CN110781677A - Medicine information matching processing method and device, computer equipment and storage medium - Google Patents
Medicine information matching processing method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110781677A CN110781677A CN201910969896.7A CN201910969896A CN110781677A CN 110781677 A CN110781677 A CN 110781677A CN 201910969896 A CN201910969896 A CN 201910969896A CN 110781677 A CN110781677 A CN 110781677A
- Authority
- CN
- China
- Prior art keywords
- medicine
- information
- matching
- drug
- entity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H70/00—ICT specially adapted for the handling or processing of medical references
- G16H70/40—ICT specially adapted for the handling or processing of medical references relating to drugs, e.g. their side effects or intended usage
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computational Linguistics (AREA)
- Epidemiology (AREA)
- Chemical & Material Sciences (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Medicinal Chemistry (AREA)
- Databases & Information Systems (AREA)
- Toxicology (AREA)
- Pharmacology & Pharmacy (AREA)
- Human Computer Interaction (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The application relates to a medicine information matching processing method and device, computer equipment and a storage medium in the field of artificial intelligence. The method comprises the following steps: receiving problem information uploaded by a terminal; calling a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix; inputting the vector matrix into a trained classification model, calculating a plurality of context vectors corresponding to the vector matrix, and identifying an intention type corresponding to the problem information according to the context vectors; identifying a drug entity in the question information in a drug dictionary repository when the intent type is an intent type associated with a drug; calling a thread scanning database to determine the medicine grade corresponding to the medicine entity; and matching the medicine information in the database according to the medicine grade, outputting medicine reply information corresponding to the question information, and returning the medicine reply information to the terminal. By adopting the method, the accuracy of medicine information matching can be improved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for matching and processing medicine information, a computer device, and a storage medium.
Background
At present, the medical resources of China are in shortage, and the problem of medication safety is increasingly severe. In order to solve the problem of medication safety, a medicine information matching processing method is provided. The traditional medicine information matching processing method is that the problem information of a user is received, and the characteristic extraction is directly carried out on the problem information, so that the user intention in the problem information is identified, the medicine entity in the problem information is identified, and the corresponding medicine reply information is searched by utilizing the medicine entity.
In the traditional medicine information matching processing method, the accuracy of medicine information matching is low due to the fact that more interference information exists in problem information. Therefore, how to improve the accuracy of matching the medicine information becomes a technical problem to be solved at present.
Disclosure of Invention
In view of the above, it is necessary to provide a medicine information matching processing method, apparatus, computer device, and storage medium capable of improving accuracy of medicine information matching by filtering interference information in problem information.
A drug information matching processing method, the method comprising:
receiving problem information uploaded by a terminal;
calling a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix;
inputting the vector matrix into a trained classification model, calculating a plurality of context vectors corresponding to the vector matrix, and identifying an intention type corresponding to the problem information according to the context vectors;
identifying a drug entity in the question information in a drug dictionary repository when the intent type is an intent type associated with a drug;
calling a thread scanning database to determine the medicine grade corresponding to the medicine entity;
and matching the medicine information in the database according to the medicine grade, outputting medicine reply information corresponding to the question information, and returning the medicine reply information to the terminal.
In one embodiment, the classification model includes a plurality of network layers, the calculating a plurality of context vectors corresponding to the vector matrix, and the identifying the intention type corresponding to the question information according to the context vectors includes:
inputting a vector matrix into a trained classification model, and extracting a plurality of word vectors in the vector matrix through an input layer of the classification model;
inputting a plurality of word vectors into an attention layer, calculating a context vector and a weight corresponding to each word vector, and generating a first extraction result;
inputting the first extraction result into a convolution layer, extracting context characteristics corresponding to the context vector, and generating a second extraction result;
inputting the second extraction result into a pooling layer, and performing dimension reduction processing on the second extraction result;
inputting the second extraction result subjected to the dimensionality reduction treatment into a full-connection layer, classifying the second extraction result subjected to the dimensionality reduction treatment to obtain a classification result, and outputting the classification result after weighting through an output layer;
and selecting the intention type with the maximum weight in the classification results output after weighting as the intention type corresponding to the question information.
In one embodiment, the identifying the drug entity in the question information in the drug dictionary repository comprises:
extracting medicine characteristic information in the question information;
accurately matching the medicine characteristic information with a medicine dictionary library;
and when the medicine entities cannot be identified by accurate matching, carrying out fuzzy matching on the medicine characteristic information and a medicine dictionary library, and determining the medicine entities corresponding to the problem information.
In one embodiment, the invoking thread scans a database and determining the drug level corresponding to the drug entity comprises:
performing character matching on the medicine entity and the medicine grade in a database;
when the character matching is not successful, carrying out pinyin matching on the medicine entity and the corresponding medicine level;
and selecting the medicine grade with the highest pinyin matching probability as the medicine grade corresponding to the medicine entity.
In one embodiment, the matching of the drug information in the database according to the drug grade comprises:
calling a thread scanning configuration file, wherein a plurality of medicine grades and corresponding matching sequences are recorded in the configuration file;
and performing character matching on the medicine entity and a database according to the matching sequence corresponding to the medicine grade to obtain medicine reply information corresponding to the question information.
A drug information matching processing apparatus, the apparatus comprising:
the communication module is used for receiving the problem information uploaded by the terminal;
the word segmentation module is used for calling a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix;
the first identification module is used for inputting the vector matrix into the trained classification model and calculating a plurality of context vectors corresponding to the vector matrix; identifying an intention type corresponding to the problem information according to the context vector;
a second identification module for identifying a drug entity in the question information in a drug dictionary repository when the intent type is an intent type associated with a drug;
the determining module is used for determining the medicine grade corresponding to the medicine entity by utilizing a thread scanning database;
and the matching module is used for matching medicines in a database according to the medicine grades and outputting medicine reply information corresponding to the question information.
In one embodiment, the first recognition module is further configured to input a word segmentation sequence into a trained classification model, and extract a plurality of word vectors in the vector matrix through an input layer of the classification model; inputting a plurality of word vectors into an attention layer, calculating a context vector and a weight corresponding to each word vector, and generating a first extraction result; inputting the first extraction result into a convolution layer, extracting context characteristics corresponding to the context vector, and generating a second extraction result; inputting the second extraction result into a pooling layer, and performing dimension reduction processing on the second extraction result; inputting the second extraction result subjected to the dimensionality reduction treatment into a full-connection layer, classifying the second extraction result subjected to the dimensionality reduction treatment to obtain a classification result, and outputting the classification result after weighting through an output layer; and selecting the intention type with the maximum weight in the classification results output after weighting as the intention type corresponding to the question information.
In one embodiment, the second identification module is further configured to extract medicine feature information in the question information; accurately matching the medicine characteristic information with a medicine dictionary library; and when the medicine entities cannot be identified by accurate matching, carrying out fuzzy matching on the medicine characteristic information and a medicine dictionary library, and determining the medicine entities corresponding to the problem information.
A computer device comprising a memory and a processor, the memory storing a computer program operable on the processor, the processor implementing the steps in the various method embodiments described above when executing the computer program.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the respective method embodiment described above.
According to the medicine information matching processing method, the medicine information matching processing device, the computer equipment and the storage medium, the problem information is subjected to word segmentation processing through the calling thread, a corresponding vector matrix is obtained, and the input into a trained classification model is facilitated. A plurality of context vectors corresponding to the vector matrix are calculated through the trained classification model, the intention type corresponding to the problem information is identified according to the context vectors, interference information in the problem information can be filtered, word focusing processing is carried out on the vector matrix, and therefore intention identification accuracy is improved. After the medicine entities in the problem information are identified, the thread scanning database is called to determine the medicine levels corresponding to the medicine entities, the medicine entities can be effectively classified, and the medicine granularity can be more finely divided. Medicine information matching is carried out in the database through the medicine level, medicine reply information corresponding to the question information is output, medicine information matching can be achieved aiming at diversified question information, and accuracy and comprehensiveness of medicine information matching are improved.
Drawings
FIG. 1 is a flow chart illustrating a method for matching drug information according to an embodiment;
FIG. 2 is a flowchart illustrating the step of identifying an intent type corresponding to the issue information in one embodiment;
FIG. 3 is a block diagram showing the configuration of a medicine information matching processing apparatus according to an embodiment;
FIG. 4 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In one embodiment, as shown in fig. 1, a medicine information matching processing method is provided, which is described by taking an example of the method applied to a computer device, and includes the following steps:
and 102, receiving the problem information uploaded by the terminal.
And step 104, calling a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix.
The computer device can receive the problem information uploaded by the terminal. The question information may be voice information or text information. And when the problem information is voice information, the computer equipment performs voice recognition on the voice information to obtain corresponding text information. After the computer equipment acquires the problem information uploaded by the terminal, word segmentation processing can be carried out on the problem information. Specifically, the computer device may first perform part-of-speech tagging on the question information. The computer equipment acquires a pre-established word bank, wherein the word bank comprises a large number of common words, specific words and corresponding part of speech tag information.
The computer equipment extracts the vocabularies in the problem information, extracts all corresponding part-of-speech tag information from a word bank according to the vocabularies, and determines the part-of-speech corresponding to each vocabulary according to the part-of-speech tag information so as to obtain a part-of-speech tagging result. And the computer equipment matches the problem information with a plurality of words in a word bank according to the part of speech tagging result, and divides the matched words into words for the problem information to obtain a plurality of problem words. And the computer equipment converts each problem vocabulary into a word vector and generates a vector matrix corresponding to the problem information according to a plurality of word vectors.
And 106, inputting the vector matrix into the trained classification model, calculating a plurality of context vectors corresponding to the vector matrix, and identifying the intention type corresponding to the problem information according to the context vectors.
And after the computer equipment obtains the vector matrix corresponding to the problem information, inputting the vector matrix into the trained classification model. Wherein the classification model may be a convolutional neural network model based on an attention layer. The classification model may include multiple network layers. For example, the classification model may include an input layer, an attention layer, a convolutional layer, a pooling layer, a fully-connected layer, and an output layer. The computer equipment extracts a plurality of word vectors in the vector matrix through the classification model, and calculates a context vector corresponding to each word vector and a weight corresponding to each word vector. And extracting the context characteristics corresponding to the context vectors by the computer equipment, and further performing dimension reduction processing on the context characteristics to obtain the main context characteristics corresponding to the context vectors. And the computer equipment classifies the context characteristics after the dimension reduction processing, weights the classification result and outputs the result. The computer equipment further selects the intention type with the largest weight from the classification results after weighted output as the intention type corresponding to the question information, and realizes the identification of the intention type corresponding to the question information.
In step 108, when the intent type is an intent type associated with a drug, a drug entity in the question type is identified in a drug dictionary repository.
And after identifying the intention type corresponding to the problem information according to the context vector, the computer equipment judges whether the intention type is related to the medicine. The types of intentions related to the drugs include indications of the drugs, adverse reactions of the drugs, usage amounts of the drugs, cautions of the drugs, interactions between drugs, and the like. The computer device may mark the type of intent associated with the drug as valid intent, the computer device performing drug information matching processing only on valid intents. When the intent type is not an intent type related to the drug, i.e., the intent type is a non-valid intent, the computer device may send a valid intent to the terminal.
When the intention type is an intention type related to the medicine, the computer device acquires a pre-established medicine dictionary library including a common name of the medicine and a medicine trade name. The drug trade name means identification information that can be accurately positioned to a corresponding commodity. Such as an approval code, a drug order code, etc. The computer device identifies the drug entity in the question information in a drug dictionary repository. The drug entity may be a generic name, a product name, and a commodity name of the drug. For example, the generic name for the drug product may be acarbose, the product name may be acarbose tablets, and the product name may be bayer apple.
Specifically, the computer equipment extracts medicine characteristic information in the question information and accurately matches the medicine characteristic information with a medicine dictionary library. When the medicine entities in the problem information cannot be identified through accurate matching, the computer equipment conducts fuzzy matching on the extracted medicine characteristic information and the medicine dictionary library, and therefore the medicine entities corresponding to the problem information are determined. The computer equipment identifies the medicine entity in the problem information by combining the precise matching mode and the fuzzy matching mode, so that the accuracy of medicine entity identification can be improved.
Step 110, invoking a thread scanning database to determine a drug level corresponding to the drug entity.
After identifying the drug entity corresponding to the problem information, the computer device calls the thread scanning database to determine the drug level corresponding to the drug entity. The database comprises common medicine specification information and medicine prescription information. For example, the drug prescription information may be the chinese national formulary collection (2010 edition). The computer device divides the common medicine instruction information and the medicine prescription information into a plurality of fields of indication, contraindication, adverse reaction, usage amount, cautionary matters, medicine interaction and the like according to the medicine grade and the intention type respectively, and stores the fields in a database. The drug grade includes the generic name grade, the generic name and dosage form grade, and the commodity grade.
And the computer equipment performs character matching on the medicine entity and the medicine grade in the database, and performs pinyin matching on the medicine entity and the corresponding medicine grade when the character matching is unsuccessful. And the computer equipment selects the medicine grade with the highest pinyin matching probability as the medicine grade corresponding to the medicine entity. In order to improve the matching accuracy of the medicine level, the computer equipment combines the character matching mode and the pinyin matching mode, and the accuracy of medicine entity identification can be further improved.
And 112, matching the medicine information in the database according to the medicine grade, outputting medicine reply information corresponding to the question information, and returning the medicine reply information to the terminal.
And after determining the medicine grade corresponding to the medicine entity, the computer equipment further performs medicine information matching in the database according to the medicine grade. Specifically, the computer device calls a thread scanning configuration file to obtain a matching sequence corresponding to the drug level. A plurality of medicine levels and corresponding matching sequences are recorded in the configuration file. The matching order may be the priority of the information matching content corresponding to the drug level. The information matching content can be medicine specification information or medicine prescription information which is commonly found in a database. The matching sequence may be to preferentially match common drug specification information in the database and then to match drug prescription information in the database. And the computer equipment performs character matching on the medicine entity and the database according to the matching sequence corresponding to the medicine grade, so as to obtain medicine reply information corresponding to the question information.
In this embodiment, the computer device performs word segmentation processing on the problem information by calling a thread to obtain a corresponding vector matrix, which is favorable for inputting into the trained classification model. The computer equipment calculates a plurality of context vectors corresponding to the vector matrix through the trained classification model, identifies the intention type corresponding to the problem information according to the context vectors, can filter interference information in the problem information, achieves word focusing processing on the vector matrix, and further improves the accuracy of intention identification. After identifying the drug entities in the problem information, the computer device calls the thread scanning database to determine the drug levels corresponding to the drug entities, so that the drug entities can be effectively classified, and drug granularity can be more finely divided. The computer equipment performs medicine information matching in the database through the medicine grade, outputs medicine reply information corresponding to the question information, can realize medicine information matching aiming at diversified question information, and further improves accuracy and comprehensiveness of medicine information matching.
In one embodiment, as shown in fig. 2, calculating a plurality of context vectors corresponding to the vector matrix, and identifying the type of intention corresponding to the question information according to the context vectors includes:
And 208, inputting the second extraction result into the pooling layer, and performing dimension reduction processing on the second extraction result.
And step 210, inputting the second extraction result subjected to the dimensionality reduction into a full-connection layer, classifying the second extraction result subjected to the dimensionality reduction to obtain a classification result, and outputting the classification result after weighting through an output layer.
And 212, selecting the intention type with the highest weight in the weighted and output classification results as the intention type corresponding to the problem information.
And after the computer equipment carries out word segmentation processing on the problem information, the word segmentation processed vector matrix is input into the trained classification model for intention recognition. The classification model may be a classification model obtained by the server training in advance with a large amount of sample data. The classification model may be a convolutional neural network model based on the attention layer. The classification model may include multiple network layers. For example, the classification model may include an input layer, an attention layer, a convolutional layer, a pooling layer, a fully-connected layer, and an output layer.
Specifically, the computer device extracts word vectors from the vector matrix through the input layer of the classification model, and inputs the extracted word vectors into the attention layer. The computer device further calculates a context vector and a weight corresponding to each word vector through an attention layer of the classification model, and generates first extraction information according to the plurality of context vectors and the weights. The computer device passes the first extraction result to the convolutional layer through an attention layer of the classification model as an input to the convolutional layer. The computer equipment extracts the context features corresponding to the context vectors through the convolution layers of the classification model and generates a second extraction result according to the context features corresponding to the context vectors. And inputting the second extraction result into the pooling layer by the convolution layer of the classification model, and performing dimension reduction processing on the second extraction result through the pooling layer. And the pooling layer of the classification model inputs the second extraction result after the dimensionality reduction treatment into the full-connection layer, and classifies the second extraction result after the dimensionality reduction treatment through the full-connection layer to obtain a classification result. Multiple intent types may be included in the classification results. And the full connection layer of the classification model inputs the classification result obtained after classification into the output layer, and the classification result is weighted and output through the output layer. And the computer equipment selects the intention type with the largest weight from the classification results as the intention type corresponding to the question information.
In this embodiment, the computer device extracts a plurality of word vectors in the vector matrix through the classification model, and calculates a context vector and a weight corresponding to each word vector. The interference information in the problem matrix can be filtered, and word focusing processing on the vector matrix is realized. The computer equipment can extract main context characteristics by performing dimension reduction processing on the context characteristics, and avoids the influence of redundant characteristics. The computer equipment classifies the context features after the dimension reduction processing, and performs weighted output on the classification result, so that the classification result can be normalized, and the accuracy of intention identification can be improved.
In one embodiment, identifying the drug entity in the question information in the drug dictionary repository includes: extracting medicine characteristic information in the question information; accurately matching the medicine characteristic information with a medicine dictionary library; and when the medicine entities cannot be identified by accurate matching, carrying out fuzzy matching on the medicine characteristic information and the medicine dictionary library, and determining the medicine entities corresponding to the problem information.
When the intent type is an intent type related to a drug. The computer device identifies the drug entity in the question information in a drug dictionary repository. The type of intent associated with a drug includes indications for the drug, adverse reactions to the drug, amounts of the drug used, precautions for the drug, interactions between drugs, and the like. The drug dictionary library includes the common names of drugs and the drug trade names. The drug trade name means identification information that can be accurately positioned to a corresponding commodity. Such as an approval code, a drug order code, etc.
And the computer equipment extracts the characteristics of the problem information and accurately matches the extracted medicine characteristic information with a medicine dictionary library. The exact match may be a character match. When the drug entities cannot be identified by accurate matching, the computer device may match the semantic keywords with keywords in a drug dictionary repository by extracting the semantic keywords in the drug feature information. And the computer equipment determines a medicine entity corresponding to the medicine information according to the keyword with the maximum similarity.
In this embodiment, the computer device is able to match more comprehensive drug entities by matching the drug information in the question information with a drug dictionary repository, since the drug dictionary repository is more comprehensive than the existing dictionary repository content. The computer equipment identifies the medicine entity in the problem information by combining the precise matching mode and the fuzzy matching mode, so that the accuracy of medicine entity identification can be improved.
In one embodiment, invoking the thread scan database and determining the drug level corresponding to the drug entity comprises: performing character matching on the medicine entity and the medicine grade in the database; when the character matching is not successful, the medicine entity is subjected to pinyin matching with the corresponding medicine level; and selecting the medicine grade with the highest pinyin matching probability as the medicine grade corresponding to the medicine entity.
After identifying the drug entities in the problem information, the computer device calls the thread scanning database to determine the drug levels corresponding to the drug entities. The drug grade includes the generic name grade, the generic name and dosage form grade, and the commodity grade. Specifically, the computer device determines whether a universal name for the drug, a dosage form for the drug, and commodity information are present in the drug entity. And if the drug entity only has the drug universal name, the drug level corresponding to the drug entity is the universal name level. If a generic name and a dosage form of the drug exist, the drug entity corresponds to a drug class that is generic name and dosage form class. And if the commodity information exists, the medicine grade corresponding to the medicine entity is the commodity grade. The commodity information includes manufacturer information, medicine commodity name, approval document number, medicine position code, etc.
For example, if only acarbose exists in the drug entity, and acarbose is a generic drug name, the drug level corresponding to the drug entity is a generic name level. The acarbose tablets exist in the medicine entities, the acarbose tablets comprise both the common names of the medicines and the dosage forms, and the medicine grades corresponding to the medicine entities are the common names and the dosage form grades. The pharmaceutical entity contains acarbose tablets produced by Bayer, which contain the common name, dosage form and manufacturer information of the medicine, and the medicine grade corresponding to the pharmaceutical entity is the commodity grade. The Lemna bicolor exists in the medicine entity and comprises the name of the medicine commodity, and the medicine grade corresponding to the medicine entity is the commodity grade.
In this embodiment, the computer device can determine the drug granularity corresponding to the drug entity by performing character matching and pinyin matching on the drug entity and the drug level, so as to improve the accuracy of drug information matching.
In one embodiment, matching drug information in the database according to drug class includes: calling a thread scanning configuration file, wherein a plurality of medicine grades and corresponding matching sequences are recorded in the configuration file; and performing character matching on the medicine entity and the database according to the matching sequence corresponding to the medicine level to obtain medicine reply information corresponding to the question information.
And after determining the medicine grade corresponding to the medicine entity, the computer equipment performs medicine information matching in the database according to the medicine grade. And the computer equipment acquires the matching sequence corresponding to the medicine grade by calling the thread scanning configuration file. The database includes the information of common medicine specification or medicine prescription information. The matching order may be to preferentially match common drug description information in the database and then to match drug prescription information in the database. And the computer equipment performs character matching on the medicine grade and the database according to the matching sequence, acquires corresponding medicine reply information and returns the medicine reply information to the terminal.
For example, when the medicine level is a generic name level, the matching sequence may be that the medicine prescription information is matched first to obtain the corresponding first prescription medicine information. And if the corresponding drug reply information does not exist in the drug prescription information, matching the drug entity with the drug instruction book information to obtain the corresponding instruction book drug information. And if the corresponding drug reply information does not exist in the drug specification information, searching the next-level drug universal name in the drug prescription information of the same drug type of the drug entity. And matching the searched medicine universal name of the next level with the corresponding medicine prescription information to obtain corresponding second prescription medicine information.
When the medicine grade is the common name and the dosage form grade, the matching sequence can be that the medicine specification information is preferentially matched, and the corresponding manufacturer information is obtained by a mode of multiple times of medicine information matching processing. And if the corresponding drug reply information does not exist in the drug specification information, matching with the drug prescription information, wherein the subsequent matching sequence is the same as the matching sequence of the universal name level drugs.
When the medicine level is the commodity level, the matching sequence of the medicine specification information can be the prior matching of the medicine specification information, and if the corresponding medicine reply information does not exist in the medicine specification information, the medicine information matching is carried out according to the matching sequence of the universal name and the dosage form level medicines.
In this embodiment, the computer device obtains a corresponding matching sequence according to the medicine level, performs character matching on the medicine entity and the database according to the matching sequence, obtains corresponding reply information, and can perform medicine information matching according to the medicine granularity in the question information, thereby further improving the accuracy of medicine information matching.
It should be understood that although the steps in the flowcharts of fig. 1 to 2 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 2-3 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performing the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 3, there is provided a medicine information matching processing apparatus including: a communication module 302, a word segmentation module 304, a first recognition module 306, a second recognition module 308, a determination module 310, and a matching module 312, wherein:
a communication module 302, configured to receive the question information uploaded by the terminal.
And the word segmentation module 304 is configured to invoke a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix.
The first identification module 306 is configured to input the vector matrix into the trained classification model, calculate a plurality of context vectors corresponding to the vector matrix, and identify an intention type corresponding to the problem information according to the context vectors.
A second identification module 308 to identify a drug entity in the question information in the drug dictionary repository when the intent type is an intent type associated with a drug.
The determining module 310 is configured to scan the database with a thread to determine a drug level corresponding to a drug entity.
And a matching module 312, configured to perform drug matching in the database according to the drug level, and output drug reply information corresponding to the question information.
In one embodiment, the first recognition module 306 is further configured to input the word segmentation sequence into a trained classification model, and extract a plurality of word vectors in a vector matrix through an input layer of the classification model; inputting a plurality of word vectors into an attention layer, calculating a context vector and a weight corresponding to each word vector, and generating a first extraction result; inputting the first extraction result into the convolution layer, extracting context characteristics corresponding to the context vector, and generating a second extraction result; inputting the second extraction result into the pooling layer, and performing dimensionality reduction on the second extraction result; inputting the second extraction result subjected to the dimensionality reduction treatment into a full-connection layer, classifying the second extraction result subjected to the dimensionality reduction treatment to obtain a classification result, and outputting the classification result after weighting through an output layer; and selecting the intention type with the maximum weight in the classification results output after weighting as the intention type corresponding to the problem information.
In one embodiment, the second identification module 308 is further configured to extract drug characteristic information from the question information; accurately matching the medicine characteristic information with a medicine dictionary library; and when the medicine entities cannot be identified by accurate matching, carrying out fuzzy matching on the medicine characteristic information and the medicine dictionary library, and determining the medicine entities corresponding to the problem information.
In one embodiment, the determination module 310 is further configured to character match the drug entity with the drug level in the database; when the character matching is not successful, the medicine entity is subjected to pinyin matching with the corresponding medicine level; and selecting the medicine grade with the highest pinyin matching probability as the medicine grade corresponding to the medicine entity.
In one embodiment, the matching module 312 is further configured to invoke a thread scan configuration file, where a plurality of drug levels and corresponding matching orders are recorded in the configuration file; and performing character matching on the medicine entity and the database according to the matching sequence corresponding to the medicine level to obtain medicine reply information corresponding to the question information.
For specific limitations of the medicine information matching processing device, reference may be made to the above limitations of the medicine information matching processing method, which are not described herein again. The modules in the medicine matching device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a computer device, the internal structure of which may be as shown in fig. 4. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing question information, drug response information and the like. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a drug information matching processing method.
Those skilled in the art will appreciate that the architecture shown in fig. 4 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, there is provided a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the drug information matching processing method provided in any one of the above-described embodiments.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A drug information matching processing method, the method comprising:
receiving problem information uploaded by a terminal;
calling a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix;
inputting the vector matrix into a trained classification model, calculating a plurality of context vectors corresponding to the vector matrix, and identifying an intention type corresponding to the problem information according to the context vectors;
identifying a drug entity in the question information in a drug dictionary repository when the intent type is an intent type associated with a drug;
calling a thread scanning database to determine the medicine grade corresponding to the medicine entity;
and matching the medicine information in the database according to the medicine grade, outputting medicine reply information corresponding to the question information, and returning the medicine reply information to the terminal.
2. The method of claim 1, wherein the classification model comprises a plurality of network layers, wherein the computing a plurality of context vectors corresponding to the vector matrix, and wherein the identifying the type of intent corresponding to the issue information from the context vectors comprises:
inputting a vector matrix into a trained classification model, and extracting a plurality of word vectors in the vector matrix through an input layer of the classification model;
inputting a plurality of word vectors into an attention layer, calculating a context vector and a weight corresponding to each word vector, and generating a first extraction result;
inputting the first extraction result into a convolution layer, extracting context characteristics corresponding to the context vector, and generating a second extraction result;
inputting the second extraction result into a pooling layer, and performing dimension reduction processing on the second extraction result;
inputting the second extraction result subjected to the dimensionality reduction treatment into a full-connection layer, classifying the second extraction result subjected to the dimensionality reduction treatment to obtain a classification result, and outputting the classification result after weighting through an output layer;
and selecting the intention type with the maximum weight in the classification results output after weighting as the intention type corresponding to the question information.
3. The method of claim 1, wherein identifying the drug entity in the question information in a drug dictionary repository comprises:
extracting medicine characteristic information in the question information;
accurately matching the medicine characteristic information with a medicine dictionary library;
and when the medicine entities cannot be identified by accurate matching, carrying out fuzzy matching on the medicine characteristic information and a medicine dictionary library, and determining the medicine entities corresponding to the problem information.
4. The method of claim 1, wherein the invoking thread scans a database and determining the drug level corresponding to the drug entity comprises:
performing character matching on the medicine entity and the medicine grade in a database;
when the character matching is not successful, carrying out pinyin matching on the medicine entity and the corresponding medicine level;
and selecting the medicine grade with the highest pinyin matching probability as the medicine grade corresponding to the medicine entity.
5. The method of claim 1, wherein matching drug information in a database according to the drug class comprises:
calling a thread scanning configuration file, wherein a plurality of medicine grades and corresponding matching sequences are recorded in the configuration file;
and performing character matching on the medicine entity and a database according to the matching sequence corresponding to the medicine grade to obtain medicine reply information corresponding to the question information.
6. A medicine information matching processing apparatus, characterized in that the apparatus comprises:
the communication module is used for receiving the problem information uploaded by the terminal;
the word segmentation module is used for calling a thread to perform word segmentation processing on the problem information to obtain a corresponding vector matrix;
the first identification module is used for inputting the vector matrix into the trained classification model and calculating a plurality of context vectors corresponding to the vector matrix; identifying an intention type corresponding to the problem information according to the context vector;
a second identification module for identifying a drug entity in the question information in a drug dictionary repository when the intent type is an intent type associated with a drug;
the determining module is used for determining the medicine grade corresponding to the medicine entity by utilizing a thread scanning database;
and the matching module is used for matching medicines in a database according to the medicine grades and outputting medicine reply information corresponding to the question information.
7. The apparatus of claim 6, wherein the first recognition module is further configured to input a word segmentation sequence into a trained classification model, and extract a plurality of word vectors in the vector matrix through an input layer of the classification model; inputting a plurality of word vectors into an attention layer, calculating a context vector and a weight corresponding to each word vector, and generating a first extraction result; inputting the first extraction result into a convolution layer, extracting context characteristics corresponding to the context vector, and generating a second extraction result; inputting the second extraction result into a pooling layer, and performing dimension reduction processing on the second extraction result; inputting the second extraction result subjected to the dimensionality reduction treatment into a full-connection layer, classifying the second extraction result subjected to the dimensionality reduction treatment to obtain a classification result, and outputting the classification result after weighting through an output layer; and selecting the intention type with the maximum weight in the classification results output after weighting as the intention type corresponding to the question information.
8. The apparatus of claim 6, wherein the second identification module is further configured to extract drug characteristic information from the question information; accurately matching the medicine characteristic information with a medicine dictionary library; and when the medicine entities cannot be identified by accurate matching, carrying out fuzzy matching on the medicine characteristic information and a medicine dictionary library, and determining the medicine entities corresponding to the problem information.
9. A computer device comprising a memory and a processor, the memory storing a computer program operable on the processor, wherein the processor implements the steps of the method of any one of claims 1 to 5 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910969896.7A CN110781677B (en) | 2019-10-12 | 2019-10-12 | Medicine information matching processing method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910969896.7A CN110781677B (en) | 2019-10-12 | 2019-10-12 | Medicine information matching processing method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110781677A true CN110781677A (en) | 2020-02-11 |
| CN110781677B CN110781677B (en) | 2023-02-07 |
Family
ID=69385264
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910969896.7A Active CN110781677B (en) | 2019-10-12 | 2019-10-12 | Medicine information matching processing method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110781677B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111985224A (en) * | 2020-08-31 | 2020-11-24 | 平安医疗健康管理股份有限公司 | Medication instruction text processing method, device, equipment and storage medium |
| CN113130038A (en) * | 2021-04-30 | 2021-07-16 | 康键信息技术（深圳）有限公司 | Medicine data matching method, device, equipment and storage medium |
| CN113223657A (en) * | 2021-06-01 | 2021-08-06 | 联仁健康医疗大数据科技股份有限公司 | Medicine information processing method and device, electronic equipment and storage medium |
| CN113742454A (en) * | 2021-09-09 | 2021-12-03 | 平安科技（深圳）有限公司 | Response corpus generation method based on artificial intelligence and related equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102810135A (en) * | 2012-09-17 | 2012-12-05 | 顾泰来 | Auxiliary processing system for drug prescription |
| CN106776711A (en) * | 2016-11-14 | 2017-05-31 | 浙江大学 | A kind of Chinese medical knowledge mapping construction method based on deep learning |
| CN108986908A (en) * | 2018-05-31 | 2018-12-11 | 平安医疗科技有限公司 | Interrogation data processing method, device, computer equipment and storage medium |
| CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技（深圳）有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
| US20190095432A1 (en) * | 2017-09-26 | 2019-03-28 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for building text classification model, and text classification method and apparatus |
| CN110176315A (en) * | 2019-06-05 | 2019-08-27 | 京东方科技集团股份有限公司 | Medical answering method and system, electronic equipment, computer-readable medium |
-
2019
- 2019-10-12 CN CN201910969896.7A patent/CN110781677B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102810135A (en) * | 2012-09-17 | 2012-12-05 | 顾泰来 | Auxiliary processing system for drug prescription |
| CN106776711A (en) * | 2016-11-14 | 2017-05-31 | 浙江大学 | A kind of Chinese medical knowledge mapping construction method based on deep learning |
| US20190095432A1 (en) * | 2017-09-26 | 2019-03-28 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for building text classification model, and text classification method and apparatus |
| CN108986908A (en) * | 2018-05-31 | 2018-12-11 | 平安医疗科技有限公司 | Interrogation data processing method, device, computer equipment and storage medium |
| CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技（深圳）有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
| CN110176315A (en) * | 2019-06-05 | 2019-08-27 | 京东方科技集团股份有限公司 | Medical answering method and system, electronic equipment, computer-readable medium |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111985224A (en) * | 2020-08-31 | 2020-11-24 | 平安医疗健康管理股份有限公司 | Medication instruction text processing method, device, equipment and storage medium |
| CN113130038A (en) * | 2021-04-30 | 2021-07-16 | 康键信息技术（深圳）有限公司 | Medicine data matching method, device, equipment and storage medium |
| CN113223657A (en) * | 2021-06-01 | 2021-08-06 | 联仁健康医疗大数据科技股份有限公司 | Medicine information processing method and device, electronic equipment and storage medium |
| CN113742454A (en) * | 2021-09-09 | 2021-12-03 | 平安科技（深圳）有限公司 | Response corpus generation method based on artificial intelligence and related equipment |
| CN113742454B (en) * | 2021-09-09 | 2023-07-21 | 平安科技（深圳）有限公司 | Response corpus generation method based on artificial intelligence and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110781677B (en) | 2023-02-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110781677B (en) | Medicine information matching processing method and device, computer equipment and storage medium | |
| CN110021439B (en) | Medical data classification method and device based on machine learning and computer equipment | |
| CN108986908B (en) | Method and device for processing inquiry data, computer equipment and storage medium | |
| CN109729383B (en) | Double-recording video quality detection method and device, computer equipment and storage medium | |
| CN109408631B (en) | Medicine data processing method, device, computer equipment and storage medium | |
| WO2021042503A1 (en) | Information classification extraction method, apparatus, computer device and storage medium | |
| CN108595695B (en) | Data processing method, data processing device, computer equipment and storage medium | |
| CN110909137A (en) | Information pushing method and device based on man-machine interaction and computer equipment | |
| CN111160017A (en) | Keyword extraction method, phonetics scoring method and phonetics recommendation method | |
| CN110457302B (en) | Intelligent structured data cleaning method | |
| CN110008250B (en) | Social security data processing method and device based on data mining and computer equipment | |
| US11164564B2 (en) | Augmented intent and entity extraction using pattern recognition interstitial regular expressions | |
| CN110674684A (en) | Micro-expression classification model generation method, micro-expression classification model generation device, micro-expression classification model image recognition method, micro-expression classification model image recognition device, micro-expression classification model image recognition equipment and micro-expression classification model image recognition medium | |
| CN112488896B (en) | Emergency plan generation method and device, computer equipment and storage medium | |
| CN110674319A (en) | Label determination method and device, computer equipment and storage medium | |
| CN110334179B (en) | Question-answer processing method, device, computer equipment and storage medium | |
| CN112015900A (en) | Medical attribute knowledge graph construction method, device, equipment and medium | |
| CN110377558A (en) | Document searching method, device, computer equipment and storage medium | |
| CN111445968A (en) | Electronic medical record query method and device, computer equipment and storage medium | |
| CN111180086B (en) | Data matching method, device, computer equipment and storage medium | |
| CN110362798B (en) | Method, apparatus, computer device and storage medium for judging information retrieval analysis | |
| WO2021043076A1 (en) | Method and apparatus for processing network data to be published, and computer device and storage medium | |
| CN112016318A (en) | Triage information recommendation method, device, equipment and medium based on interpretation model | |
| CN111783471A (en) | Semantic recognition method, device, equipment and storage medium of natural language | |
| CN109033427B (en) | Stock screening method and device, computer equipment and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40023191 Country of ref document: HK |
|
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220520 Address after: 518048 China Aviation Center 2901, No. 1018, Huafu Road, Huahang community, Huaqiang North Street, Futian District, Shenzhen, Guangdong Province Applicant after: Shenzhen Ping An medical and Health Technology Service Co.,Ltd. Address before: Room 12G, Area H, 666 Beijing East Road, Huangpu District, Shanghai 200001 Applicant before: PING AN MEDICAL AND HEALTHCARE MANAGEMENT Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |