CN116485587A

CN116485587A - Community service acquisition method, community service providing method, electronic device and storage medium

Info

Publication number: CN116485587A
Application number: CN202310458609.2A
Authority: CN
Inventors: 赵小强; 田威; 葛蒙; 邓俊杰; 韦文文; 杨世广
Original assignee: Shenzhen Rungao Intelligent Industry Co ltd
Current assignee: Shenzhen Rungao Intelligent Industry Co ltd
Priority date: 2023-04-21
Filing date: 2023-04-21
Publication date: 2023-07-25
Anticipated expiration: 2043-04-21
Also published as: CN116485587B

Abstract

The present disclosure relates to the field of intelligent communities, and in particular, to a community service obtaining method, a community service providing method, an electronic device, and a storage medium. The community service acquisition method is applied to a user terminal, service request information is required to be acquired firstly, then the service request information is analyzed to obtain service content information and service type information, key field identification is carried out on the service content information, work order information is generated, the service type of the service request is determined according to the service type information, when the service type is normal form community service, the work order information is sent to a platform service end to acquire typed feedback service, when the service type is flexible community service, the work order information is sent to a housekeeper service end to acquire personalized feedback service, and therefore rapid typed feedback service can be acquired from the platform service end, flexible personalized service can be acquired from the housekeeper service end, and flexibility and diversity of community service acquisition of residents are improved.

Description

Community service acquisition method, community service providing method, electronic device and storage medium

Technical Field

The present disclosure relates to the field of intelligent communities, and in particular, to a community service obtaining method, a community service providing method, an electronic device, and a storage medium.

Background

With the improvement of living standard of people, people pay more attention to the acquisition of community services. It should be noted that recently built emerging communities often have a relatively perfect property service system for residents to obtain community services from. However, the old communities may lack maintenance and management due to lack of matched property teams, so that community residents are difficult to enjoy basic property service guarantee, and the living is inconvenient.

In the related art, if residents of the old communities want to enjoy community services, selection and comparison are needed in massive community service information through the internet, complicated operation steps are not friendly to the residents of the old communities, the service types available from the internet are fixed at present, and some community residents cannot enjoy flexible and personalized community services. Therefore, how to determine personalized service requests from information provided by community residents and dispatch corresponding community services while providing convenient and quick community services for the residents has become a problem to be solved in the industry.

Disclosure of Invention

The present application aims to solve at least one of the technical problems existing in the prior art. Therefore, the community service acquisition method, the community service providing method, the electronic equipment and the storage medium can determine personalized service requests from information provided by community residents and dispatch corresponding community services while providing convenient and quick community services for the residents, so that flexibility and diversity of acquiring community services by the residents are improved, and residents of old communities can acquire community services convenient for the residents.

The community service acquisition method according to the embodiment of the first aspect of the application is applied to the user terminal, and comprises the following steps:

acquiring service request information, wherein the service request information is used for reflecting a service request of a target object;

analyzing the service request information to obtain service content information and service type information;

carrying out key field identification on the service content information to generate work order information;

determining the service type of the service request according to the service type information;

when the service type is a normal form community service, the work order information is sent to a platform service end, so that the platform service end provides typed feedback service for the target object according to the work order information;

And when the service type is the flexible community service, the work order information is sent to a manager service end, so that the manager service end provides personalized feedback service for the target object according to the work order information.

According to some embodiments of the present application, the obtaining service request information includes;

acquiring a mode selection instruction in an audio form, and determining a service request mode corresponding to the target object according to the mode selection instruction;

when the service request mode is a special crowd mode, acquiring sound data of the target object through a sound acquisition unit of the user terminal to obtain an audio request instruction;

and acquiring the service request information, wherein the service request information is obtained by performing voice recognition on the audio request instruction through a pre-trained audio processing model.

According to some embodiments of the present application, before the voice recognition is performed on the audio request instruction based on the pre-trained audio processing model to obtain the service request information, the method further includes pre-training the audio processing model, specifically including:

acquiring a dialect training data set, wherein the training data set comprises a plurality of dialect training samples, and each dialect training sample is configured with a corresponding dialect training label;

Inputting the dialect training data set into the original audio processing model, and performing iterative training on the audio processing model based on the dialect training sample and the dialect training label;

and when the audio processing model accords with a first preset condition in iterative training, obtaining the pre-trained audio processing model.

According to some embodiments of the present application, the sending the work order information to a housekeeping server includes:

acquiring a preset manager information base, wherein the manager information base comprises a plurality of manager identification information, each manager identification information is configured with corresponding manager tag information, and the manager identification information is used for identifying candidate community service managers;

performing relevance matching processing on the housekeeping tag information based on the work order information to obtain target tag information;

determining the manager identification information corresponding to the target tag information as target identification information;

and sending the target identification information and the work order information to the manager server side so that the manager server side distributes the work order information to the community service manager corresponding to the target identification information.

generating a manager profile information field based on the manager identification information and the manager tag information, the manager profile information field including profile information of a plurality of candidate community service managers;

displaying the manager profile information field on the user terminal, and acquiring a manager selection instruction issued by the target object based on the manager profile information field;

and determining the target identification information from a plurality of pieces of manager identification information according to the manager selection instruction.

According to some embodiments of the present application, the performing key field identification on the service content information, generating work order information includes:

acquiring a pre-trained semantic recognition model; wherein the pre-training of the semantic recognition model comprises: acquiring a semantic training data set, wherein the semantic training data set comprises a plurality of semantic training samples, and each semantic training sample is configured with a corresponding semantic training label; inputting the semantic training data set into the original semantic recognition model, and carrying out iterative training on the semantic recognition model based on the semantic training sample and the semantic training label; when the semantic recognition model accords with a second preset condition in iterative training, obtaining a pre-trained semantic recognition model;

And carrying out key field recognition on the service content information through the semantic recognition model to generate the work order information.

According to an embodiment of the second aspect of the application, the community service acquisition method is applied to a platform server, and the method comprises the following steps:

acquiring work order information from a user terminal; the user terminal acquires service request information, wherein the service request information is used for reflecting a service request of a target object, analyzing the service request information to obtain service content information and service type information, carrying out key field identification on the service content information to generate work order information, and the service type of the work order information is a normal form community service;

and providing typed feedback service for the target object according to the work order information.

According to an embodiment of the third aspect of the application, the community service acquisition method is applied to a manager server, and the method comprises the following steps:

acquiring work order information from a user terminal; the user terminal acquires service request information, wherein the service request information is used for reflecting a service request of a target object, analyzing the service request information to obtain service content information and service type information, carrying out key field identification on the service content information to generate work order information, and the service type of the work order information is a flexible community service;

And providing personalized feedback service for the target object according to the work order information.

In a fourth aspect, an embodiment of the present application provides an electronic device, including: the system comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the community service acquisition method according to any one of the embodiments of the first aspect of the application or the community service providing method according to any one of the embodiments of the second aspect when executing the computer program.

In a fifth aspect, embodiments of the present application provide a computer-readable storage medium storing a program that is executed by a processor to implement the community service acquisition method according to any one of the embodiments of the first aspect of the present application, or the community service providing method according to any one of the embodiments of the second aspect.

The community service acquisition method, the community service providing method, the electronic equipment and the storage medium have at least the following beneficial effects:

the community service acquisition method is applied to a user terminal, service request information is required to be acquired firstly, the service request information is used for reflecting a service request of a target object, then the service request information is analyzed to obtain service content information and service type information, further, key field identification is carried out on the service content information, work order information is generated, further, the service type of the service request is determined according to the service type information, when the service type is normal form community service, the work order information is sent to a platform service end, so that the platform service end provides typed feedback service for the target object according to the work order information, when the service type is flexible community service, the work order information is sent to a manager service end, and the manager service end provides personalized feedback service for the target object according to the work order information. According to the community service acquisition method, the work order information is distinguished according to the service type corresponding to the service request and is sent to the platform service end or the manager service end, so that the rapid typed feedback service can be acquired from the platform service end, the flexible individual service can be acquired from the manager service end, the individual service request can be determined from the information provided by the community residents and the corresponding community service can be distributed while the convenient and rapid community service is provided for the residents, the flexibility and the diversity of the residents for acquiring the community service are improved, and the residents of the old community can acquire the community service of the convenience for the residents.

Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.

Drawings

The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, wherein:

fig. 1 is a schematic flow chart of a community service acquisition method provided in an embodiment of the present application;

fig. 2 is a schematic flow chart of step S101 in fig. 1;

FIG. 3 is a schematic block diagram of an architecture of an audio processing model;

fig. 4 is another flow chart of the community service acquisition method provided in the embodiment of the present application;

fig. 5 is a schematic flow chart of step S103 in fig. 1;

FIG. 6 is another schematic flow chart of a community service acquisition method according to an embodiment of the present disclosure;

fig. 7 is a schematic flow chart of step S106 in fig. 1;

fig. 8 is a flow chart of step S702 in fig. 7;

fig. 9 is another flow chart of step S702 in fig. 7;

FIG. 10 is a block diagram of a user terminal implementing the community service acquisition method shown in FIG. 1, according to an embodiment of the present application;

fig. 11 is a server configuration diagram for implementing the community service acquisition method shown in fig. 1 according to an embodiment of the present application.

Detailed Description

Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are exemplary only for the purpose of explaining the present application and are not to be construed as limiting the present application.

In the description of the present application, the meaning of a number is one or more, the meaning of a number is two or more, greater than, less than, exceeding, etc. are understood to not include the present number, and the meaning of a number above, below, within, etc. are understood to include the present number. The description of the first and second is for the purpose of distinguishing between technical features only and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or implicitly indicating the precedence of the technical features indicated. It should be understood that in this application, "at least one" means one or more, and "a plurality" means two or more. "and/or" for describing the association relationship of the association object, the representation may have three relationships, for example, "a and/or B" may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.

In the description of the present application, it should be understood that the direction or positional relationship indicated with respect to the description of the orientation, such as up, down, left, right, front, rear, etc., is based on the direction or positional relationship shown in the drawings, is merely for convenience of describing the present application and simplifying the description, and does not indicate or imply that the apparatus or element to be referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present application.

In the description of the present specification, reference to the terms "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

In the description of the present application, unless explicitly defined otherwise, terms such as arrangement, installation, connection, etc. should be construed broadly and the specific meaning of the terms in the present application can be reasonably determined by a person skilled in the art in combination with the specific content of the technical solution. In addition, the following description of specific steps does not represent limitations on the order of steps or logic performed, and the order of steps and logic performed between steps should be understood and appreciated with reference to what is described in the embodiments.

The characteristics of the old community comprise the following four aspects:

firstly, the community management structure is imperfect. The community management structure of the old community is simple, community management staff is fewer, the community management system is simple, the community management capability is weak, and the safety management is not perfect.

Secondly, the safety facility is behind. The safety facilities of old communities are behind, and the number of households installing safety facilities such as an access control system, a monitoring system and the like is small, so that the overall safety protection capability of the communities is weak.

Thirdly, community management is not strict. Some old communities lack the configuration of community managers or lack strict management of community managers, so that the community management system is not perfect enough and the security management is not strict enough.

Fourth, the community environment lacks maintenance. The community environment of the old community presents more complicated and chaotic performance due to lack of maintenance, so that the community environment pollution is easy to occur, and the potential safety hazard is more.

In the related art, if residents of the old communities want to enjoy community services, selection and comparison are needed in massive community service information through the Internet, complicated operation steps are not friendly to the residents of the old communities, the types of services which can be obtained at present are relatively fixed, and some community residents cannot enjoy flexible and personalized community services. Therefore, how to determine personalized service requests from information provided by community residents and dispatch corresponding community services while providing convenient and quick community services for the residents has become a problem to be solved in the industry.

The user terminal refers to a terminal used by the target object; the platform service end refers to a server for providing typed feedback service; the personalized feedback service refers to a server for providing the personalized feedback service.

Referring to fig. 1, the community service acquisition method according to the embodiment of the present application is applied to a user terminal, and may include, but is not limited to, steps S101 to S106 described below.

Step S101, obtaining service request information, wherein the service request information is used for reflecting a service request of a target object;

step S102, analyzing the service request information to obtain service content information and service type information;

step S103, carrying out key field identification on the service content information to generate work order information;

step S104, determining the service type of the service request according to the service type information;

step S105, when the service type is a normal form community service, the work order information is sent to the platform service end, so that the platform service end provides typed feedback service for the target object according to the work order information;

and step S106, when the service type is the flexible community service, the work order information is sent to the manager server side, so that the manager server side provides personalized feedback service for the target object according to the work order information.

The community service acquisition method applied to the user terminal through steps S101 to S106 needs to acquire service request information firstly, the service request information is used for reflecting a service request of a target object, then the service request information is analyzed to obtain service content information and service type information, further, key field identification is carried out on the service content information to generate work order information, further, the service type of the service request is determined according to the service type information, when the service type is normal form community service, the work order information is sent to a platform service end, so that the platform service end provides typed feedback service for the target object according to the work order information, and when the service type is variable-flexibility community service, the work order information is sent to a manager service end, so that the manager service end provides personalized feedback service for the target object according to the work order information. By the community service acquisition method, the platform service end can provide typed feedback service, the manager service end can be prompted to provide personalized feedback service, flexibility and diversity of community service are improved, and residents of old communities can acquire community service convenient for people.

In step S101 of some embodiments of the present application, service request information is acquired, where the service request information is used to reflect a service request of a target object. It should be noted that, the service request information is sent by the target object through the input device of the user terminal, where the user terminal includes, but is not limited to, various types of terminals such as a mobile phone, a tablet computer, a Personal Computer (PC), and the like. It should be noted that, the manner of obtaining the service request information is different based on different types of the user terminal, for example, when the service request information is obtained through a mobile phone, the service request information can be obtained through an input device such as a touch screen, a microphone, etc.; when the service request information is acquired by a Personal Computer (PC), the service request information may be acquired by a key mouse device. It should be understood that the service request information is used to reflect the service request of the target object, and the service request includes the service type and the service content required by the target object.

Referring to fig. 2, according to some embodiments provided herein, step S101 may include, but is not limited to, steps S201 to S203 described below.

Step S201, a mode selection instruction is obtained, and a service request mode corresponding to a target object is determined according to the mode selection instruction;

Step S202, when the service request mode is a special crowd mode, acquiring sound data of a target object through a sound acquisition unit of a user terminal to obtain an audio request instruction;

step S203, voice recognition is performed on the audio request instruction based on the pre-trained audio processing model, and service request information is obtained.

It should be noted that, the special crowd refers to a crowd who is inconvenient to operate the user terminal through the touch screen for the reasons of old, disabled, injured, and the like. It should be noted that the number of communities in a city is very large, and some residents have inconvenient actions due to aging, disability, injury and the like, and cannot smoothly operate the user terminal through the touch screen like a general resident, and thus cannot conveniently issue service requests to the general resident through the user terminal to form service request information. When this particular group of people is the target object, there is a problem in that it is inconvenient to acquire community services. It is emphasized that in old communities with imperfect community management structure, lagged safety facilities, relaxed community management and lack of maintenance of the community environment, the difficulty of normally acquiring community services for special people is increased sharply, and residents of special people in the old communities face more inconvenient community service acquisition approaches than resident environments of the emerging communities. How to more conveniently acquire community services is a problem to be solved in the industry.

In step S201 of some embodiments of the present application, it is necessary to first obtain a mode selection instruction, and determine a service request mode corresponding to a target object according to the mode selection instruction. The service request mode refers to a usage mode of issuing a service request. When the target object is a special crowd, in order to facilitate the special crowd to issue a service request, a plurality of preset specific functions can be loaded, for example, an application system of a user terminal is converted into a compact version, so that the page layout and the font size of the application program are adjusted to be more convenient to read, the target object can more efficiently find out the service meeting the self requirements from a display interface in the process of issuing the service request, and the function is suitable for myopic resident crowds without glasses and residents with less intelligent terminal operation experience; for example, the application system of the user terminal can start an audio input mode and start an audio guiding function, under the action of audio guiding, a target object can input an audio instruction (such as 'I want to find a person to repair a window') through the user terminal, after the application system of the user terminal recognizes the audio instruction, the application system of the user terminal can be matched with corresponding service request information according to the voice recognition content of the audio instruction, so that a service request can be issued through voice, and the function is suitable for residents with less operation experience of the intelligent terminal and disabled residents with disabled upper limbs. It should be appreciated that different modes of service request are intended to be adapted to the manner in which different groups of people issue service requests, and thus facilitate the issue of service request information for different groups of people, and thus the type of service request mode is not limited to the specific embodiments illustrated above. The mode selection instruction is an instruction for determining a service request mode corresponding to the target object. It should be understood that the mode selection instruction may be acquired in various manners, such as a touch manner by a touch screen, an audio input manner, or a fingerprint recognition manner, a gesture recognition manner, etc.

In step S202 of some embodiments of the present application, when the service request mode is a special crowd mode, sound data of a target object is collected by a sound collection unit of a user terminal, so as to obtain an audio request instruction. In some exemplary embodiments of the present application, when the service request mode is a special crowd mode, the application system of the user terminal may enable the sound collection unit to collect sound data of the target object, so as to obtain an audio request instruction for issuing the service request. It should be noted that the sound data of the target object is collected by the sound collecting unit in various manners, which may be real-time collection, collection under the prompt of a preset audio guidance sentence, or collection after the audio input mode is started by the target object.

In step S203 of some embodiments of the present application, voice recognition is performed on the audio request instruction based on the pre-trained audio processing model, so as to obtain service request information. It should be noted that, the audio request instruction is an audio signal input by the target object for issuing the service request. In order to convert the audio request instructions into text-form service request information, speech recognition of the audio request instructions by means of a pre-trained audio processing model is required. It should be noted that the pre-trained audio processing model refers to a neural network model for speech recognition of audio, which is obtained by pre-training. Wherein the audio processing model is pre-trained to have speech recognition capability for audio.

Through the embodiment shown in step S201 to step S203, the audio processing model is used to obtain the audio request instruction input by the special crowd, and the service request is issued, so that a more convenient community service obtaining method can be implemented for the special crowd. The application of the community service acquisition method in the old community further embodies the value of providing convenience for residents compared with the application in the emerging community.

Referring to fig. 3, speech recognition refers to a process of converting speech signals into text, according to some embodiments provided herein. Specifically, a section of audio request instruction is input, and a text sequence (consisting of words or characters) is found so that the matching degree of the text sequence and the audio request instruction is the highest. This degree of matching is typically represented by a probability. The audio processing model may comprise the following components: a signal processing module, a language model, an acoustic model, a decoder and a text output module.

The signal processing module enhances voice by eliminating noise, channel distortion and the like according to the auditory perception characteristics of human ears, converts a voice signal from a time domain to a frequency domain, extracts proper characteristics for a following acoustic model and converts an audio request instruction into a characteristic vector sequence. The acoustic features extractable by the signal processing module may include, but are not limited to: linear predictive coding (Linear Predictive Coding, LPC), mel frequency cepstral coefficients (Mel-frequency Cepstrum Coefficients, MFCC), mel scale filter banks (Mel-scale Filter Bank, FBank), and the like.

The acoustic model integrates the knowledge of acoustics and phonology, and the feature vector sequence extracted by the signal processing module is used as input to obtain an acoustic model score. In some embodiments, the acoustic model takes the feature vector sequence extracted by the signal processing module as input, identifies and obtains a phoneme string corresponding to the voice signal, and it should be pointed out that the phonemes are minimum voice units divided according to natural attributes of the voice, and analyze the phonemes according to pronunciation actions in syllables, and one action forms one phoneme. Phonemes are divided into two major classes, vowels and consonants. For example, the Chinese syllable o (ā) has only one phoneme, the love (a i) has two phonemes, the generation (d a i) has three phonemes and the like. And then carrying out character recognition on the phoneme strings corresponding to the voice signals through a preset pronunciation dictionary to obtain a plurality of candidate character sequences, wherein the pronunciation dictionary stores the mapping relation between Chinese characters and pronunciations. Further, the candidate character sequence is input into the language model, and the reasonable degree and the smoothness degree of the candidate character sequence can be judged. In addition, after obtaining a plurality of candidate character sequences, the acoustic model needs to combine knowledge representation of differentiation of acoustics, phonetics, environment variables, speaker gender, accent and the like to predict and obtain a section of predicted phoneme string based on each candidate character sequence, and then compares and calculates the phoneme string corresponding to the voice signal with the predicted phoneme string, so that the acoustic model score of the feature vector sequence on the acoustic feature level can be obtained.

The language model realizes mathematical modeling on the context characteristics of sentences through retraining the mutual probability among corpus learning words so as to determine whether the sentences appear reasonably and smoothly. In some embodiments, by obtaining a candidate character sequence given by the acoustic model, the language model predicts a word that may appear next based on a preceding word of the candidate character sequence, and then compares the predicted word with a following word in the candidate character sequence to calculate, thereby obtaining a language model score of the candidate character sequence at the word sense compliance level. In some more specific embodiments, the language model may also be involved in the calculation of language model scores based on prior knowledge about the domain or task.

The decoder performs comprehensive calculation on the acoustic model score and the language model score, and takes the candidate character sequence with the highest total output score as a recognition result. It is emphasized that the acoustic model score is mainly combined with knowledge representation of differences of acoustics, phonetics, environment variables, speaker gender, accent and the like, and the rationality of the voice signal on acoustic characteristics is measured; the language model score is mainly used for evaluating the reasonable degree of the character sequence corresponding to the voice signal on word sequence arrangement.

In some more specific embodiments of the present application, the neural network language model (Neural Network Language Model, NNLM) has better generalization performance due to the fact that word vectors can be mapped to a low-dimensional continuous space. It should be noted that the language model may be selected from a variety of types. Modeling of long text sequences in a low-dimensional continuous space can be achieved based on a language model of a feedforward neural network (Feedforward Neural Network, FNN), however the text length that such a language model can handle is limited by the input length of the network; based on a language model of a cyclic neural network (Recurrent Neural Network, RNN), the text with infinite length can be theoretically modeled by utilizing a cyclic structure, and the performance is greatly improved compared with the language model of a feedforward neural network type; the language model based on Long Short-term memory cyclic neural network (LSTM-Term Memory Recurrent Neural Network) solves the problem that the gradient of the RNN disappears in modeling of a Long history sequence, and can achieve good effects on various tasks; the language model based on the transducer has stronger modeling capability on long text under the action of a self-attention mechanism, and better performance is achieved on a series of tasks of natural language and voice. It should be noted that the above classes of language models can be used to construct the audio processing model of the present application.

In some more specific embodiments of the present application, there are mainly two problems with acoustic models, namely variable length of the feature vector sequence and rich variability of the audio signal. The variable length feature vector sequence problem can be solved based on dynamic time planning (Dynamic Time Warping, DTW) and hidden markov model (Hidden Markov Model, HMM) methods. The rich variability of the audio signal is caused by various complex characteristics of the speaker or factors such as speaking style and speed, ambient noise, channel interference, dialect differences, etc. The acoustic model needs to be robust enough to handle the above situation. In some embodiments, the audio processing model may use Mel-cepstral coefficients (Mel-Frequency Cepstral Coefficient, MFCC) or linear perceptual prediction (Perceptual Linear Prediction, PLP) as acoustic features, and a mixed gaussian model-hidden markov model (GMM-HMM) as acoustic model. While in other embodiments, discriminative models, such as deep neural networks (Deep Neural Network, DNN), exhibit better results in modeling acoustic features. Acoustic models based on deep neural networks, such as context-dependent deep neural network-hidden markov models (CD-DNN-HMM), have vastly exceeded GMM-HMM models in the field of speech recognition. It should be noted that the above classes of acoustic models can be used to construct the audio processing model of the present application.

Referring to fig. 4, in accordance with some embodiments provided herein, prior to step S203, the community service acquisition method further includes pre-training the audio processing model. It should be emphasized that the audio processing model is pre-trained in order to provide the audio processing model with speech recognition capabilities for the audio. It should be noted that the audio processing model is essentially a neural network model that performs speech recognition, and thus the training process for the audio processing model may be pre-trained in a usual training manner for the neural network model.

However, among old communities of cities, the proportion of old residents is absolutely not small, and problems also occur in applying an audio processing model to voice recognition of the old residents. Specifically, the aged population is not much good at communicating with mandarin, and many aged population have not received the mandarin training of the system, so that the aged population communicates more with hometown dialects or is the mandarin and dialect mixed with use. At this time, when the audio processing model is used for performing voice recognition on the elderly residents and acquiring service request information, the problem of low recognition accuracy can occur.

In order to solve the difficulty of the old community residents in issuing service requests, the application provides some exemplary embodiments for pre-training an audio processing model, which specifically can include but is not limited to:

step S401, a dialect training data set is obtained, wherein the training data set comprises a plurality of dialect training samples, and each dialect training sample is configured with a corresponding dialect training label;

step S402, inputting a dialect training data set into an original audio processing model, and performing iterative training on the audio processing model based on a dialect training sample and a dialect training label;

step S403, when the audio processing model accords with the first preset condition in the iterative training, obtaining the pre-trained audio processing model.

In step S401 of some embodiments of the present application, a dialect training data set is obtained, the training data set including a plurality of dialect training samples, each dialect training sample being configured with a corresponding dialect training label. It should be noted that the dialect of chinese is various, and different cities and regions have different local languages. The dialect training data set of the embodiment of the application may include training samples corresponding to the above several dialects, may further include more types of dialect training samples, and may of course also include all types of dialect training samples. The specific dialect training data set includes which dialect training samples can be determined according to the types of dialects that are actually required to be supported, for example, if the audio processing model is required to support understanding and recognition of the dialects in the area a, the dialects in the area B and the dialects in the area C, the training corpus needs to include voice data of a plurality of dialects including the area a, voice data of a plurality of dialects including the area B and voice data of a plurality of dialects including the area C. Each dialect training sample is configured with a corresponding dialect training tag, which may include, but is not limited to, the belonging voice content, the belonging dialect category, and the dialect attribute category of the dialect training sample. It should be noted that, the voice content refers to text content corresponding to the dialect training sample, for example, "qifan" expresses the meaning of "eat" in the gan, the xiang and the southwest official speech, and then the labeling of the dialect training sample on the content characteristics is clarified according to the text content of "qifan"; the dialect category specifically refers to category labels such that the dialect training sample 1 is the dialect of the area a, the dialect training sample 2 is the dialect of the area B, and the dialect training sample 3 is the dialect of the area C; the dialect attribute categories are category labels that are partitioned based on the dialect regions. It should be noted that chinese dialects can be divided into seven large areas: official dialect, hunan dialect, ganner dialect, wu Fangyan, min dialect, guangdong dialect and Hakka dialect. It should be appreciated that, because dialects under the same dialect region have common features, they may also be configured as dialect training samples in the dialect training dataset. It is emphasized that the training data set used for pre-training the audio processing model may be varied and is not limited to the specific embodiments presented above.

In steps S402 to S403 of some embodiments of the present application, a dialect training data set is input into an original audio processing model, and an iterative training is performed on the audio processing model based on a dialect training sample and a dialect training label, and when the audio processing model meets a first preset condition in the iterative training, a pre-trained audio processing model is obtained. It should be noted that, based on the dialect training sample and the dialect training label, the audio processing model is iteratively trained, and the purpose is to iteratively update the model parameters of the audio processing model, so as to gradually improve the accuracy of the audio processing model in recognizing the dialect. The audio processing model accords with a first preset condition in iterative training, which means that the recognition accuracy of the audio processing model for dialects reaches the expected requirement, and the audio processing model of the current iteration round is output at the moment, so that the pre-trained audio processing model can be obtained.

In some more specific embodiments, a back propagation algorithm (Backpropagation Algorithm) may be employed in pre-training the audio processing model. It should be noted that the back propagation algorithm consists of two processes, forward propagation of the signal and back propagation of the error. The forward propagation of the signal refers to the process that the audio processing model receives the dialect training sample and outputs the speech recognition result of the sample. The back propagation of the error refers to the process of returning the error between the speech recognition result output by the audio processing model and the dialect training label to the input end of the audio processing model. It is clear that back propagation is one of the gradient drops and that these two terms are often used interchangeably in many textbooks. Gradient descent refers to the computation of a gradient on each weight in the neural network for each training element. Since the neural network will not output the expected value of the training element, the gradient of each weight will prompt you how to modify the weights to achieve the expected output. If the neural network does output the expected result, the gradient of each weight will be 0, indicating that no modification of the weights is required. The gradient is the derivative of the error function at the current value of the weight. The error function is used to measure the gap between the neural network output and the expected output. In practice we can use gradient descent, in which the gradient of each weight can bring the error function to a lower value. The gradient is essentially the partial derivative of the error function for each weight in the neural network. Each weight has a gradient, i.e. the slope of the error function. The weights are the connections between two neurons. Calculating the gradient of the error function may determine whether the training algorithm should increase or decrease the weight. In turn, such a determination will reduce the error of the neural network. The error is the difference between the expected output and the actual output of the neural network. Many different training algorithms, known as "propagation training algorithms", utilize gradients. Overall, the gradient tells the neural network the following information: zero gradient-the weights do not lead to errors in the neural network; negative gradient-weight should be increased to reduce error; positive gradient-the weight should be reduced to reduce the error.

According to the embodiment of the application shown in the steps S401 to S403, the audio processing model is trained through the dialect training samples and the dialect training labels in the dialect training data set, so that the recognition accuracy of the audio processing model for the dialect can be improved. When the old people in the old community acquire community services, the input of the audio request instruction can be completed by speaking the dialect to the sound acquisition unit of the user terminal, and a more convenient community service acquisition method is provided for the old people in the old community.

In step S102 of some embodiments of the present application, the service request information is parsed to obtain service content information and service type information. It should be noted that, since the service request includes the service type and the service content required by the target object, the service request information is parsed to obtain the service content information and the service type information. It should be noted that, the service content information can be obtained by converting the service content into electronic information which can be processed by the computer system, and the service type information can be obtained by converting the service type into electronic information which can be processed by the computer system.

According to other exemplary embodiments of the present application, for some conventional community services, such as shopping, housekeeping, etc., some query index information corresponding to the community services one to one may be configured in the service request information, and through the query index information, service type information may be rapidly located, so that specific service content information may be further clarified, and analysis efficiency may be improved. It should be appreciated that the query index information may be a string containing numbers, letters, or symbols.

According to some exemplary embodiments of the present application, the service type and the service content required by the target object may be hidden in the statement meaning of the service request information, for example, if the service request information issued by the target object through the user terminal is specifically text information of "key break in keyhole of burglary-resisting door in home", requesting for person maintenance ", semantic recognition may be performed on the text information, where the service type is determined to be" home facility maintenance ", and the service content is determined to be" solve the problem of key break in keyhole of burglary-resisting door ", so as to generate service content information and service type information that can be processed by the computer system, and complete analysis of the service request information.

In step S103 of some embodiments of the present application, key field identification is performed on service content information, so as to generate work order information. The key field may be a preset field for selecting the service content, and the work order information records information required for service dispatch or service provision. However, the service request information issued by different target objects is also different, and although the service request information contains the same service content, the specific expression forms may also be different.

In some embodiments provided herein, for the service requirement of maintaining the access control, some service request information may include the expression of "trouble helping us to maintain the access control of the downstairs unit building", and other service request information may also include the expression of "ask for the master to maintain the access control of the building", where these two expressions differ in text, but are semantically identical. And the key fields of access control and maintenance are extracted from the two expressions through key field identification, so that the work order information of access control and maintenance is generated according to the access control and maintenance. Therefore, by carrying out key field identification on the service content information and generating the work order information, the key field can be identified from the service request information, and then the work order information is generated according to the key field, so that the service content information is processed in a standardized manner, and the community service acquisition efficiency is improved.

In some more specific embodiments of the present application, key field identification is performed on service content information. The service content information can be subjected to text segmentation to obtain a content field sequence, and then the content field sequence is subjected to similarity matching with a preset candidate field sequence, so that the work order information is generated according to the candidate field sequence which is the most similar. The preset candidate field sequence refers to a sequence formed by preset candidate fields. For example, when the content field sequence includes { key, broken, door, keyhole, person-sending, maintenance }, and the candidate field is { key, door, maintenance }, { sanitary, clean }, { water pipe, maintenance }, the content field sequence may be subjected to similarity matching with a preset candidate field sequence, and the most similar candidate field sequence { key, door, maintenance }, thereby generating the work order information of "door lock maintenance" according to the most similar candidate field sequence { key, door, maintenance }. In some more specific embodiments, the work order information may also be configured with an object identifier, where the object identifier is used to identify a target object from which the service request is made, so that a community service that is subsequently served can be given to the target object.

It is noted that two n-dimensional vectors are represented as [ x ] ₁ ,x ₂ ,...,x _n ]And [ y ] ₁ ,y ₂ ,...,y _n ]The degree of similarity of the two vectors can be determined by the distance or the similarity between the two vectors, and obviously, the smaller the distance between the two vectors is, the higher the similarity is; the greater the distance between the two vectors, the lower the similarity. Thus, in some embodiments, similarity matching of text may be performed after vectorizing the sequence of content fields with the sequence of candidate fields. Text vectorization, also known as Word Embedding (Word Embedding), refers to representing text information as a vector capable of expressing text semantics, which is represented by a numeric vector. The manner in which text is vectorized is varied, such as One-hot (One-hot) encoding, bag of words model (BOW), N-Gram, and the like.

It should be noted that, the manner of calculating the distance between the two vectors is also various, for example, the cosine similarity (Cosine Similarity) of the two vectors is calculated:

for another example, the Euclidean distance of two vectors is calculated (Euclidean distance):

for another example, a Manhattan distance (Manhattan Distance) of the two vectors is calculated, etc.

It should be clear that the key field identification is performed on the service content information, and the work order information is generated, which is not limited to the specific embodiments described above.

Referring to fig. 5, step S103 may include, but is not limited to, steps S501 to S502 described below, according to some embodiments provided herein.

Step S501, a pre-trained semantic recognition model is obtained;

step S502, carrying out key field recognition on the service content information through a semantic recognition model to generate work order information.

It should be emphasized that some service request information may include the expression "trouble helping us to repair the entrance guard of the downstairs unit building", and other service request information may include the expression "ask for the master to repair the entrance guard of the building", where the two expressions differ in text, but are semantically identical. Thus, there may be a difference in recognition accuracy for different expressions.

In step S501 of some embodiments of the present application, a pre-trained semantic recognition model is obtained;

in step S502 of some embodiments of the present application, key field recognition is performed on service content information through a semantic recognition model, so as to generate work order information.

The service content information reflects the service content to be acquired by the target object. In order to identify key fields of service content information, the service content information needs to be identified semantically by means of a pre-trained semantic identification model. It should be noted that the pre-trained semantic recognition model refers to a neural network model which is obtained by pre-training and is used for carrying out semantic recognition on text information. The semantic recognition model is trained in advance to have the capability of carrying out semantic recognition on text information. The method includes the steps of carrying out key field recognition on service content information through a semantic recognition model, specifically, firstly carrying out text segmentation on the service content information to obtain a content field sequence, then carrying out semantic feature extraction on the content field sequence based on the semantic recognition model to obtain a semantic feature vector, and carrying out semantic recognition according to the semantic feature vector to obtain a corresponding recognition result. It should be understood that, in the embodiment of the present application illustrated in step S501 to step S502, the semantic recognition model is applied to perform key field recognition on service content information, so as to generate work order information, and improve accuracy of key field recognition, and meanwhile, further improve efficiency of obtaining community services.

It should be understood that, in this embodiment, normalization is performed for different expressions to improve accuracy of identifying the key field, which is not limited to the embodiments of the present application shown in the above steps S501 to S502.

Referring to fig. 6, according to some embodiments provided herein, before step S501, the community service acquisition method further includes pre-training the semantic recognition model, which may specifically include, but is not limited to, steps S601 to S603 described below.

Step S601, acquiring a semantic training data set, wherein the semantic training data set comprises a plurality of semantic training samples, and each semantic training sample is configured with a corresponding semantic training label;

step S602, inputting a semantic training data set into an original semantic recognition model, and carrying out iterative training on the semantic recognition model based on a semantic training sample and a semantic training label;

step S603, when the semantic recognition model accords with a second preset condition in the iterative training, a pre-trained semantic recognition model is obtained.

In step S601 of some embodiments of the present application, a semantic training data set is obtained, where the semantic training data set includes a plurality of semantic training samples, and each semantic training sample is configured with a corresponding semantic training label. It should be noted that each semantic training sample may include a plurality of training sub-samples, where the plurality of training sub-samples respectively have different expression forms for the same semantic content. For example, the training subsamples a "troublesome help us maintain the entrance guard of the downstairs unit building" ask the security master to maintain the entrance guard of the building "and the training subsamples B", which are different in text expression form, but are semantically identical, and are both "entrance guard maintenance required". It should be noted that, for pre-training of the semantic recognition model, the aim is to promote the capability of the semantic recognition model to perform semantic recognition on text information, so when each semantic training sample comprises training subsamples in a plurality of expression forms, a semantic training data set is input into an original semantic recognition model, specifically, a plurality of training subsamples of each semantic training sample and a semantic training label are input into the semantic recognition model, then iteration training is performed on the semantic recognition model based on the semantic training sample and the semantic training label, and the capability of the semantic recognition model to extract the same semantics from different expressions can be promoted.

In step S602 to step S603 of some embodiments of the present application, a semantic training data set is input into an original semantic recognition model, and iteration training is performed on the semantic recognition model based on a semantic training sample and a semantic training label, and when the semantic recognition model meets a second preset condition in the iteration training, a pre-trained semantic recognition model is obtained. It should be noted that, after the training data set is acquired, the semantic recognition model may be iteratively trained based on the semantic training samples and the semantic training labels. The iterative training aims at improving the semantic recognition capability of the semantic recognition model. According to some exemplary embodiments of the present application, in the process of performing iterative training on the semantic recognition model, model parameters of the semantic recognition model are continuously updated, so that the semantic recognition model has more and more excellent semantic recognition capability in the iterative process, common semantic features extracted from each sub-sample of each semantic training sample are set up, and mapping association is established between the common semantic features and each semantic category. And when the semantic recognition model accords with a second preset condition in the iterative training, obtaining the pre-trained semantic recognition model. It should be noted that the second semantic recognition model meets the preset condition in the iterative training, which means that the semantic recognition model under the current iteration round reaches the pre-training expectation, the recognition accuracy of the semantic recognition model reaches the degree that the semantic recognition model can be applied to the actual scene, and at the moment, the semantic recognition model under the current iteration round is output, so that the pre-trained semantic recognition model can be obtained.

It should be noted that there are two general types of parameters in the neural network model: one type of parameter is tuning parameters (Tuning Parameters) in machine learning algorithms, which need to be flexibly set according to existing or existing experience, also known as hyper parameters (hyper parameters). Such as regularization coefficient λ, the depth of the tree in the decision tree model. A hyper-parameter is also a parameter that has the property of a parameter, such as unknowns, i.e. it is not a known constant, but a manually configurable value for which it is required to specify a "correct" value, i.e. a flexibly set value, based on existing or existing experience, which is not obtained by system learning; another type of Parameter that needs to be learned and estimated from the data is called the model Parameter (Parameter), i.e., the learnable Parameter of the model itself. For example, the weighting coefficient (slope) and the deviation term (intercept) of the linear regression line are all model parameters. The learnable parameters particularly refer to parameter values learned during neural network model training, for which the values are typically updated in an iterative manner, starting from a set of random values, and then as the neural network model is learned. In fact, when the neural network model is learned, more accurate means that the parameters of the neural network model are in the process of iterative updating, and the proper values of the parameters are gradually determined. It is noted that the appropriate value may be a value that minimizes or converges the loss function. According to some more specific embodiments of the present application, the semantic recognition model meets a second preset condition in iterative training, which may be a value where the loss function reaches a minimum or converges.

According to the embodiment of the application shown in the steps S601 to S603, the audio processing model is trained through the semantic training samples and the semantic training labels in the semantic training data set, so that the accuracy of semantic recognition by the semantic recognition model can be improved. After the residents in a plurality of communities input service request information, key field identification is carried out on service content information through a pre-trained semantic training model, work order information is generated, the accuracy of key field identification can be improved, and meanwhile, the efficiency of obtaining community services by the residents in the communities can be further improved.

In step S104 of some embodiments of the present application, a service type of the service request is determined according to the service type information. The types of community services are various, for example, property services, community activities, community environment protection services, community facility maintenance services, security services for the elderly, the weak, the sick and the disabled, and the like. In some exemplary embodiments of the present application, to facilitate the distribution or provision of various types of community services, the various types of community services are re-divided into a paradigm community service and an variational community service according to their service properties. It should be noted that the paradigm community service refers to a service that a service party can provide according to a predetermined step paradigm, such as take-out, group purchase, home service, and the like. The paradigm community service often has a fixed service pattern, such as daily ordering, buying, etc. demands of community residents. Unlike the paradigm community service, the variant community service has certain flexibility and is difficult to complete according to a given step paradigm, for example, the running leg service aims to be entrusted to the walking of people to do miscellaneous matters, can be simpler to help and queue, help and deliver objects, can be slightly complex to help and maintain, do spoken language translation, is also called a hosting service, is simpler to keep objects, and is more complex to help and care for children. Flexible community services are difficult to mass-process by the service party due to their flexibility. It should be clear that, according to the service property, the embodiments of the present application re-divide various community services into a paradigm community service and an alternative community service, which aims to send work order information to different service ends based on service requirements of different service properties, so as to facilitate dispatch or provision of various community services.

In step S105 of some embodiments of the present application, when the service type is a normal form community service, the work order information is sent to the platform service end, so that the platform service end provides a typed feedback service for the target object according to the work order information. It should be noted that, because the normal form community services refer to services that can be provided by a service side according to a predetermined step form, and these normal form community services are reflected in work order information sent by a user terminal to a platform service side, the platform service side can form a corresponding typed feedback service after typed the work order information, and further, the platform service side executes the dispatch of the typed feedback service to a target object according to the predetermined step form, thereby realizing efficient service acquisition. The step of typing the work order information is to map the work order information to a specific feedback service, for example, community residents as target objects generate service request information based on daily demands such as ordering, shopping and the like, analyze the service request information to obtain service content information and service type information, identify key fields of the service content information at the user terminal side, and generate a work order information shopping delivery list: ordering an article A, an article B and an article C; delivery address: XXX ", then" shopping delivery List "work order information: ordering an article A, an article B and an article C; delivery address: XXX' is sent to a shopping platform serving as a platform service end, the shopping platform takes and packages the articles A, B and C according to the work order information, then the articles A, B and C are distributed according to the distribution address, and after the articles A, B and C are sent to the distribution address XXX for signing, the acquisition of the typed feedback service is realized. In the above embodiment, the typed feedback service is the delivery service for the article a, the article B and the article C. It should be noted that the platform service end may include multiple types of shopping platform, take-out platform, home service platform, etc., and different platform service ends also have differences corresponding to the type of feedback service that can be dispatched, which is not limited to the specific embodiments mentioned above.

In step S106 of some embodiments of the present application, when the service type is the flexible community service, the work order information is sent to the manager server, so that the manager server provides the personalized feedback service to the target object according to the work order information. It should be noted that, because the flexible community services are different from the normal form community services, the services have certain flexibility and are difficult to be completed according to the established step modes, the flexible community services are embodied in the work order information sent by the user terminal to the manager service end, the manager service end distributes the work order information to the specific community service manager, and the community service manager flexibly schedules resources and arranges strategies, thereby pertinently providing corresponding personalized feedback services for the target object so as to cope with the personalized service request of the target object. For example, community residents as target objects generate service request information based on demands such as running legs, hosting and the like, service content information and service type information are analyzed from the service request information, key field identification is carried out on the service content information at the user terminal side, and work order information 'running leg content' is generated: assisting a carpenter who finds energy nearby the community to manufacture a cabinet; principal contact means: xxxxx ", the caretaker service side sets the work order information as" leg running content: assisting a carpenter who finds energy nearby the community to manufacture a cabinet; principal contact means: xxxxx' is distributed to specific community service managers, and some community service managers find corresponding carpenter master to communicate the requirements of cabinet making by inquiring xx community store service address book, and then feed back transaction processing conditions to a principal through principal contact ways. Wherein the pre-gathered records may be gathered by a community service manager of the xx community at a pre-visit. It should be noted that the personalized feedback service in the above embodiment is "find carpenter master to make cabinets". It should be understood that the service requests of different target objects are different, and the manager server provides personalized feedback service to the target objects according to the work order information, which is not limited to the specific embodiments.

Referring to fig. 7, step S106 may include, but is not limited to, steps S701 to S703 described below, according to some embodiments provided herein.

Step S701, a preset manager information base is obtained, wherein the manager information base comprises a plurality of manager identification information, each manager identification information is configured with corresponding manager label information, and the manager identification information is used for identifying candidate community service managers;

step S702, determining target identification information according to the manager identification information and the manager tag information;

in step S703, the work order information is sent to the manager server based on the target identification information, so that the manager server distributes the work order information to the community service manager corresponding to the target identification information.

It should be emphasized that, unlike the normal form community service, the flexible community service has a certain flexibility and is difficult to be completed according to a predetermined step mode, and the flexible community service is embodied in the work order information sent by the user terminal to the manager service end, the manager service end distributes the work order information to a specific community service manager, and the community service manager flexibly schedules resources and arranges strategies, thereby pertinently providing corresponding personalized feedback service for the target object to cope with the personalized service request of the target object. However, different community service households may be adept at different areas of service, for example, some community service households have a career experience, where such households are adept at providing personalized accompanying services for the elderly, disabled persons with impaired mobility; for another example, some community service households are good at gathering information messages around the community, and then these households are good at contacting surrounding small stores, and it should be noted that, since many of the carpenter's stores, repair stores, and tailor's stores are small stores operated by individuals, the community service households of the community can provide personalized docking services to customers after searching through the information of these small stores. In view of the different characteristics of different community service managers, if the manager is allocated to the target object immediately, when the service request of the target object does not correspond to the characteristics of the community service manager, the problems of low office efficiency and poor quality of community service acquired by residents can occur.

In step S701 of some embodiments of the present application, a preset manager information base is obtained, where the manager information base includes a plurality of manager identification information, and each manager identification information is configured with corresponding manager tag information, where the manager identification information is used to identify candidate community service managers. It should be noted that, in some embodiments, the housekeeper information base is configured to store introduction information of each community service manager, where the housekeeper information base includes a plurality of housekeeper identification information, where the housekeeper identification information is used to identify candidate community service managers, and the housekeeper identification information may be information that is sufficient to identify candidate community service managers, such as a name, a housekeeper number, and the like. It should be noted that each of the housekeeping identification information is configured with corresponding housekeeping tag information, which matches with the service characteristics of the housekeeping itself, for example, if the community service housekeeping has a career experience, the housekeeping tag information may include "career" and if the community service housekeeping has good information about the surrounding community, the housekeeping tag information may include "message career" and if the community service housekeeping has rich activity planning organization experience, the housekeeping tag information may also include "activity organization passer". It should be understood that the manager server distributes the work order information to a specific community service manager, and aims to provide corresponding personalized feedback services for the target object in a targeted manner, so that optional manager tag information is various and is not limited to the above examples.

In step S702 of some embodiments of the present application, target identification information is determined according to the housekeeping identification information and the housekeeping tag information. It is required to clarify that the manager identification information, i.e., the target identification information, of the community service manager determined to handle the service request. Wherein the target identification information is selected from candidate housekeeping identification information. It should be noted that, according to the housekeeping identification information and the housekeeping tag information, the determination of the target identification information may be implemented in various manners, in some embodiments, the association matching may be performed between the housekeeping tag information and the previously generated work order information, and since the work order information records information required for service dispatch or service provision, the work order information reflects a personalized service request of the target object, and therefore, words associated with the service request may be included in the housekeeping tag information and the work order information, for example, if the work order information includes "elderly accompanying", the corresponding housekeeping tag information may be matched to "career accompanying care" according to the work order information. In other embodiments, a corresponding manager profile information field may be generated according to the manager tag information, and then a plurality of manager profile information fields may be pushed to the target object for selection, so as to determine the community service manager meeting the requirement of the target object. Note that, the embodiment of determining the target identification information based on the housekeeping identification information and the housekeeping tag information is not limited to the above examples.

Referring to fig. 8, according to some embodiments provided herein, it should be emphasized that, since the work order information describes information required for service dispatch or service provision, the work order information represents a personalized service request of a target object, and thus, there are words associated with the service request in the manager tag information and the work order information. Step S702 may therefore include, but is not limited to, steps S801 to S802 described below.

Step S801, carrying out relevance matching processing on the manager tag information based on the work order information to obtain target tag information;

step S802, determining the manager identification information corresponding to the target tag information as the target identification information.

In steps S801 to S802 of some embodiments of the present application, association matching processing needs to be performed on the housekeeping tag information based on the work order information to obtain the target tag information, and then the housekeeping tag information corresponding to the target tag information is determined as the target tag information. It is emphasized that it is necessary to clarify the manager identification information of the community service manager, that is, the target identification information, determined as the community service manager that processes the service request. Wherein the target identification information is selected from candidate housekeeping identification information. In the embodiment of the application, the target identification information is specifically selected from candidate manager identification information based on the target tag information. Through the embodiment shown in step S801 to step S802, the target tag information is determined based on the relevance matching process, and the manager identification information corresponding to the target tag information is further determined to be the target identification information, so that the work order information can be more effectively corresponding to the service request of the target object, the work order information can be conveniently distributed to the community service manager, and the work efficiency and the community service quality acquired by residents can be improved.

According to some more specific embodiments of the present application, the manner of the relevance matching process is various, and may include, but is not limited to, the following embodiments:

firstly, respectively carrying out problem segmentation on the work order information and the housekeeping tag information to obtain a work order field sequence and a housekeeping tag sequence, and then carrying out similarity matching on the work order field sequence and the housekeeping tag sequence, thereby obtaining the target tag information according to the housekeeping tag sequence which is the most similar. It should be appreciated that the above principle of similarity matching between two text sequences has been described in more detail, and will not be described in detail herein.

Secondly, problem segmentation is respectively carried out on the work order information and the housekeeping label information, after the work order field sequence and the housekeeping label sequence are obtained, the work order field sequence and the housekeeping label sequence can be respectively input into a pre-trained service classification model to obtain a service type prediction result corresponding to the work order field sequence and a service type prediction result corresponding to the housekeeping label sequence, if the two service type prediction results are consistent, the successful matching of the work order information and the housekeeping label information can be judged, and target label information is obtained, so that the accuracy of relevance matching can be further improved.

Third, the semantically associated two words do not necessarily have the same character, e.g., there is no overlap of characters between "door and window damage" and "utility maintenance", but both can be categorized as associated words. Therefore, in other embodiments, a knowledge graph of the community service domain with flexibility may be preset, and the relevance matching may be implemented based on the knowledge graph. It should be noted that, the knowledge graph is a graph-based data structure, and is composed of nodes (points) and edges (edges), each node represents an "entity", each Edge is a "relationship" between entities, and the knowledge graph is essentially a semantic network. An entity may refer to something in the real world, such as a person, place name, company, phone, animal, etc.; relationships are used to express some kind of relationship between different entities. In some embodiments, knowledge maps of the community of variabilities service domain may categorize the community of variabilities service into several primary categories, such as "home services", "maintenance services", "running leg services", "escrow services" and "skill services". Each primary classification may further include a plurality of secondary classifications, for example, "household service" may be classified into "cleaning service" and "accompanying service", and "maintenance service" may be classified into "furniture maintenance" and "electronic equipment maintenance". Each secondary classification can be further subdivided into more tertiary classifications, for example, "furniture maintenance" can be classified into "door and window maintenance", "table and chair maintenance" and "stove maintenance". Each three-level classification can preset a plurality of associated words, for example, the associated words of door and window maintenance comprise: door and window, burglary-resisting door, window, key, keeper. After the work order information and the housekeeper label information are obtained, word segmentation processing is carried out on the work order information and the housekeeper label information respectively to obtain a plurality of work order words and a plurality of housekeeper label words, and then the first class classification, the second class classification and the third class classification corresponding to the work order words or the housekeeper label words can be defined by further comparing each work order word with each housekeeper label word in a preset knowledge graph. If the work order vocabulary and the manager tag vocabulary belong to the same three-level classification, the work order information and the manager tag information are judged to be successfully matched, and the target tag information is obtained.

It should be emphasized that the implementation of performing the relevance matching process on the housekeeping tag information based on the work order information to obtain the target tag information is not limited to the specific examples set forth above.

Referring to fig. 9, step S702 may include, but is not limited to, steps S901 to S903 described below, according to some embodiments provided herein.

Step S901, generating a manager profile information field based on the manager identification information and the manager tag information;

step S902, displaying a manager profile information field on a user terminal, and acquiring a manager selection instruction issued by a target object based on the manager profile information field;

in step S903, the target identification information is determined from the plurality of housekeeping identification information according to the housekeeping selection instruction.

In steps S901 to S903 of some embodiments of the present application, a housekeeper profile information field may be generated based on the housekeeper identification information and the housekeeper label information, and then the housekeeper profile information field is displayed on the user terminal, and a housekeeper selection instruction issued by the target object based on the housekeeper profile information field is acquired, and further, the target identification information is determined from the plurality of housekeeper identification information according to the housekeeper selection instruction. The manager profile information field is used to introduce profile information of each community service manager. In some exemplary embodiments, since the housekeeping identification information is used to identify candidate community service housekeeping and the housekeeping tag information reflects the service characteristics of the housekeeping itself, a housekeeping profile information field may be generated based on the housekeeping identification information and the housekeeping tag information. It should be noted that the manager profile information field may include profile information for a plurality of candidate community service managers. The manager profile information column is displayed on the user terminal, and profile information of each community service manager is presented to the target object, so that service characteristics of each community service manager of the target object can be informed, a manager selection instruction issued by the target object based on the manager profile information column is further acquired, and then the target identification information can be determined from a plurality of candidate manager identification information according to the manager selection instruction. It should be appreciated that since the housekeeping selection instruction is issued by the target object based on the housekeeping profile information field, the target object's willingness to select for community service households may be reflected.

Through the embodiments of the present application shown in steps S901 to S903, the target identification information can be determined from the candidate housekeeper identification information based on the selection intention of the target object for the housekeeper label information in the housekeeper profile information column, so that the quality of the community service acquired by the community residents is further improved, and the requirements of the community residents for the flexible community service are met.

In step S703, the work order information is sent to the manager server based on the target identification information, so that the manager server distributes the work order information to the community service manager corresponding to the target identification information. Note that the manager identification information, i.e., the target identification information, of the community service manager determined to process the service request. Wherein the target identification information is selected from candidate housekeeping identification information. After the target identification information is determined according to the manager identification information and the manager label information, the work order information can be sent to the manager server based on the target identification information, after the manager server receives the work order information sent based on the target identification information, a designated community service manager can be allocated to a service request of a target object according to the target identification information, and the community service manager flexibly schedules resources and arranges strategies to provide corresponding personalized feedback services for the target object in a targeted manner so as to cope with the personalized service request of the target object.

Through the embodiment of the application shown in the steps S701 to S703, the work order information can be allocated according to different characteristics of different community service managers, the work order information is corresponding to the service request of the target object, the work order information is conveniently allocated to the community service manager, and the improvement of work efficiency and the improvement of community service quality acquired by residents are facilitated.

Referring to fig. 10, fig. 10 is a block diagram illustrating a portion of a user terminal implementing a community service acquisition method according to an embodiment of the present application, the user terminal including: radio Frequency (RF) circuit 1010, memory 1015, input unit 1030, display unit 1040, sensor 1050, audio circuit 1060, wireless fidelity (wireless fidelity, wiFi) module 1070, processor 1080, and power source 1090. It will be appreciated by those skilled in the art that the user terminal structure shown in fig. 10 is not limiting of a cell phone or computer and may include more or fewer components than shown, or may combine certain components, or a different arrangement of components.

The RF circuit 1010 may be used for receiving and transmitting signals during a message or a call, and particularly, after receiving downlink information of a base station, the signal is processed by the processor 1080; in addition, the data of the design uplink is sent to the base station.

The memory 1015 may be used to store software programs and modules, and the processor 1080 performs various functional applications and data processing of the subject user terminal by executing the software programs and modules stored in the memory 1015.

The input unit 1030 may be used to receive input numeric or character information and generate key signal inputs related to setting and function control of the subject user terminal. Specifically, the input unit 1030 may include a touch panel 1031 and other input devices 1032.

The display unit 1040 may be used to display input information or provided information and various menus of the subject user terminal. The display unit 1040 may include a display panel 1041.

Audio circuitry 1060, a speaker 1061, and a microphone 1062 may provide an audio interface.

In this embodiment, the processor 1080 included in the user terminal may perform the community service acquisition method of the previous embodiment.

The user terminal of the embodiment of the application comprises, but is not limited to, a mobile phone, a computer, intelligent voice interaction equipment, intelligent household appliances, vehicle-mounted user terminals, aircrafts and the like. Embodiments of the present application may be applied to a variety of scenarios including, but not limited to, data security, blockchain, data storage, information technology, and the like.

Fig. 11 is a block diagram of a part of a server that implements the community service acquisition method of the embodiment of the present application. The servers may vary widely in configuration or performance, and may include one or more central processing units (Central Processing Units, simply CPU) 1122 (e.g., one or more processors) and memory 1132, one or more storage media 2130 (e.g., one or more mass storage devices) that store applications 1142 or data 1144. Wherein the memory 1132 and the storage medium 1130 may be transitory or persistent. The program stored on the storage medium 1130 may include one or more modules (not shown), each of which may include a series of instruction operations on a server. Still further, the central processor 1122 may be provided in communication with a storage medium 1130, executing a series of instruction operations in the storage medium 1130 on a server.

The server(s) may also include one or more power supplies 1111, one or more wired or wireless network interfaces 1150, one or more input/output interfaces 1158, and/or one or more operating systems 1141, such as Windows Server, mac OS XTM, unixTM, linuxTM, freeBSDTM, etc.

The central processor 1122 in the server may be used to perform the community service acquisition method of the embodiments of the present application.

The embodiment of the application also provides a computer readable storage medium, which is used for storing program codes, and the program codes are used for executing the community service acquisition method of each embodiment.

Embodiments of the present application also provide a computer program product comprising a computer program. The processor of the computer device reads the computer program and executes it, so that the computer device executes the community service acquisition method described above.

The technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of the embodiments of the present application.

Those of ordinary skill in the art will appreciate that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, storage device storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically include computer readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transport mechanism, and may include any information delivery media. It should also be appreciated that the various embodiments provided in the embodiments of the present application may be arbitrarily combined to achieve different technical effects.

The embodiments of the present application have been described in detail, but the present application is not limited to the above embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present application, and the equivalent modifications or substitutions are included in the scope of the present application as defined in the claims.

Claims

1. A community service acquisition method, applied to a user terminal, the method comprising:

2. The method of claim 1, wherein the obtaining service request information comprises;

3. The method according to claim 2, wherein before speech recognition of the audio request instructions based on the pre-trained audio processing model, resulting in the service request instructions, the method further comprises pre-training the audio processing model, in particular comprising:

4. The method of claim 1, wherein the sending the work order information to a housekeeping server comprises:

5. The method of claim 1, wherein the sending the work order information to a housekeeping server comprises:

6. The method of claim 1, wherein the performing key field identification on the service content information to generate work order information includes:

7. A community service providing method, which is applied to a platform service end, the method comprising:

8. A community service providing method, which is applied to a manager service end, the method comprising:

9. An electronic device, comprising: a memory, a processor storing a computer program, the processor implementing the community service acquisition method of any one of claims 1 to 6 or the community service acquisition method of any one of claims 7, 8 when executing the computer program.

10. A computer-readable storage medium storing a program that is executed by a processor to implement the community service acquisition method according to any one of claims 1 to 6, or the community service acquisition method according to any one of claims 7 and 8.