Method and device for constructing wireless network data dictionary

Abstract

Provided are a method and device for constructing a wireless network data dictionary. The method comprises: on the basis of a communication protocol specification text, determining a target entity, attribute information of the target entity, an associated entity, and attribute information of the associated entity by means of a large language model (S202); and constructing a wireless network data dictionary on the basis of the target entity, the associated entity, the attribute information of the target entity, and the attribute information of the associated entity (S204).

Description

A wireless network data dictionary construction method and device

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This disclosure is based on Chinese patent application CN202311663397.8, filed on December 1, 2023, entitled “A Method and Apparatus for Constructing a Wireless Network Data Dictionary”, and claims the priority of the patent application, and all the contents disclosed therein are incorporated into this disclosure by reference. Technical Field

[0003] The present disclosure relates to the field of communication data processing technology, and in particular to a method and device for constructing a wireless network data dictionary. Background Art

[0004] The Wireless Network Data Dictionary is designed to be a key tool for efficiently understanding B5G / 6G networks, thereby promoting data uniformity, comprehensibility, and usability. Based on the Wireless Network Data Dictionary, a knowledge graph of related data can be established, revealing the coupling between different data and the deep, in-line influence relationships between them.

[0005] If the wireless network data dictionary is constructed entirely by human expert knowledge, on the one hand, there will be defects such as high labor (expert) costs and low efficiency; on the other hand, due to the subjectivity and limitations of expert knowledge, there will be the possibility of errors and omissions in the data dictionary construction process.

[0006] Summary of the Invention

[0007] Embodiments of the present disclosure provide a method and apparatus for constructing a wireless network data dictionary.

[0008] According to one embodiment of the present disclosure, a method for constructing a wireless network data dictionary is provided, comprising: determining a target entity, attribute information of the target entity, associated entities, and attribute information of the associated entities through a large language model based on a communication protocol specification text; and constructing a wireless network data dictionary based on the target entity, the associated entities, the attribute information of the target entity, and the attribute information of the associated entities.

[0009] According to another embodiment of the present disclosure, a device for constructing a wireless network data dictionary is provided, including: a determination module, configured to determine a target entity, attribute information of the target entity, associated entities, and attribute information of the associated entities through a large language model based on a communication protocol specification text; and a construction module, configured to construct a wireless network data dictionary based on the target entity, the associated entities, the attribute information of the target entity, and the attribute information of the associated entities.

[0010] According to another embodiment of the present disclosure, a computer-readable storage medium is provided, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above method embodiments when running.

[0011] According to another embodiment of the present disclosure, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any one of the above method embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] 1 is a hardware structure block diagram of a computer terminal that executes a method for constructing a wireless network data dictionary according to an embodiment of the present disclosure;

[0013] FIG2 is a flowchart of a method for constructing a wireless network data dictionary according to an embodiment of the present disclosure;

[0014] 3 is a structural block diagram of a device for constructing a wireless network data dictionary according to an embodiment of the present disclosure;

[0015] FIG4 is a schematic diagram of a process for constructing a wireless network data dictionary according to an embodiment of the present disclosure;

[0016] FIG5 is a flowchart of a method for constructing a wireless network data dictionary based on a large language model according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0017] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings and in conjunction with embodiments.

[0018] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0019] The method embodiments provided in the embodiments of the present disclosure can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a computer terminal as an example, FIG1 is a hardware structure block diagram of a computer terminal of a wireless network data dictionary construction method according to an embodiment of the present disclosure. As shown in FIG1 , the computer terminal may include one or more (only one is shown in FIG1 ) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the above-mentioned computer terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It can be understood by those skilled in the art that the structure shown in FIG1 is only for illustration and does not limit the structure of the above-mentioned computer terminal. For example, the computer terminal may also include more or fewer components than those shown in FIG1 , or have a configuration different from that shown in FIG1 .

[0020] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the wireless network data dictionary construction method in the embodiment of the present disclosure. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implementing the above-mentioned method. The memory 104 may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely located relative to the processor 102, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0021] The transmission device 106 is used to receive or transmit data via a network. A specific example of the aforementioned network may include a wireless network provided by a computer terminal's communications provider. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other computer terminals via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0022] In this embodiment, a method for constructing a wireless network data dictionary running on the above-mentioned computer terminal is provided. FIG2 is a flow chart of the method for constructing a wireless network data dictionary according to an embodiment of the present disclosure. As shown in FIG2 , the flow chart includes the following steps:

[0023] Step S202, based on the communication protocol specification text, determine the target entity, the attribute information of the target entity, the associated entity and the attribute information of the associated entity through the large language model;

[0024] In this embodiment, the communication protocol specification text includes the Third Generation Partnership Project (3 rd Generation Partnership Project (3GPP) protocol specification text.

[0025] In one embodiment, the large language model can be of various types, such as OpenAI's GPT large model, Baidu's Wenxin Qianfan large model, etc.

[0026] In one embodiment, a large language model is used to intelligently construct a wireless network data dictionary, which provides a powerful tool for further constructing a complete knowledge table and dynamically reasoning a wireless network knowledge graph, and can quickly and deeply understand the intrinsic connections between the attributes of various functional elements in the wireless network system.

[0027] In this embodiment, the attribute information includes but is not limited to: the Chinese and English names of the entity, the entity meaning, the calculation formula, the statistical method, the category to which it belongs, and the dimension.

[0028] Before step S202 of this embodiment, the method further includes: constructing a core key performance indicator (KPI) list according to the target communication scenario, wherein the core KPI list includes multiple core KPI entities.

[0029] Before step S202 of this embodiment, the method further includes: creating a cloud application based on the large language model according to dictionary requirements; and receiving the communication protocol specification text and the core KPI list through the cloud application.

[0030] In one embodiment, the communication protocol specification text and the core KPI list are input into the large language model through an application programming interface (API) interface or a related plug-in. Before inputting the large language model, if the communication protocol text needs to be screened, a manual screening input method is selected, and the communication protocol specification text is first screened, and then the screened communication protocol specification text is input into the model; if no screening is required, it can be input through an automatic input tool.

[0031] In this embodiment, it can be well combined with artificial expert knowledge, and the communication protocol specification text can be preliminarily screened based on artificial expert knowledge, thereby enhancing the accuracy and comprehensiveness of the dictionary, making it more efficient and comprehensive.

[0032] In this embodiment, the target entity is determined through a large language model, including: selecting any of the core KPI entities as the core KPI target entity; based on the core KPI target entity, performing entity alignment on the entities in the communication protocol specification text to determine the target entity corresponding to the core KPI target entity.

[0033] In this embodiment, entity alignment is performed on entities in the communication protocol specification text, including: searching for candidate entities in the communication protocol specification text whose editing distance with the core KPI target entity is less than a preset threshold, and extracting nested entities based on linguistic rules to obtain a candidate entity set; and determining the target entity from the candidate entity set through linguistic rules and C-value algorithm.

[0034] In one embodiment, searching for candidate entities whose edit distance to the core KPI target entity is less than a preset threshold includes: using a Levenshtein Distance algorithm to search for a set of candidate entities whose edit distance to the core KPI string is less than a preset threshold in the communication protocol specification text.

[0035] In this embodiment, the attribute information of the target entity is determined by a large language model, including: searching for a text paragraph corresponding to the target entity in the communication protocol specification text by the large language model; and extracting the attribute information of the target entity from the text paragraph.

[0036] In this embodiment, after the target entity is determined from the candidate entity set, the method further includes: determining associated entities based on the target entity to perform associated entity expansion.

[0037] In this embodiment, determining the associated entity based on the target entity includes: searching the communication protocol specification text for an entity directly associated with the target entity, and determining it as the first associated entity; searching the communication protocol specification text for an entity directly associated with the first associated entity, and determining it as the second associated entity; searching the communication protocol specification text for an entity directly associated with the N-1th associated entity, and determining it as the Nth associated entity; wherein N is an integer greater than 2.

[0038] In this embodiment, in the communication protocol specification text, there is no entity directly associated with the Nth associated entity.

[0039] In this embodiment, the attribute information of the target entity is determined by a large language model, including: searching for a text paragraph corresponding to the associated entity in the communication protocol specification text by the large language model; and extracting the attribute information of the associated entity from the text paragraph.

[0040] Step S204: construct a wireless network data dictionary according to the target entity, the associated entity, the attribute information of the target entity, and the attribute information of the associated entity.

[0041] Through the above steps, the large language model intelligently queries, summarizes, analyzes, and concludes the relationships and attribute information between entities in the input communication protocol specification text. This process then constructs a wireless network data dictionary based on these relationships and attribute information. This entire process eliminates the need for human intervention, effectively reducing the cost of building the wireless network data dictionary and avoiding the subjectivity and limitations inherent in manual construction. This solves the inefficient and incomplete nature of building wireless network data dictionaries based on human expert knowledge in related technologies, improving the efficiency of wireless network data dictionary construction.

[0042] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present disclosure is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), including a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or computer terminal, etc.) to execute the methods described in each embodiment of the present disclosure.

[0043] This embodiment also provides a device for constructing a wireless network data dictionary, which is used to implement the above-mentioned embodiments and preferred implementations. Details already described will not be repeated here. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0044] FIG3 is a structural block diagram of a device for constructing a wireless network data dictionary according to an embodiment of the present disclosure. As shown in FIG3 , the device includes: a determining module 32 and a constructing module 34 .

[0045] A determination module 32 is configured to determine a target entity, attribute information of the target entity, associated entities, and attribute information of the associated entities through a large language model based on a communication protocol specification text;

[0046] The construction module 34 is configured to construct a wireless network data dictionary according to the target entity, the associated entity, the attribute information of the target entity, and the attribute information of the associated entity.

[0047] It should be noted that the above modules can be implemented through software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0048] Data plays a crucial role in shaping and driving the development of future B5G / 6G wireless networks, forming the cornerstone of network intelligence. Considering the data-native characteristics and application trends of future wireless networks, the design and evolution of B5G / 6G networks will inevitably rely closely on the application of big data technologies, requiring the empowerment of the three intelligent pillars of "data, knowledge, and models."

[0049] The core concept of wireless network data native is to collect diverse information distributed across various sources from a global perspective to achieve comprehensive classification, unified storage and management of different forms of structured and unstructured data. However, B5G / 6G wireless networks have the following characteristics:

[0050] First, the generation and distribution of various types of data occurs across different devices, platforms, and layers. Data formats and models lack consistent planning, making it difficult to ensure data uniformity and availability, while also increasing the complexity of data circulation and sharing. This multi-source, diverse, and heterogeneous nature results in a wide variety of data types and a high degree of abstraction, further complicating understanding the data's meaning.

[0051] 2. The complex coupling and correlation between data also increases the difficulty of data value mining.

[0052] To address the above-mentioned issues, it is necessary to standardize and summarize the relevant information of various data attributes on the basis of unified management and real-time sharing of data information, so as to efficiently clarify the internal correlation between various types of data, thereby deeply exploring the deep connections between these relationships and clearly and accurately presenting the key value information contained in the data and relationships.

[0053] The Wireless Network Data Dictionary encompasses a wide range of data types and formats, covering both structured and unstructured data. It provides clear definitions, attributes, relationships, and examples for each data type, eliminating the barriers to understanding the diverse range of data types and the complex relationships between them. The Wireless Network Data Dictionary promotes data uniformity, understanding, and usability.

[0054] The disclosed embodiments provide a wireless network data dictionary and a method for constructing the same. In this method, the Large Language Model (LLM), which is rapidly developing in current technology, is used to intelligently query, summarize, analyze, and conclude the attributes and association relationship information of each data field in the 3GPP network, starting from the basic 3GPP protocol specification text. This can not only reduce the cost of dictionary construction, but also combine artificial expert knowledge to further enhance the accuracy and comprehensiveness of the dictionary. Compared with existing wireless network data dictionaries, the method has more efficient and comprehensive performance.

[0055] FIG4 is a schematic diagram of a process for constructing a wireless network data dictionary according to an embodiment of the present disclosure. As shown in FIG4 , the process mainly includes:

[0056] Through the large language model program interface, the 3GPP protocol specification text and core KPI indicator list are imported into the large language model for training, and the large language model is used to complete the alignment of the protocol specification content and the target entity. The target entity is the entity representing the core KPI indicator. Then, the attribute information of the target entity is output, and the expansion of the entities associated with the target entity and the output of the entity association relationship and entity attribute information are completed. The associated entity is the entity that has a direct impact on the target entity. Among them, the entity association relationship refers to the association relationship between the target entity and the associated entity, and between each associated entity. The entity attribute information includes the attribute information of the target entity and the attribute information of the associated entity.

[0057] To facilitate understanding of the technical solutions provided by the present disclosure, embodiments of specific scenarios will be described in detail below.

[0058] FIG5 is a flow chart of a method for constructing a wireless network data dictionary based on a large language model according to an embodiment of the present disclosure. In this embodiment, the communication protocol specification text is a 3GPP protocol specification text. As shown in FIG5 , the method includes the following steps:

[0059] Step S501: Input the 3GPP protocol specification text content into a large language model for training and learning.

[0060] Since current large language pre-training models usually limit the length of a single input text, segmented input can be achieved through the existing large language pre-training model API or related plug-ins.

[0061] In one embodiment, when the 3GPP protocol specification text needs to be screened first, the 3GPP protocol specification text is input into the large language model through manual screening; when the 3GPP protocol specification text does not need to be screened, the 3GPP protocol specification text is input into the large language model through manual screening using an existing text automatic input tool.

[0062] For example, on an API interface or related plug-in, the filtered 3GPP protocol specification text is input into the large language model.

[0063] In one embodiment, the large language model includes but is not limited to: OpenAI's GPT large model, Baidu's Wenxin Qianfan large model, etc.

[0064] Taking Baidu's Wenxin Qianfan model as an example, step S501 includes the following sub-steps:

[0065] Step S5011, create an intelligent cloud application (such as the cloud application of the Wenxin Qianfan large model) according to actual needs, obtain AppID, API Key, Secret Key and interface access credential access_token to call the API and select the large language model interface.

[0066] That is, when the user enters the 3GPP protocol specification text into the cloud application on the terminal, the cloud application is called and some authentication of the cloud application is required. After the authentication is passed, the subsequent steps can be carried out. Among them, the authentication requires obtaining AppID, API Key, Secret Key and other information from the cloud application, as well as the interface access credential access_token.

[0067] In this embodiment, actual requirements include but are not limited to: wireless network data dictionary entries, wireless network data dictionary content requirements; large language model interfaces include but are not limited to: ERNIE-Bot, Llama-2.

[0068] Step S5012: calculate appropriate input prompts and input user text in segments.

[0069] For example: "Assuming you are an expert in the field of wireless communications,I'll give you a split document about the 3GPP protocol, and Please WAIT until I finish sending the whole context of the document.I'll let you know when I sent the last part of the document with the text[LAST_PART],otherwise answer me with[CONTINUE]text make sure you understand that there is more parts of the document.I'll let you know how many parts of the whole document.So, you have to wait until I've finished,meantime please DON'T generate any new response rather than [CONTINUE]".

[0070] Step S502: construct a core KPI indicator list, and input the core KPI indicator list into the large language model through the large language model program interface.

[0071] Step S503: Based on the core KPI target entity, entity alignment is performed on the entities in the communication protocol specification text.

[0072] In this embodiment, multiple core KPI indicator target entities are designed for different communication scenarios, such as uplink physical layer throughput, RRC connection success rate, etc., to form a core KPI list, and an entity in the core KPI indicator list is used as the initial target entity.

[0073] Because the customized core KPI string may not be completely consistent with the agreement text, entity alignment is required, that is, determining the representation string of the core KPI in the agreement text.

[0074] In one embodiment, in step S503 of this embodiment, performing entity alignment includes the following sub-steps:

[0075] Step S5031: Use the Levenshtein Distance algorithm to search the input agreement text for a set of candidate entities C whose edit distance to the core KPI string is less than a threshold d.

[0076] Step S5032: Set the following regular expression based on linguistic rules:

[0077] Noun+Noun

[0078] (Adj|Noun)+Noun

[0079] ((Adj|Noun)+((Adj|Noun)*(Noun Prep)?(Adj|Noun)*)Noun

[0080] Among them, Noun is a noun, Adj is an adjective, and Prep is a preposition.

[0081] Of the three linguistic rules mentioned above, the first one assumes that the target entity can only consist of nouns and their noun phrases; the second one assumes that the target entity can include adjectives and nouns; and the third one is the most relaxed, assuming that the target entity can contain nouns, adjectives, and prepositions. Based on the above regular expressions, the candidate entity set is further extracted to obtain more nested term phrases.

[0082] Step S5033: Entity alignment is performed using the C-value algorithm. After classifying the candidate entity set using linguistic rules, the C-value of each candidate entity is calculated using the following formula:

[0083] Among them, t is a candidate entity, |t| is the length of the candidate entity, f(t) is the word frequency of the candidate entity in the input text, C t is the set of nested term phrases containing t. Finally, the candidate entity with a C-value greater than the threshold α or the largest candidate entity after sorting by C-value is selected as the target entity.

[0084] Step S504: Use the large language model to find relevant paragraphs in the protocol specification text that are aligned with the target entity, extract attribute information of the target entity from the paragraphs, and output the attribute information of the target entity.

[0085] The attribute information of the target entity includes at least one of the following: the Chinese and English name of the entity, the meaning of the entity, the calculation formula, the statistical method, the category and the dimension.

[0086] In this embodiment, an input prompt may be designed, for example:

[0087] "Search for relevant paragraphs from the input document,summarize and output the following information about the target entity:entity name in both Chinese and English,entity definition,calculation formula,statistical method,relevant category,and unit".

[0088] Step S505: Searching for other entities directly associated with the target entity in the protocol text as associated entities through the large language model, and performing associated entity expansion.

[0089] In this embodiment, an input prompt may be designed, for example:

[0090] "Search for other entities directly associated with the target entity from the relevant paragraphs mentioned above and treat them as target entities".

[0091] Step S505 of this embodiment also includes: outputting the attribute information of each associated entity in sequence through the method in step S504.

[0092] Step S506: Output the target entity, associated entity, attribute information of the target entity, and attribute information of the associated entity in the core KPI list through the large language model, and construct a wireless network data dictionary based on the target entity, associated entity, attribute information of the target entity, and attribute information of the associated entity.

[0093] In one embodiment, the associated entities directly associated with the target entity of the core KPI indicator are used as the first associated entity set. Step S505 is repeated to output entities directly associated with the first associated entity set other than the core KPI as the second associated entity set. Step S505 is repeated to output entities directly associated with the second associated entity set other than the first associated entity set as the third associated entity set. Finally, the large language model outputs the entities in the core KPI list (including the target entity and associated entities) and the attribute information of the first, second, and third associated entity sets, which are then organized by the large language model to form a wireless network data dictionary.

[0094] In one embodiment, if the 3GPP protocol is updated, the core KPI indicators are further screened for the updated portion and steps S501 to S506 are repeated.

[0095] Through the above-mentioned embodiments of the present disclosure, a wireless network data dictionary is intelligently constructed through a large language model, which can reduce the cost of constructing the wireless network data dictionary. It can also be combined with artificial expert knowledge to further enhance the accuracy and comprehensiveness of the wireless network data dictionary, making it more efficient and comprehensive; it provides a reliable reference tool for further constructing a wireless network knowledge graph with complete knowledge representation and dynamic reasoning, and for quickly and deeply understanding the intrinsic relationship between the attributes of various functional elements in the wireless network system.

[0096] An embodiment of the present disclosure further provides a computer-readable storage medium, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above method embodiments when run.

[0097] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0098] An embodiment of the present disclosure further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0099] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0100] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail here.

[0101] Obviously, those skilled in the art should understand that the modules or steps of the present disclosure described above can be implemented using a general-purpose computing device, they can be concentrated on a single computing device, or distributed across a network composed of multiple computing devices, they can be implemented using program code executable by the computing device, and thus, they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be performed in a different order than herein, or they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. Thus, the present disclosure is not limited to any particular combination of hardware and software.

[0102] The foregoing description is merely a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Those skilled in the art will readily appreciate that various modifications and variations of the present disclosure are possible. Any modifications, equivalent substitutions, or improvements made within the principles of the present disclosure shall be included within the scope of protection of the present disclosure.

Claims

1. A method for constructing a wireless network data dictionary, comprising: Based on the communication protocol specification text, determine the target entity, the attribute information of the target entity, the associated entity and the attribute information of the associated entity through the large language model; A wireless network data dictionary is constructed according to the target entity, the associated entity, the attribute information of the target entity and the attribute information of the associated entity.

2. The method according to claim 1, wherein: The attribute information includes at least one of the following: Chinese and English names of the entity, entity meaning, calculation formula, statistical method, category and dimension.

3. The method according to claim 1, wherein: Before determining the attribute information of the target entity, the associated entity, and the attribute information of the associated entity by using the large language model, the method further includes: Creating a cloud application based on the large language model according to dictionary requirements; The communication protocol specification text is received through the cloud application.

4. The method according to claim 1, wherein Before determining the attribute information of the target entity, the associated entity, and the attribute information of the associated entity by using the large language model, the method further includes: A core KPI list is constructed according to a target communication scenario, wherein the core KPI list includes a plurality of core KPI entities.

5. The method according to claim 4, wherein: The target entity is determined through a large language model, including: Select any of the core KPI entities as the core KPI target entity; Based on the core KPI target entity, entity alignment is performed on entities in the communication protocol specification text to determine the target entity corresponding to the core KPI target entity.

6. The method according to claim 5, wherein: Performing entity alignment on entities in the communication protocol specification text includes: In the communication protocol specification text, searching for candidate entities whose edit distance with the core KPI target entity is less than a preset threshold, and extracting nested entities based on linguistic rules to obtain a candidate entity set; The target entity is determined from the candidate entity set by using linguistic rules and C-value algorithm.

7. The method according to claim 1, wherein: The attribute information of the target entity is determined through a large language model, including: Using the large language model, searching for a text paragraph corresponding to the target entity in the communication protocol specification text; Attribute information of the target entity is extracted from the text segment.

8. The method according to claim 1, wherein: After determining the target entity through the large language model, the method further includes: Determine associated entities based on the target entity to expand the associated entities.

9. The method according to claim 8, wherein: Determining an associated entity based on the target entity includes: In the communication protocol specification text, searching for an entity directly associated with the target entity and determining it as a first associated entity; In the communication protocol specification text, searching for entities other than the target entity that are directly associated with the first associated entity, and determining them as second associated entities; In the communication protocol specification text, entities directly associated with the N-1th associated entity except the N-2th associated entity are searched and determined as the Nth associated entity; wherein N is an integer greater than 2.

10. The method according to claim 9, wherein: In the communication protocol specification text, there is no entity directly associated with the Nth associated entity.

11. The method according to claim 1, wherein: The attribute information of the target entity is determined through a large language model, including: Using the large language model, searching for a text paragraph corresponding to the associated entity in the communication protocol specification text; Attribute information of the associated entity is extracted from the text segment.

12. A wireless network data dictionary construction device, comprising: A determination module is configured to determine a target entity, attribute information of the target entity, an associated entity, and attribute information of the associated entity through a large language model based on a communication protocol specification text; The construction module is configured to construct a wireless network data dictionary according to the target entity, the associated entity, the attribute information of the target entity and the attribute information of the associated entity.

13. A computer-readable storage medium, wherein: The computer-readable storage medium stores a computer program, wherein the computer program implements the steps of the method described in any one of claims 1 to 11 when executed by a processor.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method described in any one of claims 1 to 11 are implemented.

Images