Travel guide system and use method thereof

Abstract

The invention discloses a travel guide system and a use method thereof. The method comprises the following steps: configuring a semantic large model and converting voice data input by a target object into text data through a service and decision system; deploying a semantic large model and storing business data through hardware equipment; wherein the semantic large model is used for performing response processing on the service request; connecting business services of the hardware equipment and the application applet through the interface application program, and carrying out data exchange and service calling; and obtaining a service request of the target object through the application applet, and feeding back a response processing result to the target object. According to the method, resources such as a semantic large model, hardware equipment and an application applet are integrated, tourism guide is promoted to be transformed to intelligence and automation, and the method can be widely applied to the technical field of data processing.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a tourist guide system and a method for using the same. Background Art

[0002] The tourism guide industry is currently facing the following major problems:

[0003] 1. Low efficiency of resource utilization: The collection, processing and application methods of tourism data are relatively primitive, failing to fully tap the potential of data resources and resulting in waste of resources.

[0004] 2. Insufficient service personalization: The existing system lacks in-depth analysis of tourist behaviors and preferences, resulting in the inability to provide guided tour services that truly meet personalized needs.

[0005] 3. Insufficient interactivity: The existing system has poor interactivity with tourists, which limits the tourists’ participation and the richness of their experience.

[0006] 4. Limited level of intelligence: The existing tourist guide system is not very intelligent, and it is difficult to effectively support tourist decisions and improve service quality. Summary of the invention

[0007] The present invention provides a tourist guide system and a method for using the same, aiming to solve at least one of the technical problems in the related art to a certain extent. The embodiments of the present invention can conveniently implement tourist guide.

[0008] On the one hand, an embodiment of the present invention provides a tourism guide system, the system includes a service and decision system, a hardware device, an interface application program and an application applet;

[0009] Service and decision system, used to configure the semantic big model and convert the voice data input by the target object into text data;

[0010] Hardware equipment, used to deploy the semantic big model and store business data; the semantic big model is used to respond to business requests;

[0011] Interface application, used to connect hardware devices with the business services of application applet, to exchange data and call services;

[0012] The application applet is used to obtain the business request of the target object and to feed back the result of the response processing to the target object.

[0013] Optionally, the service and decision-making system includes:

[0014] The corpus collection and processing module is used to collect pre-selected tourism data, pre-process the tourism data, and obtain model training data; pre-processing includes data cleaning, data labeling, and data segmentation;

[0015] A semantic big model training module, used to train a preset semantic big model using model training data;

[0016] The semantic large model optimization module is used to optimize the trained semantic large model; the optimization process includes model fine-tuning, model compression, incremental learning and error analysis;

[0017] Model deployment and service module, used to deploy the trained and optimized semantic large model to the cloud server of hardware devices;

[0018] The speech recognition module is used to perform background denoising and speech recognition on the speech data input by the target object to obtain text data.

[0019] Optionally, the speech recognition module performs background denoising and speech recognition on the speech data input by the target object to obtain text data, which is specifically used to perform the following operations:

[0020] Perform noise detection on speech data;

[0021] When the result of the noise detection indicates that the speech data contains noise information, a real component and an imaginary component are obtained based on the spectrum data of the speech data;

[0022] Based on the real and imaginary components, a complex domain ideal ratio mask is generated using a preset convolutional recurrent connectionist temporal classification network;

[0023] Based on the spectrum data of speech data, clean speech information is obtained through complex domain ideal ratio masking processing;

[0024] Based on clean voice information, speech recognition is performed using a preset audio model to obtain text data.

[0025] Optionally, the hardware device includes a load balancer and multiple servers, and the servers include an API interface server, a business processing server, and a database server;

[0026] Load balancer, used to handle the access of business requests and the load distribution of each server;

[0027] API interface server, used to process the routing of business requests initiated by application applet;

[0028] Business processing server, used to deploy the semantic big model and respond to various business requests based on the semantic big model;

[0029] Database server, used for storage and management of business data.

[0030] Optionally, the service request includes a question-and-answer service request, a conversation service request, an itinerary planning service request, a route service request, a scenic spot service request, a hotel and homestay service request, a catering and food service request, and a public resource service request; and the interface application includes:

[0031] Question and answer API interface, used to connect and process question and answer business requests;

[0032] Session API interface, used to connect and process session service requests;

[0033] Itinerary API interface, used to connect and process itinerary planning business requests;

[0034] Travel route API interface, used to connect and process route business requests;

[0035] Scenic spot API interface, used to connect and process scenic spot business requests;

[0036] Hotel and homestay API interface, used to connect and process hotel and homestay business requests;

[0037] Catering and food API interface, used to connect and process catering and food business;

[0038] Public resource recommendation API interface, used to connect and process public resource business requests.

[0039] Optionally, the application applet includes:

[0040] The dialogue module is used to obtain a business request in response to an input operation of a target object, and then call the semantic big model based on the business request to respond to the business request; and feed back the result of the response processing to the target object.

[0041] Optionally, the application applet also includes a companion module, a discovery module and a personal center module;

[0042] The tour companion module is used to call the voice explanation service for guided tours based on the result of response processing;

[0043] A discovery module, for displaying thematic information to the target object in response to the target object's browsing operation;

[0044] The person-centric model is used to store the associated information of the target object and adjust the associated information in response to the operation instruction of the target object; the associated information includes the historical record of the response processing.

[0045] On the other hand, an embodiment of the present invention provides a method for using a tourist guide system, comprising the following steps:

[0046] Configure the semantic big model through the service and decision system and convert the voice data input by the target object into text data;

[0047] The semantic big model is deployed through hardware devices and business data is stored; wherein the semantic big model is used to respond to business requests;

[0048] Connect hardware devices and application applet business services through interface applications to exchange data and call services;

[0049] The target object's business request is obtained through the application applet, and the result of the response processing is fed back to the target object.

[0050] On the other hand, an embodiment of the present invention provides an electronic device, including: a processor and a memory; the memory is used to store programs; the processor executes the program to implement the method of using the above-mentioned tourism guide system.

[0051] On the other hand, an embodiment of the present invention provides a computer storage medium, in which a program executable by a processor is stored. When the program executable by the processor is executed by the processor, it is used to implement the method of using the above-mentioned tourism guide system.

[0052] The embodiment of the present invention configures a semantic big model through a service and decision system and converts the voice data input by the target object into text data; deploys a semantic big model through a hardware device and stores business data; wherein the semantic big model is used to respond to business requests; connects the hardware device with the business service of the application applet through an interface application to exchange data and call services; obtains the business request of the target object through the application applet, and feeds back the result of the response processing to the target object. The embodiment of the present invention includes the following beneficial effects:

[0053] Efficient response to business requests: By configuring a semantic big model, the technology can quickly understand and respond to the business requests of the target object. This intelligent processing method not only improves the response speed, but also ensures the accuracy of the response, thereby improving the user experience.

[0054] Realize seamless data exchange: The interface application serves as a bridge to seamlessly connect the business services between hardware devices and application applet. This design makes data exchange efficient and convenient, ensuring the smooth progress of business processes.

[0055] Business data storage and management: By deploying semantic big models and storing business data on hardware devices, this technology enables comprehensive management and efficient use of business data. This helps enterprises better understand business dynamics, optimize business processes, and improve operational efficiency.

[0056] Enhance the flexibility of application applet: As the front-end interface for interaction with users, application applet can easily obtain the business request of the target object and feedback the response processing results to the user. This design makes application applet more flexible and easy to use, improving user satisfaction and loyalty.

[0057] The present invention promotes the transformation of tourist guidance towards intelligence and automation by integrating resources such as semantic big models, hardware devices and application applets. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.

[0059] Figure 1 A schematic diagram of the overall architecture of the tourist guide system provided by an embodiment of the present invention;

[0060] Figure 2 A schematic diagram of the expansion logic of background denoising and speech recognition provided by an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of the speech denoising technology process provided by an embodiment of the present invention;

[0062] Figure 4 A schematic diagram of an improved CR-CED network model structure provided by an embodiment of the present invention;

[0063] Figure 5 A schematic diagram of a detailed system architecture of a tourist guide system provided by an embodiment of the present invention;

[0064] Figure 6 A schematic diagram of a business process of an example of using the tourist guide system provided by an embodiment of the present invention;

[0065] Figure 7 A schematic diagram of parameter examples for STOI comparison of different models on a self-made dataset provided by an embodiment of the present invention;

[0066] Figure 8 A schematic diagram of an implementation environment for using a tourist guide system provided by an embodiment of the present invention;

[0067] Fig. 9 A schematic diagram of a flow chart of a method for using a tourist guide system provided by an embodiment of the present invention;

[0068] Fig.10 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0070] It should be noted that, although the functional modules are divided in the system schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first / S100", "second / S200", etc. in the specification, claims and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0071] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0072] To facilitate the understanding of the technical solution of the present invention, the technical feature terms that may appear in the embodiments of the present invention are first explained:

[0073] Semantic Big Model: The semantic big model is a large-scale language model based on deep learning and natural language processing technology, which can understand and generate natural language text. Its core lies in training through a large number of corpora, so as to master the complex relationships in language such as grammar, semantics, and context, and has strong language understanding and generation capabilities.

[0074] Retrieval-Augmented Generation (RAG): Retrieval-Augmented Generation is a technical method that combines knowledge retrieval with language model generation. In simple terms, it does not rely solely on the internal knowledge and training data of the language model itself to generate answers, but retrieves relevant knowledge information from external large-scale knowledge bases or databases during the generation process, and integrates the retrieved knowledge into the generation process of the language model.

[0075] Agent: Agents can be understood as entities with certain autonomy, perception, decision-making and action capabilities, capable of performing tasks in a specific environment and interacting with the environment. Agents can perceive the state and changes of the surrounding environment, make decisions by analyzing and processing the collected information, and take corresponding actions to achieve specific goals. They can continuously improve their behavior according to preset rules, strategies or through learning and adaptation.

[0076] Convolutional Recurrent-Connectionist Temporal Classification (CR-CED): The CR-CED network model is a deep learning model for speech denoising, which effectively combines the advantages of convolutional neural network (CNN) and recurrent neural network (RNN) to process noise in speech signals.

[0077] Short-Time Fourier Transform (STFT): STFT is a critical tool in speech analysis. It divides the speech signal into many short time segments and then performs Fourier transform on each time segment.

[0078] On the one hand, if Figure 1 As shown, an embodiment of the present invention provides a tourism guide system 100, which may include a service and decision system 110, a hardware device 120, an interface application 130 and an application applet 140; wherein:

[0079] Service and decision system, used to configure the semantic big model and convert the voice data input by the target object into text data;

[0080] It should be noted that, in some embodiments, the service and decision-making system may include:

[0081] The corpus collection and processing module is used to collect pre-selected tourism data, pre-process the tourism data, and obtain model training data; pre-processing includes data cleaning, data labeling, and data segmentation;

[0082] For example, in some specific implementations, the corpus collection and processing module is used to obtain and preprocess data for training the semantic big model. Specifically, the following functional logic can be implemented:

[0083] 1. Data collection: Gathering resident travel data, scenic spot travel data, cultural and tourism gathering area customer sentiment data, rural travel data, railway station passenger flow data, holiday customer sentiment data, etc. from China Telecom's own signaling data; collecting industry-public cultural, scenic spot, transportation, hotel, catering data, etc.; collecting Internet travel-related data, smart city travel-related data, tourism customer service data, tour guide data, travel strategy data, etc. This step brings together data from different sources for unified analysis and processing, providing corpus support for the semantic big model.

[0084] 2. Data cleaning: Clean the collected data to remove noise, duplication and irrelevant information to ensure high data quality. Prevent misleading conclusions from being output by large models due to data quality issues, which can affect tourists’ decision-making.

[0085] 3. Data labeling: Label the data, including entity recognition, intent recognition, sentiment analysis, etc., to generate high-quality training data sets.

[0086] 4. Data segmentation: Split the data into model training data, including training set, validation set and test set, to ensure the effectiveness of model training and the accuracy of evaluation.

[0087] A semantic big model training module, used to train a preset semantic big model using model training data;

[0088] For example, in some specific implementations, the semantic big model training module is used to train the semantic big model using preprocessed data. Specifically, the following functional logic can be implemented:

[0089] 1. Model selection: Combine the Telecom Xingchen big model with the open source big model, Embedding model, and Re-rank model to complete tasks in different links and improve the overall performance of the system. Open source models include Qwen, bce-embedding-base_v1, bce-reranker-base_v1, etc.

[0090] 2. Retrieval-Augmented Generation (RAG): Use inverted indexing technology to build indexes for the cultural and tourism industry knowledge base and scenic spot ticketing system built by Telecom.

[0091] 3. Training process: In a high-performance computing environment, use the pre-processed tourism scene data to pre-train the large model to improve the robustness and generalization ability of the large model in tourism scenarios.

[0092] 4. Agent: When the user enters "Are there any children's tickets for the Forbidden City after 3 pm tomorrow?", the engine searches for the keywords "Forbidden City", "tickets", "tomorrow 3 pm", and "children's tickets" to quickly obtain relevant ticket inventory information. When the user makes a ticket booking request, the agent first obtains the necessary user information (such as the number of people, identity, etc.) from the semantic big model, and then sends a ticket booking request to the scenic spot ticketing system through the scenic spot ticketing API interface. If the ticket booking is successful, the agent will feedback the result to the user and provide relevant ticket collection or entry prompts; if the ticket booking fails, the agent analyzes the reasons for the failure (such as insufficient inventory, incorrect identity information, etc.) and informs the user.

[0093] 5. Model evaluation: Use the validation set to evaluate the model and measure its performance in semantic understanding, text generation, etc. to ensure that the model meets the expected performance indicators.

[0094] The semantic large model optimization module is used to optimize the trained semantic large model; the optimization process includes model fine-tuning, model compression, incremental learning and error analysis;

[0095] For example, in some specific implementations, the semantic big model optimization module is used to improve the performance and adaptability of the semantic big model. Specifically, the following functional logic can be implemented:

[0096] 1. Fine-tune the model: Fine-tune the basic model and further train the model using corpus from the cultural and tourism field as well as cross-domain historical, geographical, and transportation data to improve the model's performance in the field of tourism guidance.

[0097] 2. Model compression: Use the AutoAWQ framework to quantize the model to 8 bits, while retaining the original capabilities, reducing the model's video memory usage and improving inference efficiency.

[0098] 3. Incremental learning: Continuously collect new tourism data and perform incremental training on the model to ensure that the model always maintains the latest knowledge and capabilities.

[0099] 4. Error analysis: Based on the user feedback information transmitted by the front-end applet, adjust the content with poor results and conduct targeted training.

[0100] Model deployment and service module, used to deploy the trained and optimized semantic large model to the cloud server of hardware devices;

[0101] For example, in some specific implementations, model deployment and services are used to apply the trained semantic big model to actual business scenarios. Specifically, the following functional logic can be implemented:

[0102] 1. Model deployment: Deploy the trained model to Tianyi Cloud Server to ensure that the model can run stably and efficiently.

[0103] 2. Interface development: Develop an API interface for calling the semantic big model to facilitate front-end applications and other services to call the model's functions.

[0104] 3. Performance monitoring: Real-time monitoring of the model’s operating performance to ensure the stability and efficiency of the model in practical applications.

[0105] 4. User feedback: Collect user feedback through application mini-programs to continuously improve model performance and user experience.

[0106] A speech recognition module is used to perform background denoising and speech recognition on the speech data input by the target object to obtain text data;

[0107] In some preferred embodiments, Figure 2 As shown, the speech recognition module performs background denoising and speech recognition on the speech data input by the target object to obtain text data, which is specifically used to perform the following operations: T1, perform noise detection on the speech data; T2, when the result of the noise detection indicates that the speech data contains noise information, obtain the real component and the imaginary component based on the spectrum data of the speech data; T3, based on the real component and the imaginary component, generate the complex domain ideal ratio masking by using the preset convolutional recurrent connectionist time classification network; T4, based on the spectrum data of the speech data, obtain clean speech information by complex domain ideal ratio masking processing; T5, based on the clean speech information, perform speech recognition by using the preset audio macro model to obtain text data.

[0108] For example, in some specific implementations, the speech recognition module is responsible for audio content detection and recognizes the user's input speech as text that can be understood by the semantic large model. The present invention innovatively proposes an improved CR-CED-based speech signal background denoising method in the direction of the tourism guide large model speech input, which is used to reduce the impact of noisy background sound on the accuracy of the system speech input during tourism. The speech recognition module can implement the following functional logic:

[0109] 1. Users can operate the application applet and click the "Press and Talk" button in the dialogue section or the companion section, and the system will call the voice recognition API interface to record voice.

[0110] 2. Send the speech to the speech signal background denoising module to remove the background sound. The technical flow chart is as follows Figure 3 As shown, the implementation of the speech signal background denoising module specifically includes the following steps:

[0111] 2.1. A noise detector module is added to the front end of CR-CED (Convolutional Recurrent Connectionist Time Classification Network). By performing binary classification training on the collected audio signal to determine whether it contains noise, it serves as the basis for judging whether the subsequent speech denoiser is enabled, and improves the efficiency of the system through intermittent processing.

[0112]

[0113] S t,f =S r +iS i (Formula 2)

[0114] In the formula, S t,f Represents the time-frequency unit of the spectrogram, |S t,f | and Represent the amplitude and phase of the time-frequency unit respectively. In the whole processing process, the Cartesian expression is used to represent the time-frequency unit (i.e., Formula 2). r and S i Represent the real and imaginary parts of the time-frequency unit respectively.

[0115] 2.2. In the previous step, the speech signal that has passed through the noise detector can more accurately determine whether the current speech segment contains noise. In order not to increase the burden on system resources, the noise detector does not run continuously, so the signal input to the speech denoiser will contain clean speech and noisy speech. The noisy speech signal is no longer input as a single spectrum amplitude. Based on the phase perception principle, the real and imaginary parts of the spectrum are input into two independent channels for training.

[0116] The training goal of the speech denoising network model proposed in the present invention is no longer to directly predict the spectrogram of the clean speech signal, but to learn the optimal complex domain ideal ratio mask (cIRM) corresponding to the input noisy speech signal. According to the phase perception theory of Ephraim et al., Formula 3 can be obtained.

[0117] C t,f =M t,f ×N t,f (Formula 3)

[0118] In the formula, C t,f Represents the complex spectrum of the clean speech signal, M t,f Represents the complex domain value of cIRM, N t,fRepresents the complex spectrum of the noisy speech signal. In formula 2.3.3, the complex spectrum of each signal is expressed in the form of a Cartesian coordinate system. The complex spectrum representing the clean speech signal is equal to the product of the cIRM value and the complex spectrum of the noisy speech. The expression formula for cIRM can be inverted. In formula 3, M represents the cIRM value, and the subscripts r and i represent the real and imaginary components of the complex spectrum, respectively.

[0119]

[0120] As above, the real part of the noisy speech spectrum N r and the imaginary component N i By inputting different channels of the speech denoising model respectively, the real and imaginary parts of cIRM can be obtained at the output of the model respectively. Then, the output value M of the model is converted into t,f With the input spectrum N t,f Multiplying together can get the clean speech signal C t,f Spectrum diagram of .

[0121] 2.3. The improved CR-CED network model structure is as follows Figure 4 As shown. The STFT operation of the CR-CED network uses a Hamming window with a length of 256 and a displacement of 64. The network output can only obtain a clean audio signal of 64 samples after each conversion. Therefore, for a 16kHz audio signal, the network model makes at least 5 judgments on the audio within a duration of 20ms. For the currently widely used speech signal with a sampling rate of 44.1kHz, it means that the denoising network model must run 15 judgments within 20ms. For the entire recognition system, it increases power consumption and also affects the stability of recognition. Unlike the CR-CED network, the improved CR-CED network uses a Hamming window with a length of 512 and an overlap of 75% for STFT, and in order to ensure that the real and imaginary parts of the STFT are extracted simultaneously without occupying too many resources, only 4 historical fragments are selected. In addition, because the model uses the phase component of the noisy signal spectrum, the input of the model is no longer just a real number, so the activation function in the output unit of the convolutional layer is replaced by LeakyReLU from ReLU. The LeakyReLU activation function is shown in the following figure.

[0122] As shown in formula 5:

[0123]

[0124] 3. The speech after background and noise removal is sent to the preset audio model (such as Qwen-Audio model) for speech recognition, and the recognized text is sent to the semantic model for semantic understanding.

[0125] Hardware equipment, used to deploy the semantic big model and store business data; the semantic big model is used to respond to business requests;

[0126] For example, in some specific implementations, the hardware devices can all use preset cloud servers, with a total of 6 servers and 1 load balancer, including 2 API interface servers, 2 business processing servers, and 2 database servers, and the database server adopts master-slave deployment. All servers are installed with Centos7.4 operating system, business processing servers and interface servers are installed with Nginx+php environment, and database servers are installed with Mysql8.0. All servers can be linearly expanded to carry the increasing business needs in the future.

[0127] It should be noted that in some embodiments, the hardware device may include a load balancer and multiple servers, and the servers include an API interface server, a business processing server, and a database server; wherein:

[0128] Load balancer, used to handle the access of business requests and the load distribution of each server;

[0129] Exemplarily, in some specific implementations, the load balancer is used to process service request access and load distribution, and ensures load balancing among servers by dynamically allocating resources, thereby improving the overall performance and reliability of the system.

[0130] API interface server, used to process the routing of business requests initiated by application applet;

[0131] Exemplarily, in some specific implementations, the API interface server is used to process the business request routing initiated by the application applet, and by parsing and forwarding API requests, it achieves seamless connection with the front-end system and other services, and provides efficient interface services.

[0132] Business processing server, used to deploy the semantic big model and respond to various business requests based on the semantic big model;

[0133] Exemplarily, in some specific implementations, the business processing server is responsible for processing various business requests to ensure that user requests are responded to and processed in a timely manner.

[0134] Database server, used for storage and management of business data;

[0135] For example, in some specific implementations, the database server is used for data storage and management to ensure the consistency, integrity and efficient access of system data. The master database server is responsible for data writing and partial query, undertakes the storage and management tasks of key data, and ensures the real-time and consistency of data. The slave database server focuses on data query, and improves the system's response speed and query efficiency by sharing the query pressure.

[0136] Interface application, used to connect hardware devices with the business services of application applet, to exchange data and call services;

[0137] Exemplarily, in some specific implementations, the interface application is used to provide seamless connection between front-end operations and business services, ensuring efficient data exchange and service calls.

[0138] It should be noted that the service request includes a question-and-answer service request, a conversation service request, an itinerary planning service request, a route service request, a scenic spot service request, a hotel and homestay service request, a catering service request, and a public resource service request; in some embodiments, the interface application may include:

[0139] Question and answer API interface, used to connect and process question and answer business requests;

[0140] Exemplarily, in some specific implementations, the question-and-answer API interface is used to process large-model question-and-answer business requests, including a popular question list interface, a text question interface, a speech recognition interface, a large-model answer question interface, a card information acquisition interface, and a stop answering interface.

[0141] Session API interface, used to connect and process session service requests;

[0142] Exemplarily, in some specific implementations, the session API interface is used to process session service requests, including a session record interface, a session detail interface, a session evaluation interface, a session record saving interface, and a single historical session deletion interface.

[0143] Itinerary API interface, used to connect and process itinerary planning business requests;

[0144] Exemplarily, in some specific implementations, the itinerary API interface is used to process itinerary planning-related business requests, including an itinerary list viewing interface, an itinerary detailed information viewing interface, an itinerary addition interface, an itinerary change interface, an itinerary reminder interface, an itinerary deletion interface, an itinerary planning change interface, an interface for obtaining element IDs and types based on keywords, an itinerary planning modification interface without using a large model, and an itinerary planning subscription message interface.

[0145] Travel route API interface, used to connect and process route business requests;

[0146] Exemplarily, in some specific implementations, the travel route API interface is used to process travel route status service requests, including an interface for obtaining daily routes.

[0147] Scenic spot API interface, used to connect and process scenic spot business requests;

[0148] Exemplarily, in some specific implementations, the scenic spot API interface is used to process information service requests for the scenic spot and its surrounding areas and provide scenic spot ticket reservation, subscription, refund and other services, including a scenic spot list interface, a scenic spot details interface, a nearby scenic spot voice interface, a scenic spot surrounding recommendation interface, and a scenic spot ticketing API interface.

[0149] Hotel and homestay API interface, used to connect and process hotel and homestay business requests;

[0150] Exemplarily, in some specific implementations, the hotel and homestay API interface is used to process hotel and homestay and hotel and homestay surrounding information service requests, including a hotel and homestay list interface, a hotel and homestay details interface, and a hotel and homestay surrounding recommendation interface.

[0151] Catering and food API interface, used to connect and process catering and food business;

[0152] Exemplarily, in some specific implementations, the catering and food API interface is used to process catering and food and catering and food surrounding information service requests, including a catering and food list interface, a catering and food details interface, and a catering and food surrounding recommendation interface.

[0153] Public resource recommendation API interface, used to connect and process public resource business requests;

[0154] Exemplarily, in some specific implementations, the public resource recommendation API interface is used to process public service resource recommendation service requests, including obtaining a public resource interface.

[0155] The application applet is used to obtain the business request of the target object and to feed back the result of the response processing to the target object.

[0156] Exemplarily, in some specific implementations, the application applet is used to provide users with a convenient access path and interactive platform to ensure the efficiency and user-friendliness of the tourism guide service.

[0157] It should be noted that, in some embodiments, the application applet may include:

[0158] The dialogue module is used to obtain a business request in response to an input operation of a target object, and then call the semantic big model based on the business request to respond to the business request; and feed back the result of the response processing to the target object.

[0159] Exemplarily, in some specific implementations, the dialogue module is used to conduct large-model dialogues to achieve functions such as intelligent question and answer, personalized itinerary planning, interactive intelligent explanations, and convenient service recommendations.

[0160] Among them, intelligent dialogue allows users to ask questions through voice or text, and can evaluate the answers obtained. Users can choose to like to express satisfaction, or click to provide feedback. In addition, users can easily copy the answer text for saving or sharing. The system also has a keyword annotation function, which can intelligently identify specific information points in the answer, such as scenic spots. Users only need to click on these keywords, and the system will provide more in-depth explanations and related information, so that users can quickly obtain the required knowledge and enrich their travel experience. At the same time, if users are dissatisfied after the tour, they can make a complaint through intelligent dialogue. The large model provides users with complaint telephone numbers, complaint addresses, and complaint form filling services.

[0161] The itinerary planning function collects the user's key travel preferences, such as travel dates, travel groups (such as family, friends, couples, etc.), travel types (such as natural scenery, history and culture, city exploration, etc.), the number of people, transportation methods, and scenic spots, hotels, homestays, and catering that the user is interested in, to customize an exclusive itinerary for the user, providing a comprehensive and flexible travel planning tool. Users can not only enjoy the itinerary recommended by the system, but also manually modify the itinerary points according to personal needs, add or delete certain restaurants or attractions, and ensure that the itinerary fully meets their expectations and interests. In addition, itinerary planning also supports viewing itinerary details, and users can easily obtain specific information for each itinerary point, including recommended reasons, precautions, recommended duration of play, etc., so as to better arrange and adjust their travel plans. The present invention provides users with personalized pre-tour itinerary planning based on a large model through detailed collection of user travel preferences, and customizes exclusive itineraries. The recommended reasons, precautions, recommended duration of play, etc. generated by the large model are recorded in the itinerary details, which greatly enriches the itinerary information.

[0162] The intelligent explanation function greatly enriches the user's learning experience by popping up various explanation cards such as video explanations and graphic explanations. These cards not only provide detailed audio-visual materials, but also make the information transmission more intuitive and vivid through the combination of pictures and texts. Video explanation cards lead users to deeply understand the history and culture of the scenic spot through dynamic pictures and sounds, while graphic explanation cards help users capture every detail and deepen their memory through carefully designed pictures.

[0163] The service recommendation function is committed to providing users with a comprehensive travel service experience. It not only recommends public services at the destination, such as visitor centers and parking lots, but also covers a variety of travel elements such as attractions, restaurants, hotels, etc. This function provides personalized service options by accurately analyzing the needs and preferences of users, ensuring that users can receive thoughtful care and convenient experience in every aspect of their travels.

[0164] In some preferred implementations, the application applet may further include a companion module, a discovery module, and a personal center module; wherein:

[0165] The tour companion module is used to call the voice explanation service for guided tours based on the result of response processing;

[0166] For example, in some specific implementations, the companion module is used to provide all-round guidance during the tour, realizing functions such as walkman, voice interaction, voice explanation, map guidance, one-key navigation, etc. The companion section is a highlight function of the mini program, providing tourists with walkman and intelligent customer service based on semantic big model and voice recognition technology during the tour.

[0167] The AI ​​real-time tour companion function provides tourists with a new interactive experience with its advanced walkman technology. When tourists authorize the location and turn on the walkman function, once they enter the scenic area within 500 meters, the system can automatically call the large model to trigger the voice explanation service without any manual operation. This intelligent triggering mechanism not only provides users with instant information services, but also greatly enhances the convenience and immersion of travel. At the same time, the AI ​​real-time tour companion also supports voice interaction, which means that users can communicate with the system directly through voice during the tour, just like talking to a knowledgeable tour guide. Whether you are interested in the historical background of a certain attraction or want to understand the local cultural characteristics, users only need to simply ask questions, and the AI ​​system can quickly provide accurate and vivid answers.

[0168] A discovery module, for displaying thematic information to the target object in response to the target object's browsing operation;

[0169] Exemplarily, in some specific implementations, the discovery module is used for displaying thematic information and realizing the functions of thematic information classification and thematic information question and answer jump.

[0170] The Discovery function is a window to explore the unknown. It stimulates the user's curiosity through visual guidance. When the user clicks on a picture during browsing, the system will respond immediately and jump to a carefully designed dialogue topic page. Users can not only obtain rich question and answer information related to the picture, but also gain an in-depth understanding of the cultural stories and historical background behind it.

[0171] A person-centric model is used to store the associated information of the target object and adjust the associated information in response to the operation instructions of the target object; the associated information includes the historical record of the response processing

[0172] For example, in some specific implementations, the personal center module is used to store and display user information, and realize historical conversation recording and deletion, my travel recording and deletion, personal information modification, travel reminder, and complaint feedback functions.

[0173] The Personal Center is the user's private space on the platform, which is used to display and organize personal travel information and settings. It includes historical conversation records, where users can review previous exchanges, and each conversation is properly saved; the My Itinerary section summarizes all the user's travel plans and historical itineraries, which are convenient for users to view and manage at any time; the itinerary reminder function ensures that users will not miss previously planned travel arrangements; the personal information area allows users to edit and manage their account information, ensuring that privacy and preference settings are properly handled; the login and logout function provides users with convenient account management options. The complaint feedback function provides users with an entry point for post-travel complaints. The Personal Center is designed to provide a comprehensive, personalized and easy-to-operate control panel, allowing users to easily manage their travel life while enjoying travel services.

[0174] In order to explain the principle of the technical solution of the present invention in detail, the overall process of the present invention is described below in combination with some specific embodiments. It is easy to understand that the following is an explanation of the technical principle of the present invention and cannot be regarded as a limitation of the present invention.

[0175] First of all, it should be noted that the existing "one-machine tour" tourist service platform for all-region tourism, although providing basic services for tourists, still has limitations in terms of resource concentration, personalized services, interactivity, and intelligence. In order to meet these challenges, improve the intelligence level of tourist guides, optimize the tourist experience, and promote the digital transformation of the cultural and tourism industry, the present invention proposes an innovative tourist guide system based on a semantic big model and its use method to reconstruct the "one-machine tour" tourist service platform for all-region tourism. The system integrates multi-source data and uses advanced artificial intelligence technology for real-time analysis and intelligent decision-making, aiming to provide high-quality guide services. This can not only meet the personalized needs of tourists, but also promote the efficient use and sustainable development of tourism resources, realize the efficient sharing and intelligent analysis of tourism information, and provide strong support for tourism decision-making.

[0176] The implementation of this invention is expected to significantly improve the intelligence and personalization level of tourist guide services, bring tourists a richer, more convenient and accurate travel experience, and inject new vitality into the high-quality development of the cultural and tourism industry.

[0177] To achieve the above-mentioned purpose, the present invention provides a system for tourist guidance, which introduces the powerful content generation capability and conversational interactive experience of a large model. It is different from the traditional "one-machine tour" tourist service platform that only displays the entrances to various resources. It meets the basic requirements of tourist guidance towards resource concentration, personalized services, interactivity, and intelligent development. The system includes an intelligent tourism service and decision-making system, hardware equipment, interface applications, and application applets.

[0178] The intelligent tourism service and decision-making system is deployed on a cloud server, and innovatively builds a system architecture that focuses on a semantic big model, retrieval knowledge enhancement (RAG), agent intelligence, and speech recognition model. The semantic big model is responsible for understanding the user's text input content and generating preliminary answers; retrieval knowledge enhancement searches and filters information closely related to the task from the cultural and tourism industry knowledge base built by telecommunications and the scenic spot ticketing system, and then guides the subsequent answer generation or text construction process based on this information, aiming to improve the accuracy and reliability of the model output; the agent intelligence control module coordinates the work of the model and the retrieval engine, and handles the interaction with the scenic spot ticketing interface; the speech recognition model is responsible for audio content detection, recognizing the user's input speech as text that can be understood by the speech big model.

[0179] All hardware devices are cloud servers, with a total of 6 servers and 1 load balancer, including 2 API interface servers, 2 business processing servers, and 2 database servers. The database server adopts master-slave deployment. All servers are installed with Centos7.4 operating system, business processing servers and interface servers are installed with Nginx+php environment, and database servers are installed with Mysql8.0. All servers can be linearly expanded to accommodate the increasing business needs in the future.

[0180] Interface applications are used to provide seamless connection between front-end operations and business services, ensuring efficient data exchange and service calls.

[0181] The application mini program is used to provide users with convenient access and interactive platforms, ensuring the efficiency and user-friendliness of tourism guide services.

[0182] In some specific application scenarios, such as Figure 5 As shown, the system architecture of the tourism guide system of the embodiment of the present invention is composed as follows:

[0183] 1. Intelligent tourism service and decision-making system:

[0184] The intelligent tourism semantic service and decision-making system includes a corpus collection and processing module, a semantic large model training module, a semantic large model optimization module, a semantic large model deployment and service module, and a speech recognition module.

[0185] Corpus collection and processing module: including data collection, data cleaning, data labeling and data segmentation.

[0186] Semantic big model training module: The Telecom Star big model and the open source big model, Embedding model, and Re-rank model are selected for high-performance training, and the accuracy of model answers is improved by combining retrieval knowledge enhancement (RAG) and agent intelligent technology. Finally, the model performance is evaluated through the validation set.

[0187] Semantic large model optimization module: includes fine-tuning, model compression, incremental learning, and error analysis to improve the performance of the model in the field of tourism guidance.

[0188] Semantic large model deployment and service module: realize the stable operation of the model on Tianyi cloud server, develop large model dialogue API interface, monitor performance in real time, and collect user feedback.

[0189] Speech recognition module: In the direction of speech input of the tourist guide model, an improved CR-CED-based speech signal background denoising method is innovatively proposed. The details are as follows: ① A noise detector module is added to the front end of the CR-CED network. Through binary classification training of whether the collected audio signal contains noise, it is used as the basis for judging whether the subsequent speech denoiser is enabled, and the efficiency of the system is improved through intermittent processing; ② The noisy speech signal is no longer input as a single spectrum amplitude. According to the phase perception principle, the real and imaginary parts of the spectrum are input into two independent channels for training; ③ The speech denoiser no longer aims to predict the clean signal spectrum, but to obtain the best ideal ratio masking in the complex domain. At the same time, the activation function of the convolutional layer is replaced by LeakyReLU from ReLU to meet the negative input requirements of the model.

[0190] 2. Hardware equipment:

[0191] The hardware equipment includes load balancer, API interface server, business processing server, and database server. The database server adopts master-slave deployment, and all devices are installed with Centos7.4 operating system. The interface and business processing servers are installed with Nginx+php environment, and the database server is installed with Mysql8.0. All servers can be linearly expanded.

[0192] 3. Interface application:

[0193] Provide efficient data exchange and service calls to ensure seamless integration with front-end operations and business services.

[0194] It includes question and answer API interface, conversation API interface, itinerary API interface, travel route API interface, scenic spot API interface, hotel and homestay API interface, catering and food API interface, and public resource recommendation API interface.

[0195] 4. Application applet:

[0196] Provide users with convenient access and interactive platforms to ensure the efficiency and user-friendliness of tourism guide services.

[0197] It includes dialogue, travel companion, discovery, and personal center modules, which respectively realize intelligent question and answer, all-round guide, special information display, and user information storage display. The large model covers the entire scene before, during, and after the tour, providing users with personalized itinerary planning, AI real-time travel companion, complaint feedback and other functions.

[0198] In an alternative implementation scenario, such as Figure 6 The business process diagram is shown as follows:

[0199] This flowchart shows the complete process of users experiencing the tourist guide service through the mobile application applet. First, the user launches the application and authorizes information such as WeChat mobile phone number, and then uses this information to log in. The system checks the integrity and validity of the user information. If it is invalid, it outputs an error message and ends the process; if it is valid, the user can choose functions such as itinerary planning, intelligent explanation, service recommendation, discovery, and companion recommendation. The itinerary planning function generates itinerary plans and reminders; the intelligent explanation function provides explanation content; the service recommendation function generates service cards and supports scenic spot ticket purchases; the discovery function displays related content pages; and the companion recommendation function provides real-time voice interaction. All selected functions eventually generate corresponding results and output them to the user, completing the entire experience.

[0200] In summary, the purpose of the present invention is to provide a tourist guide system based on a semantic big model, reconstruct the "one-machine tour" tourist service platform for global tourism, improve the intelligence level of the system, and provide all-round intelligent escort and guide services. Using advanced semantic big model reasoning technology, the tourist attractions and related service data are deeply integrated and optimized to provide itinerary planning, intelligent explanations, service recommendations, escort services, etc. Solve the problems of fragmented information acquisition before the tour and homogeneous services, poor interactivity during the tour, and difficulty in post-tour complaint feedback, and enhance the overall travel experience of tourists.

[0201] Compared with the prior art, the present invention has at least the following beneficial effects:

[0202] 1. Improved CR-CED-based speech signal background denoising method: The present invention uses a public dataset to compare the STOI scores under different models, such as Figure 7 As shown, Figure 7The bold part of the data in the table represents the optimal value of different networks under the same conditions. From the data in the table, it can be seen that under different SNR conditions, the STOI value of the improved CR-CED model is improved in most cases compared with the CR-CED model, and the performance of the audio signal collected in the actual tourism scene can also be improved by 6.8%, which verifies the effectiveness of the denoising function of the improved CR-CED network model.

[0203] 2. Advanced model training and optimization: In the field of tourist navigation, we innovatively propose a system architecture of semantic big model, retrieval knowledge enhancement (RAG), agent intelligence, and speech recognition model to improve the adaptability and performance of the model in the direction of tourist navigation; through model compression and incremental learning, we continuously improve the computing and storage efficiency of the model to maintain the latest knowledge and capabilities.

[0204] 3. Stable and efficient model deployment and service: Through professional deployment and monitoring mechanisms, ensure the stable operation of the model in actual business, conduct performance monitoring, and provide high-quality service experience; develop API interfaces to realize convenient calling of model functions and ensure seamless connection between front-end operations and business services.

[0205] 4. Comprehensive interface application and application app support: Provide rich API interfaces, support a variety of tourism-related business requests, ensure efficient data exchange and service calls; provide users with convenient access and interactive platforms through application applets, and provide users with personalized itinerary planning based on large models, AI real-time companionship, complaint feedback and other functions, covering the entire scene before, during and after the tour. This full-scene coverage service model not only improves the consistency and personalization of the tourism experience, but also optimizes resource allocation through intelligent means, improves service efficiency and customer satisfaction, and demonstrates the innovation and foresight in the field of tourism services.

[0206] 5. System scalability: All servers can be expanded linearly to accommodate the ever-increasing business demands in the future, ensuring that the system can adapt to business growth and changes in user needs.

[0207] On the other hand, the embodiment of the present invention also provides a method for using a tourist guide system. It can be understood that the method for using a tourist guide system provided by the embodiment of the present invention can be applied to any computer device with data processing and computing capabilities, and this computer device can be various types of terminals or servers. When the computer device in the embodiment is a server, the server is an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Optionally, the terminal is a smart phone, a tablet computer, a laptop computer, a desktop computer, etc., but is not limited to this.

[0208] like Figure 8 FIG. 1 is a schematic diagram of an implementation environment provided by an embodiment of the present invention. Figure 8 The implementation environment includes at least one terminal 102 and a server 101. The terminal 102 and the server 101 can be connected to a network wirelessly or wired to complete data transmission and exchange.

[0209] Server 101 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0210] In addition, the server 101 can also be a node server in the blockchain network. Blockchain is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm.

[0211] The terminal 102 may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal 102 and the server 101 may be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiment of the present invention.

[0212] Based on the example Figure 8In the implementation environment shown, an embodiment of the present invention provides a method for using a tourist guide system. The following is an example of the method for using the tourist guide system applied in the terminal 102. It can be understood that the method for using the tourist guide system can also be applied in the server 101.

[0213] Reference Fig. 9 , Fig. 9 The flowchart of the method for using the tourist guide system applied to the terminal provided by the embodiment of the present invention, the execution subject of the method for using the tourist guide system can be any of the aforementioned computer devices (including servers or terminals). Fig. 9 , the method comprises the following steps:

[0214] S100, configuring a semantic big model through a service and decision system and converting the voice data input by the target object into text data;

[0215] S200, deploying a semantic big model and storing business data through hardware devices; wherein the semantic big model is used to respond to business requests;

[0216] S300, connecting the hardware device with the business service of the application applet through the interface application program to perform data exchange and service call;

[0217] S400: Obtain the service request of the target object through the application applet, and feed back the result of the response processing to the target object.

[0218] It should be noted that the contents of the system embodiment of the present invention are applicable to the method embodiment, the process steps specifically implemented by the method embodiment are the same as those of the above system embodiment, and the beneficial effects achieved are also the same as those achieved by the above system. It should be understood that the functional logics implemented by the specific modules of each part of the above system embodiment all belong to the protection scope of the method embodiment of the present invention, and will not be described one by one here.

[0219] On the other hand, an embodiment of the present invention further provides an electronic device, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the method for using the above-mentioned tourist guide system when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0220] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0221] like Fig.10 As shown, Fig.10A specific example of the hardware structure of an electronic device 1000 according to an embodiment is shown. The electronic device 1000 includes:

[0222] The processor 1001 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present invention.

[0223] The memory 1002 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1002 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 1002, and the processor 1001 calls and executes the network node population optimization method of the embodiment of the present invention;

[0224] Input / output interface 1003, used to implement information input and output;

[0225] The communication interface 1004 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0226] A bus 1005 , which transmits information between various components of the device (e.g., the processor 1001 , the memory 1002 , the input / output interface 1003 , and the communication interface 1004 );

[0227] The processor 1001 , the memory 1002 , the input / output interface 1003 and the communication interface 1004 are connected to each other in communication within the device via the bus 1005 .

[0228] The electronic device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0229] The contents of the method embodiments of the present invention are all applicable to the electronic device embodiments. The functions specifically implemented by the electronic device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0230] Another aspect of an embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the above method.

[0231] It should be noted that the computer-readable medium shown in the embodiment of the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read to Only Memory, CD to ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0232] The contents of the method embodiments of the present invention are all applicable to the computer-readable storage medium embodiments. The functions specifically implemented by the computer-readable storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0233] The embodiment of the present invention also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the above method.

[0234] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0235] It should be noted that, although several modules of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0236] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD to ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation of the present invention.

[0237] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.

[0238] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0239] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the solution of the present invention, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0240] The logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be embodied in any computer-readable medium for use by an instruction execution device, apparatus or device (such as a computer-based device, a device including a processor, or other device that can fetch instructions from an instruction execution device, apparatus or device and execute the instructions), or in conjunction with such instruction execution device, apparatus or device. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution device, apparatus or device, or in conjunction with such instruction execution device, apparatus or device.

[0241] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0242] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution device. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0243] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0244] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0245] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. A tourist guide system, characterized in that: Includes service and decision systems, hardware devices, interface applications and application applets; The service and decision system is used to configure the semantic big model and convert the voice data input by the target object into text data; The hardware device is used to deploy the semantic big model and store business data; wherein the semantic big model is used to respond to business requests; The interface application is used to connect the hardware device with the business service of the application applet to perform data exchange and service calls; The application applet is used to obtain the service request of the target object and to feed back the result of the response processing to the target object.

2. The tourist guide system according to claim 1, characterized in that: The service and decision-making system includes: A corpus collection and processing module is used to collect pre-selected tourism data, pre-process the tourism data, and obtain model training data; the pre-processing includes data cleaning, data labeling, and data segmentation; A semantic big model training module, used for training a preset semantic big model using the model training data; A semantic large model optimization module, used for optimizing the semantic large model after training; the optimization process includes model fine-tuning, model compression, incremental learning and error analysis; A model deployment and service module, used to deploy the semantic large model after training and optimization to the cloud server of the hardware device; The speech recognition module is used to perform background denoising and speech recognition on the speech data input by the target object to obtain the text data.

3. The tourist guide system according to claim 2, characterized in that: The speech recognition module performs background denoising and speech recognition on the speech data input by the target object to obtain the text data, which is specifically used to perform the following operations: Performing noise detection on the speech data; When the result of the noise detection indicates that the speech data contains noise information, obtaining a real component and an imaginary component based on the spectrum data of the speech data; Based on the real component and the imaginary component, a complex domain ideal ratio mask is generated using a preset convolutional recurrent connectionist temporal classification network; Based on the spectrum data of the speech data, clean speech information is obtained by performing the complex domain ideal ratio masking process; Based on the clean voice information, the voice recognition is performed using a preset audio model to obtain the text data.

4. The tourist guide system according to claim 1, characterized in that: The hardware device includes a load balancer and multiple servers, and the servers include an API interface server, a business processing server and a database server; The load balancer is used to process the access of the service request and the load distribution of each server; The API interface server is used to process the routing of the service request initiated by the application applet; The business processing server is used to deploy the semantic big model and perform the response processing on each business request based on the semantic big model; The database server is used for storing and managing the business data.

5. The tourist guide system according to claim 1, characterized in that: The business requests include question-and-answer business requests, conversation business requests, itinerary planning business requests, route business requests, scenic spot business requests, hotel and homestay business requests, catering and food business requests, and public resource business requests; The interface application includes: Question and answer API interface, used to connect and process the question and answer service request; Session API interface, used to connect and process the session service request; Itinerary API interface, used to connect and process the itinerary planning business request; Travel route API interface, used to connect and process the route service request; Scenic spot API interface, used to connect and process the scenic spot business request; Hotel and homestay API interface, used to connect and process the hotel and homestay business requests; Catering and food API interface, used to connect and process the catering and food business; The public resource recommendation API interface is used to connect and process the public resource business request.

6. The tourist guide system according to claim 1, characterized in that: The application applet includes: The dialog module is used to obtain the business request in response to the input operation of the target object, and then call the semantic big model to perform the response processing on the business request based on the business request; and feed back the result of the response processing to the target object.

7. The tourist guide system according to claim 6, characterized in that: The application applet also includes a companion module, a discovery module and a personal center module; The companion module is used to call the voice explanation service to perform guided explanation based on the result of the response processing; The discovery module is used to display thematic information to the target object in response to the browsing operation of the target object; The person-centric model is used to store the associated information of the target object and adjust the associated information in response to the operation instruction of the target object; the associated information includes the historical record of the response processing.

8. A method for using a tourist guide system, characterized in that: The following steps are involved: Configure the semantic big model through the service and decision system and convert the voice data input by the target object into text data; The semantic big model is deployed and the business data is stored through hardware devices; wherein the semantic big model is used to respond to business requests; Connecting the hardware device with the business service of the application applet through an interface application to exchange data and call services; The service request of the target object is obtained through the application applet, and the result of the response processing is fed back to the target object.

9. An electronic device, characterized in that: including a processor and a memory; The memory is used to store programs; The processor executes the program to implement the method according to claim 8.

10. A computer storage medium storing a program executable by a processor, characterized in that: The program executable by the processor is used to implement the method according to claim 8 when executed by the processor.

Images