Information processing method and device, storage medium and electronic equipment

Abstract

The invention discloses an information processing method and device, a storage medium and electronic equipment, the method is applied to the technical field of network operation and maintenance, and the method comprises the following steps: receiving text information input by a target user; processing the text information by adopting a large language model, and identifying the user intention of the target user; obtaining at least one piece of equipment information involved by the user intention through the target model, and encoding the at least one piece of equipment information to obtain encoded equipment information; and generating a solution strategy of the problem information based on the encoded equipment information through the target model, and executing the solution strategy. Through application of the method and the device, the problem that the efficiency of solving the user problem is relatively low due to the fact that people need to participate in the solving process when the communication network problem of the user is solved through an artificial intelligence technology in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the field of network operation and maintenance technology, and specifically, to an information processing method, device, storage medium and electronic device. Background Art

[0002] With the rapid development of new-generation information technologies such as 5G, Internet of Things, cloud computing, big data, and artificial intelligence, network connection objects have expanded from connections between people, people and things, and things and things to connections between people, things and network services. The substantial increase in network connection objects and connection dimensions has put higher requirements on network resources, operation and maintenance efficiency, and user experience.

[0003] At present, network operation and maintenance mainly relies on manual experience, with low efficiency, long time consumption and high cost. During the operation and maintenance process, network managers need to monitor, configure, locate faults and handle faults for a large number of devices and systems in the network. The operation and maintenance work of network managers is complex and tedious, and the operation and maintenance efficiency is low. In the existing technology, network operation and maintenance can also respond to network failures reported by users based on question-and-answer models. However, due to the large number of devices and systems involved in the network operation and maintenance process, the limited operation and maintenance knowledge of the question-and-answer model, and the complexity of network failures, when a network failure occurs, it is difficult for the model to quickly locate the cause and point of the failure, and it is even more difficult to quickly formulate a fault handling plan based on network operation and maintenance experience. Human participation in the solution process is required, which affects the efficiency and quality of fault handling.

[0004] Regarding the problem that when using artificial intelligence technology to solve users' communication network problems in related technologies, human participation in the solution process is required, resulting in low efficiency in solving user problems. No effective solution strategy has been proposed yet. Summary of the invention

[0005] The main purpose of the present application is to provide an information processing method, device, storage medium and electronic device to solve the problem in the related art that when solving user communication network problems through artificial intelligence technology, human participation in the solution process is required, resulting in low efficiency in solving user problems.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, there is provided an information processing method, the method comprising: receiving text information input by a target user, wherein the text information includes question information raised by the target user regarding a communication network abnormality; processing the text information using a large language model to identify the user intention of the target user; obtaining at least one device information involved in the user intention through a target model, and encoding the at least one device information to obtain the encoded device information, wherein the device information is parameter information of a network device involved in the user intention; generating a solution strategy for the problem information based on the encoded device information through the target model, and executing the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task comprising determining the device information involved in the user intention, and encoding the device information to obtain the encoded device information, and the second task comprising determining the cause information of the communication network abnormality based on the encoded device information, and generating the solution strategy based on the cause information.

[0007] Furthermore, before using the large language model to process the text information and identify the user intention of the target user, the method also includes: determining historical text information and annotation information corresponding to the historical text information, and constructing a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information contains multiple preset user intentions; and using the first training set to train the preset large language model to obtain the large language model.

[0008] Furthermore, before obtaining at least one device information involved in the user intention through the target model, the method also includes: determining at least one device information involved in the multiple preset user intentions, and determining at least one first information corresponding to each device information in the at least one device information, wherein the device information at least includes: device status information, device function information, and the first information in the at least one first information is the minimum fine-grained parameter information; determining the interface call information of each first information in the at least one first information, and obtaining the interface call information corresponding to the multiple preset user intentions, wherein the interface call information of each first information contains the interface identification information corresponding to each first information; configuring a first mapping relationship between the multiple preset user intentions and the interface call information corresponding to the multiple preset user intentions; and training the preset deep learning model using the multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions, and the first mapping relationship.

[0009] Furthermore, encoding the at least one device information to obtain the encoded device information includes: determining an enumeration value of each device information in the at least one device information, and encoding the enumeration value of each device information to obtain first encoded information, wherein the device information includes at least: device status information, device function information; determining the business scenario intended by the user, and encoding the first encoded information according to multiple device information involved in the business scenario to obtain second encoded information.

[0010] Furthermore, a solution strategy for the problem information is generated based on the encoded device information through the target model, and the solution strategy is executed, including: inputting the encoded device information into the target model to obtain cause information of the communication network abnormality, wherein the cause information includes encoded information consisting of abnormal states of network devices; determining the solution strategy based on the cause information, and determining interface call information corresponding to the solution strategy; and executing the solution strategy based on the interface call information corresponding to the solution strategy.

[0011] Furthermore, before determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy, the method also includes: determining a second mapping relationship between historical cause information and a preset solution strategy, and determining the interface call information of the preset solution strategy; determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy, including: determining the solution strategy based on the second mapping relationship and the cause information, and the interface call information corresponding to the solution strategy; or when the cause information contains third coding information, receiving the solution strategy of the cause information, and the interface call information corresponding to the solution strategy, wherein the third coding information is coding information not included in the historical cause information.

[0012] Furthermore, after executing the solution strategy, the method also includes: receiving feedback from the target user regarding the solution strategy; adjusting the second information based on the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: the text information, the user intention, and at least one device information involved in the user intention; adjusting the solution strategy based on the adjusted second information to obtain an adjusted solution strategy; adjusting the third information using the adjusted second information and the adjusted solution strategy, wherein the third information includes at least one of the following: the historical cause information, the preset solution strategy, and the second mapping relationship.

[0013] In order to achieve the above-mentioned purpose, according to another aspect of the present application, an information processing device is provided, which includes: a first receiving unit, which is used to receive text information input by a target user, wherein the text information includes question information raised by the target user regarding a communication network anomaly; an identification unit, which is used to process the text information using a large language model to identify the user intention of the target user; an encoding unit, which is used to obtain at least one device information involved in the user intention through a target model, and encode the at least one device information to obtain the encoded device information, wherein the device information is parameter information of a network device involved in the user intention; an execution unit, which is used to generate a solution strategy for the problem information based on the encoded device information through the target model, and execute the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task includes determining the device information involved in the user intention, and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the communication network anomaly based on the encoded device information, and generating the solution strategy based on the cause information.

[0014] Furthermore, the device also includes: a construction unit, which is used to determine historical text information and annotation information corresponding to the historical text information before using the large language model to process the text information and identify the user intention of the target user, and construct a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information contains multiple preset user intentions; a first training unit, which is used to train the preset large language model using the first training set to obtain the large language model.

[0015] Furthermore, the device also includes: a first determination unit, used to determine at least one device information involved in the multiple preset user intentions before obtaining at least one device information involved in the user intention through the target model, and determine at least one first information corresponding to each device information in the at least one device information, wherein the device information at least includes: device status information, device function information, and the first information in the at least one first information is the smallest fine-grained parameter information; a second determination unit, used to determine the interface call information of each first information in the at least one first information, and obtain the interface call information corresponding to the multiple preset user intentions, wherein the interface call information of each first information contains the interface identification information corresponding to each first information; a configuration unit, used to configure a first mapping relationship between the multiple preset user intentions and the interface call information corresponding to the multiple preset user intentions; a second training unit, used to train the preset deep learning model using the multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions and the first mapping relationship.

[0016] Furthermore, the encoding unit includes: a first encoding subunit, used to determine the enumeration value of each device information in the at least one device information, and encode the enumeration value of each device information to obtain first encoded information, wherein the device information includes at least: device status information, device function information; a second encoding subunit, used to determine the business scenario intended by the user, and encode the first encoded information according to multiple device information involved in the business scenario to obtain second encoded information.

[0017] Furthermore, the execution unit includes: a processing subunit, used to input the encoded device information into the target model to obtain cause information of the communication network abnormality, wherein the cause information includes encoded information consisting of abnormal status of the network device; a first determination subunit, used to determine the solution strategy based on the cause information, and determine the interface call information corresponding to the solution strategy; an execution subunit, used to execute the solution strategy based on the interface call information corresponding to the solution strategy.

[0018] Furthermore, the execution unit also includes: a second determination subunit, which is used to determine a second mapping relationship between historical cause information and a preset solution strategy, and determine the interface call information of the preset solution strategy before determining the solution strategy based on the cause information and determining the interface call information corresponding to the solution strategy; the third determination subunit includes: determining the solution strategy and the interface call information corresponding to the solution strategy based on the second mapping relationship and the cause information; or a receiving subunit, which is used to receive the solution strategy of the cause information and the interface call information corresponding to the solution strategy when the cause information contains third coding information, wherein the third coding information is coding information not included in the historical cause information.

[0019] Furthermore, the device also includes: a second receiving unit, used to receive feedback from the target user on the solution strategy after executing the solution strategy; a first adjustment unit, used to adjust the second information according to the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: the text information, the user intention, and at least one device information involved in the user intention; a second adjustment unit, used to adjust the solution strategy according to the adjusted second information to obtain an adjusted solution strategy; a third adjustment unit, used to adjust the third information using the adjusted second information and the adjusted solution strategy, wherein the third information includes at least one of the following: the historical cause information, the preset solution strategy, and the second mapping relationship.

[0020] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a computer program product is provided, including a computer program, which, when executed by a processor, implements any one of the above-mentioned information processing methods, and which, when executed by a processor, implements the steps of the information processing methods described in each embodiment of the present application.

[0021] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes stored computer instructions, wherein when the computer instructions are executed by a processor, any one of the above-mentioned information processing methods is implemented.

[0022] In order to achieve the above-mentioned purpose, according to one aspect of the present application, an electronic device is provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement any one of the above-mentioned information processing methods.

[0023] Through the present application, the following steps are adopted: receiving text information input by a target user, wherein the text information includes question information raised by the target user regarding a communication network anomaly; using a large language model to process the text information and identify the user intention of the target user; obtaining at least one device information involved in the user intention through a target model, and encoding the at least one device information to obtain the encoded device information, wherein the device information is parameter information of a network device involved in the user intention; generating a solution strategy for the problem information based on the encoded device information through the target model, and executing the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task including determining the device information involved in the user intention, and encoding the device information to obtain the encoded device information, the second task including determining the cause information of the communication network anomaly based on the encoded device information, and generating the solution strategy based on the cause information, thereby solving the problem that when solving a user's communication network problem through artificial intelligence technology in the related art, manual participation in the solution process is required, resulting in low efficiency in solving user problems. By processing the text information input by the user through a large language model, the user's real intention can be accurately identified and understood, thereby achieving a precise understanding of the user's intention, so that the relevant device information can be identified based on the user's intention, and the information can be encoded for digital parsing and processing. At the same time, the encoded device information is inferred through the target model to generate a solution strategy for the user's problem, and this strategy can be automatically executed to complete the diagnosis and repair of network problems, reducing the need for manual intervention, reducing operation and maintenance costs, and significantly shortening the time for problem diagnosis and resolution, thereby achieving the technical effect of quickly responding to user needs and improving operation and maintenance and service efficiency, and further achieving the technical effect of improving user satisfaction with the service and service efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0025] Figure 1 is a flowchart of an information processing method provided according to Embodiment 1 of the present application;

[0026] Figure 2 This is a flowchart of an optional method for analyzing network status and generating a solution strategy according to the first embodiment of the present application;

[0027] Figure 3 is a flowchart of an optional method for learning the association between user intention and device information provided according to the first embodiment of the present application;

[0028] Figure 4 is a flowchart of optional digital encoding of device information provided according to Embodiment 1 of the present application;

[0029] Figure 5 It is a schematic diagram of an optional generation solution strategy provided according to the first embodiment of the present application;

[0030] Figure 6 is a flowchart of an optional generation solution strategy provided according to Embodiment 1 of the present application;

[0031] Figure 7 is a schematic diagram of an information processing device provided according to Embodiment 2 of the present application;

[0032] Figure 8 It is a schematic diagram of an information processing electronic device provided according to Embodiment 5 of the present application. DETAILED DESCRIPTION

[0033] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0034] It should be noted that the user information (including but not limited to user device information, user personal information, collected data, used data, generated data, processed data, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, collected information, used information, generated information, processed information, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data are in compliance with the relevant laws, regulations and standards of relevant countries and regions, necessary confidentiality measures are taken, and public order and good morals are not violated, and corresponding operation portals are provided for users to choose to authorize or refuse. For example, an interface is set between this system and relevant users or organizations. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or organization.

[0035] It should be noted that this application provides users with corresponding operation entrances for them to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.

[0036] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0037] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0038] Embodiment 1

[0039] The present invention is described below in conjunction with preferred implementation steps. Figure 1 is a flowchart of an information processing method provided according to Embodiment 1 of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0040] Step S101: receiving text information input by a target user, wherein the text information includes question information raised by the target user regarding communication network anomalies.

[0041] The information processing method provided in the first embodiment of the present application is applied to a maintenance system of a network device. Through the maintenance system, the user intention is correctly grasped and matched, the network situation is analyzed and a solution strategy is generated. Therefore, in the face of complex communication network situations, the analysis method can be used in a standard and efficient manner to achieve the learning and iterative capabilities of complex communication network situation analysis, solution arrangement, and operation verification in home broadband scenarios, thereby improving the efficiency of self-service.

[0042] In the first embodiment, it is necessary to receive the original input from the target user, which is in the form of text. For example, the target user may submit network abnormality information or problem description through SMS, chat application or self-service platform. For example, the text message sent by the user may be: "My network connection is unstable and often disconnected." or "Why is my network speed so slow?"

[0043] Step S102: Use a large language model to process the text information and identify the user intention of the target user.

[0044] In the first embodiment, an advanced large language model is used to understand and parse the needs or questions expressed by users in text form, so as to accurately identify the user's true intention. For example, when a user enters "My network connection is unstable and sometimes disconnects" through a service channel, the large language model can understand and identify that the user's intention is "network connection problem".

[0045] The big language model can be the Xingchen big model or the Qiming network big model independently developed by China Telecom, which can process various text information generated by users during the interaction process, such as user consultations, feedback or complaints, etc. Through the semantic understanding ability of the model, these unstructured text information can be converted into structured user intentions, providing a clear direction for the subsequent network situation analysis and the generation of solution strategies.

[0046] Step S103, obtaining at least one piece of equipment information involved in the user's intention through the target model, and encoding the at least one piece of equipment information to obtain encoded equipment information, wherein the equipment information is parameter information of the network equipment involved in the user's intention.

[0047] In the first embodiment of the present invention, the target model can determine at least one network device (i.e., device) related to the user intent, and the device information of each network device (i.e., the above-mentioned device information) based on the identified user intent. The device information includes, but is not limited to, the device's attribute information, operation interface, performance indicators, status information, and device supported function information, wherein the attribute information includes the device's model, manufacturer, serial number, physical location, management IP address, hardware configuration (such as CPU, memory, storage capacity, etc.) and software version (operating system, firmware version, etc.), which helps to identify and locate the device, and also provides a basis for the compatibility and performance evaluation of the device; the operation interface refers to the interface provided by the network device for communicating and controlling it, for example, a physical interface, such as an Ethernet port, a fiber optic interface, etc., or a logical interface, such as one based on SNMP (Simple Network Management Protocol), NETCONF (Network Configuration Protocol), REST API and other protocols, through which maintenance personnel can perform equipment configuration, status query, troubleshooting and other operations; performance indicators may include throughput, latency, packet loss rate, CPU utilization, memory utilization, etc. These indicators are very important for evaluating network service quality, equipment load and network fault points, and are also an important basis for network optimization; status information refers to the current operating status of network equipment, such as online, offline, fault, maintenance mode, or the connection status of each interface of the device (such as UP, DOWN), service status (such as start, stop), system log information, etc. Status information is crucial for real-time monitoring of the health status of network equipment and timely discovery and handling of problems; functional information supported by the device includes all services and functions that the device can provide, such as routing function, switching function, firewall, NAT (network address translation), QoS (quality of service) control, VLAN (virtual local area network) division, etc. Device function information helps to reasonably plan network architecture and service configuration to ensure that the network can meet user needs.

[0048] In order to enable the target model to learn and understand the network status more effectively, the device information of at least one device needs to be digitally encoded. For example, the enumeration value of the device status is digitally encoded in standard 0, 1, 2, 3, etc., where 0 usually represents a normal state and other numbers represent different types of abnormal states.

[0049] By associating user intent with device information and encoding the device information, the intelligent agent can understand user needs more accurately and quickly diagnose network problems, providing a basis for subsequent solution strategy orchestration and operation, achieving the technical effect of improving operation and maintenance efficiency and accuracy, while also providing data support for the continuous learning and optimization of the intelligent agent.

[0050] Step S104, generating a solution strategy for the problem information based on the encoded device information through the target model, and executing the solution strategy, wherein the target model is a model trained based on the first task and the second task, the first task includes determining the device information involved in the user's intention, and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the communication network abnormality based on the encoded device information, and generating a solution strategy based on the cause information.

[0051] In the first embodiment, the target model generates and executes a solution strategy corresponding to the user input problem based on the encoded device information, and executes the solution strategy. The solution strategy may include a series of operations, such as adjusting network parameters, restarting devices, optimizing resource allocation, etc., in order to restore the normal operation of the network and solve the problem reported by the user. The execution process may require calling the API interface of the network device to realize the actual control of the network.

[0052] The target model is a model trained by deep learning or machine learning technology, which can analyze the network situation according to the status of network devices (i.e. the above-mentioned device information) and then generate a solution strategy. The training of the model is divided into two stages, corresponding to the learning tasks of the two stages: in the first stage, the model learning task is to identify the network device information related to the user's intention and encode the device information, that is, convert the device status value into a digital form that is easy for the machine to process, and understand the network situation in the form of structured data; in the second stage, the model learning task is to identify the pattern of abnormal state combination according to the encoded device information, diagnose the cause information of the communication network abnormality, and generate an effective solution strategy based on the cause information.

[0053] Through the training of the above two tasks, the target model can realize the automation of the whole process from understanding user problems to diagnosing network problems, and then to the generation and execution of solution strategies, which greatly improves the efficiency of network maintenance and services, while also reducing the need for manual intervention and improving customer satisfaction. Especially in complex home broadband installation and maintenance scenarios, problems can be solved more quickly and accurately.

[0054] In summary, the information processing method provided in Example 1 of the present application receives text information input by a target user, wherein the text information includes question information raised by the target user regarding a communication network abnormality; uses a large language model to process the text information and identifies the user intent of the target user; obtains at least one device information involved in the user intent through a target model, and encodes at least one device information to obtain the encoded device information, wherein the device information is parameter information of a network device involved in the user intent; generates a solution strategy for the problem information based on the encoded device information through a target model, and executes the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task includes determining the device information involved in the user intent, and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the communication network abnormality based on the encoded device information, and generating a solution strategy based on the cause information, thereby solving the problem in the related art that when solving a user's communication network problem through artificial intelligence technology, manual participation in the solution process is required, resulting in low efficiency in solving user problems. By processing the text information input by the user through a large language model, the user's real intention can be accurately identified and understood, thereby achieving a precise understanding of the user's intention, so that the relevant device information can be identified based on the user's intention, and the information can be encoded for digital parsing and processing. At the same time, the encoded device information is inferred through the target model to generate a solution strategy for the user's problem, and this strategy can be automatically executed to complete the diagnosis and repair of network problems, reducing the need for manual intervention, reducing operation and maintenance costs, and significantly shortening the time for problem diagnosis and resolution, thereby achieving the technical effect of quickly responding to user needs and improving operation and maintenance and service efficiency, and further achieving the technical effect of improving user satisfaction with the service and service efficiency.

[0055] Optionally, in the information processing method provided in Example 1 of the present application, before using a large language model to process text information and identify the user intent of the target user, the method further includes: determining historical text information and annotation information corresponding to the historical text information, and constructing a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information includes multiple preset user intents; and using the first training set to train the preset large language model to obtain the large language model.

[0056] In the first embodiment, in order to understand the user's intention, the large language model can be trained using a training set consisting of historical text information to enhance its ability to understand the user's intention and make corresponding network status analysis.

[0057] Specifically, various types of historical text information input by users in the past period of time (i.e., the above-mentioned preset time period, for example, within 1 month, 6 months, 3 years, 10 years, etc. from the current time) can be collected and organized. This information can be the user's questions, requests, or the content of the interactive dialogue. At the same time, for each piece of historical text information, there needs to be corresponding annotation information. These annotation information are determined by manual or pre-set rules, indicating the user intention corresponding to the historical text information. The user intention in the annotation information is a pre-set standard user intention, such as asking about network failures, requesting network optimization, or consulting service information.

[0058] Then, the historical text information is paired with their corresponding annotation information to form a first training set containing text and user intent, which is used as input for training the large language model. This training set can help the large language model learn the user intent patterns hidden behind different text expressions, thereby improving the accuracy of the large language model in identifying user intent.

[0059] Finally, the large language model is trained using the constructed first training set, so that it can gradually master the ability to parse intent from user input by learning historical text information and its annotated user intent, and can more accurately convert users' natural language descriptions into standard user intent, thereby guiding subsequent network status queries, abnormal cause diagnosis, and solution strategy generation processes, thereby improving overall service efficiency and user experience.

[0060] By making full use of historical data, not only can the accuracy and robustness of the model be improved, but the model performance can also be continuously optimized as data accumulates, enabling it to more accurately identify user intent.

[0061] Optionally, in the information processing method provided in Example 1 of the present application, before obtaining at least one device information involved in the user intention through the target model, the above method also includes: determining at least one device information involved in multiple preset user intentions, and determining at least one first information corresponding to each device information in the at least one device information, wherein the device information includes at least: device status information, device function information, and the first information in the at least one first information is the smallest fine-grained parameter information; determining the interface call information of each first information in the at least one first information, and obtaining the interface call information corresponding to multiple preset user intentions, wherein the interface call information of each first information contains the interface identification information corresponding to each first information; configuring a first mapping relationship between multiple preset user intentions and the interface call information corresponding to the multiple preset user intentions; and training the preset deep learning model using multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions, and the first mapping relationship.

[0062] In the first embodiment, in order to solve the problem raised by the target user, it is necessary to identify and understand the user's intention, determine the network device (or system) information related to the user's intention, and diagnose the cause of the network abnormality and generate a solution strategy based on the device information. For example, when the target user reports that the broadband connection is unstable, it is necessary to identify that the intention of "unstable broadband connection" is related to devices such as home routers, ONUs (optical network units), and OLTs (optical line terminals), and determine the cause of the abnormality and generate a solution strategy based on these device information.

[0063] In an optional embodiment, the device status information involved in each user intention and the device function information involved in each user intention are further decomposed into finer-grained parameter information, namely the first information mentioned above. The first information of the device status information is the most basic element for monitoring network devices, such as the signal strength of the router, the bit error rate of the ONU, etc. The first information of the device function information is the most basic element for controlling network devices, such as the connection model of the wireless access point, the working model of the device, the configuration interface of the device, etc. It should be noted that the above-mentioned device information includes at least one of the following: device status information, device function information, and may also include device attribute information, operation interface, performance indicators and other information. By separately determining the first information of the device status information and the first information of the device function information, the target model can understand and operate the network status more accurately.

[0064] Then, the application program interface (hereinafter referred to as API) call information for querying or operating each first information involved in the user's intention is determined, including the interface identification information corresponding to each first information. For example, querying the signal strength of a router may require calling a specific API interface, and this interface will have a unique identification for accurate calling.

[0065] Secondly, the user intention is mapped with the corresponding device information and the interface call information of each first information in the device information to obtain the above-mentioned first mapping relationship. This configuration establishes a bridge between the user intention and the calling device interface, so that the maintenance system can automatically call the corresponding device interface and perform operations according to the user intention.

[0066] Finally, the preset deep learning model is trained using the configured first mapping relationship, as well as the collected user intent and interface call information. Through a large amount of training data, the target model will learn the process of identifying the correct device information according to the user intent and calling the corresponding interface, so that the target model can continuously improve its ability to understand and process network services, thereby responding to user requests more efficiently in actual applications.

[0067] Optionally, in the information processing method provided in the first embodiment of the present application, at least one device information is encoded to obtain the encoded device information, including: determining an enumeration value of each device information in the at least one device information, and encoding the enumeration value of each device information to obtain first encoded information, wherein the device information includes at least: device status information, device function information; determining the business scenario intended by the user, and encoding the first encoded information based on multiple device information involved in the business scenario to obtain second encoded information.

[0068] In the first embodiment, in order to enable the target model to understand and process these states, it is necessary to convert at least one device information involved in the user's intention into a digital code that can be recognized and processed by the computer.

[0069] In a communication network, the status information of each device can contain multiple values, such as "normal", "fault", "maintenance", etc. These status representations are called enumeration values ​​in computer science, which are a predefined series of specific values. For example, "normal" can be encoded as 0, "fault" can be encoded as 1, "maintenance" can be encoded as 2, and so on. The functional information of each device can contain multiple values. For example, the working mode of the device includes a normal connection model or a wired connection mode, where the normal connection model can be encoded as "1" and the wired connection mode can be encoded as "2", etc.; or, the working mode of the device includes a custom configuration and a factory default configuration, where the custom configuration can be encoded as "1" and the factory default configuration can be encoded as "2", etc.

[0070] In an optional embodiment, each device status information and each device function information in at least one device information can be converted into an enumeration value, and after these enumeration values ​​are converted into digital codes, the above-mentioned first coded information is generated, which reflects the real-time status of the current network. It should be noted that the above-mentioned device information includes at least one of the following: device status information, device function information, and may also include device attribute information, operation interface, performance indicators and other information.

[0071] Then, the expression of the user's intention may be related to a specific business scenario. After determining the business scenario of the user's intention, the encoding information of the device information related to the business scenario in the first encoding information may be further processed, for example, performing combined encoding or specific encoding conversion. For example, the user inputs "My network connection is unstable", and the user's current network belongs to home broadband, so it is necessary to understand the device information involved in the home broadband scenario, determine the enumeration value of the device information involved in the home broadband scenario, obtain the encoding information in the home broadband scenario, and combine the above-mentioned first encoding information with the encoding information in the home broadband scenario to obtain the above-mentioned second encoding information.

[0072] Alternatively, if the status of device A and device B are both abnormal (coded as 1 and 2 respectively), and the status of device C is normal (coded as 0), in the home broadband service scenario, these status codes may need to be combined to form a composite code that reflects the network status in the scenario, namely the second coding information mentioned above. The second coding information can provide a more comprehensive view of the network situation, helping the target model to more accurately understand the nature of the problem and formulate corresponding solutions.

[0073] Through the above steps, the target model can not only identify the status of each independent device in the network, but also comprehensively analyze the network situation according to the specific needs of users and the complex business scenarios of the network, so as to make more accurate solution strategy judgments.

[0074] Optionally, in the information processing method provided in Example 1 of the present application, a solution strategy for problem information is generated based on the encoded device information through a target model, and the solution strategy is executed, including: inputting the encoded device information into the target model to obtain cause information of the communication network abnormality, wherein the cause information includes encoded information consisting of abnormal states of network devices; determining a solution strategy based on the cause information, and determining interface call information corresponding to the solution strategy; and executing the solution strategy based on the interface call information corresponding to the solution strategy.

[0075] In an optional embodiment, after analyzing the input coded information, the target model outputs cause information, which includes the code of the identified abnormal state of the network device or the combination of codes of the abnormal state, that is, the device information in the abnormal state determined by the target model, and the possible causes or patterns of these abnormal states.

[0076] After determining the cause of the communication network anomaly, in order to restore the normal operation of the network or optimize network performance, it is necessary to determine the specific measures taken by the maintenance system, that is, the above-mentioned solution strategy, which may include restarting the device, adjusting parameters, implementing repair operations, automated device operations, system configuration adjustments, etc. In order to execute the solution strategy, it is necessary to determine the called API interface, which is connected to the network device or other business system, allowing the maintenance system to directly or indirectly control the device or obtain necessary information. The process of executing the solution strategy may include multiple steps, and its operation can be dynamically adjusted according to the current network status and feedback information of the solution strategy until the abnormal status is resolved or the network performance is optimized.

[0077] By learning and understanding the network status, automatically analyzing the cause of the anomaly, and formulating and implementing solution strategies, the efficiency and effectiveness of network maintenance can be improved, manual intervention can be reduced, and the effect of improving operation and maintenance efficiency can be achieved.

[0078] Optionally, in the information processing method provided in the first embodiment of the present application, before determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy, the above method also includes: determining a second mapping relationship between the historical cause information and the preset solution strategy, and determining the interface call information of the preset solution strategy; determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy, including: determining the solution strategy based on the second mapping relationship and the cause information, and the interface call information corresponding to the solution strategy; or when the cause information contains third coding information, receiving the solution strategy of the cause information, and the interface call information corresponding to the solution strategy, wherein the third coding information is the coding information not included in the historical cause information.

[0079] In an optional embodiment, the historical cause information of past network failures (i.e., the network status, device status, etc. when the failure occurred) can be matched with the known solution strategy to obtain the above-mentioned second mapping relationship. By learning the second mapping relationship, the target model can identify the association between a specific network status or device status and the corresponding solution strategy, so that the corresponding solution strategy can be quickly located after determining the abnormal cause of the target user. After determining the solution strategy, it is necessary to further determine the interface call information required to execute the solution strategy, such as restarting the device, updating the configuration, checking the connection, etc.

[0080] In an optional embodiment, when the determined cause information is new coded information that has not appeared in the historical data (i.e., the third coded information described above, which is equivalent to the emergence of a new network state or fault type), a solution strategy and its interface call information manually input by an expert (or staff) are received to handle these newly emerged network problems. This approach allows the system to obtain solutions even when encountering unknown problems.

[0081] By analyzing historical fault data and preset solution strategies, and determining the solution strategies corresponding to the cause information that has not appeared, the solution strategies for network faults can be quickly determined and executed, improving the efficiency and accuracy of fault diagnosis and resolution.

[0082] Optionally, in the information processing method provided in Example 1 of the present application, after executing the solution strategy, the above method also includes: receiving feedback from the target user on the solution strategy; adjusting the second information based on the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: text information, user intention, and at least one device information related to the user intention; adjusting the solution strategy based on the adjusted second information to obtain an adjusted solution strategy; adjusting the third information using the adjusted second information and the adjusted solution strategy, wherein the third information includes at least one of the following: historical cause information, a preset solution strategy, and a second mapping relationship.

[0083] In an optional embodiment, after the solution strategy is pushed to the target user and the solution strategy is executed, the target user's feedback on the solution strategy can be collected. Based on the target user's feedback, if it is determined that the user intent recognition is incorrect, the user intent can be modified; if it is determined that the device information involved in the target model analysis of the user intent is incorrect, the corresponding device information or the first mapping relationship can be adjusted; or the text information can be directly corrected to more accurately reflect the target user's real needs and the network status that needs to be processed, in order to improve the pertinence and effectiveness of the subsequent solution strategy.

[0084] After the second information is adjusted, the solution strategy can be adjusted according to the adjusted second information, that is, the details of the solution strategy are fine-tuned, or the direction of the entire solution strategy is changed to better match the real intention of the target user and the current state of the network. When the solution strategy is adjusted, this change can be fed back to the third information, for example, updating the solution record of the historical cause information, modifying the preset solution strategy, or adjusting the second mapping relationship between the historical cause information and the preset solution strategy.

[0085] In addition, secondary intelligent orchestration can be performed based on the combined solution strategy and feedback to deduce a solution strategy with a shorter path, that is, to determine the number of API interfaces that require fewer calls.

[0086] By continuously collecting and analyzing user feedback, we can gradually optimize the target model's understanding of user intent, more accurately analyze network status, and provide more effective solution strategies, which will help improve user satisfaction, reduce on-site service time, increase the proportion of user self-service, and improve overall service efficiency.

[0087] Optionally, Figure 2 It is a flowchart for analyzing network situation and generating solution strategies, in which: Figure 2 The intelligent agent in is equivalent to the target model in the first embodiment of the present invention. In this process, the original user voice can be used as input, and through text translation and system mapping matching, it is parsed into user intent by the large language model intention recognition learning module. Subsequently, the intention and device mapping learning module maps the user intent to a specific device, namely a network device or system, platform, etc. The device status analysis module further analyzes the current status of the device through network status analysis learning and device solution strategy learning, determines whether there is an abnormality, and forms a solution strategy for the abnormal status. Finally, the intelligent agent performs orchestration and scheduling execution, and realizes autonomous intelligent orchestration and processing of network anomalies based on multiple rounds of interactive calls of the orchestration learning scheme. In this process, through the interactive capabilities of the large language model (such as Figure 2The intention mapping device call, network status anomaly diagnosis output and automatic processing and execution of the program orchestration) enhance the interactivity between the intelligent agent and the user or system. Through continuous interaction and calling, the intelligent agent can learn and optimize its ability to analyze the network situation and the orchestration strategy of the solution strategy, realize self-processing execution, and interact with the user after verification to form a closed loop of multiple rounds of intelligent processing. The key to the intelligent agent's autonomous deduction of intelligent orchestration lies in the multiple rounds of calling based on the orchestration learning plan, combined with the stored successful experience data of the orchestration execution. The intelligent agent can autonomously perform intelligent orchestration for multi-scenario anomalies, significantly improving the efficiency and accuracy of home broadband maintenance.

[0088] Optionally, Figure 3 It is a flowchart for learning the relationship between user intention and device information. The flowchart mainly includes four steps. First, the standardized management of user intention. The input original speech text is converted into a unified intention identifier through standardized management. For example, the original voice of broadband connection failure is converted to YTBS0001, and the broadband-weak signal is converted to YTBS0005. Next, the ability to query the device status is decomposed into multiple atomic API interfaces, such as broadband access type query interface CX_WY0001, arrears status interface CX_WY0004, and Lianchuang authentication status interface CX_WY0003. Subsequently, the standardized user intention and the encapsulated atomic API interface are mapped to form an association relationship between the intention and the device status query call. Finally, the atomic dimension device parameter API call mapping configuration is used to realize the query call of the device status. By learning this mapping relationship, the intelligent agent can autonomously realize the API call, realize the analysis of the network situation and the arrangement of the solution strategy, thereby building the relationship between the user intention and the device information or device interface, and realizing the accurate mapping of the intention to the device status query. Through this process, the intelligent agent can understand user problems more accurately, quickly locate network anomalies, and take effective measures to resolve them, thereby improving the response speed of broadband services and user satisfaction.

[0089] Optionally, Figure 4This is a flowchart for digitally encoding device information. Taking broadband unavailable as an example, the agent first maps to device information based on user intent, for example, CX_WY0003 queries resource status, CX_WY0010 queries Lianchuang account status, CX_WY0011 queries ONU operating status, etc. Subsequently, the device status digital encoder system converts the enumeration value of the device status into a digital code to form a network status digital code, which is convenient for the agent to understand and process. For example, the ONU operating status is coded as 0 for normal, online status, coded as 1 for offline (Loss Of Signal, LOS, signal loss), coded as 2 for power failure, coded as 3 for offline (DOWN, device offline), and coded as 4 for acquisition failure. Then, the agent performs situation analysis learning based on the coding results to determine whether the network is in an abnormal state. Finally, the combined coding output combines multiple status codes to form a combined code representing the overall network status, for example, 01000 represents arrears status, and 00100 represents Lianchuang shutdown status. Through digital coding, intelligent agents can process network status information more efficiently, quickly identify problems, and call corresponding solution strategies, thereby achieving rapid response and processing of network anomalies, improving the stability of broadband services and user experience.

[0090] Optionally, Figure 5 Schematic diagram of generating solution strategies. The coded values ​​generated by the device status digital encoder system are sent to the device status solution strategy management system, which stores the solution strategies for network abnormal states, such as payment reminders and payment operations for arrears, resource verification and correction operations for unregistered ONUs, etc. Then, the intelligent agent performs multi-state combination arrangement review and correction. According to the combined coding of the device status, the intelligent agent learns and arranges the combination of solution strategies to form a solution strategy for a specific network status. For example, the state combination 01000 indicates that the user is in arrears shutdown state, and the solution strategies that the intelligent agent may arrange include arrears reminders, payment operations, etc. In this process, the intelligent agent conducts deep learning based on the combined solution strategy and the expert verification results to achieve secondary intelligent arrangement of the solution strategy, that is, to adjust the combination and operation sequence of the solution strategy through learning to optimize the processing efficiency and the success rate of problem solving. The key to this process lies in the learning ability of the intelligent agent and the flexible arrangement of the solution strategy. Through intelligent solution strategy arrangement, the intelligent agent can respond to various network anomalies more accurately, achieve rapid problem location and resolution, and improve the quality of broadband services and user satisfaction.

[0091] Optionally, Figure 6It is a flowchart for generating a solution strategy. The process consists of four steps: First, the solution strategy is arranged. The intelligent agent (i.e., the target model mentioned above) learns network situation analysis and solution strategy arrangement, integrates the relationship between user intention, device status, digital coding, and solution strategy, and conducts deep learning and optimization. Next, the solution strategy is executed, i.e., the intelligent agent calls the API interface, and the execution of the solution strategy contains the API interface corresponding to the operation. Subsequently, the solution strategy is verified, i.e., the execution result of the solution strategy generated by the intelligent agent is verified. Through user feedback and manual annotation, the solution strategy is modified and optimized to achieve iterative optimization of the intelligent agent. Finally, the intelligent agent verifies the iteration module. Through multiple rounds of verification, feedback, and iteration, the intelligent agent learns and optimizes its solution strategy arrangement strategy to improve processing efficiency and accuracy. The key to this process is to establish a feedback mechanism and continuous iterative optimization. Through the close cooperation between the production application system and the intelligent agent, the solution strategy can be quickly verified and corrected to ensure that the performance of the intelligent agent is continuously improved, adapt to the ever-changing network environment, and provide a stable and efficient service experience for home broadband users.

[0092] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0093] Embodiment 2

[0094] The second embodiment of the present application also provides an information processing device. It should be noted that the information processing device of the second embodiment of the present application can be used to execute the information processing method provided in the first embodiment of the present application. The information processing device provided in the second embodiment of the present application is introduced below.

[0095] Figure 7 Schematic diagram of an information processing device according to Embodiment 2 of the present application. Figure 7 As shown, the device includes: a first receiving unit 701, a recognition unit 702, an encoding unit 703 and an execution unit 704.

[0096] Specifically, the first receiving unit 701 is used to receive text information input by a target user, wherein the text information includes question information raised by the target user regarding the abnormality of the communication network.

[0097] The recognition unit 702 is used to process the text information using a large language model to recognize the user intention of the target user.

[0098] The encoding unit 703 is used to obtain at least one device information involved in the user intention through the target model, and encode the at least one device information to obtain the encoded device information, wherein the device information is parameter information of the network device involved in the user intention.

[0099] The execution unit 704 is used to generate a solution strategy for the problem information based on the encoded device information through the target model, and execute the solution strategy, wherein the target model is a model trained based on the first task and the second task, the first task includes determining the device information involved in the user's intention, and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the communication network abnormality based on the encoded device information, and generating a solution strategy based on the cause information.

[0100] The information processing device provided in the second embodiment of the present application receives text information input by a target user through a first receiving unit 701, wherein the text information includes question information raised by the target user regarding the abnormality of the communication network; the identification unit 702 uses a large language model to process the text information and identify the user intention of the target user; the encoding unit 703 obtains at least one device information involved in the user intention through the target model, and encodes the at least one device information to obtain the encoded device information, wherein the device information is parameter information of the network device involved in the user intention; the execution unit 704 generates a solution strategy for the problem information based on the encoded device information through the target model, and executes the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task includes determining the device information involved in the user intention, and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the abnormality of the communication network based on the encoded network element device information, and generating a solution strategy based on the cause information, which solves the problem in the related art that when solving the user's communication network problem through artificial intelligence technology, manual participation in the solution process is required, resulting in low efficiency in solving the user's problem. By processing the text information input by the user through a large language model, the user's real intention can be accurately identified and understood, thereby achieving a precise understanding of the user's intention, so that the relevant device information can be identified based on the user's intention, and the information can be encoded for digital parsing and processing. At the same time, the encoded device information is inferred through the target model to generate a solution strategy for the user's problem, and this strategy can be automatically executed to complete the diagnosis and repair of network problems, reducing the need for manual intervention, reducing operation and maintenance costs, and significantly shortening the time for problem diagnosis and resolution, thereby achieving the technical effect of quickly responding to user needs and improving operation and maintenance and service efficiency, and further achieving the technical effect of improving user satisfaction with the service and service efficiency.

[0101] Optionally, in the information processing device provided in Example 2 of the present application, the above-mentioned device also includes: a construction unit, which is used to determine historical text information and annotation information corresponding to the historical text information before using the large language model to process the text information and identify the user intention of the target user, and construct a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information includes multiple preset user intentions; a first training unit, which is used to train the preset large language model using the first training set to obtain the large language model.

[0102] Optionally, in the information processing device provided in Example 2 of the present application, the above-mentioned device also includes: a first determination unit, used to determine at least one device information involved in multiple preset user intentions before obtaining at least one device information involved in the user intention through the target model, and determine at least one first information corresponding to each device information in the at least one device information, wherein the device information includes at least: device status information, device function information, and the first information in the at least one first information is the smallest fine-grained parameter information; a second determination unit, used to determine the interface call information of each first information in the at least one first information, and obtain the interface call information corresponding to multiple preset user intentions, wherein the interface call information of each first information contains the interface identification information corresponding to each first information; a configuration unit, used to configure a first mapping relationship between multiple preset user intentions and the interface call information corresponding to the multiple preset user intentions; a second training unit, used to train the preset deep learning model using multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions, and the first mapping relationship.

[0103] Optionally, in the information processing device provided in Embodiment 2 of the present application, the above-mentioned encoding unit 703 includes: a first encoding subunit, used to determine the enumeration value of each device information in at least one device information, and encode the enumeration value of each device information to obtain first encoded information, wherein the device information includes at least: device status information, device function information; a second encoding subunit, used to determine the business scenario intended by the user, and encode the first encoded information according to multiple device information involved in the business scenario to obtain second encoded information.

[0104] Optionally, in the information processing device provided in Example 2 of the present application, the above-mentioned execution unit 704 includes: a processing subunit, used to input the encoded device information into the target model to obtain cause information of the communication network abnormality, wherein the cause information includes encoded information consisting of the abnormal state of the network device; a first determination subunit, used to determine a solution strategy based on the cause information, and determine the interface call information corresponding to the solution strategy; an execution subunit, used to execute the solution strategy based on the interface call information corresponding to the solution strategy.

[0105] Optionally, in the information processing device provided in Example 2 of the present application, the above-mentioned execution unit 704 also includes: a second determination sub-unit, which is used to determine the solution strategy based on the cause information, and before determining the interface call information corresponding to the solution strategy, determine the second mapping relationship between the historical cause information and the preset solution strategy, and determine the interface call information of the preset solution strategy; the third determination sub-unit includes: a fourth determination sub-unit, which is used to determine the solution strategy based on the second mapping relationship and the cause information, and the interface call information corresponding to the solution strategy; or a receiving sub-unit, which is used to receive the solution strategy of the cause information and the interface call information corresponding to the solution strategy when the cause information contains third coding information, wherein the third coding information is coding information not included in the historical cause information.

[0106] Optionally, in the information processing device provided in Example 2 of the present application, the above-mentioned device also includes: a second receiving unit, used to receive feedback from the target user on the solution strategy after executing the solution strategy; a first adjustment unit, used to adjust the second information based on the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: text information, user intention, and at least one device information involved in the user intention; a second adjustment unit, used to adjust the solution strategy based on the adjusted second information to obtain the adjusted solution strategy; a third adjustment unit, used to adjust the third information using the adjusted second information and the adjusted solution strategy, wherein the third information includes at least one of the following: historical cause information, a preset solution strategy, and a second mapping relationship.

[0107] The information processing device includes a processor and a memory. The above-mentioned first receiving unit 701, identification unit 702, encoding unit 703 and execution unit 704 are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0108] The processor contains a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the efficiency of solving user problems can be improved by adjusting kernel parameters.

[0109] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0110] Embodiment 3 of the present invention provides a computer-readable storage medium on which a program is stored. When the program is executed by a processor, an information processing method is implemented.

[0111] A fourth embodiment of the present invention provides a processor, which is used to run a program, wherein the information processing method is executed when the program is running.

[0112] like Figure 8 As shown, embodiment 5 of the present invention provides an electronic device, the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and the processor implements the following steps when executing the program: receiving text information input by a target user, wherein the text information includes question information raised by the target user regarding a communication network anomaly; using a large language model to process the text information and identify the user intent of the target user; obtaining at least one device information involved in the user intent through a target model, and encoding at least one device information to obtain encoded device information, wherein the device information is parameter information of a network device involved in the user intent; generating a solution strategy for the problem information based on the encoded device information through a target model, and executing the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task including determining the device information involved in the user intent, and encoding the device information to obtain the encoded device information, and the second task including determining the cause information of the communication network anomaly based on the encoded device information, and generating the solution strategy based on the cause information.

[0113] When the processor executes the program, the following steps are also implemented: before using the large language model to process the text information and identify the user intention of the target user, the above method also includes: determining historical text information and annotation information corresponding to the historical text information, and constructing a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information contains multiple preset user intentions; using the first training set to train the preset large language model to obtain the large language model.

[0114] When the processor executes the program, the following steps are also implemented: before obtaining at least one device information involved in the user intention through the target model, the above method also includes: determining at least one device information involved in multiple preset user intentions, and determining at least one first information corresponding to each device information in the at least one device information, wherein the device information at least includes: device status information, device function information, and the first information in the at least one first information is the smallest fine-grained parameter information; determining the interface call information of each first information in the at least one first information, and obtaining the interface call information corresponding to multiple preset user intentions, wherein the interface call information of each first information contains the interface identification information corresponding to each first information; configuring a first mapping relationship between multiple preset user intentions and the interface call information corresponding to the multiple preset user intentions; and using multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions, and the first mapping relationship to train the preset deep learning model.

[0115] When the processor executes the program, the following steps are also implemented: encoding at least one device information to obtain encoded device information, including: determining an enumeration value of each device information in the at least one device information, and encoding the enumeration value of each device information to obtain first encoded information, wherein the device information includes at least: device status information, device function information; determining the business scenario intended by the user, and encoding the first encoded information based on multiple device information involved in the business scenario to obtain second encoded information.

[0116] When the processor executes the program, the following steps are also implemented: generating a solution strategy for the problem information based on the encoded device information through the target model, and executing the solution strategy, including: inputting the encoded device information into the target model to obtain the cause information of the communication network abnormality, wherein the cause information includes the encoded information composed of the abnormal state of the network device; determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy; and executing the solution strategy based on the interface call information corresponding to the solution strategy.

[0117] When the processor executes the program, the following steps are also implemented: before determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy, the above method also includes: determining a second mapping relationship between the historical cause information and the preset solution strategy, and determining the interface call information of the preset solution strategy; determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy, including: determining the solution strategy based on the second mapping relationship and the cause information, and the interface call information corresponding to the solution strategy; or when the cause information contains third coding information, receiving the solution strategy of the cause information, and the interface call information corresponding to the solution strategy, wherein the third coding information is coding information not included in the historical cause information.

[0118] When the processor executes the program, the following steps are also implemented: After executing the solution strategy, the above method also includes: receiving feedback from the target user on the solution strategy; adjusting the second information based on the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: text information, user intention, and at least one device information related to the user intention; adjusting the solution strategy based on the adjusted second information to obtain an adjusted solution strategy; using the adjusted second information and the adjusted solution strategy to adjust the third information, wherein the third information includes at least one of the following: historical cause information, a preset solution strategy, and a second mapping relationship.

[0119] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0120] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with the following method steps: receiving text information input by a target user, wherein the text information includes question information raised by the target user regarding a communication network anomaly; using a large language model to process the text information and identify the user intent of the target user; obtaining at least one device information involved in the user intent through a target model, and encoding at least one device information to obtain encoded device information, wherein the device information is parameter information of a network device involved in the user intent; generating a solution strategy for the problem information based on the encoded device information through a target model, and executing the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task including determining the device information involved in the user intent, and encoding the device information to obtain the encoded device information, and the second task including determining the cause information of the communication network anomaly based on the encoded device information, and generating the solution strategy based on the cause information.

[0121] When executed on a data processing device, it is also suitable for executing a program that initializes the following method steps: before using a large language model to process text information and identify the user intention of a target user, the above method also includes: determining historical text information and annotation information corresponding to the historical text information, and constructing a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information contains multiple preset user intentions; using the first training set to train the preset large language model to obtain the large language model.

[0122] When executed on a data processing device, it is also suitable for executing a program that initializes the following method steps: before obtaining at least one device information related to a user intention through a target model, the above method also includes: determining at least one device information related to multiple preset user intentions, and determining at least one first information corresponding to each device information in the at least one device information, wherein the device information at least includes: device status information, device function information, and the first information in the at least one first information is the smallest fine-grained parameter information; determining the interface call information of each first information in the at least one first information, and obtaining the interface call information corresponding to multiple preset user intentions, wherein the interface call information of each first information contains the interface identification information corresponding to each first information; configuring a first mapping relationship between multiple preset user intentions and the interface call information corresponding to the multiple preset user intentions; and using multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions, and the first mapping relationship to train the preset deep learning model.

[0123] When executed on a data processing device, it is also suitable for executing an initialized program having the following method steps: encoding at least one device information to obtain encoded device information, including: determining an enumeration value of each device information in at least one device information, and encoding the enumeration value of each device information to obtain first encoded information, wherein the device information includes at least: device status information, device function information; determining a business scenario intended by the user, and encoding the first encoded information based on multiple device information involved in the business scenario to obtain second encoded information.

[0124] When executed on a data processing device, it is also suitable for executing an initialized program having the following method steps: generating a solution strategy for problem information based on the encoded device information through a target model, and executing the solution strategy, including: inputting the encoded device information into the target model to obtain the cause information of the communication network abnormality, wherein the cause information includes encoded information consisting of the abnormal state of the network device; determining the solution strategy based on the cause information, and determining the interface call information corresponding to the solution strategy; and executing the solution strategy based on the interface call information corresponding to the solution strategy.

[0125] When executed on a data processing device, it is also suitable for executing a program that is initialized with the following method steps: before determining a solution strategy based on cause information, and determining the interface call information corresponding to the solution strategy, the above method also includes: determining a second mapping relationship between historical cause information and a preset solution strategy, and determining the interface call information of the preset solution strategy; determining a solution strategy based on cause information, and determining the interface call information corresponding to the solution strategy, including: determining the solution strategy based on the second mapping relationship and the cause information, and the interface call information corresponding to the solution strategy; or when the cause information contains third coding information, receiving the solution strategy of the cause information, and the interface call information corresponding to the solution strategy, wherein the third coding information is coding information not included in the historical cause information.

[0126] When executed on a data processing device, it is also suitable for executing a program initialized with the following method steps: after executing the solution strategy, the above method also includes: receiving feedback from the target user on the solution strategy; adjusting the second information based on the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: text information, user intention, and at least one device information related to the user intention; adjusting the solution strategy based on the adjusted second information to obtain an adjusted solution strategy; adjusting the third information using the adjusted second information and the adjusted solution strategy, wherein the third information includes at least one of the following: historical cause information, a preset solution strategy, and a second mapping relationship.

[0127] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0128] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0129] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0131] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0132] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0133] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0134] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0135] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0136] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. An information processing method, characterized in that: include: Receiving a text message input by a target user, wherein the text message includes question information raised by the target user regarding a communication network anomaly; Using a large language model to process the text information to identify the user intention of the target user; Obtaining at least one piece of equipment information involved in the user intention through a target model, and encoding the at least one piece of equipment information to obtain encoded equipment information, wherein the equipment information is parameter information of a network device involved in the user intention; A solution strategy for the problem information is generated based on the encoded device information through the target model, and the solution strategy is executed, wherein the target model is a model trained based on a first task and a second task, the first task includes determining the device information involved in the user intention, and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the communication network abnormality based on the encoded device information, and generating the solution strategy based on the cause information.

2. The method according to claim 1, characterized in that Before using the large language model to process the text information and identify the user intention of the target user, the method further includes: Determine historical text information and annotation information corresponding to the historical text information, and construct a first training set based on the historical text information and the annotation information, wherein the historical text information is text information input by historical users within a preset time period, and the annotation information includes multiple preset user intentions; The first training set is used to train a preset large language model to obtain the large language model.

3. The method according to claim 2, characterized in that Before obtaining at least one device information related to the user intention through the target model, the method further includes: Determine at least one device information involved in the multiple preset user intentions, and determine at least one first information corresponding to each device information in the at least one device information, wherein the device information at least includes: device status information, device function information, and the first information in the at least one first information is the smallest fine-grained parameter information; Determine the interface calling information of each first information in the at least one first information, and obtain the interface calling information corresponding to the multiple preset user intentions, wherein the interface calling information of each first information includes the interface identification information corresponding to each first information; configuring a first mapping relationship between the plurality of preset user intentions and the interface call information corresponding to the plurality of preset user intentions; The preset deep learning model is trained using the multiple preset user intentions, the interface call information corresponding to the multiple preset user intentions, and the first mapping relationship.

4. The method according to claim 1, characterized in that: Encoding the at least one device information to obtain the encoded device information includes: Determine an enumeration value of each piece of device information in the at least one piece of device information, and encode the enumeration value of each piece of device information to obtain first coded information, wherein the device information at least includes: device status information and device function information; Determine the business scenario of the user's intention, and encode the first coded information according to multiple device information involved in the business scenario to obtain second coded information.

5. The method according to claim 1, characterized in that Generating a solution strategy for the problem information based on the encoded device information through the target model, and executing the solution strategy, including: Inputting the encoded device information into the target model to obtain the cause information of the abnormality of the communication network, wherein the cause information includes the encoded information composed of the abnormal state of the network device; Determine the solution strategy according to the cause information, and determine the interface call information corresponding to the solution strategy; The solution strategy is executed according to the interface call information corresponding to the solution strategy.

6. The method according to claim 5, characterized in that Before determining the solution strategy based on the cause information and determining the interface call information corresponding to the solution strategy, the method further includes: determining a second mapping relationship between the historical cause information and the preset solution strategy, and determining the interface call information of the preset solution strategy; Determine the solution strategy based on the cause information, and determine the interface call information corresponding to the solution strategy, including: determining the solution strategy based on the second mapping relationship and the cause information, and the interface call information corresponding to the solution strategy; or when the cause information contains third coding information, receive the solution strategy of the cause information, and the interface call information corresponding to the solution strategy, wherein the third coding information is the coding information not included in the historical cause information.

7. The method according to claim 6, characterized in that After executing the solution strategy, the method further includes: receiving feedback from the target user regarding the solution strategy; Adjusting the second information according to the feedback to obtain adjusted second information, wherein the second information includes at least one of the following: the text information, the user intention, and at least one device information involved in the user intention; Adjusting the solution strategy according to the adjusted second information to obtain an adjusted solution strategy; The third information is adjusted using the adjusted second information and the adjusted solution strategy, wherein the third information includes at least one of the following: the historical cause information, the preset solution strategy, and the second mapping relationship.

8. An information processing device, characterized in that: include: A first receiving unit, configured to receive text information input by a target user, wherein the text information includes question information raised by the target user regarding the abnormality of the communication network; an identification unit, configured to process the text information using a large language model to identify the user intention of the target user; an encoding unit, configured to obtain at least one piece of equipment information involved in the user intention through a target model, and encode the at least one piece of equipment information to obtain encoded equipment information, wherein the equipment information is parameter information of a network equipment involved in the user intention; An execution unit is used to generate a solution strategy for the problem information based on the encoded device information through the target model and execute the solution strategy, wherein the target model is a model trained based on a first task and a second task, the first task includes determining the device information involved in the user intention and encoding the device information to obtain the encoded device information, and the second task includes determining the cause information of the communication network abnormality based on the encoded device information and generating the solution strategy based on the cause information.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes stored computer instructions, wherein when the computer instructions are executed by a processor, the information processing method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the information processing method described in any one of claims 1 to 7.

Images