text message to the operation and maintenance personnel's mobile phone, so that the operation and maintenance personnel can promptly detect abnormal situations and perform maintenance, so that the online service can resume normal operation as soon as possible. However, during the alarm connection process, due to the different regions of the alarm personnel, dialects from various places will appear, resulting in the operator being unable to distinguish the key information, and thus unable to quickly distinguish the alarm content, which greatly affects the efficiency of alarm information processing. Summary of the invention

[0003] Purpose of the invention: The purpose of the present invention is to provide an alarm information aggregation method, system, device and storage medium to solve one or more problems existing in the prior art.

[0004] Technical solution: A method for gathering alarm information of the present invention comprises:

[0005] S1. Acquire the voice data of multiple original alarm messages;

[0006] S2. Convert the voice data into text data, optimize the text data through natural language processing methods, and extract keywords from the optimized text data; for alarm content with dialects or strong accents, use Gaussian mixture model GMM to estimate the probability of speech features, and use Bayesian inference method to calculate the probability of occurrence of keywords in combination with dialect data in the language database, and select the keyword with the highest probability of occurrence according to the maximum a posteriori probability rule MAP;

[0007] S3. Based on the extracted keywords, perform dependency grammar analysis on the text data optimized by natural language processing to extract feature information;

[0008] S4. Based on the feature information, multiple original alarm information in the same time period and the same location are aggregated to obtain a complete aggregated alarm information.

[0009] Furthermore, step S1 includes: assuming that the original alarm voice signal is S(t), using fast Fourier transform FFT to convert the original alarm voice signal S(t) from a time domain signal to a frequency domain signal S(f) for subsequent voice feature extraction and analysis.

[0010] Further, step S2 includes:

[0011] S201: Conversion of voice data into text data

[0012] Use a pre-trained language model that combines a hidden Markov model HMM and a deep neural network DNN to convert the frequency domain signal S(f) into text data W(t);

[0013] S202. Text data processing and keyword recognition

[0014] The converted text data W(t) is subjected to natural language processing to remove redundant information, perform spelling correction and language generation optimization; then the language model is used in combination with a predefined vocabulary to extract keywords K from the text data optimized by natural language processing. i ;

[0015] S203, Dialect and Accent Processing

[0016] For alarm contents with dialects or strong accents, the Gaussian mixture model (GMM) is used to estimate the probability of speech features, and the Bayesian inference method is used to calculate the keyword K based on the dialect data in the language database. i Probability of occurrence:

[0017] ;

[0018] P(A|B) represents the probability of A under the condition of B;

[0019] S204. Select the most likely keywords

[0020] According to the maximum a posteriori probability rule MAP, the keyword with the highest probability of occurrence is selected.

[0021] Furthermore, step S204 also includes: inputting the selected keywords into a Gaussian mixture model GMM for further analysis and inference to optimize the final keyword extraction result.

[0022] Furthermore, step S2 further includes:

[0023] S205, output and subsequent processing

[0024] If the dialect causes keyword recognition errors, a context-based error correction algorithm is used to correct the keywords.

[0025] Furthermore, step S3 includes: the feature information extraction process adopts a rule-based pattern matching algorithm, combined with a long short-term memory network model LSTM to predict the contextual relationship:

[0026] ;

[0027] Among them, F j represents feature information; W(t) represents text, C(t) represents context information, is the feature extraction function.

[0028] Furthermore, step S4 includes: clustering the multiple original alarm information by using a K-means clustering algorithm.

[0029] Based on the same inventive concept, an alarm information aggregation system of the present invention comprises:

[0030] Telephone recording module, used to obtain voice data of multiple original alarm messages;

[0031] The keyword determination module is used to convert voice data into text data, optimize the text data through natural language processing methods, and extract keywords from the optimized text data; for alarm content with dialects or strong accents, the Gaussian mixture model GMM is used to estimate the probability of voice features, and the Bayesian inference method is used to calculate the probability of keyword occurrence in combination with the dialect data in the language database, and the keyword with the highest probability of occurrence is selected according to the maximum a posteriori probability rule MAP;

[0032] A feature information extraction module is used to perform dependency grammar analysis on the text data optimized by natural language processing based on the extracted keywords to extract feature information;

[0033] The aggregation module is used to aggregate multiple original alarm information in the same time period and the same location based on feature information to obtain a complete aggregated alarm information;

[0034] The display module is used to display the aggregated alarm information text.

[0035] Based on the same inventive concept, an alarm information aggregation device of the present invention includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the device implements the steps of the above-mentioned alarm information aggregation method.

[0036] Based on the same inventive concept, a computer-readable storage medium of the present invention stores a computer program thereon, and when the program is executed by a processor, the steps of the above-mentioned alarm information aggregation method are implemented.

[0037] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: the present invention processes and aggregates dialect alarm information through text analysis and speech recognition, converts the dialect into Mandarin text, and thus generates standardized alarm information, which solves the problem that the current dialect alarm content is difficult to identify and improves the efficiency of alarm information processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flowchart of an alarm information aggregation method disclosed in an embodiment of the present invention;

[0039] Figure 2 It is a structural schematic diagram of an alarm information aggregation system disclosed in an embodiment of the present invention;

[0040] Figure 3 It is a structural schematic diagram of an alarm information aggregation device disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It will be understood by those skilled in the art that the purposes and advantages that can be achieved with the present invention are not limited to the specific description of the above beneficial effects, and the above and other purposes that can be achieved by the present invention will be more clearly understood based on the following detailed description.

[0042] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed in the present invention can be implemented in hardware, software or a combination of the two. Whether it is implemented in hardware or software depends on the specific application, design and conditions of the technical solution. Professional and technical personnel can use different methods to implement the functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0043] Reference to "embodiments" in the present invention 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 invention. The appearance of the phrase in various places 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.

[0044] Example 1: Please refer to Figure 1 , Figure 1 1 is a flow chart of a method for gathering alarm information disclosed in an embodiment of the present invention. The method for gathering alarm information may include the following operations:

[0045] S1. Acquire the voice data of multiple original alarm messages;

[0046] The alarm receiving system receives the alarm calls made by users. The alarm receiving system is available and can support calls from different communication devices, including landline phones, mobile phones, etc., and can identify the source of the phone signal. The alarm receiving system can also handle multiple alarm requests at the same time to ensure timely response during peak hours.

[0047] The voice data of the original alarm information obtained includes the conversation content, voice features and dialect or local accent used by the alarm personnel. The voice features include habitual pauses, modal particles, stuttering, etc.

[0048] Assuming that the original alarm voice signal is S(t), the fast Fourier transform FFT is used to convert the original alarm voice signal S(t) from a time domain signal to a frequency domain signal S(f) for subsequent voice feature extraction and analysis.

[0049] S2. Convert the voice data into text data, optimize the text data through natural language processing methods, and extract keywords from the optimized text data; for alarm content with dialects or strong accents, use Gaussian mixture model GMM to estimate the probability of speech features, and use Bayesian inference method to calculate the probability of occurrence of keywords in combination with dialect data in the language database, and select the keyword with the highest probability of occurrence according to the maximum a posteriori probability rule MAP;

[0050] S201: Conversion of voice data into text data

[0051] Use a pre-trained language model that combines the hidden Markov model HMM and the deep neural network DNN to convert the frequency domain signal S(f) into text data W(t).

[0052] The hidden Markov model HMM and the deep neural network DNN complement each other. The deep neural network DNN enhances the acoustic modeling capability of the traditional hidden Markov model HMM, while the hidden Markov model HMM provides the advantage of time series dependency.

[0053] S202. Text data processing and keyword recognition

[0054] The converted text data W(t) is subjected to natural language processing (NLP) to remove redundant information (for example, in the task of generating text, it may be necessary to remove repeated sentences or words), perform spelling correction and language generation optimization (for example, ensure that the generated text is grammatically correct and as consistent as possible with the context). Then, the language model is used in combination with the predefined vocabulary to extract the keywords K from the text data optimized by natural language processing. i , such as event type (medical emergency, traffic accident, public safety incident, natural disaster, technical failure, etc.), location, time, etc.

[0055] S203, Dialect and Accent Processing

[0056] For alarm contents with dialects or strong accents, the Gaussian mixture model (GMM) is used to estimate the probability of speech features, and the Bayesian inference method is used to calculate the keyword K based on the dialect data in the language database. i Probability of occurrence:

[0057] ;

[0058] P(A|B) represents conditional probability, that is, the probability of A under the condition of B.

[0059] S204. Select the most likely keywords

[0060] According to the maximum a posteriori probability rule MAP, the keywords with the highest probability of occurrence are selected. These keywords are input into the Gaussian mixture model GMM for further analysis and inference to optimize the final keyword extraction results.

[0061] S205, output and subsequent processing

[0062] If the dialect causes keyword recognition errors, a context-based error correction algorithm is used to correct the keywords to ensure the accuracy of the keywords.

[0063] S3. Based on the extracted keywords, perform dependency grammar analysis on the text data optimized by natural language processing to extract feature information;

[0064] The feature information extraction process uses a rule-based pattern matching algorithm combined with a long short-term memory network model LSTM to predict contextual relationships:

[0065] ;

[0066] Among them, F j represents feature information; W(t) represents text, C(t) represents context information, Represents the feature extraction function.

[0067] After identifying the keywords, further correlation analysis will be conducted on the keywords to determine the characteristic information in the alarm call. The characteristic information includes key factors such as event type, location, time, severity, etc. These characteristic information are compared with the initial alarm data to ensure its accuracy.

[0068] The urgency of the alarm is determined based on the voice characteristics of the alarm personnel, such as speech speed and tone, and the alarm information is prioritized. For example, the urgency of the alarm information that mentions the keywords "medical emergency" or "traffic accident" is increased accordingly.

[0069] S4. Based on the feature information, multiple original alarm information in the same time period and the same location are aggregated to obtain a complete aggregated alarm information.

[0070] After obtaining multiple original alarm information, these alarm information are aggregated based on feature information. For multiple alarm records in the same time period and the same location, the K-means clustering algorithm is used to cluster the multiple original alarm information, integrate them into one alarm record, and eliminate redundant information.

[0071] The aggregated alarm information will include complete event description, location, time, event type, urgency, etc. The aggregated alarm information will be sent to the relevant operation and maintenance department to achieve the fastest possible alarm response.

[0072] Example 2: Please refer to Figure 2 , Figure 2 : is a structural diagram of an alarm information aggregation system disclosed in an embodiment of the present invention. The system can realize alarm information aggregation, specifically including:

[0073] Telephone recording module, used to obtain voice data of multiple original alarm messages;

[0074] The keyword determination module is used to convert voice data into text data, optimize the text data through natural language processing methods, and extract keywords from the optimized text data; for alarm content with dialects or strong accents, the Gaussian mixture model GMM is used to estimate the probability of voice features, and the Bayesian inference method is used to calculate the probability of keyword occurrence in combination with the dialect data in the language database, and the keyword with the highest probability of occurrence is selected according to the maximum a posteriori probability rule MAP;

[0075] A feature information extraction module is used to perform dependency grammar analysis on the text data optimized by natural language processing based on the extracted keywords to extract feature information;

[0076] The aggregation module is used to aggregate multiple original alarm information in the same time period and the same location based on feature information to obtain a complete aggregated alarm information;

[0077] The display module is used to display the aggregated alarm information text.

[0078] Example 3: Please refer to Figure 3 , Figure 3 The present invention is a schematic diagram of the structure of an alarm information aggregation device disclosed in an embodiment of the present invention. The device may include a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the device implements the steps of the method described in the above embodiment and can achieve the same technical effect as the above method.

[0079] The memory may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The device may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the memory may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as a "hard drive"). A program / utility having a set (at least one) of program modules may be stored in, for example, the memory, such program modules including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include the implementation of a network environment. The program modules typically perform the functions and / or methods in the embodiments described herein.

[0080] The processor executes various functional applications and data processing by running the programs stored in the memory, such as implementing the method provided in the embodiment of the present invention.

[0081] Embodiment 4: Embodiment 4 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the method described in the above embodiment are implemented and the technical effect consistent with the above method can be achieved.

[0082] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0083] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0084] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0085] Computer program code for performing the operation of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0086] Of course, the storage medium containing computer executable instructions provided by an embodiment of the present invention, whose computer executable instructions are not limited to the above method operations, can also execute related operations in the method provided by any embodiment of the present invention.

[0087] The specific embodiments 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 embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for gathering alarm information, characterized in that: include: S1. Acquire the voice data of multiple original alarm messages; The original alarm voice signal S(t) is converted from a time domain signal to a frequency domain signal S(f) using Fast Fourier Transform (FFT). S2, converting the voice data into text data, optimizing the text data through a natural language processing method, and extracting keywords from the optimized text data; S201, using a pre-trained language model that combines a hidden Markov model HMM and a deep neural network DNN, converting the frequency domain signal S(f) into text data W(t); S202, perform natural language processing on the converted text data W(t), remove redundant information, perform spelling correction and language generation optimization; then use the language model, combined with a predefined vocabulary, to extract keywords K from the text data optimized by natural language processing. i ; S203: For alarm contents with dialects or strong accents, the Gaussian mixture model (GMM) is used to estimate the probability of speech features, and the Bayesian inference method is used to calculate the keyword K based on the dialect data in the language database. i The probability of occurrence; S204, selecting the keyword with the highest probability of occurrence according to the maximum a posteriori probability rule MAP; S3. Based on the extracted keywords, perform dependency grammar analysis on the text data optimized by natural language processing to extract feature information; S4. Based on the feature information, multiple original alarm information in the same time period and the same location are aggregated to obtain a complete aggregated alarm information.

2. The alarm information aggregation method according to claim 1, characterized in that: Step S204 also includes: inputting the selected keywords into a Gaussian mixture model GMM for further analysis and inference to optimize the final keyword extraction result.

3. The alarm information aggregation method according to claim 2, characterized in that: Step S2 also includes: S205, output and subsequent processing If the dialect causes keyword recognition errors, a context-based error correction algorithm is used to correct the keywords.

4. The alarm information aggregation method according to claim 1, characterized in that: Step S3 includes: the feature information extraction process uses a rule-based pattern matching algorithm combined with a long short-term memory network model LSTM to predict contextual relationships: ； Among them, F j represents feature information; W(t) represents text, C(t) represents context information, is the feature extraction function.

5. The alarm information aggregation method according to claim 1, characterized in that: Step S4 includes: clustering the multiple original alarm information by using a K-means clustering algorithm.

6. An alarm information aggregation system, implementing the alarm information aggregation method as claimed in any one of claims 1 to 5, characterized in that: The alarm information aggregation system comprises: Telephone recording module, used to obtain voice data of multiple original alarm messages; The keyword determination module is used to convert voice data into text data, optimize the text data through natural language processing methods, and extract keywords from the optimized text data; for alarm content with dialects or strong accents, the Gaussian mixture model GMM is used to estimate the probability of voice features, and the Bayesian inference method is used to calculate the probability of keyword occurrence in combination with the dialect data in the language database, and the keyword with the highest probability of occurrence is selected according to the maximum a posteriori probability rule MAP; A feature information extraction module is used to perform dependency grammar analysis on the text data optimized by natural language processing based on the extracted keywords to extract feature information; The aggregation module is used to aggregate multiple original alarm information in the same time period and the same location based on feature information to obtain a complete aggregated alarm information; The display module is used to display the aggregated alarm information text.

7. An alarm information aggregation device, characterized in that: The device comprises a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the device implements the steps of the alarm information aggregation method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the alarm information aggregation method as claimed in any one of claims 1 to 5.

Images