Medicine database model construction method and device, electronic equipment and storage medium

Abstract

The embodiment of the application provides a kind of drug database model construction method and device, electronic equipment and storage medium, belong to artificial intelligence technical field.The method comprises: obtaining drug knowledge text;The field extraction is carried out to drug knowledge text, and the drug attribute field of at least two drug attribute dimensions is obtained;Obtain database simulation model;Wherein, database simulation model includes at least two simulation coordinate axes, and simulation coordinate axis includes multiple simulation intervals;According to the simulation coordinate axis of drug attribute dimension, target coordinate axis is obtained, and according to the simulation interval of target coordinate axis of drug attribute field of drug attribute dimension, target interval is obtained, drug attribute field is stored to target interval, and target database model is obtained.The embodiment of the application can improve the reuse rate of drug knowledge, to realize more efficient, accurate knowledge transmission.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method and apparatus for constructing a drug database model, an electronic device, and a storage medium. Background Technology

[0002] In the field of traditional Chinese medicine (TCM) process development, pharmaceutical company employee training primarily relies on learning pre-established standardized operating procedures (SOPs) and apprenticeship systems. This leads to fragmented TCM-related process knowledge and practical experience. Different dimensions of technical data are stored in separate systems; for example, process parameters are stored in databases, quality standards in document systems, and equipment maintenance knowledge in enterprise management platforms. Employees must sift through a vast amount of knowledge files across platforms to obtain the content that meets their needs, or passively wait for mentorship in practice. This forces R&D personnel to spend a significant amount of time acquiring knowledge data, resulting in low knowledge reuse rates. Therefore, improving the reuse rate of pharmaceutical knowledge has become an urgent technical problem to be solved. Summary of the Invention

[0003] The main objective of this application is to propose a method and apparatus for constructing a drug database model, an electronic device, and a storage medium, which aims to improve the reusability of drug knowledge and achieve more efficient and accurate knowledge transfer.

[0004] To achieve the above objectives, a first aspect of this application proposes a method for constructing a drug database model, the method comprising: Obtain drug knowledge text; Field extraction is performed on the drug knowledge text to obtain drug attribute fields with at least two drug attribute dimensions; Obtain a database simulation model; wherein the database simulation model includes at least two simulation coordinate axes, and the simulation coordinate axes include multiple simulation intervals; The simulation coordinate axis is obtained based on the drug attribute dimension to obtain the target coordinate axis. The simulation interval of the target coordinate axis is obtained based on the drug attribute field of the drug attribute dimension to obtain the target interval. The drug attribute field is stored in the target interval to obtain the target database model.

[0005] In some embodiments, obtaining the simulation coordinate axis based on the drug attribute dimension to obtain the target coordinate axis, and obtaining the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension to obtain the target interval, includes: Obtain the coordinate axis dimension attribute of all the simulated coordinate axes; For each of the drug attribute dimensions, the simulation coordinate axis that has the same coordinate axis attribute dimension as the drug attribute dimension is used as the reference coordinate axis. Use at least two of the reference coordinate axes as the target coordinate axes; The simulation interval of the target coordinate axis is obtained based on the drug attribute field of the drug attribute dimension, thus obtaining the target interval.

[0006] In some embodiments, the drug attribute dimension includes a process stage attribute dimension, an industry stage attribute dimension, and a drug value attribute dimension. Obtaining the simulation interval of the target coordinate axis based on the drug attribute fields of the drug attribute dimension to obtain the target interval includes: Process step identification is performed on the drug attribute fields whose drug attribute dimension is the process stage attribute dimension to obtain the process step type; The industry stage node is identified for the drug attribute field whose drug attribute dimension is the industry stage attribute dimension, and the industry node type is obtained. Value category identification is performed on the drug attribute fields whose drug attribute dimension is the drug value attribute dimension to obtain the drug value type; Obtain the preset attribute types for all simulation intervals of the target coordinate axis; Based on the preset attribute type, the simulation interval corresponding to the process step type, the industry node type, and the drug value type is determined as the target interval.

[0007] In some embodiments, after storing the drug attribute fields into the target interval to obtain the target database model, the method further includes: Obtain the object profile of the target object; For each of the drug attribute dimensions, knowledge reading behavior features matching the drug attribute dimension are extracted from the object profile; The job type is extracted from the object profile, and the job type is queried according to the preset weight mapping relationship to obtain feature weight data matching each of the drug attribute dimensions. The target behavior feature vector is obtained by concatenating the feature weight data and the knowledge reading behavior features. Based on the target behavior feature vector, data is filtered in the target database model to obtain target drug knowledge; The target drug knowledge is pushed to the target object.

[0008] In some embodiments, the step of filtering the target database model based on the target behavior feature vector to obtain target drug knowledge includes: For each interval of the target database model, the drug attribute field is used to obtain attribute semantic features by semantic feature encoding of the drug attribute field; Based on the drug attribute dimension, the semantic features of the attribute are vectorized to obtain the field feature vector; Calculate the similarity score between the field feature vector and the target behavior feature vector; Based on the similarity score, target attribute fields are selected from the drug attribute fields, and the drug knowledge text corresponding to the target attribute fields is determined as the target drug knowledge.

[0009] In some embodiments, after extracting job types from the object profile and querying the job types according to a preset weight mapping relationship to obtain feature weight data matching each of the drug attribute dimensions, the method further includes: Obtain the behavior timestamp of the knowledge reading behavior characteristics, and obtain the current timestamp; The decay index is determined based on the duration between the behavior timestamp and the current timestamp; Based on the attenuation index and the preset attenuation rate, feature attenuation is calculated on the feature weight data to obtain reference weight data; The feature weight data is updated based on the reference weight data.

[0010] In some embodiments, after storing the drug attribute fields into the target interval to obtain the target database model, the method further includes: Obtain candidate knowledge texts and obtain the first creation timestamp of the candidate knowledge texts; Field extraction is performed on the candidate knowledge text to obtain at least two candidate attribute fields for the drug attribute dimensions; Candidate coordinate axes are determined from the simulation coordinate axes based on the drug attribute dimension, and the simulation interval of the candidate coordinate axes is obtained based on the drug attribute field of the drug attribute dimension, thus obtaining the candidate interval; Obtain the drug attribute field in the candidate interval from the target database model, and obtain the second creation timestamp of the drug attribute field; If the second creation timestamp is earlier than the first creation timestamp, the candidate attribute field replaces the drug attribute field.

[0011] To achieve the above objectives, a second aspect of this application provides a drug database model construction apparatus, the apparatus comprising: The text acquisition module is used to acquire drug knowledge text. The field extraction module is used to extract fields from the drug knowledge text to obtain drug attribute fields with at least two drug attribute dimensions. A model acquisition module is used to acquire a database simulation model; wherein, the database simulation model includes at least two simulation coordinate axes, and the simulation coordinate axes include multiple simulation intervals; The drug database model construction module is used to obtain the simulation coordinate axis based on the drug attribute dimension, obtain the target coordinate axis, obtain the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension, obtain the target interval, store the drug attribute field in the target interval, and obtain the target database model.

[0012] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method described in the first aspect.

[0013] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0014] The method, apparatus, electronic device, and storage medium for constructing a drug database model proposed in this application acquire drug knowledge text and identify drug attribute fields corresponding to drug attribute dimensions. Then, based on the simulation interval of the simulation coordinate axis, the drug attribute fields are combined with multiple drug attribute dimensions to determine the target interval from the simulation interval, and the drug attribute fields are stored in the target interval, thereby constructing a target database model. This achieves accurate classification and storage of drug knowledge, avoiding the fragmentation problem of traditional knowledge storage. Compared with existing technologies, the target database model constructed by the method of this application for drug knowledge acquisition solves the problem of low efficiency in traditional knowledge transfer, where the target object needs to select the required knowledge from fragmented information. This enables more efficient and accurate knowledge sharing and dissemination, thereby helping to improve the learning efficiency of pharmaceutical company employees. Attached Figure Description

[0015] Figure 1 This is a flowchart of the drug database model construction method provided in the embodiments of this application; Figure 2 This is a schematic diagram of an optional database simulation model provided in an embodiment of this application; Figure 3 yes Figure 1 The flowchart of step S104 in the process; Figure 4yes Figure 3 The flowchart of step S304 in the process; Figure 5 This is another flowchart of the drug database model construction method provided in the embodiments of this application; Figure 6 This is another flowchart of the drug database model construction method provided in the embodiments of this application; Figure 7 yes Figure 6 The flowchart of step S605 in the process; Figure 8 This is a schematic diagram of the structure of the drug database model construction device provided in the embodiments of this application; Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0017] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0019] First, let's analyze some of the terms used in this application: Artificial intelligence (AI) is a new branch of computer science that studies, develops, and applies theories, methods, technologies, and systems to simulate, extend, and expand human intelligence. It aims to understand the essence of intelligence and produce intelligent machines that can react in a way similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thought. Furthermore, AI utilizes digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceiving the environment, acquiring knowledge, and using that knowledge to achieve optimal results.

[0020] Natural Language Processing (NLP): NLP uses computers to process, understand, and utilize human language (such as Chinese and English). It is a branch of artificial intelligence and an interdisciplinary field of computer science and linguistics, often referred to as computational linguistics. NLP includes syntactic analysis, semantic analysis, and discourse understanding. It is commonly used in machine translation, handwritten and printed character recognition, speech recognition and text-to-speech conversion, intent recognition, information extraction and filtering, text classification and clustering, sentiment analysis, and opinion mining. It involves data mining, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research, and linguistic research related to language computation.

[0021] Information extraction is a text processing technique that extracts factual information such as entities, relationships, and events from natural language text and outputs it as structured data. Information extraction is a technique for extracting specific information from text data. Text data is composed of specific units, such as sentences, paragraphs, and chapters. Text information is composed of smaller, specific units, such as characters, words, phrases, sentences, paragraphs, or combinations of these units. Extracting noun phrases, names of people, and place names from text data is an example of text information extraction. Of course, text information extraction techniques can extract information of various types.

[0022] In the field of traditional Chinese medicine (TCM) process development, pharmaceutical company employee training primarily relies on learning pre-established standardized operating procedures (SOPs) and apprenticeship systems. This leads to fragmented TCM-related process knowledge and practical experience. Different dimensions of technical data are stored in separate systems; for example, process parameters are stored in databases, quality standards in document systems, and equipment maintenance knowledge in enterprise management platforms. Employees must sift through a vast amount of knowledge files across platforms to obtain the content that meets their needs, or passively wait for mentorship in practice. This forces R&D personnel to spend a significant amount of time acquiring knowledge data, resulting in low knowledge reuse rates. Therefore, improving the reuse rate of pharmaceutical knowledge has become an urgent technical problem to be solved.

[0023] Based on this, embodiments of this application provide a method and apparatus for constructing a drug database model, an electronic device and a storage medium, aiming to improve the reuse rate of drug knowledge and achieve more efficient and accurate knowledge transfer.

[0024] The drug database model construction method, apparatus, electronic device, and storage medium provided in this application are specifically described through the following embodiments. First, the drug database model construction method in this application embodiment is described.

[0025] The drug database model construction method provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the drug database model construction method, but is not limited to the above forms.

[0026] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0027] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0028] Figure 1 This is an optional flowchart of the drug database model construction method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S104.

[0029] Step S101: Obtain the drug knowledge text.

[0030] Step S102: Extract fields from the drug knowledge text to obtain drug attribute fields of at least two drug attribute dimensions.

[0031] Step S103: Obtain the database simulation model; wherein the database simulation model includes at least two simulation coordinate axes, and the simulation coordinate axes include multiple simulation intervals.

[0032] Step S104: Obtain the simulation coordinate axis based on the drug attribute dimension to obtain the target coordinate axis, and obtain the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension to obtain the target interval. Store the drug attribute field in the target interval to obtain the target database model.

[0033] Steps S101 to S104 of this embodiment involve acquiring drug knowledge text and identifying drug attribute fields corresponding to drug attribute dimensions. Then, based on the simulation interval of the simulation coordinate axis, the drug attribute fields are combined with multiple drug attribute dimensions to determine the target interval from the simulation interval, and the drug attribute fields are stored in the target interval, thereby constructing a target database model. This achieves accurate classification and storage of drug knowledge, avoiding the fragmentation problem of traditional knowledge storage. Compared with existing technologies, the target database model constructed by the method of this embodiment for drug knowledge acquisition solves the problem of low efficiency in traditional knowledge transfer, where the target object needs to select the required knowledge from fragmented information. This enables more efficient and accurate knowledge sharing and dissemination, thereby helping to improve the learning efficiency of pharmaceutical company employees.

[0034] In step S101 of some embodiments, pharmaceutical knowledge text refers to various types of information and data about pharmaceuticals, which may include the name, basic attributes, manufacturing process, quality control standards, clinical applications, pharmacological effects, etc., and may include process flow documents from the pharmaceutical production process, clinical research reports, etc. Pharmaceutical knowledge text can be obtained through manual input or automatic retrieval from existing literature databases, which may include, but are not limited to, pharmaceutical instructions, industry literature, internal system reports, research reports, etc.

[0035] In step S102 of some embodiments, the drug attribute dimension refers to the characteristic classification of a drug in different dimensions. In the embodiments of this application, the drug attribute dimension includes the process stage attribute dimension, the industry stage attribute dimension, the drug value attribute dimension, etc.

[0036] Drug attribute fields refer to the specific data fields corresponding to drug attribute dimensions. For example, the process stage attribute dimension reflects the specific steps in the drug's processing, and the corresponding drug attribute fields might include specific technical parameters such as "temperature is 80℃," "pressure," and "reaction time is 1 hour." The industry stage attribute dimension reflects the specific stage in the drug's production process, and the corresponding drug attribute fields might include "planting cycle," "transportation process," and "inspection standards." Field extraction can be automatically performed using entity recognition technology in Natural Language Processing (NLP).

[0037] In step S103 of some embodiments, the database simulation model refers to a database model used to store and simulate drug knowledge, which categorizes the knowledge according to multiple dimensions using simulation coordinate axes. Simulation coordinate axes are axes in the simulation model used to represent different drug attribute dimensions, such as process stage attribute dimensions, industry stage attribute dimensions, etc. Simulation intervals refer to specific intervals on each coordinate axis, used to divide the specific attributes of the drug at each stage. For example, please refer to... Figure 2 , Figure 2 This is a schematic diagram of an optional database simulation model provided in this application embodiment. In this embodiment, the database simulation model includes three simulation coordinate axes: the z-axis represents the drug value attribute dimension, the x-axis represents the process stage attribute dimension, and the y-axis represents the industry stage attribute dimension. Each simulation coordinate axis includes multiple simulation intervals, each used to reflect the specific stage or state of the current drug attribute dimension. Specifically, each interval of the x-axis represents a specific process stage type, which, along the positive x-axis, are: granulation stage, tableting stage, coating stage, and aluminum-plastic stage. Each simulation interval of the y-axis represents a specific industry stage type, which, along the positive y-axis, are: planting stage (involving medicinal herb planting, cultivation, and field management), supply stage (including raw material procurement, transportation coordination, and inventory optimization), manufacturing stage (focusing on processing and production process control), distribution stage (managing distribution networks and logistics), and sales stage (responsible for market promotion, channel construction, and customer relationship maintenance). Each interval on the z-axis represents a dimension for evaluating the value of a drug. Along the positive z-axis, these dimensions are: quality (the rate at which the drug meets relevant production standards), cost (which may include value cost and energy consumption costs required for the production process), health (product efficacy verification, user health promotion, and efficacy tracking), and safety (the degree of risk associated with medication use). It is understood that the specific type of each simulation interval is preset.

[0038] In step S104 of some embodiments, the target coordinate axis refers to the simulation coordinate axis corresponding to the drug attribute dimension of the drug attribute field. The target interval refers to a specific interval on the target coordinate axis corresponding to the specific stage type of the drug attribute field. The drug attribute field will be stored in this target interval, ultimately forming the target database model.

[0039] Please see Figure 3 In some embodiments, step S104 may include, but is not limited to, steps S301 to S304: Step S301: Obtain the coordinate axis attribute dimensions of all simulation coordinate axes.

[0040] Step S302: For each drug attribute dimension, use the simulation coordinate axis with the same coordinate axis attribute dimension as the drug attribute dimension as the reference coordinate axis.

[0041] Step S303: Select at least two of the reference coordinate axes as the target coordinate axes.

[0042] Step S304: Obtain the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension to obtain the target interval.

[0043] In step S301 of some embodiments, the coordinate axis attribute dimension refers to the dimensional feature information of the simulated coordinate axis, such as... Figure 2 In the illustrated embodiment, the x-axis coordinate attribute dimension is the process stage attribute dimension, the y-axis coordinate attribute dimension is the industry stage attribute dimension, and the z-axis coordinate attribute dimension is the drug value attribute dimension.

[0044] In step S302 of some embodiments, since the number of drug attribute dimensions obtained from the drug knowledge text is greater than or equal to 2, the corresponding number of reference coordinate axes is the same as the number of drug attribute dimensions obtained in step S102, which is also greater than or equal to 2.

[0045] In step S303 of some embodiments, when there are more than 2 reference coordinate axes, any two coordinate axes can be selected as target coordinate axes to be combined into a two-dimensional knowledge graph in subsequent steps. However, this two-dimensional storage structure has the disadvantage that it is difficult to obtain multi-dimensional related data simultaneously. For example, a certain processing technology knowledge contains multi-dimensional features such as "operation steps (process dimension)," "temperature / time parameters (numerical dimension)," "related equipment model (entity dimension)," and "quality impact (result dimension)," while the two-dimensional knowledge graph can only be classified according to the single dimension of "process type," and the graph can only display preset entity relationships, resulting in incomplete knowledge expression and multi-dimensional related queries (such as "the production qualification rate that simultaneously satisfies 'wine-processing,' 'temperature 120℃,' and 'related equipment A'") cannot be realized. Therefore, in step S303 of other embodiments, more than 2 reference coordinate axes are selected as target coordinate axes to construct a multi-dimensional knowledge-related drug knowledge database in subsequent steps. In this embodiment, 3 reference coordinate axes are selected as target coordinate axes (such as... Figure 2 (As shown).

[0046] In step S304 of some embodiments, please refer to Figure 4 In some embodiments, the drug attribute dimension includes the process stage attribute dimension, the industry stage attribute dimension, and the drug value attribute dimension. Step S304 may include, but is not limited to, steps S401 to S405: Step S401: Perform process step identification on the drug attribute fields whose drug attribute dimension is the process stage attribute dimension to obtain the process step type.

[0047] Step S402: Identify the industry stage nodes for drug attribute fields whose drug attribute dimension is the industry stage attribute dimension, and obtain the industry node type.

[0048] Step S403: Identify the value category of the drug attribute field whose drug attribute dimension is the drug value attribute dimension to obtain the drug value type.

[0049] Step S404: Obtain the preset attribute types of all simulation intervals of the target coordinate axis.

[0050] Step S405: Based on the preset attribute types, determine the simulation interval corresponding to the process step type, industry node type, and drug value type as the target interval.

[0051] In step S401 of some embodiments, the process step type refers to the specific production steps and operation types that the drug undergoes in each process stage. The process step type can be determined by identifying the process flow steps corresponding to the drug attribute field using techniques such as deep learning networks or rule matching. For example, assuming the drug attribute field to be identified is "grinding raw materials into powder," a pre-trained process flow classification model is used for classification and identification, determining that the process step type corresponding to this drug attribute field is "granulation stage." It is understood that process flow classification models typically utilize deep learning methods, such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs), to classify and identify drug attribute fields by training a large number of relationships between text fields and process steps. Alternatively, traditional machine learning methods, such as support vector machines (SVMs) and decision trees, can be used to train a process flow classification model using labeled data.

[0052] In step S402 of some embodiments, the industry node type refers to a specific node type in the industry chain. The relationship between text fields and production steps can be learned using techniques such as deep learning networks or rule matching to identify the production stage corresponding to the drug attribute field, thereby determining the industry node type. For example, assuming the drug attribute field to be identified is "cultivate seeds in a dark place, with a planting cycle of 180 days," a pre-trained deep learning network is used for classification and identification, determining the industry node type corresponding to this drug attribute field as "planting stage."

[0053] In step S403 of some embodiments, the drug value attribute dimension refers to the risk factors reflecting the drug during production and use, encompassing assessments of multiple aspects such as quality, energy consumption costs, environmental sustainability, health, and the safety of the manufacturing process. The relationship between text fields and drug value types can be learned using techniques such as deep learning networks or rule matching to identify the drug value type corresponding to the drug attribute field. For example, assuming the drug attribute field to be identified is "the drug retains 95% of its efficacy after being stored at high temperatures for 6 months," a pre-trained deep learning network is used for classification and identification, determining that the drug value type corresponding to this attribute field is "health."

[0054] Furthermore, in other embodiments, after obtaining the specific drug value type, a pre-trained risk assessment model can be used to evaluate the risk score of the corresponding drug value type to reflect the potential risk impact of drug knowledge text under the corresponding drug value type. The risk assessment model can quantitatively assess the risk in a specific context by analyzing historical data and known risk factors. During drug production and use, the risk assessment model can predict potential risks and provide risk scores based on various drug attributes and known variables in the production and use process. This model can be trained using machine learning algorithms, such as deep learning networks, decision trees, or support vector machines, and by learning from a large amount of historical data and expert knowledge, it helps assess the risk level of a drug under specific conditions. For example, when conducting a "quality risk" assessment, the production process of the drug is evaluated to determine whether it strictly conforms to standards and whether there are potential quality defects, resulting in a quality score; the higher the score, the greater the probability of quality defects. An "energy cost" assessment can also be conducted to evaluate the energy consumption of the equipment required to manufacture the drug, resulting in an energy consumption score; the higher the score, the greater the energy consumption. The average score of multiple risk factor assessments is calculated, and the drug risk level can be determined based on the final average and a preset score range. Assuming the final average value is 20, the corresponding low-risk score range is [0, 30), then the drug risk level is determined to be low risk.

[0055] In step S404 of some embodiments, the preset attribute type refers to a specific attribute type defined in each simulation interval of the simulation coordinate axis. In this embodiment, for a target coordinate axis whose attribute dimension is a process stage attribute dimension, its preset attribute types include granulation stage, tableting stage, coating stage, and aluminum-plastic stage. For a target coordinate axis whose attribute dimension is an industry stage attribute dimension, its preset attribute types include planting stage, supply stage, manufacturing stage, distribution stage, and sales stage. For a target coordinate axis whose attribute dimension is a drug value attribute dimension, its preset attribute types include quality, cost, health, and safety.

[0056] In step S405 of some embodiments, the simulation intervals corresponding to the three types are determined as target intervals.

[0057] Steps S401 to S405 as shown in the embodiments of this application, through multi-dimensional drug attribute recognition technology, can accurately identify the process steps, industrial nodes and value types of drugs, and intelligently select simulation intervals based on these recognition results, thus deeply exploring the correlation of drug knowledge in different drug attribute dimensions.

[0058] Steps S301 to S304 of this embodiment of the application select the target coordinate axis by accurately mapping the simulation coordinate axis and the drug attribute dimension, and obtain the accurate simulation interval according to the drug attribute dimension. This solves the problem of information fragmentation in traditional knowledge transfer methods. The method of this embodiment enhances the intelligence and refinement of knowledge management.

[0059] Pharmaceutical manufacturing knowledge is constantly being updated with the application of new extraction processes, the release of new industry standards, and the rise of personalized formulations. However, in current technology, the classification, coding, and updating of knowledge still rely on manual operations and lack the ability to be dynamically updated. Please refer to... Figure 5 In some embodiments, after step S104, steps S501 to S505 may also be included, but are not limited to: Step S501: Obtain candidate knowledge text and obtain the first creation timestamp of the candidate knowledge text.

[0060] Step S502: Extract fields from the candidate knowledge text to obtain candidate attribute fields for at least two drug attribute dimensions.

[0061] Step S503: Determine candidate coordinate axes from the simulation coordinate axes based on the drug attribute dimension, and obtain the simulation interval of the candidate coordinate axes based on the drug attribute field of the drug attribute dimension to obtain the candidate interval.

[0062] Step S504: Obtain the drug attribute field in the candidate interval from the target database model, and obtain the second creation timestamp of the drug attribute field.

[0063] Step S505: If the second creation timestamp is earlier than the first creation timestamp, replace the drug attribute field with the candidate attribute field.

[0064] In step S501 of some embodiments, candidate knowledge text refers to newly added drug-related knowledge text at the current moment. The first creation timestamp refers to the time marked when the candidate knowledge text is created; it indicates the time when the text is made public, which is different from the time the text is obtained or included in the system. For example, if the candidate knowledge text is a new standard document in the industry, and the candidate knowledge text was obtained on January 10th but was published at 00:00 on January 1st, then the first creation timestamp should be based on the creation at 00:00 on January 1st.

[0065] In some embodiments, step S502 is implemented using the same logic as step S102, except that the object is transformed from drug knowledge text to candidate knowledge text, which will not be elaborated here. The final extracted subtext is the candidate attribute field.

[0066] In step S503 of some embodiments, the candidate coordinate axis refers to the simulation coordinate axis corresponding to the candidate attribute field and attribute dimension of the drug attribute field. The candidate interval refers to the specific interval on the candidate coordinate axis corresponding to the specific stage type of the candidate attribute field. The implementation logic of this step is the same as that of the relevant embodiments of step S104, and will not be repeated here.

[0067] In step S504 of some embodiments, the second creation timestamp refers to the creation timestamp of the existing drug attribute field in the candidate interval position in the target database model. It represents the time when the drug attribute field stored in the database is first created or made public.

[0068] In some embodiments, after field extraction, the drug knowledge text is identified and encoded. A preset classification and recognition model identifies the drug type, process stage, and text type (text type may include technical summary reports, quality standard descriptions, internal training materials, etc.). This information, along with the interval identifiers corresponding to different coordinate axes in the target database model and the second creation timestamp, serves as the identifier for the drug knowledge text (e.g., [Drug A, Drying, Technical Summary Report, x01, y01, z01, 202501010000], where x01 identifies the first simulation interval on the x-axis, y01 identifies the first simulation interval on the y-axis, z01 identifies the first simulation interval on the z-axis, and 202501010000 represents 00:00 on January 1, 2025). This identifier is stored together with the drug attribute fields.

[0069] In step S505 of some embodiments, if the second creation timestamp is earlier than the first creation timestamp, it indicates that the version of the drug attribute field originally stored in the target database model is older and will be replaced with the latest candidate attribute field. If the second creation timestamp is later than the first creation timestamp, it indicates that the currently obtained candidate knowledge text version is older and does not require updating the target database model.

[0070] Steps S501 to S505, as illustrated in this embodiment, achieve timely updates and automatic iteration of drug knowledge by dynamically acquiring and comparing timestamps of knowledge text. This solves the problems of response lag and high error rates caused by manual management in the prior art. The method of this embodiment ensures that drug attribute fields in the target database always reflect the latest industry dynamics and technological advancements, improving the efficiency and accuracy of knowledge management.

[0071] Please see Figure 6 Following step S104 in some embodiments, the drug database model construction method provided in this application embodiment further includes, but is not limited to, steps S601 to S606: Step S601: Obtain the object profile of the target object.

[0072] Step S602: For each drug attribute dimension, extract the knowledge reading behavior features that match the drug attribute dimension from the object profile.

[0073] Step S603: Extract the job type from the object profile, and query the job type according to the preset weight mapping relationship to obtain the feature weight data matching each drug attribute dimension.

[0074] Step S604: Based on the feature weight data and knowledge reading behavior features, feature concatenation is performed to obtain the target behavior feature vector.

[0075] Step S605: Based on the target behavior feature vector, perform data filtering on the target database model to obtain target drug knowledge.

[0076] Step S606: Push the target drug knowledge to the target audience.

[0077] In step S601 of some embodiments, the target object refers to the user who receives the drug knowledge text, which may be R&D personnel, production personnel, or quality control personnel of a pharmaceutical company. The object profile is a description of the target object's characteristics, typically including information such as the object's job position, department, experience level, and knowledge needs. Object profiles can be obtained through data collection, user behavior analysis, etc., and are usually analyzed using system logs, historical records, user input, and other data.

[0078] In step S602 of some embodiments, knowledge reading behavior characteristics refer to behavioral data characteristics extracted from the knowledge reading activities of the target object. Specifically, knowledge reading behavior characteristics include indicators such as the target object's level of attention to specific drug knowledge, reading frequency, reading duration, and reading content type.

[0079] In step S603 of some embodiments, the job type refers to the responsibilities and role definition of the target object's job, which may include R&D personnel, quality control personnel, production management personnel, etc. Each job type has different knowledge requirements, so the different job types need to be considered when recommending knowledge. The weight mapping relationship refers to the weight mapping rules set according to the degree of correlation between the target object's job type and the drug attribute dimension. The feature weight data matched to each drug attribute dimension refers to the feature weight corresponding to each drug attribute dimension obtained by combining the job type with the drug attribute dimension. For example, R&D personnel may have a higher weight in the "production process" dimension, while quality control personnel may have a higher weight in the "industry stage" dimension. For example, the feature weight data of process R&D personnel can be represented as [0.7 (corresponding to the process stage attribute dimension), 0.2 (corresponding to the industry stage attribute dimension), 0.1 (corresponding to the drug value attribute dimension)].

[0080] In some embodiments, after step S603, the method further includes updating the feature weight data. The specific steps are as follows: 1. Obtain the behavior timestamp of the knowledge reading behavior feature and obtain the current timestamp. The behavior timestamp refers to the time when the target object performs a knowledge reading operation. The current timestamp refers to the current time when this step is performed.

[0081] 2. Determine the decay index based on the duration between the behavior timestamp and the current timestamp. The decay index is a measure of the timeliness of the target object's reading behavior, calculated based on the time difference between the behavior timestamp and the current timestamp. In this embodiment, the decay index can be the actual number of days from the behavior timestamp to the current timestamp, rounded up to one day if less than one day.

[0082] 3. Perform feature decay calculations on the feature weight data based on the decay exponent and the preset decay rate to obtain reference weight data. Reference weight data refers to the decay result of the feature weight data calculated based on the decay exponent. The feature decay calculation process can be referenced in the following analytical formula: (1), in, This indicates the reference weight data. This represents feature weight data. This represents the attenuation rate, which can be 0.15. t represents the attenuation exponent.

[0083] 4. Update the feature weight data based on the reference weight data, that is, replace the feature weight data with the reference weight data.

[0084] In step S604 of some embodiments, the weights corresponding to each drug attribute dimension in the feature weight data are weighted and calculated with the knowledge reading behavior features corresponding to the drug attribute dimensions to obtain the calculation result. Then, the weighted calculation results corresponding to all drug attribute dimensions are concatenated to obtain the target behavior feature vector.

[0085] In step S605 of some embodiments, please refer to Figure 7 Step S605 may include, but is not limited to, steps S701 to S704: Step S701: For each interval of the drug attribute field in the target database model, semantic features are obtained by encoding the semantic features of the drug attribute fields.

[0086] Step S702: Construct vectors for the semantic features of attributes based on the drug attribute dimension to obtain field feature vectors.

[0087] Step S703: Calculate the similarity score between the field feature vector and the target behavior feature vector.

[0088] Step S704: Based on the similarity score, the target attribute field is selected from the drug attribute fields, and the drug knowledge text corresponding to the target attribute field is determined as the target drug knowledge.

[0089] In step S701 of some embodiments, the attribute semantic feature refers to the semantic information corresponding to the drug attribute field. The text information in the drug attribute field can be converted into a vector representation using text analysis techniques in Natural Language Processing (NLP), such as Word2Vec or the BERT model. This vector representation is the attribute semantic feature.

[0090] In step S702 of some embodiments, the field feature vector refers to concatenating the features corresponding to each drug attribute dimension to construct a high-dimensional vector. In this embodiment, the field feature vector includes three elements, corresponding to the attribute semantic features of the process stage attribute dimension, the attribute semantic features of the industry stage attribute dimension, and the attribute semantic features of the drug value attribute dimension, respectively.

[0091] In step S703 of some embodiments, the similarity score is a numerical value used to measure the degree of similarity between two vectors, which can be calculated using the cosine similarity formula.

[0092] In step S704 of some embodiments, a preset number of drug attribute fields with the highest similarity scores can be selected. These filtered drug attribute fields are the target attribute fields, and the preset number can be 10. Alternatively, a similarity score threshold can be set for filtering. For example, the similarity score threshold can be 0.65; fields with scores below this value are not selected as target attribute fields. If a drug attribute field has a similarity score of 0.7, it will be identified as a target attribute field; if the score is 0.6, it will not be recommended. Finally, the corresponding drug knowledge text is determined based on the filtered target attribute fields and pushed to the target audience.

[0093] Steps S701 to S704, as illustrated in this embodiment, involve semantic processing and vectorization of the drug attribute fields in the target database model. This enables efficient filtering and delivery of drug knowledge based on the actual needs of the target audience. The method in this embodiment achieves accurate delivery of drug knowledge, not only improving the efficiency of drug knowledge acquisition but also dynamically responding to changes in the needs of the target audience, significantly enhancing the efficiency of drug knowledge transmission and the accuracy of delivery.

[0094] In step S606 of some embodiments, target drug knowledge refers to specific drug knowledge filtered based on the needs of the target object after model querying. Target drug knowledge may include drug manufacturing processes, quality control standards, process optimization suggestions, etc. For example, a certain object profile may require a query for "quality control standards in the raw material processing stage," and the query result will be relevant drug knowledge. The query results are pushed to the target object based on the target object's profile characteristics to ensure that the recommended knowledge is accurate and meets its needs.

[0095] "Pushing target drug knowledge to target users" can utilize push engines or message queues to deliver target drug knowledge to target users on demand. Push methods can include system notifications, emails, instant messages, etc., to ensure that target users can obtain the necessary knowledge in a timely manner.

[0096] Steps S601 to S606, as illustrated in this embodiment, combine object profiling and drug attribute dimensions, utilizing feature weight data of knowledge reading behavior characteristics and job types to achieve precise drug knowledge delivery. By extracting object profiling and mapping job types, personalized drug knowledge recommendations can be made based on the responsibilities and needs of different positions. The introduction of feature splicing technology makes knowledge recommendations more aligned with the actual needs of the target audience. Finally, through the selection of target behavior feature vectors and database models, the recommended drug knowledge is ensured to be both efficient and accurate, significantly improving the efficiency and quality of drug knowledge acquisition.

[0097] Please see Figure 8This application also provides a drug database model construction apparatus, which can implement the above-described drug database model construction method. The apparatus includes: Text acquisition module 801 is used to acquire drug knowledge text; The field extraction module 802 is used to extract fields from drug knowledge text to obtain drug attribute fields with at least two drug attribute dimensions. The model acquisition module 803 is used to acquire the database simulation model; wherein, the database simulation model includes at least two simulation coordinate axes, and the simulation coordinate axes include multiple simulation intervals; The database model building module 804 is used to obtain the simulation coordinate axis based on the drug attribute dimension, obtain the target coordinate axis, obtain the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension, obtain the target interval, store the drug attribute field into the target interval, and obtain the target database model.

[0098] The specific implementation of this drug database model construction device is basically the same as the specific implementation of the drug database model construction method described above, and will not be repeated here.

[0099] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described drug database model construction method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0100] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 902 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 902 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and is called and executed by the processor 901 using the drug database model construction method of the embodiments of this application. The input / output interface 903 is used to implement information input and output; The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904); The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0101] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described drug database model construction method.

[0102] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0103] The drug database model construction method, device, electronic device, and storage medium provided in this application embodiment acquire drug knowledge text and identify drug attribute fields corresponding to drug attribute dimensions. Then, based on the simulation interval of the simulation coordinate axis, the drug attribute fields are combined with multiple drug attribute dimensions to determine the target interval from the simulation interval, and the drug attribute fields are stored in the target interval, thereby constructing a target database model. This achieves accurate classification and storage of drug knowledge, avoiding the fragmentation problem of traditional knowledge storage. Compared with existing technologies, researchers using the target database model constructed by the method in this application embodiment for drug knowledge acquisition can solve the problem of low knowledge acquisition efficiency in traditional knowledge transmission, where the target object needs to select the required knowledge from fragmented information. This enables more efficient and accurate knowledge sharing and dissemination, thereby helping to improve the learning efficiency of pharmaceutical company employees.

[0104] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0105] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0107] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0108] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0109] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0111] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0113] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for constructing a drug database model, characterized in that, The method includes: Obtain drug knowledge text; Field extraction is performed on the drug knowledge text to obtain drug attribute fields with at least two drug attribute dimensions; Obtain a database simulation model; wherein the database simulation model includes at least two simulation coordinate axes, and the simulation coordinate axes include multiple simulation intervals; The simulation coordinate axis is obtained based on the drug attribute dimension to obtain the target coordinate axis. The simulation interval of the target coordinate axis is obtained based on the drug attribute field of the drug attribute dimension to obtain the target interval. The drug attribute field is stored in the target interval to obtain the target database model.

2. The method according to claim 1, characterized in that, The process of obtaining the simulation coordinate axis based on the drug attribute dimension to obtain the target coordinate axis, and obtaining the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension to obtain the target interval, includes: Obtain the coordinate axis dimension attribute of all the simulated coordinate axes; For each of the drug attribute dimensions, the simulation coordinate axis that has the same coordinate axis attribute dimension as the drug attribute dimension is used as the reference coordinate axis. Use at least two of the reference coordinate axes as the target coordinate axes; The simulation interval of the target coordinate axis is obtained based on the drug attribute field of the drug attribute dimension, thus obtaining the target interval.

3. The method according to claim 2, characterized in that, The drug attribute dimensions include process stage attribute dimensions, industry stage attribute dimensions, and drug value attribute dimensions. The step of obtaining the simulation interval of the target coordinate axis based on the drug attribute fields of the drug attribute dimensions, to obtain the target interval, includes: Process step identification is performed on the drug attribute fields whose drug attribute dimension is the process stage attribute dimension to obtain the process step type; The industry stage node is identified for the drug attribute field whose drug attribute dimension is the industry stage attribute dimension, and the industry node type is obtained. Value category identification is performed on the drug attribute fields whose drug attribute dimension is the drug value attribute dimension to obtain the drug value type; Obtain the preset attribute types for all simulation intervals of the target coordinate axis; Based on the preset attribute type, the simulation interval corresponding to the process step type, the industry node type, and the drug value type is determined as the target interval.

4. The method according to claim 1, characterized in that, After storing the drug attribute fields into the target range to obtain the target database model, the method further includes: Obtain the object profile of the target object; For each of the drug attribute dimensions, knowledge reading behavior features matching the drug attribute dimension are extracted from the object profile; The job type is extracted from the object profile, and the job type is queried according to the preset weight mapping relationship to obtain feature weight data matching each of the drug attribute dimensions. The target behavior feature vector is obtained by concatenating the feature weight data and the knowledge reading behavior features. Based on the target behavior feature vector, data is filtered in the target database model to obtain target drug knowledge; The target drug knowledge is pushed to the target object.

5. The method according to claim 4, characterized in that, The process of filtering data from the target database model based on the target behavior feature vector to obtain target drug knowledge includes: For each interval of the target database model, the drug attribute field is used to obtain attribute semantic features by semantic feature encoding of the drug attribute field; Based on the drug attribute dimension, the semantic features of the attribute are vectorized to obtain the field feature vector; Calculate the similarity score between the field feature vector and the target behavior feature vector; Based on the similarity score, target attribute fields are selected from the drug attribute fields, and the drug knowledge text corresponding to the target attribute fields is determined as the target drug knowledge.

6. The method according to claim 4, characterized in that, After extracting the job type from the object profile and querying the job type according to a preset weight mapping relationship to obtain feature weight data matching each of the drug attribute dimensions, the method further includes: Obtain the behavior timestamp of the knowledge reading behavior characteristics, and obtain the current timestamp; The decay index is determined based on the duration between the behavior timestamp and the current timestamp; Based on the attenuation index and the preset attenuation rate, feature attenuation is calculated on the feature weight data to obtain reference weight data; The feature weight data is updated based on the reference weight data.

7. The method according to any one of claims 1 to 6, characterized in that, After storing the drug attribute fields into the target range to obtain the target database model, the method further includes: Obtain candidate knowledge texts and obtain the first creation timestamp of the candidate knowledge texts; Field extraction is performed on the candidate knowledge text to obtain at least two candidate attribute fields for the drug attribute dimensions; Candidate coordinate axes are determined from the simulation coordinate axes based on the drug attribute dimension, and the simulation interval of the candidate coordinate axes is obtained based on the drug attribute field of the drug attribute dimension, thus obtaining the candidate interval; Obtain the drug attribute field in the candidate interval from the target database model, and obtain the second creation timestamp of the drug attribute field; If the second creation timestamp is earlier than the first creation timestamp, the candidate attribute field replaces the drug attribute field.

8. A drug database model construction device, characterized in that, The device includes: The text acquisition module is used to acquire drug knowledge text. The field extraction module is used to extract fields from the drug knowledge text to obtain drug attribute fields with at least two drug attribute dimensions. A model acquisition module is used to acquire a database simulation model; wherein, the database simulation model includes at least two simulation coordinate axes, and the simulation coordinate axes include multiple simulation intervals; The drug database model construction module is used to obtain the simulation coordinate axis based on the drug attribute dimension, obtain the target coordinate axis, obtain the simulation interval of the target coordinate axis based on the drug attribute field of the drug attribute dimension, obtain the target interval, store the drug attribute field in the target interval, and obtain the target database model.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images