Drug searching method and device and computer readable storage medium

Abstract

The invention provides a medicine searching method and device and a computer readable storage medium. The medicine searching method comprises the steps of obtaining a query statement; semantic analysis is conducted on the query statement through the trained inference model to obtain multiple pieces of medicine dimension information, and different pieces of medicine dimension information represent different types of medicine information; according to the multiple pieces of medicine dimension information, a medicine database is inquired to obtain an inquiry result, and the inquiry result comprises basic information of the target medicine. The invention provides a scheme for improving drug search accuracy.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a drug search method and apparatus, and a computer-readable storage medium. Background Technology

[0002] With the development of mobile terminals and internet technology, people have gradually become accustomed to searching for relevant knowledge through the internet, such as searching for the ingredients and usage methods of medicines.

[0003] Existing drug search methods typically use keyword matching to search for drug information in drug databases.

[0004] However, with the number of drugs increasing exponentially, the number of similar drugs in drug databases also grows. When users input numerous search terms or use natural language descriptions, existing drug search methods struggle to accurately identify drug information. This leads to inaccurate search results, or even the inability to find any data, severely impacting the user experience. Summary of the Invention

[0005] This application provides a drug search method and apparatus, and offers a solution to improve the accuracy of drug search.

[0006] To achieve the above objectives, this application provides the following technical solution:

[0007] Firstly, a drug search method is provided, comprising: obtaining a query statement; performing semantic analysis on the query statement using a trained inference model to obtain multiple drug dimension information, wherein different drug dimension information represents different types of drug information; and querying a drug database based on the multiple drug dimension information to obtain query results, wherein the query results include basic information of the target drug.

[0008] Optionally, the multiple drug dimension information includes at least one of the following: generic name, brand, approval number, specifications, dosage form, pharmacy name, manufacturer, disease name, and indication.

[0009] Optionally, the trained inference model includes a first model and a second model, wherein the training time of the first model is later than the training time of the second model, and the semantic analysis of the query statement using the trained inference model includes: performing semantic analysis of the query statement using the first model to obtain a first drug dimension information combination; and performing semantic analysis of the query statement using the second model in response to the accuracy of the query result using the first drug dimension information combination being lower than a threshold.

[0010] Optionally, the inference model can be trained in the following way: acquiring training data, each training data including training statements and their corresponding standard drug dimension information; constructing the inference model; and training the inference model using the training data.

[0011] Optionally, the drug search method further includes: calculating the loss rate of the inference model based on the combination of drug dimension information of the training data and the standard drug dimension information; and evaluating the inference model based on the loss rate.

[0012] Optionally, the drug search method further includes: sorting the query results according to business sorting rules to obtain a first order; and displaying the query results according to the first order.

[0013] Optionally, before performing semantic analysis on the query statement using the trained inference model, the method further includes: preprocessing the query statement, wherein the inference model is capable of recognizing the transformed query statement.

[0014] Secondly, this application also discloses a drug search device, which includes: an acquisition module for acquiring a query statement; a semantic analysis module for performing semantic analysis on the query statement using a trained inference model to obtain multiple drug dimension information, wherein different drug dimension information represents different types of drug information; and a query module for querying a drug database based on the multiple drug dimension information to obtain query results, wherein the query results include basic information of the target drug.

[0015] Thirdly, a computer-readable storage medium is provided having a computer program stored thereon, the computer program being executed by a processor to perform any of the methods provided in the first aspect.

[0016] Fourthly, a communication device is provided, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to perform any of the methods provided in the first aspect.

[0017] Fifthly, a computer program product is provided, on which a computer program is stored, the computer program being executed by a processor to perform any of the methods provided in the first aspect.

[0018] Sixthly, embodiments of this application also provide a chip that stores a computer program, which, when executed by the chip, implements the steps of the above-described method.

[0019] In a seventh aspect, embodiments of this application also provide a system chip for use in a terminal. The system chip includes at least one processor and an interface circuit. The interface circuit and the at least one processor are interconnected via a line. The at least one processor is used to execute instructions to perform any one of the methods provided in the first or second aspect.

[0020] Compared with the prior art, the technical solution of this application has the following beneficial effects:

[0021] In this technical solution, a semantic analysis of the query statement is performed using an inference model to obtain multiple drug-related dimensions of the query statement. These dimensions represent the query statement's intent regarding the target drug, meaning the user's search intent is deduced from the query statement and mapped to multiple dimensions of drug information for retrieval. This approach allows for a better understanding of user intent, thereby improving the accuracy of drug searches. Furthermore, it can handle complex queries, significantly enhancing the user experience.

[0022] Furthermore, the multiple drug-related information dimensions include at least one of the following: generic name, brand, approval number, specifications, dosage form, pharmacy name, manufacturer, disease name, and indication. This application's technical solution lays the foundation for improving subsequent search accuracy by accurately identifying key information such as drug name, approval number, and specifications.

[0023] Furthermore, when constructing the training data for the inference model, each training data point includes a training statement and its corresponding standard drug dimension information. By defining the training data set in multiple dimensions, the inference model's ability to understand drug information is improved, thereby enhancing the accuracy of subsequent search results. Attached Figure Description

[0024] Figure 1 This is a flowchart of a drug search method provided in an embodiment of this application;

[0025] Figure 2 This is a flowchart of another drug search method provided in the embodiments of this application;

[0026] Figure 3 This is a schematic diagram illustrating an application scenario for drug search provided in an embodiment of this application;

[0027] Figure 4 This is a schematic diagram of the structure of a drug search device provided in an embodiment of this application;

[0028] Figure 5 This is a schematic diagram of another drug search device provided in an embodiment of this application;

[0029] Figure 6 This is a schematic diagram of the hardware structure of a drug search device provided in an embodiment of this application. Detailed Implementation

[0030] As described in the background section, with the number of drugs increasing exponentially, the number of similar drugs in drug databases also grows. When users input numerous search terms or use natural language descriptions, existing drug search methods struggle to accurately identify drug information. This leads to inaccurate search results, or even the inability to find any data, severely impacting the user experience.

[0031] In this technical solution, a semantic analysis of the query statement is performed using an inference model to obtain multiple drug-related dimensions of the query statement. These dimensions represent the query statement's intent regarding the target drug, meaning the user's search intent is deduced from the query statement and mapped to multiple dimensions of drug information for retrieval. This approach allows for a better understanding of user intent, thereby improving the accuracy of drug searches. Furthermore, it can handle complex queries, significantly enhancing the user experience.

[0032] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, specific embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0033] This invention provides a drug search method, referring to... Figure 1 The following will provide a detailed explanation through specific steps.

[0034] It is understood that, in specific implementations, the drug search method can be implemented using software programs, which run in a processor integrated within a chip or chip module. This method can also be implemented using a combination of software and hardware; this application does not impose any restrictions.

[0035] In step 101, the query statement is obtained.

[0036] In practice, users can enter a query to search for the desired medication. The query can include information about the medication (also known as the drug) to be searched for.

[0037] Specifically, the content of drug information included in the query statement varies in different application scenarios. For example, query statement 1 only includes drug name-related content, while query statement 2 includes drug name, brand, specifications, and other related content.

[0038] In step 102, the trained inference model is used to perform semantic analysis on the query statement to obtain multiple drug dimension information, which represent different types of drug information.

[0039] In practice, the inference model can analyze the search intent in the query statement, specifically by representing the search intent through multiple dimensions of drug information. In other words, the inference model can analyze and categorize the semantics of the query statement to obtain relevant drug information across multiple dimensions.

[0040] Furthermore, multiple drug-related information dimensions include at least one of the following: generic name, brand, approval number, specifications, dosage form, pharmacy name, manufacturer, disease name, and indication.

[0041] The generic name represents the name of the drug to be queried, obtained through semantic analysis of the query statement by the inference model. The brand name represents the brand name of the drug to be queried. Since the same drug may have different brands, identifying the brand helps to find more accurate drugs. The approval number is a legal certificate issued by the drug regulatory authority to a specific manufacturer for a particular drug according to legal standards, production processes, and production conditions. Each manufacturer and each product has a specific approval number. The drug specification indicates the weight (or potency) or percentage (%) of the active ingredient contained in each vial, tablet, or other unit of preparation. The dosage form indicates the form suitable for medical and preventative applications, also known as a pharmaceutical preparation. The pharmacy name indicates the store that can provide or sell the drug. The manufacturer indicates the manufacturer of the drug, including basic information such as the name and address of the production site. The indications indicate the scope and standards for the drug's appropriate use, also known as indications. Furthermore, since each drug can be used for at least one indication, each drug can also correspond to at least one disease. Therefore, the inference model can also analyze the query statement to obtain the names of diseases that may be included in the statement, to assist in finding the target drug.

[0042] By setting the above-mentioned drug dimension information, the embodiments of this application can characterize the drug to be queried from multiple dimensions, so that subsequent queries can be performed from more dimensions, which helps to accurately determine the target drug.

[0043] In step 103, the drug database is queried based on multiple drug dimension information to obtain query results.

[0044] Compared to existing technologies that directly use user queries for fuzzy field searches, this application uses multiple drug dimension information output by an inference model to query a drug database. Because multiple drug dimension information categorizes drug-related information and can represent drugs from more dimensions, queries based on these multiple drug dimension information can obtain richer and more comprehensive information.

[0045] In practice, the drug database can include basic information about each drug. For example, the drug database can include drug name, drug price, drug manufacturer, pharmacy, and store address. The drug database can be pre-set by the user; this application does not impose any restrictions on this.

[0046] Accordingly, the query results include basic information about the target drug, which is obtained by querying the drug database based on multiple drug-related dimensions.

[0047] For example, for a user's query, the inference model obtains the following drug-related information through semantic analysis:

[0048] It should be noted that the sequence number of each step in this embodiment does not represent a limitation on the execution order of each step.

[0049] It is understood that, in specific implementations, the drug search method can be implemented using software programs, which run in a processor integrated within a chip or chip module. This method can also be implemented using a combination of software and hardware; this application does not impose any restrictions.

[0050] This application embodiment performs semantic analysis on the query statement through an inference model to obtain multiple drug dimension information of the query statement. It represents the query statement for the drug to be searched from multiple dimensions, that is, it infers the user's search intent from the query statement and maps it to multiple dimensions of drug information for retrieval. This can better understand the user's intent, thereby improving the accuracy of drug search and handling complex queries.

[0051] In a non-limiting embodiment, the inference model is a pre-trained model. The training data used to train the inference model specifically includes at least one standard drug dimension information. This standard drug dimension information includes at least one of the following: generic name, brand, approval number, specifications, dosage form, pharmacy name, manufacturer, disease name, and indication.

[0052] Specifically, when constructing training data, a dataset with at least one standard drug dimension information can be selected from the user's search terms (i.e., frequently used words) to form training statements. The training statements are then labeled to obtain the corresponding standard drug dimension information and basic information about the target drug. The training data includes the training statements, the corresponding standard drug dimension information, and the basic information about the target drug.

[0053] Furthermore, the training data can be processed to convert the natural language in the training data into digital language for the inference model to recognize.

[0054] Furthermore, the training data can be divided into a training dataset and a validation dataset. The training dataset is used to train the inference model, and the validation dataset is used to verify the performance of the inference model. For example, 80% of the training data can be used as the training dataset, and 20% as the validation dataset.

[0055] In one specific embodiment, the inference model can be a Natural Language Processing (NLP) model. Specifically, it can be a Large-Scale Knowledge Enhanced Pre-Training for Language Understanding and Generation model, also known as the Ernie model.

[0056] It should be noted that the inference model can also be any other implementable NLP model, and this application does not impose any restrictions on it.

[0057] In one specific embodiment, the loss rate of the inference model during the training process can be calculated by combining drug dimension information to comprehensively evaluate the model's performance and ensure that a better inference model is selected.

[0058] Specifically, please refer to Figure 2 , Figure 2 A method for selecting inference models is shown.

[0059] In step 201, the loss rate of the inference model is calculated based on the combination of drug dimension information of the training data and standard drug dimension information.

[0060] In practice, the inference model outputs a combination of drug dimension information based on the training data. This combination includes at least one drug dimension information. By comparing the combination of drug dimension information output by the inference model with the standard drug dimension information, a loss function is used to calculate the loss rate, thus determining the analytical accuracy of the inference model. In other words, the loss rate reflects the inference accuracy of the model across the aforementioned multiple drug dimension information dimensions.

[0061] In step 202, the inference model is evaluated based on the loss rate.

[0062] Specifically, the loss rate measures the degree of difference between the predictions and actual values ​​of an inference model. The smaller the loss rate, the more accurate the model's predictions are, and the better its performance.

[0063] Furthermore, the optimal inference model can be determined using precision and recall. Specifically, the F1 score can be calculated based on precision and recall, and after multiple training iterations, the inference model with the highest F1 score can be selected as the final inference model.

[0064] Specifically, the precision calculation formula is: Precision = TP / (TP + FP), where TP represents the number of samples that are actually positive and are predicted as positive, and FP represents the number of samples that are actually negative and are predicted as positive. Specifically, the training data corresponding to the drug dimension information combination and the standard drug dimension information are considered positive samples, while the training data where the drug dimension information combination does not correspond to the standard drug dimension information are considered parent samples.

[0065] The formula for calculating recall is: Recall = TP / (TP + FN), where FN represents the number of samples that were actually positive but were predicted as negative.

[0066] The formula for calculating the F1 score is: F1 = 2 × (Precision × Recall) / (Precision + Recall) = 2 × TP / (2 × TP + FP + FN).

[0067] In this embodiment, since the inference model needs to output multiple dimensions of drug information, and the calculation of the F1 score can reflect the difference between the actual value and the standard value of multiple dimensions of drug information, the performance of the inference model can be comprehensively measured.

[0068] The embodiments of this application can significantly improve the accuracy and efficiency of drug information retrieval, providing users with intelligent drug and detailed drug information query services.

[0069] In a non-limiting embodiment, please refer to Figure 3 The trained inference model 30 includes a first model 301 and a second model 302, with the training time of the first model 301 being later than that of the second model 302. Embodiments of this application can switch inference models according to actual needs to ensure search accuracy.

[0070] In this embodiment, the first model 301 is used to perform semantic analysis on the query statement to obtain the first drug dimension information combination. That is, the latest inference model (i.e., the first model 301) is preferentially used to perform semantic analysis on the query statement, and the first model 301 can output the first drug dimension information combination. The first drug dimension information combination includes at least one drug dimension information.

[0071] A query based on the first drug dimension information combination can be performed in the drug database to obtain query results. If the accuracy of this query result is below a threshold, a second model 302 is used to perform semantic analysis on the query statement. The second model 302 can output a second drug dimension information combination. This second drug dimension information combination can be used for subsequent queries in the drug database.

[0072] In practice, multiple training sessions can yield multiple inference models. If the inference performance of the latest inference model is unsatisfactory, the system can switch back to the original inference model (i.e., the second model 302) to continue semantic analysis, thus ensuring the accuracy of drug search.

[0073] In a non-limiting embodiment, after obtaining the query results, the query results can be sorted according to certain rules to further meet user needs and improve user experience.

[0074] Specifically, the query results are sorted according to the business sorting rules to obtain the first order; the query results are then displayed in the first order.

[0075] Specifically, when the basic information of the target drug includes its price, the basic information of the target drug can be sorted by price; when the basic information of the target drug includes the distance between the pharmacy providing the target drug and the user, the basic information of the target drug can be sorted by distance.

[0076] It should be noted that the query results can also be sorted according to any other feasible business sorting rules, and this application does not impose any restrictions on this.

[0077] In a non-limiting embodiment, the query statement may be preprocessed before being input into the inference model so that the inference model can recognize the transformed query statement.

[0078] Specifically, the language of the query statement can be converted. For example, when the query statement is in natural language, it can be converted into numerical language so that the inference model can recognize and process the query statement more promptly, thereby improving search efficiency.

[0079] In a non-limiting embodiment, the inference model can also process the query statement based on semantic reasoning capabilities, specifically by merging the number of repeated words and removing irrelevant search terms, thereby accelerating the retrieval speed.

[0080] In this embodiment of the application, high-quality vocabulary is obtained through semantic reasoning of the reasoning model, which reduces the number of searches, reduces the retrieval of large amounts of data, and improves the search speed.

[0081] Please refer to Figure 4 , Figure 4 A drug search device 40 is shown, which may include:

[0082] Module 401 is used to retrieve the query statement;

[0083] The semantic analysis module 402 is used to perform semantic analysis on the query statement using a trained reasoning model to obtain multiple drug dimension information, with different drug dimension information representing different types of drug information.

[0084] The query module 403 is used to query the drug database based on multiple drug dimension information to obtain query results, which include basic information about the target drug.

[0085] In specific implementations, the aforementioned drug search device 40 may correspond to a chip in a computing device that has a drug search function, such as a system-on-a-chip (SOC), a baseband chip, etc.; or to a chip module in a computing device that includes a drug search function; or to a chip module that has a data processing function; or to a computing device.

[0086] Other relevant descriptions of the drug search device 40 can be found in the descriptions in the foregoing embodiments, and will not be repeated here.

[0087] Please refer to Figure 5 , Figure 5 A structural diagram of another drug search system is shown. This drug search system may include a client 501 and a server 502.

[0088] Client 501 includes a drug data query module 5011, which provides users with an entry point for drug information search. Users can input query statements through the drug data query module 5011.

[0089] Client 501 may also include a model data labeling module 5012 for data preparation and labeling of drug information classification data. Specifically, administrators can use the model data labeling module 5012 to label training statements and determine the standard drug dimension information corresponding to the training statements. In this case, client 501 is for use by administrators.

[0090] The server 502 may include a model data storage module 5021 for storing the trained inference model and the training data.

[0091] The server 502 may include a drug data training module 5022, which is used to call the training data in the model data storage module 5021 to train the inference model.

[0092] The server 502 may include a drug search module 5023 to support the search service for the client 501, returning query results to the client 501. The drug search module 5023 is equivalent to... Figure 4 The acquisition module 401 and the query module 403 are shown in the figure.

[0093] The server 502 may include an inference module 5024, which provides the search module 5023 with intelligent intent parsing and search term recognition services, returning multiple drug-related dimension information to the search module 5023. The inference module 5024 is equivalent to... Figure 4 The semantic analysis module 402 shown.

[0094] In one specific embodiment, a user initiates a search task request, including a query statement, to the drug search module 5023 on the server 502 through the drug data query module 5011 on the client 501. Upon receiving the task, the drug search module 5023 calls the inference module 5024 to perform semantic analysis on the query statement. During inference, the inference module 5024 queries the currently active inference model in the model data storage module 5021, performs inference based on the currently active inference model, and returns multiple drug dimension information to the drug search module 5023 after the inference is complete. The drug search module 5023 then searches the drug database based on the obtained drug dimension information, performs a secondary sorting according to business sorting rules, and returns the query results to the drug data query module 5011. The drug data query module 5011 displays the query results.

[0095] Furthermore, the backend training operator obtains the training statements from the model data storage module 5021 through the model data labeling module 5012, labels the training statements, and submits the results. After labeling is completed, the training data is recorded and stored in the model data labeling module 5012. The drug data training module 5022 starts running, calling the training data in the model data labeling module 5012 for training. After completion, it stores the latest trained inference model in the model data labeling module 5012.

[0096] Regarding the modules / units included in the various devices and products described in the above embodiments, they can be software modules / units, hardware modules / units, or a combination of both. For example, for devices and products applied to or integrated into a chip, all modules / units can be implemented using hardware methods such as circuits, or at least some modules / units can be implemented using software programs running on a processor integrated within the chip, while the remaining (if any) modules / units can be implemented using hardware methods such as circuits. For devices and products applied to or integrated into a chip module, all modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components of the chip module, or at least some modules / units can be implemented using hardware methods such as circuits. The implementation is achieved through a software program that runs on the processor integrated within the chip module. The remaining modules / units (if any) can be implemented using hardware methods such as circuits. For various devices or products applied to or integrated into computing devices, each of their modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components within the computing device. Alternatively, at least some modules / units can be implemented through a software program that runs on the processor integrated within the computing device, while the remaining modules / units (if any) can be implemented using hardware methods such as circuits.

[0097] This application also discloses a storage medium, which is a computer-readable storage medium storing a computer program thereon. When the computer program is executed, it can perform the steps of the method shown in the foregoing embodiments. The storage medium may include read-only memory (ROM), random access memory (RAM), a magnetic disk, or an optical disk, etc. The storage medium may also include non-volatile memory or non-transitory memory, etc.

[0098] Please refer to Figure 6 This application also provides a schematic diagram of the hardware structure of a communication device. The device includes a processor 601, a memory 602, and a transceiver 603.

[0099] Processor 601 can be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program according to the present application. Processor 601 may also include multiple CPUs, and processor 601 can be a single-core processor or a multi-core processor. Here, processor can refer to one or more devices, circuits, or processing cores used to process data (e.g., computer program instructions).

[0100] The memory 602 can be a ROM or other type of static storage device capable of storing static information and instructions, RAM or other type of dynamic storage device capable of storing information and instructions, or it can be an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer. This application embodiment does not impose any limitations on this. The memory 602 can exist independently (in this case, the memory 602 can be located outside or inside the device) or it can be integrated with the processor 601. The memory 602 may contain computer program code. The processor 601 is used to execute the computer program code stored in the memory 602, thereby implementing the method provided in this application embodiment.

[0101] The processor 601, memory 602, and transceiver 603 are connected via a bus. The transceiver 603 is used to communicate with other devices or communication networks. Optionally, the transceiver 603 may include a transmitter and a receiver. The device in the transceiver 603 that implements the receiving function can be considered as a receiver, which is used to perform the receiving steps in the embodiments of this application. The device in the transceiver 603 that implements the transmitting function can be considered as a transmitter, which is used to perform the transmitting steps in the embodiments of this application.

[0102] when Figure 6The schematic diagram shown illustrates the structure of the computing device involved in the above embodiments. The processor 601 is used to control and manage the operations of the computing device. For example, the processor 601 supports the computing device in performing actions in other processes described in the embodiments of this application. The processor 601 can communicate with other network entities via the transceiver 603, for example, with the aforementioned network device. The memory 602 is used to store the program code and data of the computing device.

[0103] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article indicates that the preceding and following related objects have an "or" relationship.

[0104] In the embodiments of this application, "multiple" refers to two or more.

[0105] The descriptions of "first," "second," etc., appearing in the embodiments of this application are for illustrative purposes and to distinguish the objects being described. They have no order and do not indicate any special limitation on the number of devices in the embodiments of this application, nor do they constitute any limitation on the embodiments of this application.

[0106] In this application, the term "connection" refers to various connection methods, such as direct connection or indirect connection, to achieve communication between devices. This application does not impose any limitations on this.

[0107] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means.

[0108] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0109] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and other division methods may exist in actual implementation; 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, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0110] The units described 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.

[0111] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can be physically included separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0112] The integrated unit implemented as a software functional unit described above can be stored in a computer-readable storage medium. This software functional unit, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute some steps of the methods described in the various embodiments of this application.

[0113] While this application discloses the above information, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of this application; therefore, the scope of protection of this application shall be determined by the scope defined in the claims.

Claims

1. A drug search method, characterized in that, include: Retrieve the query statement; The trained reasoning model is used to perform semantic analysis on the query statement to obtain multiple drug dimension information, and different drug dimension information represents different types of drug information; The drug database is queried based on the multiple drug dimension information to obtain query results, which include basic information about the target drug.

2. The drug search method according to claim 1, characterized in that, The multiple drug-related dimensions of information include at least one of the following: Generic name, brand, approval number, specifications, dosage form, pharmacy name, manufacturer, disease name, indications.

3. The drug search method according to claim 1, characterized in that, The trained inference model includes a first model and a second model, wherein the training time of the first model is later than that of the second model, and the semantic analysis of the query statement using the trained inference model includes: The first model is used to perform semantic analysis on the query statement to obtain a combination of first drug dimension information; If the accuracy of the query results using the first drug dimension information combination is lower than the threshold, the second model is used to perform semantic analysis on the query statement.

4. The drug search method according to claim 1, characterized in that, The inference model is trained using the following method: Acquire training data, each training data set including training statements and their corresponding standard drug dimension information; Construct the inference model; The inference model is trained using the training data.

5. The drug search method according to claim 4, characterized in that, Also includes: The loss rate of the inference model is calculated based on the combination of drug dimension information of the training data and the standard drug dimension information. The inference model is evaluated based on the loss rate.

6. The drug search method according to claim 1, characterized in that, Also includes: The query results are sorted according to the business sorting rules to obtain the first order; The query results are displayed in the first order.

7. The drug search method according to claim 1, characterized in that, Before performing semantic analysis of the query statement using the trained inference model, the following steps are also included: The query statement is preprocessed, and the inference model is able to recognize the transformed query statement.

8. A drug search device, characterized in that, include: The retrieval module is used to retrieve query statements; The semantic analysis module is used to perform semantic analysis on the query statement using a trained reasoning model to obtain multiple drug dimension information, with different drug dimension information representing different types of drug information. The query module is used to query the drug database based on the multiple drug dimension information to obtain query results, which include basic information about the target drug.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when run by a processor, performs the steps of the drug search method according to any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that, When executed by a processor, the computer program / instruction implements the steps of the drug search method according to any one of claims 1 to 7.

11. A computing device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor runs the computer program, it performs the steps of the drug search method according to any one of claims 1 to 7.

Images