CN118072973A - Intelligent inquiry method and system based on medical knowledge base - Google Patents
Intelligent inquiry method and system based on medical knowledge base Download PDFInfo
- Publication number
- CN118072973A CN118072973A CN202410445186.5A CN202410445186A CN118072973A CN 118072973 A CN118072973 A CN 118072973A CN 202410445186 A CN202410445186 A CN 202410445186A CN 118072973 A CN118072973 A CN 118072973A
- Authority
- CN
- China
- Prior art keywords
- inquiry
- symptom
- intelligent
- result
- data
- 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
- 238000000034 method Methods 0.000 title claims abstract description 26
- 208000024891 symptom Diseases 0.000 claims abstract description 141
- 230000007246 mechanism Effects 0.000 claims abstract description 29
- 238000000605 extraction Methods 0.000 claims description 15
- 201000010099 disease Diseases 0.000 claims description 10
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 10
- 238000005457 optimization Methods 0.000 claims description 8
- 230000011218 segmentation Effects 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 208000010668 atopic eczema Diseases 0.000 description 7
- 208000003251 Pruritus Diseases 0.000 description 6
- 230000007803 itching Effects 0.000 description 5
- 210000000689 upper leg Anatomy 0.000 description 5
- 201000004624 Dermatitis Diseases 0.000 description 4
- 230000000172 allergic effect Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 208000010201 Exanthema Diseases 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 201000005884 exanthem Diseases 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 206010037844 rash Diseases 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 206010020751 Hypersensitivity Diseases 0.000 description 1
- 208000024780 Urticaria Diseases 0.000 description 1
- 230000007815 allergy Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000035807 sensation Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The application discloses an intelligent inquiry method and system based on a medical knowledge base, wherein the method comprises the steps of acquiring inquiry data input by a user; inputting the extracted inquiry data into an intelligent medical inquiry model, extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism, and matching the extracted symptom characteristics with standardized symptoms by adopting a symptom alignment module to obtain an initial inquiry result; carrying out detail questioning to obtain symptom detail information; and obtaining a final inquiry result according to the symptom detail information and the initial inquiry result. The method and the system have the advantages that the data are marked and the symptoms are analyzed, the multi-head self-attention mechanism and the alignment mechanism are adopted for carrying out symptom identification and matching on the data described by the patient, then the specific questioning is carried out according to the preliminary predicted questioning result, the detail information is obtained, and finally the questioning result is obtained.
Description
Technical Field
The invention relates to the technical field of intelligent questioning and answering, in particular to an intelligent questioning and diagnosing method and system based on a medical knowledge base.
Background
With the rapid development of artificial intelligence technology, the use and effect of artificial intelligence have shown explosive growth trend, and artificial intelligence technology can be used for more and more tasks. The medical inquiry is a work with the characteristics of professional, diversity, wide range, large demand and the like, and because people have different main descriptions of the physical symptoms of the people, the accurate and complete extraction of the related symptoms from the main descriptions of the patients is a challenging task. At present, a technology capable of rapidly and accurately analyzing key symptoms from patient illness complaints is urgently needed, the key symptoms are used for carrying out expansion questioning to obtain specific information of a patient, and the medical knowledge base is used for confirming diseases, so that the accuracy of intelligent medical consultation is greatly improved.
There are two-stage, rule-based and information-based extraction methods for medical symptom extraction algorithms, among which rule-based algorithms and information-based extraction algorithms are widely used. However, these algorithms have certain drawbacks in processing more orally-spoken query data, and rule-based algorithms are less robust to missing symptoms for orally-spoken data. The algorithm based on information extraction lacks the ability to normalize the symptoms of the spoken language, which is inconvenient for the next inquiry. The invention discloses a method for improving intelligent inquiry by using a model with the capability of extracting normalized symptoms.
Disclosure of Invention
It is therefore an object of the present invention to provide a method that enables accurate identification of symptoms in inquiry data.
In order to achieve the above purpose, the invention provides an intelligent inquiry method based on a medical knowledge base, which comprises the following steps:
S1, acquiring inquiry data input by a user;
s2, inputting the extracted inquiry data into an intelligent medical inquiry model, extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism, and matching the extracted symptom characteristics with standardized symptoms by adopting a symptom alignment module to obtain an initial inquiry result;
S3, carrying out detail questioning according to the initial questioning result to obtain symptom detail information;
s4, obtaining a final inquiry result according to the symptom detail information and the initial inquiry result.
Further preferably, in the intelligent medical inquiry model, when symptom feature extraction is performed on inquiry data, training is performed according to the following steps:
S201, forming standardized symptoms from common symptoms in inquiry data;
s202, marking inquiry data according to formed standardized symptoms;
S203, inputting the marked inquiry data into an intelligent medical inquiry model, mapping the input data into a plurality of Gao Weizi spaces through multidimensional transformation by adopting a multi-head self-attention mechanism, so that each head of the multi-head self-attention mechanism adopts symptom areas in different ranges, and extracting symptom features in different scales;
s204, repeating the steps S201-S203, inputting a plurality of pieces of inquiry data into the intelligent medical inquiry model, and repeating training to enable the symptom area of the current self-attention head of the intelligent medical inquiry model to be selected to be the optimal range.
Further preferably, in S201, the step of forming standardized symptoms from common symptoms in the query data includes performing sub-word segmentation on the query data, forming sub-word sequences with granularity between words and single words from the segmented query data, marking according to the standardized symptoms, generating a word list, and serializing the marked sub-word sequences according to the word list to form a text sequence for inputting an intelligent medical query model.
Further preferably, the multi-headed self-attention mechanism is expressed by the following formula:
Wherein S (m-n) represents the attention modifying function of the subword m and the subword n, Similarity between the subword m and the subword n; /(I)Similarity between the subword m and the subword p; s (m-p) represents the attention modifying function of the subword m and the subword p.
Further preferably, a layer of class mask is added to the sub word sequence obtained after segmentation, and the symptom class of the current sub word is predicted, wherein the mask comprises: location, manifestation, disease.
Further preferably, the symptom alignment module is composed of a double-layer multi-head self-attention mechanism network, and further comprises the following steps:
And taking the text sequence, a category mask obtained by predicting the text sequence and symptom characteristics extracted by a multi-head self-attention mechanism as input parameters of a symptom alignment module, and calculating according to the following formula to obtain a prediction result:
wherein, R is the predicted result, Is a text sequence,/>For symptom features extracted by multi-headed self-attention mechanism,/>Is a category mask;
And converting the obtained prediction result into symptom description by a decoder to obtain an initial inquiry result.
The invention also provides an intelligent inquiry system based on the medical knowledge base, which comprises a data acquisition module, an intelligent medical inquiry model and a result output module;
The data acquisition module is used for acquiring inquiry data input by a user;
The intelligent medical inquiry model comprises a symptom gradient optimization module (Som) and a symptom alignment module, wherein the symptom gradient optimization module (Som) is used for taking extracted inquiry data as input and extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism; the symptom alignment module is used for matching the extracted symptom characteristics with the standardized symptoms to obtain an initial inquiry result;
The result output module is used for carrying out detail questioning according to the initial questioning result to obtain symptom detail information; and obtaining a final inquiry result according to the symptom detail information and the initial inquiry result.
The present invention also provides an electronic device including: a memory storing computer program instructions; and a processor, which when executed by the processor, implements the steps of the medical knowledge base-based intelligent interrogation method described above.
The present invention also provides a computer readable storage medium for storing instructions that, when executed on a computer, cause the computer to perform the steps of the intelligent interrogation method described above based on a medical knowledge base.
Compared with the prior art, the intelligent inquiry method and system based on the medical knowledge base have the following advantages:
The application establishes a medical knowledge base of common symptoms and diseases, carries out symptom identification and matching on data described by patients by carrying out data labeling and symptom analysis and adopting a multi-head self-attention mechanism and an alignment mechanism, carries out targeted questioning according to a preliminary predicted questioning result to obtain detailed information, and finally obtains a questioning result.
Drawings
Fig. 1 is a schematic flow chart of an intelligent inquiry method based on a medical knowledge base.
Fig. 2 is a schematic diagram of the overall network structure of the intelligent inquiry system based on the medical knowledge base.
FIG. 3 is a simulated view of the symptom extraction process according to the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and the detailed description.
As shown in fig. 1, an intelligent inquiry method based on a medical knowledge base according to an embodiment of the present invention includes the following steps:
S1, acquiring inquiry data input by a user;
s2, inputting the extracted inquiry data into an intelligent medical inquiry model, extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism, and matching the extracted symptom characteristics with standardized symptoms by adopting a symptom alignment module to obtain an initial inquiry result;
S3, carrying out detail questioning according to the initial questioning result to obtain symptom detail information;
s4, obtaining a final inquiry result according to the symptom detail information and the initial inquiry result.
Further preferably, in the intelligent medical inquiry model, when symptom feature extraction is performed on inquiry data, training is performed according to the following steps:
S201, forming standardized symptoms from common symptoms in inquiry data; further, in S201, the step of forming standardized symptoms from common symptoms in the query data includes performing sub-word segmentation on the query data, forming sub-word sequences with granularity between words and single words from the segmented query data, marking according to the standardized symptoms, generating word lists, and serializing the marked sub-word sequences according to the word lists to form text sequences to be used as input of an intelligent medical query model.
The common parts, symptoms and disease names in the medical inquiry are standardized, the standardized symptoms are obtained, and the standardized symptoms are added into a medical knowledge base. Some standardized symptom examples for the location, symptoms and diseases are shown in table 1.
Table 1 medical knowledge base standardized symptom example
S202, marking inquiry data according to formed standardized symptoms; in the application, 10w pieces of high-quality clinical inquiry data are screened, parts, symptoms and diseases appearing in the symptom description of the patient are marked, and the medical knowledge base in the step 1 is used as a standard for marking.
Examples of the labeling results are shown in Table 2.
Table 2 training data annotation examples
S203, inputting the marked inquiry data into an intelligent medical inquiry model, mapping the input data into a plurality of Gao Weizi spaces through multidimensional transformation by adopting a multi-head self-attention mechanism, so that each head of the multi-head self-attention mechanism adopts symptom areas in different ranges, and extracting symptom features in different scales; further, the multi-headed self-attention mechanism is expressed by the following formula:
Wherein S (m-n) represents the attention modifying function of the subword m and the subword n, Similarity between the subword m and the subword n; /(I)Similarity between the subword m and the subword p; s (m-p) represents the attention modifying function of the subword m and the subword p.
Attention modifying function:
r is the radius of the extent of the symptomatic region, Representing the distance between two subwords.
S204, repeating the steps S201-S203, inputting a plurality of pieces of inquiry data into the intelligent medical inquiry model, and repeating training to enable the symptom area of the current self-attention head of the intelligent medical inquiry model to be selected to be the optimal range.
Further preferably, a layer of class mask is added to the sub word sequence obtained after segmentation, and the symptom class of the current sub word is predicted, wherein the mask comprises: location, manifestation, disease.
Further preferably, the symptom alignment module is composed of a double-layer multi-head self-attention mechanism network, and further comprises the following steps:
And taking the text sequence, a category mask obtained by predicting the text sequence and symptom characteristics extracted by a multi-head self-attention mechanism as input parameters of a symptom alignment module, and calculating according to the following formula to obtain a prediction result:
wherein, R is the predicted result, Is a text sequence,/>For symptom features extracted by multi-headed self-attention mechanism,/>Is a category mask;
And converting the obtained prediction result into symptom description by a decoder to obtain an initial inquiry result.
As shown in fig. 2, the invention further provides an intelligent inquiry system based on the medical knowledge base, which comprises a data acquisition module, an intelligent medical inquiry model and a result output module;
The data acquisition module is used for acquiring inquiry data input by a user;
The intelligent medical inquiry model comprises a symptom gradient optimization module (Som) and a symptom alignment module, wherein the symptom gradient optimization module (Som) is used for taking extracted inquiry data as input and extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism; the symptom alignment module is used for matching the extracted symptom characteristics with the standardized symptoms to obtain an initial inquiry result;
The result output module is used for carrying out detail questioning according to the initial questioning result to obtain symptom detail information; and obtaining a final inquiry result according to the symptom detail information and the initial inquiry result.
The present invention also provides an electronic device including: a memory storing computer program instructions; and a processor, which when executed by the processor, implements the steps of the medical knowledge base-based intelligent interrogation method described above.
The present invention also provides a computer readable storage medium for storing instructions that, when executed on a computer, cause the computer to perform the steps of the intelligent interrogation method described above based on a medical knowledge base.
The following examples are given: as shown in fig. 3, for a patient's description of his own condition: "I have recently itched at the root of the thigh, have a reddish spot, is not eczema? The Chinese segmentation tool is used for segmentation to obtain a subword sequence, and a symptom interval is obtained through symptom gradient optimization: the symptoms of the symptom interval are predicted by the thigh root pruritus, the reddish spot and the eczema, and the predicted part is: "thigh root"; the expression is as follows: "itching" and "reddish dots"; disease: eczema. The results of "thigh", "itching" and "rash" are obtained by modeling the symptoms, wherein "eczema" is a guess of the patient and cannot be used as a diagnostic basis, so that the disease information is abandoned.
In this step, the application proposes two innovation points of a symptom gradient optimization module (Som) and a progressive symptom extraction module (Pse), and in order to verify the improvement of the symptom extraction effect by the innovation points, we have conducted a comparison experiment with a traditional multi-head self-attention model (transducer):
And step 4, analyzing the symptom result extracted in the step 3, asking questions about the related details of the symptoms extracted in the step 3, and enriching symptom information. For example: in step 3, the symptoms of "thigh", "itching" and "rash" are extracted, and we can ask further:
(1) How does the degree of itching?
(2) The size and morphology of the red dots? Is skin morphology altered?
(3) Is there a history of allergies? Is it exposed to allergic substances?
And 5, performing preliminary diagnosis on the illness state through a medical knowledge base according to the symptom information obtained in the step 3 and the step 4, and obtaining a diagnosis result. For example: in the step 4, through questioning, the following steps are obtained: (1) severe itching sensation (2) irregular-shaped wind clusters (3) with different sizes have allergic history and allergic substance contact history, and the urticaria is primarily estimated.
In summary, the invention provides an intelligent inquiry method and system based on a medical knowledge base, which greatly improves the accuracy of symptom extraction by means of an independently developed symptom extraction model, achieves the purpose of optimizing a primary diagnosis result, and has the advantages of symptom extraction accuracy of 99% and symptom extraction recall rate of 97%.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. While still being apparent from variations or modifications that may be made by those skilled in the art are within the scope of the invention.
Claims (9)
1. An intelligent inquiry method based on a medical knowledge base is characterized by comprising the following steps:
S1, acquiring inquiry data input by a user;
s2, inputting the extracted inquiry data into an intelligent medical inquiry model, extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism, and matching the extracted symptom characteristics with standardized symptoms by adopting a symptom alignment module to obtain an initial inquiry result;
S3, carrying out detail questioning according to the initial questioning result to obtain symptom detail information;
s4, obtaining a final inquiry result according to the symptom detail information and the initial inquiry result.
2. The intelligent inquiry method based on the medical knowledge base according to claim 1, wherein in the intelligent medical inquiry model, when symptom feature extraction is performed on inquiry data, training is performed according to the following steps:
S201, forming standardized symptoms from common symptoms in inquiry data;
s202, marking inquiry data according to formed standardized symptoms;
S203, inputting the marked inquiry data into an intelligent medical inquiry model, mapping the input data into a plurality of Gao Weizi spaces through multidimensional transformation by adopting a multi-head self-attention mechanism, so that each head of the multi-head self-attention mechanism adopts symptom areas in different ranges, and extracting symptom features in different scales;
s204, repeating the steps S201-S203, inputting a plurality of pieces of inquiry data into the intelligent medical inquiry model, and repeating training to enable the symptom area of the current self-attention head of the intelligent medical inquiry model to be selected to be the optimal range.
3. The intelligent inquiry method based on the medical knowledge base according to claim 2, wherein in S201, the step of forming standardized symptoms from common symptoms in inquiry data includes performing sub-word segmentation on the inquiry data, forming sub-word sequences with granularity between words and single words from the segmented inquiry data, marking according to the standardized symptoms, generating word lists, and serializing the marked sub-word sequences according to the word lists, and forming text sequences to serve as input of an intelligent medical inquiry model.
4. The medical knowledge base based intelligent consultation method of claim 3 wherein the multi-headed self-attention mechanism is expressed by the following formula:
Wherein S (m-n) represents the attention modifying function of the subword m and the subword n, Similarity between the subword m and the subword n; /(I)Similarity between the subword m and the subword p; s (m-p) represents the attention modifying function of the subword m and the subword p.
5. The intelligent medical knowledge base based consultation method according to claim 4, characterized in that,
The method further comprises the steps of adding a layer of category mask to the sub word sequence obtained after segmentation, and predicting the symptom category of the current sub word, wherein the mask comprises the following steps: location, manifestation, disease.
6. The intelligent inquiry method based on medical knowledge base according to claim 5, wherein the symptom alignment module is composed of a double-layer multi-head self-attention mechanism network, and the steps of matching the extracted symptom characteristics with standardized symptoms to obtain an initial inquiry result are further included:
And taking the text sequence, a category mask obtained by predicting the text sequence and symptom characteristics extracted by a multi-head self-attention mechanism as input parameters of a symptom alignment module, and calculating according to the following formula to obtain a prediction result:
wherein, R is the predicted result, Is a text sequence,/>For symptom features extracted by multi-headed self-attention mechanism,/>Is a category mask;
And converting the obtained prediction result into symptom description by a decoder to obtain an initial inquiry result.
7. An intelligent inquiry system based on a medical knowledge base comprises a data acquisition module, an intelligent medical inquiry model and a result output module;
The data acquisition module is used for acquiring inquiry data input by a user;
The intelligent medical inquiry model comprises a symptom gradient optimization module and a symptom alignment module, wherein the symptom gradient optimization module is used for taking extracted inquiry data as input and extracting symptom characteristics of the inquiry data by adopting a multi-head self-attention mechanism; adopting a symptom alignment module to match the extracted symptom characteristics with standardized symptoms to obtain an initial inquiry result;
The result output module is used for carrying out detail questioning according to the initial questioning result to obtain symptom detail information; and obtaining a final inquiry result according to the symptom detail information and the initial inquiry result.
8. An electronic device, comprising: a memory storing computer program instructions;
A processor, which when executed by the processor, implements the steps of the medical knowledge base based intelligent interrogation method of any of claims 1 to 6.
9. A computer readable storage medium for storing instructions which, when executed on a computer, cause the computer to perform the steps of the medical knowledge base based intelligent interrogation method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410445186.5A CN118072973B (en) | 2024-04-15 | Intelligent inquiry method and system based on medical knowledge base |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410445186.5A CN118072973B (en) | 2024-04-15 | Intelligent inquiry method and system based on medical knowledge base |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118072973A true CN118072973A (en) | 2024-05-24 |
| CN118072973B CN118072973B (en) | 2024-06-28 |
Family
ID=
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717336A (en) * | 2019-09-23 | 2020-01-21 | 华南理工大学 | Scene text recognition method based on semantic relevance prediction and attention decoding |
| CN110889404A (en) * | 2019-11-21 | 2020-03-17 | 山东浪潮人工智能研究院有限公司 | Irregular text recognition system and method based on correction network |
| CN111709233A (en) * | 2020-05-27 | 2020-09-25 | 西安交通大学 | Intelligent diagnosis guiding method and system based on multi-attention convolutional neural network |
| CN113033249A (en) * | 2019-12-09 | 2021-06-25 | 中兴通讯股份有限公司 | Character recognition method, device, terminal and computer storage medium thereof |
| WO2022085533A1 (en) * | 2020-10-20 | 2022-04-28 | 国立研究開発法人情報通信研究機構 | Text classifier for answer identification, background knowledge expression generator, training device therefor, and computer program |
| CN114579723A (en) * | 2022-03-02 | 2022-06-03 | 平安科技（深圳）有限公司 | Interrogation method and apparatus, electronic device, and storage medium |
| CN114613489A (en) * | 2022-02-21 | 2022-06-10 | 哈尔滨工业大学（深圳） | Deep reinforcement learning-based inquiry method and system |
| CN117332784A (en) * | 2023-09-28 | 2024-01-02 | 卓世科技(海南)有限公司 | Intelligent knowledge enhancement method based on hierarchical graph attention and dynamic meta-learning |
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110717336A (en) * | 2019-09-23 | 2020-01-21 | 华南理工大学 | Scene text recognition method based on semantic relevance prediction and attention decoding |
| CN110889404A (en) * | 2019-11-21 | 2020-03-17 | 山东浪潮人工智能研究院有限公司 | Irregular text recognition system and method based on correction network |
| CN113033249A (en) * | 2019-12-09 | 2021-06-25 | 中兴通讯股份有限公司 | Character recognition method, device, terminal and computer storage medium thereof |
| CN111709233A (en) * | 2020-05-27 | 2020-09-25 | 西安交通大学 | Intelligent diagnosis guiding method and system based on multi-attention convolutional neural network |
| WO2022085533A1 (en) * | 2020-10-20 | 2022-04-28 | 国立研究開発法人情報通信研究機構 | Text classifier for answer identification, background knowledge expression generator, training device therefor, and computer program |
| US20230385558A1 (en) * | 2020-10-20 | 2023-11-30 | National Institute Of Information And Communications Technology | Text classifier for answer identification, background knowledge representation generator and training device therefor, and computer program |
| CN114613489A (en) * | 2022-02-21 | 2022-06-10 | 哈尔滨工业大学（深圳） | Deep reinforcement learning-based inquiry method and system |
| CN114579723A (en) * | 2022-03-02 | 2022-06-03 | 平安科技（深圳）有限公司 | Interrogation method and apparatus, electronic device, and storage medium |
| CN117332784A (en) * | 2023-09-28 | 2024-01-02 | 卓世科技(海南)有限公司 | Intelligent knowledge enhancement method based on hierarchical graph attention and dynamic meta-learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109086805B (en) | Clustering method based on deep neural network and pairwise constraints | |
| WO2020237869A1 (en) | Question intention recognition method and apparatus, computer device, and storage medium | |
| CN109635122A (en) | Intelligent disease inquiry method, apparatus, equipment and storage medium | |
| US11468989B2 (en) | Machine-aided dialog system and medical condition inquiry apparatus and method | |
| CN111949759A (en) | Method and system for retrieving medical record text similarity and computer equipment | |
| CN106407443A (en) | Structured medical data generation method and device | |
| CN114334070A (en) | Auxiliary prescription system based on medical big data | |
| WO2022252378A1 (en) | Method and apparatus for generating medical named entity recognition model, and computer device | |
| CN107887025A (en) | A kind of medical brain system | |
| WO2022227203A1 (en) | Triage method, apparatus and device based on dialogue representation, and storage medium | |
| CN111460175A (en) | SNOMED-CT-based medical noun dictionary construction and expansion method | |
| CN114358001A (en) | Method for standardizing diagnosis result, and related device, equipment and storage medium thereof | |
| CN116975218A (en) | Text processing method, device, computer equipment and storage medium | |
| CN117877660A (en) | Medical report acquisition method and system based on voice recognition | |
| CN111627561B (en) | Standard symptom extraction method, device, electronic equipment and storage medium | |
| CN118072973B (en) | Intelligent inquiry method and system based on medical knowledge base | |
| CN116468043A (en) | Nested entity identification method, device, equipment and storage medium | |
| CN118072973A (en) | Intelligent inquiry method and system based on medical knowledge base | |
| CN113241198B (en) | User data processing method, device, equipment and storage medium | |
| CN116258685A (en) | Multi-organ segmentation method and device for simultaneous extraction and fusion of global and local features | |
| CN113643825B (en) | Medical case knowledge base construction method and system based on clinical key feature information | |
| CN114582449A (en) | Electronic medical record named entity standardization method and system based on XLNet-BiGRU-CRF model | |
| CN113761899A (en) | Medical text generation method, device, equipment and storage medium | |
| CN105989094B (en) | Image retrieval method based on middle layer expression of hidden layer semantics | |
| CN115223001A (en) | Medical image identification method based on transfer learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |