Microbial entity relationship extraction method, system and equipment and storage medium

Abstract

The invention relates to a microbial entity relationship extraction method, system and device and a storage medium. The method comprises the following steps: acquiring a microorganism text; inputting the microorganism text into a first pre-training model, and determining entity information in the microorganism text; determining prompt word information based on the entity information; determining question information based on the prompt word information; carrying out splicing processing on the question information and the microorganism text information, and determining a question prompt word; and inputting the question prompt word into a second pre-training model, and determining the microbial entity relationship information. According to the method, the microbial entity relationship can be accurately determined from the non-structural text information, and the working efficiency and experience of a user are improved.

Description

Technical Field

[0001] The present application relates to the cross-technical field of bioinformatics and natural language, and specifically, to a method, system, device and storage medium for extracting microbial entity relationships. Background Art

[0002] In microbiology research, the acquisition and analysis of microbial entity information is crucial for microbial classification, identification and functional research. However, traditional microbial entity acquisition methods usually rely on experts to manually collect and organize entity information from published literature. This process is time-consuming and labor-intensive, resulting in the entity information of a large number of microbial groups still not being fully mined. With the continuous increase in biological literature, text mining methods based on natural language processing (NLP) technology have begun to be applied to the field of biological information extraction. NLP can effectively extract biological knowledge of interest from unstructured literature through automated entity recognition and relationship extraction technology. However, existing NLP systems still face many problems when dealing with microbial entity extraction, such as synonym processing, long-distance dependency of entity relationships, and incomplete recognition of domain-specific terms, which affect the accuracy and comprehensiveness of the extraction results.

[0003] In addition, there are also many deficiencies in the entity relationship extraction task in vertical fields. Generally speaking, the general paradigm of entity relationship extraction technology in vertical fields is: collection and preprocessing of unstructured text corpus, annotation of entity relationships (construction of training sets), model construction and training, and model optimization and iteration. Among them, "annotation of entity relationships (construction of training sets)" and "model construction and training" are the most time-consuming and important links. "Annotation of entity relationships (construction of training sets)" requires a large number of human resources with domain knowledge or an existing collection of ontology libraries. At present, the automatic annotation systems on the market are basically based on existing knowledge bases. For example, the patent "A method and system for automatic annotation of data sets for relationship extraction" automatically annotates data based on a preset domain corpus. However, in actual vertical field applications, it is difficult to obtain or even non-existent ontology library collections in specific fields. If a self-built domain database is used, there will be a series of problems such as personnel annotation coordination, annotation format specification, and annotation data collection, processing, storage, and use. In addition, in the process of "model construction and training", different model structures often have different advantages and disadvantages for different scenarios. Developers are required to do a lot of trial and error and debugging, which greatly increases the communication costs between vertical field personnel and developers.

[0004] Therefore, current microbial entity relationship extraction methods still need to be improved. Summary of the invention

[0005] The present application aims to solve one of the technical problems in the related art at least to a certain extent. To this end, one purpose of the present application is to provide a method for quickly and accurately extracting microbial entity relationships.

[0006] Specifically, this application proposes the following technical solutions:

[0007] In the first aspect, the present application proposes a method for extracting microbial entity relationships. According to an embodiment of the present application, the method includes: obtaining a microbial text; inputting the microbial text into a first pre-training model to determine entity information in the microbial text; determining prompt word information based on the entity information; determining question information based on the prompt word information; splicing the question information and the microbial text information to determine the question prompt word; inputting the question prompt word into a second pre-training model to determine the microbial entity relationship.

[0008] In some examples of the present application, the aforementioned method can accurately determine the relationship between microbial entities from unstructured text information, thereby improving user work efficiency and experience.

[0009] In the second aspect, the present application proposes a microbial entity relationship extraction system. According to an embodiment of the present application, the system includes: a text information acquisition module for acquiring microbial text; an entity information extraction module for inputting the microbial text into a first pre-trained model to determine entity information in the microbial text; a prompt word information determination module for determining prompt word information based on the entity information; a question information determination module for determining question information based on the prompt word information; a question prompt word generation module for splicing the question information and the microbial text information to determine the question prompt word; a microbial entity relationship determination module for inputting the question prompt word into a second pre-trained model to determine the microbial entity relationship.

[0010] In some examples of the present application, the aforementioned system improves the efficiency and accuracy of entity relationship extraction and reduces manual intervention by systematizing the modules of microbial text acquisition, entity information extraction, prompt word information determination, question information determination, question prompt word generation and microbial entity relationship determination, thereby ensuring that entity relationships are obtained more accurately and comprehensively, improving the information processing capabilities and extraction effects of the entire system, and being suitable for intelligent analysis and entity relationship extraction of large-scale microbial data.

[0011] In the third aspect of the present application, the present application proposes a computer program product. According to an embodiment of the present application, the computer program product includes: computer instructions; when part or all of the computer instructions are run on a computer, the microbial entity relationship extraction method described in the first aspect of the present application is executed.

[0012] In a fourth aspect of the present application, the present application proposes a computing device. According to an embodiment of the present application, the computing device comprises: a processor and a memory; the memory is used to store a computer program; the processor is used to execute the computer program to implement the microbial entity relationship extraction method as described in the first aspect of the present application.

[0013] In a fifth aspect of the present application, the present application proposes a computer-readable storage medium. According to an embodiment of the present application, the computer-readable storage medium stores computer instructions or programs, and when the computer instructions or programs are run on a computer, the microbial entity relationship extraction method described in the first aspect of the present application is executed.

[0014] The aforementioned computer program product, computing device, and computer-readable storage medium automatically execute the microbial entity relationship extraction method through computer instructions, thereby improving the efficiency and accuracy of entity relationship extraction, reducing manual intervention, and achieving a higher degree of automation. Secondly, the characteristics based on instructions make the above method highly consistent, stable, and reliable in different application scenarios.

[0015] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 A schematic diagram of a process for extracting microbial entity relationships provided for a specific embodiment of the present application;

[0018] Figure 2 A schematic diagram of microbial text entity extraction provided for a specific implementation of this application;

[0019] Figure 3 A schematic diagram of triple relationships obtained based on microbial entity extraction provided for a specific implementation of this application;

[0020] Figure 4 Schematic diagram of a microbial entity relationship extraction system provided for a specific embodiment of the present application;

[0021] Figure 5 Schematic diagram of a microbial entity relationship extraction platform provided for a specific embodiment of this application;

[0022] Figure 6 A schematic diagram of an electronic device provided for a specific implementation of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions 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 making creative work are within the scope of protection of this application.

[0024] 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 sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In the present application, 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 server 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. In the description of the present application, unless otherwise specified, "multiple" refers to two or more than two.

[0025] The current vertical domain entity relationship extraction technology faces the problems of time-consuming training set annotation, difficulty in obtaining domain ontology libraries, and complex model construction, which leads to high development costs and human resource requirements. This application proposes a microbial entity relationship extraction method, a microbial entity relationship extraction system, a computer program product, a computing device, and a computer-readable storage medium. They are described below:

[0026] Microbial Entity Relationship Extraction Method

[0027] On the one hand, this application proposes a method for extracting microbial entity relationships, referring to Figure 1 ,include:

[0028] S1, Get microbial text.

[0029] In some examples of the present application, the aforementioned microorganisms can be selected from fungi, bacteria or viruses.

[0030] In some examples of the present application, the aforementioned microbial texts include: microbial research literature, Bergey's Manual and other microbial non-structured text corpora.

[0031] S2, inputting the microbial text into a first pre-training model to determine entity information in the microbial text.

[0032] The first pre-trained model is used to identify and extract entity information in microbial text. In some examples of the present application, the first pre-trained model is obtained by training with data from a predetermined field based on a natural language processing tool. The predetermined field may optionally be a microbial field.

[0033] In some specific examples of the present application, the first pre-trained model is obtained by supervised training of the first training model using known microbial text entity information.

[0034] In some examples of the present application, the aforementioned natural language processing tool includes: at least one of SpaCy, Stanford NER and NLTK. Using the aforementioned toolkit, a first pre-trained model capable of identifying and extracting entity information in microbial text is constructed.

[0035] After obtaining the microbial text in step S1, the microbial text is further input into the first pre-training model to determine entity information in the aforementioned microbial text, such as at least one of the microbial name, classification, phenotype, ecology and environmental information. The microbial entity information includes microbial color, size, shape, carbon source and nitrogen source, etc.

[0036] For example, Figure 2 As shown, the microbial text is: Acinetobacter radioresistens are rods0.6-0.7×0.8-1.0μm.Isolated from cotton and from soils......

[0037] Its physical information is: Acinetobacter radioresistens; rods; 0.6-0.7×0.8-1.0μm; cotton; soils, etc.

[0038] In some examples of the present application, triple information ( Figure 3 , entity 1-relationship-entity 2) for storage.

[0039] It is understandable that part of the entity information to be extracted can be selected based on the first pre-trained model, or all the entity information in the microbial text can be extracted.

[0040] It is understandable that in the acquired microbial text, the microbial type may be one or more. Partial entity information of a certain microorganism may be extracted, or all entity information of a certain microorganism may be extracted; partial entity information of all microorganisms may be extracted, or all entity information of all microorganisms may be extracted.

[0041] Exemplarily, the first pre-trained model (Bert+BiLSTM+CRF) is selected to extract all entity information of all microorganisms in the input microorganism text.

[0042] S3, based on the aforementioned entity information, determine prompt word information.

[0043] After the microorganism entity information is obtained in step S2, prompt word information is further determined, and the so-called prompt word information is a classification set of relationships between different entities.

[0044] To facilitate understanding of the aforementioned prompt information, a specific example is provided below for detailed explanation:

[0045] 1) Get microbial text

[0046] "Escherichia coliis a common bacterium found in the intestines of humans and animals. It can cause various infections, including urinary tractinfections, and plays a crucial role in the digestion process."

[0047] 2) Determine microbial entity information

[0048] Microorganism name: Escherichia coli

[0049] Habitat: Intestines of humans and animals

[0050] Disease association: Urinary tract infections

[0051] Function: Plays a role in digestion

[0052] 3) Select a classification set based on entity information

[0053] Category collection 1: habitat

[0054] Classification set 2: disease

[0055] Category collection 3: function

[0056] S4, determining question information based on the prompt word information;

[0057] Based on the relationship classification set information obtained in step S3, the question information is determined, that is, the question is asked. For example, "What is the natural habitat of Escherichia coli?", or "In which environments does Lactobacillus thrive?".

[0058] S5, concatenating the question information and the microorganism text information to determine a question prompt word.

[0059] In some examples of the present application, the aforementioned splicing process is implemented by a TextChunks method.

[0060] The aforementioned term "TextChunks method" refers to dividing a long text into smaller fragments (chunks) in text processing for the convenience of analysis, processing or model training. In this application, the microbial text information and the question information are spliced ​​and processed by the "TextChunks method" to generate structured prompt word information. This process helps to integrate and format the extracted microbial entity information and related questions, thereby generating clear and unambiguous query content.

[0061] Specifically, the "TextChunks method" can divide long texts or complex information into manageable small segments, so that key information can be extracted and utilized more effectively during subsequent data analysis and processing. This block processing method can improve the pertinence and accuracy of questions, and ensure that the generated questions are well correlated with the extracted microbial entity information. For example, after splicing, specific questions such as "What is the natural habitat of Escherichia coli?" can be formed to facilitate subsequent information retrieval or knowledge mining.

[0062] In some examples of the present application, after the aforementioned splicing process, the method further includes: performing a format check on the aforementioned question prompt words. The aforementioned format check includes: punctuation marks, whether the sentences are coherent, whether the entity relationship expression is correct, etc.

[0063] S6, inputting the question prompt words into the second pre-training model to determine the microbial entity relationship.

[0064] In some examples of the present application, the aforementioned second pre-trained model is selected from at least one of ChatGLM3, GPT-3, BERT, Text-to-Text Transfer Transformer, Claude, and LLaMA.

[0065] In some examples of the present application, the second pre-trained model is obtained by supervised training of the second training model using known microbial entity relationships.

[0066] In other examples of the present application, the second pre-trained model is further fine-tuned using text information in a predetermined technical field so that its answers are more in line with the expression of the predetermined technical field.

[0067] Based on the above method, comprehensive, accurate and easy-to-use microbial entity relationships can be obtained, which can be directly applied in multiple fields such as biology, ecology, medicine and related fields.

[0068] Microbial Entity Relationship Extraction System

[0069] On the other hand, this application proposes a microbial entity relationship extraction system, referring to Figure 4 The system comprises:

[0070] The text information acquisition module 100, the entity information extraction module 200, the prompt word information determination module 300, the question information determination module 400, the question prompt word generation module 500 and the microorganism entity relationship determination module 600.

[0071] 100 modules for obtaining microbiological texts;

[0072] Module 200 is used to input the microbial text into a first pre-trained model to determine entity information in the microbial text;

[0073] In some examples of the present application, the first pre-trained model is obtained by training with data from a predetermined field based on a natural language processing tool, wherein the natural language processing tool includes at least one of SpaCy, Stanford NER, and NLTK.

[0074] In some examples of the present application, the first pre-trained model is obtained by supervised training of the first training model using known microbial text entity information.

[0075] In some examples of the present application, the entity information includes: at least one of microorganism name, classification, phenotype, ecology and environmental information.

[0076] Module 300 is used to determine prompt word information based on the entity information;

[0077] In some examples of the present application, the prompt word information is a classified set of relationships between different entities.

[0078] Module 400 is used to determine question information based on the prompt word information;

[0079] Module 500 is used to splice the question information and the microorganism text information to determine the question prompt word;

[0080] In some examples of the present application, the splicing process is implemented by a TextChunks method.

[0081] In some examples of the present application, this module further includes: a format checking submodule, which is used to perform format checking on the question prompt words.

[0082] Module 600 is used to input the question prompt words into the second pre-training model to determine the microbial entity relationship.

[0083] In some examples of the present application, the second pre-trained model is selected from at least one of ChatGLM3, GPT-3, BERT, Text-to-Text Transfer Transformer, Claude, and LLaMA.

[0084] In some examples of the present application, this module further includes: a model fine-tuning submodule, used to fine-tune the second pre-trained model.

[0085] In some examples of the present application, the second pre-trained model is obtained by supervised training of the second training model using known microbial entity relationships.

[0086] In some examples of the present application, the aforementioned system may further include: a data storage module, used to store the microorganism entity relationship generated by the 600 module.

[0087] Those skilled in the art will appreciate that the aforementioned modules may be connected via hardware or via a network.

[0088] In some examples of the present application, the above system can also be implemented through the following network architecture setting, including a model training layer, a model kernel layer, a data storage layer, an access layer, an optional feature training interface, a configurable reasoning interface and a microbial metabolic vocabulary interface.

[0089] The various connections in this system are as follows: the model training layer and the model kernel layer are directly connected through the data interface. The model training layer is mainly responsible for the parameter configuration of the automatic model training, and completes the human-computer interaction through the optional feature training interface. Through the optional feature training interface, the user can specify the training model and training data, and set hyperparameters such as training time, rounds, and learning rate. In the initialization stage, the triple extraction data for model initialization is completed by the open source LLM. The initial extracted data is of low quality and poor accuracy. Therefore, the model training layer uses the B / S architecture through the interface to build a multi-person simultaneous online review platform (the actual application effect of the platform is as follows Figure 5 As shown in Figure 2.1, the model training layer is the core module of this system. Its role is to complete the interaction with users, realize the construction and training of entity relationship models, and save the trained models to the model kernel layer. The main function of the model kernel layer is to complete model storage through a unified mode.

[0090] The data storage layer stores the model and manually annotated data into the database through different status codes. The data storage layer includes two types of databases, Mysql and Neo4J, so that data can be called according to different application types. The data storage layer is directly connected to the model training layer, the model kernel layer and the access layer through interfaces. In the data access layer, the entity relationship extracted by the model is recorded as a "0" status code, and the entity relationship that has been manually reviewed is recorded as a "1" status code. Different status codes are used to distinguish data quality. In the data annotation stage, the MicroLLM system will obtain data with a status code of "0" from the data storage layer. After the manual review is completed, the data status code will be updated to "1". In the data training stage, when the manually reviewed data reaches the training set capacity preset by the administrator, the system will automatically obtain data from the data storage layer and start model training, and update the entity relationship extraction model to obtain a more accurate output structure in the model reasoning stage. In addition, the access layer is also directly connected to the data storage layer. The data content is displayed through the feasibility set by the administrator. For example, the feasibility of manually reviewed data is "1", and the credibility of the entity relationship proposed by the model is between [0,1]. Therefore, the data storage layer is the core module of the MicroLLM system construction, and the core purpose of the MicroLLM system is to efficiently and accurately build an ontology database in the vertical field.

[0091] The access layer mainly uses and displays database data through interfaces. In this system, the access layer is directly used as a microservice for external calls, and realizes the prediction of microbial phenotypic traits based on the extracted entity relationships. For applications in other vertical fields, the access layer can quickly use the extracted entity relationships to complete tasks in specific fields.

[0092] The configurable reasoning interface mainly completes the extraction of specific features by users, as shown in the figure (the actual application effect of the platform is as follows Figure 5 The main interaction process is as follows: ① The user selects the feature table to be extracted. ② The user selects the entity features to be extracted. ③ The user selects a suitable entity extraction model. ④ Submit to complete the entity extraction. ⑤ The user selects microorganisms and feature entities, and after submission, the second pre-trained model outputs the relationship between the microorganisms and feature entities. In addition, users can perform one-click extraction through the default entity relationship extraction model.

[0093] The microbial metabolism word library interface is mainly responsible for external calls. The microbial metabolism word library interface can be used to quickly obtain the microbial feature ontology library. In addition to training the model within this platform, the model can also be trained outside the system, providing vertical industry practitioners with a more flexible way of use. In addition, using the microbial metabolism word library interface, this system has built a word segmentation tool based on microbial phenotypic traits.

[0094] The optional feature training interface incorporates model training into the system. It builds and trains the model in a configurable way, eliminating the process of model construction and configuration, greatly saving training time and improving the efficiency of phenotypic feature extraction.

[0095] The aforementioned system uses a mechanism based on the kernel atomic combination call algorithm to efficiently and accurately identify, analyze and extract key information in sentences from unstructured text information describing prokaryotes. In this process, the system can identify microbial entities, related attributes, and the relationships between them, and integrate this information into the form of triples. Finally, a high-dimensional microbial characteristic matrix is ​​obtained, which captures a variety of key features of microorganisms. It supports multiple people to review online at the same time. It supports one-click automatic completion of annotation data format specification, collection, processing and storage. It provides a set of general entity relationship extraction tasks for various vertical fields.

[0096] Another obvious advantage of the aforementioned system is that it can be started with zero samples. In the initial stage, users do not need to perform data annotation and model training. They can complete the initial model screening through the prompt word engineering of the large language model, and then use manual review to perform a small amount of extraction correction. As the number of corrections increases, the system's built-in named entity recognition (NER) and large language model can learn the content that users are interested in, thereby greatly reducing manual participation, and finally realizing a microbial phenotypic information extraction system based on large language models and NER.

[0097] It should be understood that the system embodiment and the method embodiment may correspond to each other, and similar descriptions may refer to the method embodiment. To avoid repetition, they will not be described here. Specifically, Figure 4The system shown can execute the above-mentioned embodiment of the microbial phenotype information extraction method, and the operations and / or functions performed by each module in the system correspond to the method embodiment, which will not be described here for brevity.

[0098] The system of the embodiment of the present application is described above from the perspective of the functional module in conjunction with the accompanying drawings. It should be understood that the functional module can be implemented in hardware form, can be implemented by instructions in software form, and can also be implemented by a combination of hardware and software modules. Specifically, the steps of the method embodiment in the embodiment of the present application can be completed by the hardware integrated logic circuit and / or software form instructions in the processor, and the steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to perform, or a combination of hardware and software modules in the decoding processor to perform. Optionally, the software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory, and completes the steps in the above method embodiment in conjunction with its hardware.

[0099] Computer program product, computing device, and computer readable storage medium

[0100] In another aspect of the present application, the present application proposes a computer program product, a computing device, and a computer-readable storage medium. Based on the aforementioned computer program product, computing device, or computer-readable storage medium, the aforementioned microbial entity relationship extraction method is executed.

[0101] Descriptions of computer program products, computing devices, or computer-readable storage media may refer to each other. This article takes electronic devices as an example for detailed description. The so-called electronic devices are intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Computing devices can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0102] like Figure 6As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory) 502 or a computer program loaded from a storage unit 508 to a RAM (Random Access Memory) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An I / O (Input / Output) interface 505 is also connected to the bus 504.

[0103] A number of components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0104] The computing unit 501 can be various general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, CPU (Central Processing Unit), GPU (Graphic Processing Units), various dedicated AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSP (Digital Signal Processor), and any appropriate processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the microbial entity relationship extraction method. For example, in some embodiments, the microbial entity relationship extraction method can be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the method described above can be executed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute the aforementioned microorganism entity relationship extraction method in any other appropriate manner (eg, by means of firmware).

[0105] In the present application, the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the aforementioned program may be printed, because the aforementioned program may be obtained electronically, for example, by optically scanning the paper or other medium, and then editing, interpreting or processing in other suitable ways as necessary, and then storing it in a computer memory. The various computer-readable storage media described in the present invention may represent one or more devices and / or other machine-readable storage media for storing information. The term "machine-readable storage medium" may include, but is not limited to, wireless channels and various other media capable of storing, containing and / or carrying instructions and / or data.

[0106] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0107] A person of ordinary skill in the art may understand that all or part of the steps in the method of the above-mentioned embodiment may be completed by instructing related hardware through a program, and the aforementioned program may be stored in a computer-readable storage medium, which, when executed, includes one of the steps of the method embodiment or a combination thereof.

[0108] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0109] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0110] It should be noted that the features and technical effects described in this article for different aspects can be used as reference for each other and will not be repeated here.

[0111] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0112] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in the field can change, modify, replace and modify the above embodiments within the scope of the present application without departing from the principles and purpose of the present application.

Claims

1. A method for extracting microbial entity relationships, characterized in that: include: Get microbial text; Inputting the microbial text into a first pre-trained model to determine entity information in the microbial text; Based on the entity information, determining prompt word information; Determine question information based on the prompt word information; The question information and the microorganism text information are spliced ​​together to determine a question prompt word; The question prompt words are input into the second pre-training model to determine the microbial entity relationship.

2. The method according to claim 1, characterized in that: The first pre-trained model is obtained by training with data from a predetermined field based on a natural language processing tool; Optionally, the natural language processing tool comprises: at least one of SpaCy, Stanford NER and NLTK; Preferably, the first pre-trained model is obtained by using known microbial text entity information to conduct supervised training on the first training model.

3. The method according to claim 2, characterized in that The entity information includes: at least one of the microorganism name, classification, phenotype, ecology and environmental information.

4. The method according to claim 1, characterized in that The prompt word information is a classification set of relationships between different entities.

5. The method according to claim 1, characterized in that The second pre-trained model is selected from at least one of ChatGLM3, GPT-3, BERT, Text-to-Text Transfer Transformer, Claude, and LLaMA; Optionally, the second pre-trained model is further fine-tuned.

6. The method according to claim 1, characterized in that The splicing process is implemented by the TextChunks method; Optionally, the method further includes: performing format checking on the question prompt words.

7. The method according to any one of claims 1 to 6, characterized in that: The second pre-training model is obtained by using known microbial entity relationships to conduct supervised training on the second training model.

8. A microbial entity relationship extraction system, characterized in that: include: A text information acquisition module, used to acquire microbial text; An entity information extraction module, used for inputting the microbial text into a first pre-training model to determine entity information in the microbial text; A prompt word information determination module, used to determine the prompt word information based on the entity information; A question information determination module, used to determine question information based on the prompt word information; A question prompt word generation module is used to splice the question information and the microorganism text information to determine the question prompt word; The microorganism entity relationship determination module is used to input the question prompt words into the second pre-training model to determine the microorganism entity relationship.

9. A computing device, characterized in that include: Processor and memory; The memory is used to store computer programs; The processor is used to execute the computer program to implement the microbial entity relationship extraction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions or programs. When the computer instructions or programs are executed on a computer, the method for extracting microbial entity relationships according to any one of claims 1 to 7 is executed.

Images