CN113539520A - Method, device, computer equipment and storage medium for implementing inquiry session - Google Patents
Method, device, computer equipment and storage medium for implementing inquiry session Download PDFInfo
- Publication number
- CN113539520A CN113539520A CN202110838720.5A CN202110838720A CN113539520A CN 113539520 A CN113539520 A CN 113539520A CN 202110838720 A CN202110838720 A CN 202110838720A CN 113539520 A CN113539520 A CN 113539520A
- Authority
- CN
- China
- Prior art keywords
- corpus
- target
- syndrome
- symptom
- information
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 208000024891 symptom Diseases 0.000 claims abstract description 163
- 208000011580 syndromic disease Diseases 0.000 claims abstract description 100
- 238000012216 screening Methods 0.000 claims abstract description 61
- 238000012549 training Methods 0.000 claims description 32
- 230000008506 pathogenesis Effects 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 17
- 238000013507 mapping Methods 0.000 claims description 12
- 238000005516 engineering process Methods 0.000 claims description 5
- 230000009193 crawling Effects 0.000 claims description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 abstract description 3
- 239000013598 vector Substances 0.000 description 19
- 239000003814 drug Substances 0.000 description 9
- 230000006870 function Effects 0.000 description 9
- 238000003748 differential diagnosis Methods 0.000 description 7
- 206010011224 Cough Diseases 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 201000010099 disease Diseases 0.000 description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 4
- 238000001914 filtration Methods 0.000 description 3
- 230000014509 gene expression Effects 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 206010019233 Headaches Diseases 0.000 description 2
- 206010062717 Increased upper airway secretion Diseases 0.000 description 2
- 206010028748 Nasal obstruction Diseases 0.000 description 2
- 208000002193 Pain Diseases 0.000 description 2
- 238000007664 blowing Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 231100000869 headache Toxicity 0.000 description 2
- 230000036039 immunity Effects 0.000 description 2
- 239000000411 inducer Substances 0.000 description 2
- 208000026435 phlegm Diseases 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 206010039101 Rhinorrhoea Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 208000010753 nasal discharge Diseases 0.000 description 1
- 201000009240 nasopharyngitis Diseases 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- 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
- G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
-
- 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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Animal Behavior & Ethology (AREA)
- Probability & Statistics with Applications (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The application is applicable to the field of artificial intelligence, and provides a method, a device, computer equipment and a storage medium for realizing an inquiry session, wherein the method for realizing the inquiry session comprises the following steps: according to symptom information sent by a user through a terminal, determining syndrome information from a preset stored symptom data table; wherein the syndrome information is used to characterize a category of the condition associated with the symptom information; candidate corpora screening is carried out according to the syndrome information by utilizing a corpus screening model trained in advance to obtain a target corpus; and sending the target question generated based on the target corpus to a terminal for loading. By applying the technical scheme provided by the embodiment of the application, the problems of low inquiry precision and poor user use feeling in the conventional inquiry session process are solved, and the effects of good user use feeling and improvement of the precision of the inquiry session are brought.
Description
Technical Field
The invention belongs to the field of artificial intelligence, and particularly relates to a method, a device, computer equipment and a storage medium for realizing an inquiry session.
Background
With the development of artificial intelligence technology and natural language processing technology, the intelligent dialogue system has a wide development prospect. The current intelligent dialogue system has wide application in traditional Chinese medicine inquiry. However, the existing intelligent dialogue system lacks attention to symptoms in inquiry, and cannot make correct judgment on similar symptoms so as to obtain accurate questions. In the process of implementing an inquiry session, the intelligent dialog system cannot perform inquiry related to symptoms according to the symptoms input by the user, and the inquired questions are not understood by the user using professional medical terms. Therefore, the problems of low inquiry precision and poor use feeling of users exist in the existing inquiry session process.
Disclosure of Invention
The invention aims to provide a method, a device, computer equipment and a storage medium for realizing an inquiry session, so as to solve the problems of low inquiry accuracy and poor user feeling in the conventional inquiry session process.
A first aspect of an embodiment of the present application provides a method for implementing an inquiry session, including:
according to symptom information sent by a user through a terminal, determining syndrome information from a preset stored symptom data table; wherein the syndrome information is used to characterize a category of the condition associated with the symptom information;
candidate corpora screening is carried out according to the syndrome information by utilizing a corpus screening model trained in advance to obtain a target corpus;
sending the target question generated based on the target corpus to a terminal for loading; and the target question is used for confirming whether the syndrome corresponding to the target corpus is the same as the syndrome corresponding to the symptom information input by the user.
A second aspect of an embodiment of the present application provides an apparatus for implementing an inquiry session, including:
the determining module is used for determining syndrome information from a preset stored symptom data table according to the symptom information sent by the user through the terminal; wherein the syndrome information is used to characterize a category of condition associated with the symptom information;
the screening module is used for screening candidate corpora according to the syndrome information by utilizing a pre-trained corpus screening model to obtain target corpora; the target corpus is used for screening out a candidate corpus with the largest attention weight value according to the symptom information from the candidate corpus;
the generating module is used for sending the target question generated based on the target corpus to a terminal for loading; and the target question is used for confirming whether the syndrome corresponding to the target corpus is the same as the syndrome corresponding to the symptom information input by the user.
A third aspect of embodiments of the present application provides a computer device, including a memory, a processor and a computer program stored in the memory and executable on the computer device, wherein the processor implements the steps of the method for implementing an inquiry session provided by the first aspect when executing the computer program.
A fourth aspect of the embodiments of the present application provides a storage medium storing a computer program that, when executed by a processor, implements the steps of the method for implementing an interrogation session provided by the first aspect.
The method, the device, the computer equipment and the storage medium for realizing the inquiry session have the following beneficial effects that:
the embodiment of the application provides a method for realizing an inquiry session, which comprises the steps of utilizing a pre-trained corpus screening model, inputting syndrome information corresponding to symptoms, wherein the syndrome information is used for representing the category of symptoms related to the symptom information. Because the syndrome information and the linguistic data have a mapping relation, the candidate linguistic data corresponding to the syndrome information can be obtained according to the syndrome information, and the candidate linguistic data with the highest input symptom weight, namely the target linguistic data, is screened out from the candidate linguistic data. And generating a target problem from the target corpus and sending the target problem to the terminal for loading. Because the linguistic data is the natural language expression of the syndrome, the questions are generated by using the linguistic data, the user can simply understand the questions to bring good use effect to the user, and the precision of the inquiry conversation is improved from the symptom.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the embodiments or drawings used in the prior art description, and obviously, the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a flowchart illustrating an implementation of a method for implementing an inquiry session according to an embodiment of the present application;
FIG. 2 is a flow chart illustrating the training of a corpus screening model according to an embodiment of the present disclosure;
fig. 3 is a block diagram illustrating an apparatus for implementing an inquiry session according to an embodiment of the present disclosure;
fig. 4 is a block diagram of a computer device according to an embodiment of the present disclosure.
Detailed Description
In order to make the technical problems, technical solutions and advantageous effects to be solved by the present invention more clearly apparent, the present invention is further described in 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 invention and are not intended to limit the invention.
The method for realizing the inquiry session provided by the embodiment of the invention can be applied to computer equipment. The computer device executes each step of the method for implementing an inquiry session provided by the embodiment of the invention.
Referring to fig. 1, fig. 1 illustrates a method for implementing an inquiry session according to an embodiment of the present application, including:
s11: according to symptom information sent by a user through a terminal, determining syndrome information from a preset stored symptom data table; wherein the syndrome information is used to characterize a category of the condition associated with the symptom information.
In step S11, the symptom information sent by the user through the terminal is information obtained by describing the patient 'S own symptoms during the inquiry process, and may be information obtained by inputting the patient' S language, image, video or audio. And the user sends the symptom information through the terminal, and determines the syndrome information corresponding to the symptom information from a preset symptom data table according to the sent symptom information. And obtaining the linguistic data corresponding to the syndrome through the attention mechanism in the linguistic data screening model according to the determined syndrome information, and generating candidate questions according to the linguistic data, so that the precision of the questions in the inquiry process is improved.
In this embodiment, the syndrome information determined from the preset stored symptom data table is a symptom data table that is well organized manually and is stored in advance on the system. The symptom data table of symptoms and syndromes collects symptom data and establishes a symptom data table with associated symptoms and syndromes according to the mapping relation between the symptoms and the syndromes. The syndrome is the general term for a series of associated symptoms in traditional Chinese medicine, one symptom corresponds to at least one syndrome, and a plurality of symptoms correspond to the same syndrome.
Collecting symptom data is to collect the symptoms of a patient, and when a plurality of symptoms appear on the patient, the symptoms may correspond to a plurality of syndromes, so as to collect the symptom data as much as possible, classify the symptoms, and establish a data table with a large number of symptoms. Symptom data may be derived from medical books, historical questionnaires, or the internet, among others. The symptoms are classified by putting similar symptoms together, such as the words of "headache" and "headache", which are the same expressions, so as to facilitate finding out the corresponding symptoms later.
S12: using a pre-trained corpus screening model to screen candidate corpora according to the syndrome information to obtain target corpora; the corpus screening model is obtained by training a training sample constructed according to the pathogenesis of a syndrome sample and the symptom description.
In step S12, the corpus filtering model trained in advance is a model constructed and trained in advance for filtering candidate corpuses from the syndrome information. The method comprises the steps of obtaining a training sample from the Internet, and establishing a mapping relation between a syndrome sample and a corpus sample in the training sample. And (4) building a corpus screening model, learning the parameter weight of the multimode characteristic, and learning the weight of the corpus sample through an attention mechanism. Typically we use a transformer model that assigns attention based on an attention mechanism. Through an encoder and a decoder in a transform model, the output of the encoder is used as the input of the decoder, and words are input to the encoder to obtain word vectors and the weights of the word vectors are learned. Specifically, the input words are converted into word vectors, query vectors, key vectors and value vectors are obtained according to the word vectors, vector matrixes of the query vectors, the key vectors and the value vectors are calculated, the obtained vector matrixes are transmitted to a decoder, and the decoder predicts the positions of the words. And the corpus screening model obtains attention weight among sentences through inputting a vector at a sentence level so as to obtain characteristic representation of the context. Specifically, a target sentence is input into the corpus screening model, a target sentence query vector, each sentence key vector in the context and an original value vector of each sentence in the context are obtained, the weights are normalized by using a softmax function, and the attention weight of the target sentence is obtained, so that the characteristic representation of the context is obtained. The corpus sample contains a plurality of sentences such as the pathogenesis and symptom description of the syndrome sample. The target corpus is obtained by using a corpus screening model from the candidate corpus to obtain the weight of symptoms in the candidate corpus, and the candidate corpus with the highest symptom weight is selected as the target corpus.
Referring to fig. 2, fig. 2 is a flowchart illustrating a training process of a corpus screening model according to an embodiment of the present application.
As shown in fig. 2, in an embodiment, the corpus filtering model trained in step S12 in advance is obtained by training as follows:
s100: and (4) crawling the evidence sample and the corpus sample from the network resource by using a crawler technology.
In step S100, a crawler technology is used to crawl a traditional Chinese medicine knowledge map in the internet, symptoms in a traditional Chinese medicine resource network, and a description related to the symptoms, wherein the description related to the symptoms includes pathogenesis, symptom description, differential diagnosis, and the like. The linguistic data is the detailed description of the syndrome, and the linguistic data comes from the natural language expression of the syndrome such as network, medical books, etc. The language material can be converted into questions to further ask questions of the patient. The problem generated according to the linguistic data is natural, because the linguistic data is not limited to professional medical nouns, the patient can understand the linguistic data, and is not limited to generating the problem only according to the symptoms of the patient any longer, and can generate the problem according to the pathogenesis and the like of the disease related to the syndrome, and the problem is more accurately close to the possible diseases of the patient from multiple aspects. Based on the above reasons, the crawler obtains syndrome samples and corpus samples, wherein the syndrome samples are used for establishing mapping samples with the corpus samples, and the corpus samples are used for describing pathogenesis inducers, symptom descriptions and treatment schemes, i.e. the corpus samples include the pathogenesis inducers, symptom descriptions and treatment schemes of the syndrome.
S101: and establishing a mapping relation between the syndrome sample and the corpus sample to obtain a training sample. In step S101, the mapping relationship between the syndrome sample and the corpus sample is established such that one syndrome has a corresponding syndrome corpus, and the syndrome sample and the corpus sample have a one-to-one correspondence relationship. And training the corpus screening model by using a training sample consisting of the syndrome sample and the corpus sample. As shown in table 1 below, there are syndrome samples and corpus samples in table 1, and there is a mapping relationship between the syndrome samples and the corpus samples in table 1. An example of the syndrome sample and corpus sample is shown in Table 1. For example, the pattern samples of wind-cold type cold in table 1 correspond to the wind-cold type cold caused by wind blowing and catching cold, which is more common in autumn and winter. The symptoms are aching pain all over the body, nasal obstruction, nasal discharge and cough with phlegm. The treatment can be carried out by western medicine and Chinese medicine, or by dietotherapy. The preventive measures are to pay attention to warm keeping, strengthen physical exercise and improve immunity. According to the contents of the corpus samples of the wind-cold type common cold syndrome, the contents of the corpus samples can be further classified. The content of the corpus sample of the wind-cold type cold syndrome can be divided into the causes of the diseases, the wind-cold type cold is the cold caused by wind blowing and catching cold, and the occurrence in autumn and winter is more; the symptoms are described as aching pain all over the body, nasal obstruction, running nose and phlegm cough. The traditional Chinese medicine can be used for treating both the Chinese medicine and the western medicine, and a dietotherapy and treatment scheme can be adopted, and the prevention measures are to pay attention to heat preservation, strengthen physical exercise and improve immunity.
TABLE 1
S102: and training the language material screening model by using the training sample to obtain a pre-trained language material screening model.
In step S102, training the corpus screening model by using the training samples is to obtain the training samples by crawling the syndrome samples and the corpus samples and establishing a mapping relationship therebetween, and train the model to screen the corpus and learn the weight of the corpus content corresponding to the syndrome. The pre-trained corpus screening model is used for receiving syndrome data by utilizing the trained corpus screening model, obtaining a corpus corresponding to the received syndrome and screening required corpora from the corpus.
As an embodiment of the present application, step S102 may include:
s1021: and performing feature representation on the corpus samples in the training samples to obtain the corpus samples after feature representation.
In step S1021, the feature representation of the corpus samples in the training samples is to represent the pathogenesis, symptom description, treatment plan, and the like contained in the content of the corpus samples in the corpus samples.
As an embodiment of the present application, step S1021 may further include:
establishing context feature representation for a corpus sample in a training sample or performing feature representation on the position of a disease causing cause, symptom description and treatment scheme in the corpus sample.
In an embodiment, establishing the context feature representation for the corpus samples in the training samples may use an existing SBERT model to obtain a vector representation of each sentence in the corpus samples through an encode function. The SBERT model is used for training by utilizing large-scale unmarked corpora to obtain semantic representation of texts containing rich semantic information. The occurrence causes, symptom descriptions and treatment schemes in the corpus samples are characterized and expressed in the positions of the corpus samples by applying a DeBERTA model, and each sentence in the corpus samples is expressed by two vectors which respectively encode the content and the positions of the two vectors. The similarity degree between the model and each sentence in the original corpus is calculated by using the linear matrix of the model according to the content and the relative position of the sentence respectively to predict the onset cause of the corpus, symptom description and the initial position in the corpus sample in the treatment scheme, thereby improving the understanding of the corpus screening model on Chinese medical knowledge.
S1022: and training a corpus screening model by using the corpus sample after the characteristic representation, and enabling the corpus screening model to carry out candidate corpus screening on the basis of the pathogenesis cause, the symptom description and the attention weight of the treatment scheme in the corpus sample to obtain the trained corpus screening model.
In step S1022, the corpus screening model performs corpus candidate screening based on the onset cause, the symptom description and the attention weight of the treatment plan in the corpus sample by calculating the onset cause, the symptom description and the attention weight of the treatment plan in the corpus sample, and calculating the onset cause including symptoms, the attention weight corresponding to each sentence in the corpus of the symptom description and the treatment plan. Since the attention weight represents the correlation magnitude between sentences and the sum of the attention weights is 1, the probability magnitude, i.e., confidence, of the occurrence of the sentence at each position in the corpus can be equivalent, and the correlation is larger as the attention weight is larger. The method for calculating the attention weight may use a softmax function, i.e., a normalized exponential function, in which the result of multi-classification is presented in the form of probability.
As an embodiment of the present application, step S12 includes:
s121: and inputting syndrome information, and screening a syndrome sample and a corpus sample with a mapping relation in a model by utilizing the corpus to obtain candidate corpuses corresponding to the syndrome information.
In step S121, the input syndrome information is syndrome information obtained by corresponding to the symptom information, and the input syndrome information obtains candidate corpora by using the corpus screening model, because the syndrome sample and the corpus sample having a mapping relationship exist in the corpus screening model, the candidate corpora corresponding to the syndrome information can be obtained as long as the syndrome information is input. The candidate corpus comprises the pathogenesis, symptom description, differential diagnosis and the like of the input syndrome.
S122: and calculating the attention weight of the symptom information input by the user in the candidate corpus to obtain the attention weight of each symptom information in the candidate corpus.
In step S122, the attention weight of the symptom information in the corpus candidate is calculated to obtain the attention degree of the symptom in the corpus sentence, and the scope of the corpus candidate including a plurality of different symptoms is narrowed. Calculating the attention weight of the symptom information in the corpus candidate may further include performing sentence-level vectorization on the corpus candidate by using a corpus screening model, and normalizing the weight by using a softmax function to obtain the attention weight.
S123: and comparing to obtain each attention weight, and selecting the candidate corpus with the highest symptom information attention weight in the candidate corpus to obtain a target corpus.
In step S123, attention weights of the symptom information on the corpus candidates are compared, and a corpus candidate with the highest weight of the symptom information on the corpus candidate is selected from the corpus candidates. The higher the weight is, the more relevant the symptom information is in the candidate linguistic data, so that the syndrome which corresponds to the symptom most can be obtained, the most corresponding syndrome has the corresponding linguistic data to obtain the target linguistic data, the target linguistic data is used for generating the target question, and the target question generated according to the target linguistic data can be closer to the syndrome corresponding to the symptom input by the user.
S13: sending the target question generated based on the target corpus to a terminal for loading; the target corpus is generated into the target question according to the different content in the target corpus by deleting the same content in the target corpus, and the target question is used for confirming whether the syndrome corresponding to the target corpus is the same as the syndrome corresponding to the symptom information input by the user.
In step S13, the target question is generated according to the content of the target corpus, which includes the relevant description of the corresponding syndrome of the target corpus, including but not limited to pathogenesis, symptom description, differential diagnosis. And generating the target problem according to the contents of pathogenesis, symptom description, differential diagnosis and the like, wherein the generated target problem is vivid and natural in language. When generating the target problem, a plurality of pathogenesis causes, pathogenesis, symptom description and differential diagnosis existing in the target corpus content are screened. The target corpus is a target problem generated according to the difference content in the target corpus by deleting the same content in the target corpus. For example, the symptom description in the target corpus has contents of a plurality of coughs, and the target question is generated by deleting the plurality of coughs and only one symptom description about the coughs is reserved. The target problem is generated based on the syndrome, because the symptom input by the user has the corresponding syndrome, and the corpus corresponding to the syndrome has various symptoms, the user extracts the target corpus according to the various symptoms of the corpus to obtain the target problem. And sending the generated target question to a terminal for loading, and answering by a user based on the target question loaded by the terminal.
As an embodiment of the present application, step S13 may include:
deleting the same content in the target corpus content to obtain a processed target corpus; generating a target question according to the content of the processed target corpus; and sending the target problem to a terminal for loading.
In a specific embodiment, the deleting of the same content in the target corpus is that the target corpus contains the related description of the corresponding syndrome including pathogenesis inducement, pathogenesis, symptom description and differential diagnosis, the target corpus contains the same symptom description because the target corpus is obtained according to the symptom input by the user, the deleting of the same content in the target corpus includes but is not limited to pathogenesis inducement, pathogenesis, symptom description and differential diagnosis, only the content of different target corpuses is left, and the target problem is generated according to the content of the left target corpus, the description of the target problem is natural, and the target problem is not limited to symptom description, and the target problem also asks about the pathogenesis, the pathogenesis and the like related to the symptom of the patient. And sending the target questions to a terminal for loading, enabling the patient to perform a new round of inquiry session according to the questions displayed by the terminal, answering the target questions, inputting new symptoms to obtain the next round of target questions, and finally obtaining the questions most close to the symptoms of the patient.
As an implementation manner of this embodiment, after step S13, the method further includes:
receiving new symptom information input by a user through a terminal, wherein the new symptom information is input by the user according to a target question loaded on the terminal; collecting a symptom information set input by all terminals, wherein the symptom information set comprises symptom information sent by previous users through the terminals and the new symptom information; selecting the symptom information set to obtain a subset of symptom information, wherein the subset of symptom information is the symptom information with the highest confidence level in the symptom information set; generating a diagnostic question from the subset of symptom information using an NLG algorithm, the diagnostic question to narrow the range of selectable symptoms from the subset of symptom information to obtain a more accurate question.
In one embodiment, a target question is displayed on a terminal, a user performs a new round of inquiry session according to the target question displayed on the terminal, the terminal receives information input by the user, the information includes new symptom information, and a symptom information set including the new symptom information is obtained according to the previous round and the obtained symptom information in the previous round. And similarly, extracting the corresponding syndrome information set in the symptom information set by using a corpus screening model to obtain a corresponding corpus set, calculating the corpus with high weight in the corpus set, obtaining a subset of symptom information contained in the corpus, and generating a diagnosis question according to the subset of the symptom information. The resulting diagnostic problem narrows the range of alternative symptoms, and is a more accurate and closer problem to the patient's symptom association.
Corresponding to the above method embodiment, the embodiment of the present invention further provides an apparatus 40 for implementing an inquiry session, as shown in fig. 3, the apparatus may include the following modules:
a determining module 10, configured to determine syndrome information from a preset stored symptom data table according to symptom information sent by a user through a terminal; wherein the syndrome information is used to characterize a category of the condition associated with the symptom information;
the screening module 20 is configured to perform candidate corpus screening according to the syndrome information by using a corpus screening model trained in advance to obtain a target corpus; the target corpus is used for screening out a candidate corpus with the maximum attention weight value according to the symptom information from the candidate corpus;
a generating module 30, configured to send a target question generated based on the target corpus to a terminal for loading; and the target question is used for confirming whether the syndrome corresponding to the target corpus is the same as the syndrome corresponding to the symptom information input by the user.
It should be understood that, in the structural block diagram of the apparatus for implementing an inquiry session shown in fig. 3, each module is used to execute each step in the embodiment corresponding to fig. 1 and fig. 2, while each step in the embodiment corresponding to fig. 1 and fig. 2 has been explained in detail in the above embodiment, specifically please refer to fig. 1 and fig. 2 and the related description in the embodiment corresponding to fig. 1 and fig. 2, and no further description is provided here.
Fig. 4 is a block diagram of a computer device according to another embodiment of the present application. As shown in fig. 4, the computer apparatus 50 of this embodiment includes: a processor 51, a memory 52 and a computer program 53 stored in said memory 52 and executable on said processor 51, such as a program implementing an interrogation session method. The processor 51, when executing the computer program 53, implements the steps of the above-described embodiments of the method for implementing an interrogation session, such as S11-S13 shown in fig. 1. Alternatively, when the processor 51 executes the computer program 53, the functions of the modules in the embodiment corresponding to fig. 3, for example, the functions of the modules 10 to 30 shown in fig. 3, are implemented, for which reference is specifically made to the relevant description in the embodiment corresponding to fig. 3, and details are not repeated here.
Illustratively, the computer program 53 may be partitioned into one or more modules that are stored in the memory 52 and executed by the processor 51 to accomplish the present application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution of the computer program 53 in the computer device 50. For example, the computer program 53 may be divided into a determination module, a screening module, and a generation module, each module having the specific functions as described above.
The turntable device may include, but is not limited to, a processor 51, a memory 52. Those skilled in the art will appreciate that fig. 4 is merely an example of a computer device 50 and is not intended to limit the computer device 50 and may include more or fewer components than those shown, or some components may be combined, or different components, e.g., the turntable device may also include input output devices, network access devices, buses, etc.
The Processor 51 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 52 may be an internal storage unit of the computer device 50, such as a hard disk or a memory of the computer device 50. The memory 52 may also be an external storage device of the computer device 50, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), or the like, provided on the computer device 50. Further, the memory 52 may also include both internal storage units and external storage devices of the computer device 50. The memory 52 is used for storing the computer program and other programs and data required by the turntable device. The memory 52 may also be used to temporarily store data that has been output or is to be output.
In one embodiment, a storage medium is provided, on which a computer program is stored, which, when being executed by a processor, implements the method of implementing an interrogation session in 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 above embodiments may be implemented by a computer program, which can be stored in a storage medium and executed to implement the processes of the above embodiments of the methods. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory. 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 above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the present disclosure, and are intended to be included within the scope thereof.
Claims (10)
1. A method of implementing an interrogation session, comprising:
according to symptom information sent by a user through a terminal, determining syndrome information from a preset stored symptom data table; wherein the syndrome information is used to characterize a category of the condition associated with the symptom information;
candidate corpora screening is carried out according to the syndrome information by utilizing a corpus screening model trained in advance to obtain a target corpus; the corpus screening model is obtained by training a training sample constructed according to the pathogenesis and symptom description of a syndrome sample;
sending the target question generated based on the target corpus to a terminal for loading; the target corpus is generated into the target question according to the different content in the target corpus by deleting the same content in the target corpus, and the target question is used for confirming whether the syndrome corresponding to the target corpus is the same as the syndrome corresponding to the symptom information input by the user.
2. The method according to claim 1, wherein the pre-trained corpus screening model is trained by:
crawling a syndrome sample and a corpus sample from a network resource by using a crawler technology;
establishing a mapping relation between the syndrome sample and the corpus sample to obtain a training sample;
and training the corpus screening model by using the training samples to obtain a pre-trained corpus screening model.
3. The method according to claim 2, wherein the corpus sample is used to describe pathogenesis, symptom description and treatment plan;
the training of the corpus screening model by using the training samples to obtain the pre-trained corpus screening model comprises the following steps:
performing feature representation on the corpus sample in the training sample to obtain the corpus sample after feature representation;
and training a corpus screening model by using the corpus sample after the characteristic representation, and enabling the corpus screening model to carry out candidate corpus screening on the basis of the pathogenesis cause, the symptom description and the attention weight of the treatment scheme in the corpus sample to obtain the trained corpus screening model.
4. The method according to claim 3, wherein the performing feature representation on the corpus sample in the training sample to obtain the corpus sample after feature representation comprises:
establishing context characteristic representation for the corpus samples in the training samples;
or characterizing the pathogenesis, symptom description and treatment plan in the corpus sample.
5. The method according to claim 1, wherein the obtaining of the target corpus by using the pre-trained corpus screening model and performing corpus candidate screening according to the syndrome information comprises:
inputting syndrome information, and screening a syndrome sample and a corpus sample with a mapping relation in a model by utilizing a corpus to obtain a candidate corpus corresponding to the syndrome information;
calculating the attention weight of the symptom information input by the user in the candidate corpus to obtain the attention weight of each symptom information in the candidate corpus;
and comparing to obtain each attention weight, and selecting the candidate corpus with the highest symptom information attention weight in the candidate corpus to obtain a target corpus.
6. The method according to claim 1, wherein the sending the target question generated based on the target corpus to a terminal for loading comprises:
deleting the same content in the target corpus content to obtain a processed target corpus;
generating a target question according to the content of the processed target corpus;
and sending the target problem to a terminal for loading.
7. The method according to claim 1, wherein after the step of sending the target question generated based on the target corpus to a terminal for loading, the method further comprises:
receiving new symptom information input by a user through a terminal, wherein the new symptom information is input by the user according to a target question loaded on the terminal;
collecting a symptom information set input by all terminals, wherein the symptom information set comprises symptom information sent by previous users through the terminals and the new symptom information;
selecting the symptom information set to obtain a subset of symptom information, wherein the subset of symptom information is the symptom information with the highest confidence level in the symptom information set;
generating a diagnostic question from the subset of symptom information using an NLG algorithm, the diagnostic question to narrow the range of selectable symptoms from the subset of symptom information to obtain a more accurate question.
8. An apparatus for implementing an interrogation session, comprising:
the determining module is used for determining syndrome information from a preset stored symptom data table according to the symptom information sent by the user through the terminal; wherein the syndrome information is used to characterize a category of the condition associated with the symptom information;
the screening module is used for screening candidate corpora according to the syndrome information by utilizing a pre-trained corpus screening model to obtain target corpora; the target corpus is used for screening out a candidate corpus with the largest attention weight value according to the symptom information from the candidate corpus;
the generating module is used for sending the target question generated based on the target corpus to a terminal for loading; and the target question is used for confirming whether the syndrome corresponding to the target corpus is the same as the syndrome corresponding to the symptom information input by the user.
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of implementing the method according to any one of claims 1 to 7 when executing the computer program.
10. A storage medium storing a computer program, characterized in that the computer program realizes the steps of the method according to any one of claims 1 to 7 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110838720.5A CN113539520A (en) | 2021-07-23 | 2021-07-23 | Method, device, computer equipment and storage medium for implementing inquiry session |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110838720.5A CN113539520A (en) | 2021-07-23 | 2021-07-23 | Method, device, computer equipment and storage medium for implementing inquiry session |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113539520A true CN113539520A (en) | 2021-10-22 |
Family
ID=78089307
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110838720.5A Pending CN113539520A (en) | 2021-07-23 | 2021-07-23 | Method, device, computer equipment and storage medium for implementing inquiry session |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113539520A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114141387A (en) * | 2021-11-25 | 2022-03-04 | 泰康保险集团股份有限公司 | Interactive information recommendation method, device and equipment in internet medical session |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110504028A (en) * | 2019-08-22 | 2019-11-26 | 上海软中信息系统咨询有限公司 | A kind of disease way of inquisition, device, system, computer equipment and storage medium |
| CN111834012A (en) * | 2020-07-14 | 2020-10-27 | 中国中医科学院中医药信息研究所 | Traditional Chinese medicine syndrome diagnosis method and device based on deep learning and attention mechanism |
| CN112287068A (en) * | 2020-10-29 | 2021-01-29 | 康键信息技术（深圳）有限公司 | Artificial intelligence-based inquiry dialogue data processing method and device |
| CN112768091A (en) * | 2021-01-26 | 2021-05-07 | 北京搜狗科技发展有限公司 | Method, device and medium for processing inquiry information |
| CN112863683A (en) * | 2021-02-19 | 2021-05-28 | 平安科技（深圳）有限公司 | Medical record quality control method and device based on artificial intelligence, computer equipment and storage medium |
-
2021
- 2021-07-23 CN CN202110838720.5A patent/CN113539520A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110504028A (en) * | 2019-08-22 | 2019-11-26 | 上海软中信息系统咨询有限公司 | A kind of disease way of inquisition, device, system, computer equipment and storage medium |
| CN111834012A (en) * | 2020-07-14 | 2020-10-27 | 中国中医科学院中医药信息研究所 | Traditional Chinese medicine syndrome diagnosis method and device based on deep learning and attention mechanism |
| CN112287068A (en) * | 2020-10-29 | 2021-01-29 | 康键信息技术（深圳）有限公司 | Artificial intelligence-based inquiry dialogue data processing method and device |
| CN112768091A (en) * | 2021-01-26 | 2021-05-07 | 北京搜狗科技发展有限公司 | Method, device and medium for processing inquiry information |
| CN112863683A (en) * | 2021-02-19 | 2021-05-28 | 平安科技（深圳）有限公司 | Medical record quality control method and device based on artificial intelligence, computer equipment and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114141387A (en) * | 2021-11-25 | 2022-03-04 | 泰康保险集团股份有限公司 | Interactive information recommendation method, device and equipment in internet medical session |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110929515B (en) | Reading understanding method and system based on cooperative attention and adaptive adjustment | |
| CN111128394B (en) | Medical text semantic recognition method and device, electronic equipment and readable storage medium | |
| JP2021089705A (en) | Method and device for evaluating translation quality | |
| CN113724695B (en) | Electronic medical record generation method, device, equipment and medium based on artificial intelligence | |
| CN107978311A (en) | A kind of voice data processing method, device and interactive voice equipment | |
| CN110990543A (en) | Intelligent conversation generation method and device, computer equipment and computer storage medium | |
| CN112508334A (en) | Personalized paper combining method and system integrating cognitive characteristics and test question text information | |
| WO2021139257A1 (en) | Method and apparatus for selecting annotated data, and computer device and storage medium | |
| CN112487139A (en) | Text-based automatic question setting method and device and computer equipment | |
| CN113656547B (en) | Text matching method, device, equipment and storage medium | |
| WO2022007723A1 (en) | Sentence type recognition method and apparatus, electronic device and storage medium | |
| CN110717021A (en) | Input text and related device for obtaining artificial intelligence interview | |
| CN110929532B (en) | Data processing method, device, equipment and storage medium | |
| CN111241397A (en) | Content recommendation method and device and computing equipment | |
| CN114647713A (en) | Knowledge graph question-answering method, device and storage medium based on virtual confrontation | |
| CN112364125A (en) | Text information extraction system and method combining reading course learning mechanism | |
| CN115309910A (en) | Language piece element and element relation combined extraction method and knowledge graph construction method | |
| CN113539520A (en) | Method, device, computer equipment and storage medium for implementing inquiry session | |
| CN114168743A (en) | Network defense intention recognition and analysis method, equipment and medium based on knowledge graph | |
| WO2023124837A1 (en) | Inquiry processing method and apparatus, device, and storage medium | |
| CN117009456A (en) | Medical query text processing method, device, equipment, medium and electronic product | |
| CN114358579A (en) | Evaluation method, evaluation device, electronic device, and computer-readable storage medium | |
| CN114155957A (en) | Text determination method and device, storage medium and electronic equipment | |
| CN117009532B (en) | Semantic type recognition method and device, computer readable medium and electronic equipment | |
| CN117313943A (en) | Test question accuracy prediction method, system, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |