CN112836059A - Medical map establishing method and device and medical map inquiring method and device - Google Patents
Medical map establishing method and device and medical map inquiring method and device Download PDFInfo
- Publication number
- CN112836059A CN112836059A CN201911164404.3A CN201911164404A CN112836059A CN 112836059 A CN112836059 A CN 112836059A CN 201911164404 A CN201911164404 A CN 201911164404A CN 112836059 A CN112836059 A CN 112836059A
- Authority
- CN
- China
- Prior art keywords
- node
- data
- query
- text
- 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 59
- 230000002452 interceptive effect Effects 0.000 claims abstract description 32
- 238000012545 processing Methods 0.000 claims abstract description 19
- 230000003993 interaction Effects 0.000 claims abstract description 5
- 239000013598 vector Substances 0.000 claims description 23
- 238000000605 extraction Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 201000010099 disease Diseases 0.000 description 13
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 13
- 238000004891 communication Methods 0.000 description 11
- 238000006243 chemical reaction Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 238000003062 neural network model Methods 0.000 description 7
- 206010037660 Pyrexia Diseases 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 206010011224 Cough Diseases 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 230000005236 sound signal Effects 0.000 description 4
- 206010062717 Increased upper airway secretion Diseases 0.000 description 3
- 238000012937 correction Methods 0.000 description 3
- 239000003814 drug Substances 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 208000026435 phlegm Diseases 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 208000035473 Communicable disease Diseases 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003825 pressing Methods 0.000 description 2
- 208000035143 Bacterial infection Diseases 0.000 description 1
- 206010017533 Fungal infection Diseases 0.000 description 1
- 208000031888 Mycoses Diseases 0.000 description 1
- 206010036790 Productive cough Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 208000022362 bacterial infectious disease Diseases 0.000 description 1
- 230000036760 body temperature Effects 0.000 description 1
- 210000001520 comb Anatomy 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 208000017574 dry cough Diseases 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000001717 pathogenic effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000000241 respiratory effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 230000008733 trauma Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- 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
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Abstract
The invention discloses a medical atlas establishing method and a medical atlas establishing device, wherein the method comprises the following steps: collecting a large amount of complete inquiry dialogue data; processing each dialogue data into a data format of alternate interaction between a user and a doctor to obtain an interactive data segment corresponding to each dialogue data; extracting node characteristics of each section of data in the interactive data section; the method comprises the steps of taking the first section of data which begins in an interactive data section corresponding to each conversation data as a root node, taking the next section of data as child nodes, carrying out one-way linkage according to the sequence to construct a medical map, merging the nodes with the same or similar node characteristics, recording the characteristic data of each child node and the characteristic data of all father nodes of the child node, wherein the characteristic data comprises text data and node characteristics. The invention also discloses a medical map query method and a medical map query device. The medical map established by the invention has rich content, higher practicability and universality and provides accurate and effective query results for users.
Description
Technical Field
The invention relates to the field of data processing, in particular to a medical map establishing method and device, and further relates to a medical map inquiring method and device.
Background
The knowledge graph is a network for describing the relationship between different entities, and regularly combs complex mass data, so that a user can easily and conveniently inquire required information. However, the existing medical knowledge maps are simple in structure and mostly constructed based on medical knowledge, and the knowledge maps cannot effectively provide effective suggestions for problems actually described by users, so that the practicability and universality are poor. Even though some medical knowledge maps can provide some recommended results of departments, medicines and the like, the existing medical knowledge maps still cannot provide accurate query results for users due to the fact that the diseases, the illnesses and the diagnosis and treatment means in the medical field are in complicated and complicated relations.
Disclosure of Invention
The embodiment of the invention provides a medical map establishing method and device on the one hand, so that the practicability and universality of the map are improved.
The embodiment of the invention also provides a medical atlas query method and a medical atlas query device, which can provide accurate and effective query results for users.
Therefore, the invention provides the following technical scheme:
a medical atlas creation method, the method comprising:
collecting a large amount of complete inquiry dialogue data;
processing each dialogue data into a data format of alternate interaction between a user and a doctor to obtain an interactive data segment corresponding to each dialogue data;
extracting node characteristics of each section of data in the interactive data section;
the method comprises the steps of taking the first section of data which begins in an interactive data section corresponding to each conversation data as a root node, taking the next section of data as child nodes, carrying out one-way linkage according to the sequence to construct a medical map, merging the nodes with the same or similar node characteristics, recording the characteristic data of each child node and the characteristic data of all father nodes of the child node, wherein the characteristic data comprises text data and node characteristics.
Optionally, the node characteristics include any one or more of the following:
a text-based node feature, the text-based node feature being the dialog text itself;
node features based on text features, wherein the node features based on the text features are vectors of dialog texts;
the node characteristics based on the entity information are the entity information extracted from the dialogue text;
and the node feature based on the entity feature is a vector of entity information extracted from the dialog text.
A medical atlas query method, the method comprising:
receiving query information input by a user;
extracting patient characteristic information from the query information;
determining nodes in a medical map that are related to the patient characteristic information;
if only one relevant node exists, taking the relevant node as a query node;
otherwise, respectively calculating the similarity between the text data of each node and the query information, and taking the node with the highest similarity as a query node;
determining a query result according to the text data of the child nodes of the query node;
and outputting the query result.
Optionally, the patient characteristic information comprises: keyword information and entity information.
Optionally, the determining a query result according to the text data of the child node of the query node includes:
if the query node has a plurality of child nodes, respectively calculating the matching degree of the text of each child node and the query information;
and taking the text of the child node with the highest matching degree as a query result.
Optionally, the outputting the query result includes:
and if the query result comprises a plurality of clauses, sequentially outputting each clause, and outputting one clause each time.
Optionally, the outputting the query result includes:
converting the query result into a natural language text;
and outputting the natural language text.
Optionally, the converting the query result into natural language text comprises:
filling the query result into a preset template to obtain a natural language text; or
And inputting the query result into a pre-established text conversion model, and obtaining a natural language text according to the output of the text conversion model.
A medical atlas creation apparatus, the apparatus comprising:
the data collection module is used for collecting a large amount of complete inquiry dialogue data;
the preprocessing module is used for processing each piece of dialogue data into a data format in which a user and a doctor interact in turn to obtain an interactive data segment corresponding to each piece of dialogue data;
the feature extraction module is used for extracting node features from each segment of data in the interactive data segment;
the graph generation module is used for taking the first section of data which begins in the interactive data section corresponding to each piece of session data as a root node, taking the next section of data as child nodes, performing one-way link according to the sequence to construct a medical graph, merging the nodes with the same or similar node characteristics, and recording the characteristic data of each child node and the characteristic data of all father nodes thereof, wherein the characteristic data comprises text data and node characteristics.
Optionally, the node characteristics include any one or more of the following:
a text-based node feature, the text-based node feature being the dialog text itself;
node features based on text features, wherein the node features based on the text features are vectors of dialog texts;
the node characteristics based on the entity information are the entity information extracted from the dialogue text;
and the node feature based on the entity feature is a vector of entity information extracted from the dialog text.
A medical atlas query apparatus, the apparatus comprising:
the receiving module is used for receiving query information input by a user;
the information extraction module is used for extracting patient characteristic information from the query information;
the query module is used for determining each node related to the patient characteristic information in the medical map;
the judging module is used for judging whether only one related node exists or not;
the query node determining module is used for taking the related node as a query node when only one related node exists; otherwise, respectively calculating the similarity between the text data of each node and the query information, and taking the node with the highest similarity as a query node;
the query result determining module is used for determining a query result according to the text data of the child nodes of the query node;
and the output module is used for outputting the query result.
Optionally, the patient characteristic information comprises: keyword information and entity information.
Optionally, the query result determination module includes:
the matching degree calculation unit is used for calculating the matching degree of the text of each child node and the query information when the query node has a plurality of child nodes;
and the selecting unit is used for taking the text of the child node with the highest matching degree as a query result.
Optionally, when the query result includes a plurality of clauses, the output module outputs each clause in sequence, and outputs one clause each time.
Optionally, the output module includes:
the text conversion unit is used for converting the query result into a natural language text;
and a text output unit which outputs the natural language text.
Optionally, the text conversion unit includes:
the template filling unit is used for filling the inquiry information or the inquiry result into a preset template to obtain a natural language text; or
And the text generation unit is used for inputting the inquiry information or the inquiry result into a pre-established inquiry model and obtaining a natural language text according to the output of the inquiry model.
A computer device, comprising: one or more processors, memory;
the memory is configured to store computer-executable instructions and the processor is configured to execute the computer-executable instructions to implement the method described above.
A readable storage medium having stored thereon instructions which are executed to implement the foregoing method.
The medical map establishing method and device provided by the embodiment of the invention are based on a large amount of complete inquiry dialogue data, arrange the inquiry dialogue data into a data format in which a user and a doctor interact in turn to obtain an interactive data segment corresponding to each dialogue data, extract the node characteristics of each segment of data, then use the first segment of data starting in the interactive data segment corresponding to each dialogue data as a root node and the next segment of data as child nodes, construct a medical map by unidirectional linking according to the sequence, merge the nodes with the same or similar node characteristics, and record the characteristic data of each child node and the characteristic data of all father nodes thereof, thereby ensuring the richness of the content of the medical map and having higher practicability and universality.
According to the medical map query method and device provided by the embodiment of the invention, the patient characteristic information is extracted from the query information input by the user, each node related to the patient characteristic information in the medical map is determined, the node with the highest similarity is used as the query node, the query result is determined according to the text data of the child nodes of the query node, and the query result is output to the user. Because the medical map is established based on the inquiry dialogue data, the naturalness and the accuracy of the query result are ensured. And each child node not only records the characteristic data of the child node, but also records the characteristic data of all father nodes of the child node, thereby effectively avoiding the interference of similar conversations caused by different conversation sequences.
Drawings
In order to more clearly illustrate the embodiments of the present application or technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings can be obtained by those skilled in the art according to the drawings.
FIG. 1 is a flow chart of a medical atlas creation method according to an embodiment of the invention;
FIG. 2 is a flow chart of a medical atlas query method of an embodiment of the invention;
fig. 3 is a block diagram of a medical atlas creation apparatus according to an embodiment of the invention;
fig. 4 is a block diagram of a medical atlas query apparatus according to an embodiment of the invention;
FIG. 5 is a block diagram illustrating an apparatus for a medical atlas query method, according to an example embodiment;
fig. 6 is a schematic structural diagram of a server in an embodiment of the present invention.
Detailed Description
In order to make the technical field of the invention better understand the scheme of the embodiment of the invention, the embodiment of the invention is further described in detail with reference to the drawings and the implementation mode.
The embodiment of the invention provides a medical map establishing method and device, based on a large amount of complete inquiry dialogue data, the inquiry dialogue data are arranged into a data format in which a user and a doctor interact in turn to obtain an interactive data segment corresponding to each dialogue data, the node characteristics of each segment of data are extracted, then the first segment of data starting in the interactive data segment corresponding to each dialogue data is used as a root node, the next segment of data is used as a child node, one-way linking is carried out according to the sequence to construct a medical map, nodes with the same or similar node characteristics are merged, and each child node records the text data of the child node and the text data of all father nodes of the child node.
As shown in fig. 1, it is a flowchart of a medical atlas establishing method according to an embodiment of the present invention, and the method includes the following steps:
And 102, processing each piece of dialogue data into a data format in which the user and the doctor interact in turn to obtain an interactive data segment corresponding to each piece of dialogue data.
In the embodiment of the invention, the data of each complete inquiry is used as one piece of dialogue data. Usually, in the inquiry dialogue data, a patient often says several sentences continuously, and the doctor replies; or a state in which the doctor continuously speaks. For the situation, several sentences which are continuously spoken by the patient or the doctor are integrated into one segment to form a data format for the patient and the doctor to interact alternately.
And 103, extracting node characteristics of each segment of data in the interactive data segment.
And step 104, taking the first section of data which begins in the interactive data section corresponding to each piece of session data as a root node, taking the next section of data as child nodes, performing unidirectional linking according to the sequence to construct a medical map, merging the nodes with the same or similar node characteristics, and recording the characteristic data of each child node and the characteristic data of all father nodes thereof by each child node, wherein the characteristic data comprises text data and node characteristics.
The judgment of similarity of node features may be determined according to a distance between texts corresponding to two nodes, such as a cosine distance, and the specific calculation may be performed by using the prior art, which is not described in detail herein.
In the embodiment of the present invention, the node characteristics may be any one or more of the following combinations:
(1) a text-based node feature, the text-based node feature being the dialog text itself;
(2) node features based on text features, wherein the node features based on the text features are vectors of dialog texts;
(3) the node characteristics based on the entity information are the entity information extracted from the dialogue text;
(4) and the node feature based on the entity feature is a vector of entity information extracted from the dialog text.
The extraction manner of the entity information is similar to the extraction manner of the keywords in the prior art, and is not described in detail here.
Thus, the obtained medical map can have a plurality of trees, each tree has a root node, and related nodes exist among different trees, that is, one child node can trace to different root nodes. In addition, when the medical map is constructed, nodes with the same or similar node characteristics are combined, for example, the node characteristics of 'fever' and 'fever' based on entity information are similar; as another example, the text-based node feature "I am as if I have a fever" is similar to "I may have a fever". Therefore, there are a plurality of texts corresponding to such nodes.
It can be seen that each node in the medical map may have one or more of the above-mentioned combined node features, and each node may correspond to one or more texts. In addition, each child node records not only its own text data but also the text data of all its parents.
In addition, it should be noted that the text data of each node may be the original text data in the interactive data segment, or may be the text data after being processed by semantic understanding and the like, for example, the original text of the patient confirms cough and rejects expectoration, and the text after being processed is: "Dry cough" or "cough without phlegm".
The medical map establishing method provided by the embodiment of the invention is based on a large amount of complete inquiry dialogue data, the inquiry dialogue data is arranged into a data format in which a user and a doctor interact in turn to obtain an interactive data segment corresponding to each dialogue data, the node characteristics of each segment of data are extracted, then the first segment of data starting from the interactive data segment corresponding to each dialogue data is used as a root node, the next segment of data is used as a child node, one-way linkage is carried out according to the sequence to construct the medical map, nodes with the same or similar node characteristics are merged, and each child node records the characteristic data of the child node and the characteristic data of all father nodes thereof, so that the richness of the content of the medical map is ensured, and the medical map establishing method has higher practicability and universality.
In the medical map of the embodiment of the invention, each child node records the characteristic data of the child node and the characteristic data of all father nodes, and compared with the case that the child node only records the characteristic data of the child node, the medical map has the following advantages that:
1. when the medical map is constructed, because the nodes are combined, the child nodes construct the map by inheriting the information of the father node, so that the combined nodes have more child nodes, the selection space is increased during subsequent query, and the quality of query results can be improved.
2. The method can be directly matched into a certain node during subsequent query, thereby saving computing resources and better processing complex combination conditions. For example, changing the order in which users state disease or compressing the turns of states does not affect this map.
When a user queries by using the medical map constructed based on the inquiry dialogue data, more accurate and effective query results can be obtained, the output query results are more in line with the conventional doctor-patient dialogue form, the query results are more natural, the use of the user is facilitated, and inquiry auxiliary reference can be provided for doctors.
Fig. 2 is a flowchart of a medical atlas query method according to an embodiment of the present invention, including the following steps:
In an embodiment of the present invention, the patient characteristic information includes: keyword information and entity information. The keyword information may be a spoken expression, and the entity information is a standard medical language. Keywords are broader in scope than entities, and multiple keywords may be synonymous with one entity. For example, the keyword is "body temperature higher than normal", and the corresponding entity is "fever". For another example, the keyword is "cough with phlegm", and the corresponding entity is "phlegm".
Of course, for the aforementioned query of the medical map created based on the text-based node features, the patient feature information may be the query information itself.
And step 203, determining nodes related to the patient characteristic information in the medical atlas.
Specifically, nodes containing the patient characteristic information may be looked up in the medical atlas as nodes related to the patient characteristic information.
As mentioned above, the node characteristics of each node in the medical map may be any one or more of the following combinations:
(1) a text-based node feature, the text-based node feature being the dialog text itself;
(2) node features based on text features, wherein the node features based on the text features are vectors of dialog texts;
(3) the node characteristics based on the entity information are the entity information extracted from the dialogue text;
(4) and the node feature based on the entity feature is a vector of entity information extracted from the dialog text.
Accordingly, when searching for each node containing the patient characteristic information, the searching manner may be different according to the different node characteristics.
It should be noted that, for node features based on node features or entity features, an embedded vector corresponding to the query information or the patient feature information may be obtained by using a pre-trained neural network model. And calculating the distance between the embedded vector and the vector of the node characteristics of each node by using the embedded vector and the vector of the node characteristics of each node, and determining each node related to the patient characteristic information, for example, each node with the distance larger than a set threshold value is taken as a related node.
Further, in another embodiment of the present invention, the diseases may be classified in advance, and the disease categories may be classified according to different standards, such as hospital departments, national standards, international standards, and the like, which is not limited in this embodiment of the present invention. For example, the disease can be classified into eyes, hands, feet, heart, etc. according to the location of the disease; according to the department, the medicine can be divided into ophthalmology, internal medicine, respiratory department and the like; diseases of different age groups and different sexes can be divided according to the pathogenic population; according to the causes of the disease, the disease can be divided into bacterial infection, fungal infection, trauma and the like; it can be classified into high infectious diseases, low infectious diseases, etc. according to infectivity.
Accordingly, the neural network model includes a plurality of models respectively corresponding to different disease categories, that is, a neural network model corresponding to the disease category is established for each disease category.
Correspondingly, when determining each node related to the patient feature information in the medical map, for the node features based on the node features or the entity features, a neural network model corresponding to the disease category needs to be acquired, then the query information or the entity information is input into the neural network model to obtain a corresponding embedded vector, and then each node related to the patient feature information is determined according to the embedded vector and the vector of the node features of each node.
When searching for each node containing the patient feature information, for each child node, not only the node feature of itself but also the node features of all the parent nodes recorded by the child node need to be considered.
Since each node in the medical map may have one or more of the combined node features described above, when searching for each node containing the embedded vector or text information (the query information or the patient feature information or the entity information), a precise matching or partial matching manner may be adopted, and specifically, the searching may be performed in order of priority of the node features.
And step 205, taking the related node as a query node. Then, step 207 is performed.
And step 206, respectively calculating the similarity between the text data of each node and the query information, and taking the node with the highest similarity as a query node.
If a plurality of related nodes exist, the most related node, namely the query node, is obtained through the judgment of the similarity between the text data of each node and the query information.
It should be noted that, in the above description, when the medical map is established, nodes with the same or similar node characteristics are merged, and therefore, there are a plurality of self texts corresponding to the merged nodes. In the embodiment of the present invention, in the case that the node, i.e., the relevant node, in the medical map determined in step 203, which is relevant to the patient feature information corresponds to a plurality of self texts, the similarity between each text data of the node and the current input information may be respectively calculated, and the maximum value thereof may be selected as the similarity between the text data of the node and the query information.
Because the medical map is established based on the inquiry dialogue data, and each piece of data in the interactive data segment of each piece of dialogue data is used as a node, the parent node and the child node in the medical map form a dialogue relation. Therefore, the text data of the child nodes of the query node is used as the query result, so that the query result has more naturalness and practicability.
Further, if the query node has a plurality of child nodes, the matching degree of the text data of each child node and the query information can be respectively calculated; and taking the text data of the child node with the highest matching degree as a query result.
Similarly, if the child node corresponds to a plurality of own texts, the matching degree of each text data of the node with the current input information can be calculated respectively, and the maximum value of the matching degrees is selected as the matching degree of the text data of the node with the query information.
The matching degree can be calculated by using a pre-established matching model, and the specific training process of the matching model can be obtained by using the prior art, for example, by using data training in an inquiry database. The inquiry database is built by collecting some inquiry and answer data from the network, and is structured in such a manner that one question (i.e., inquiry sentence) corresponds to one or more answers (i.e., answer text) as a set of data. Mainly comprises the online questions and answers of the patients and doctors related to the medical treatment. Specifically, score labeling is carried out on different answers of the same question according to data in the inquiry database, and a matching model is trained by using the data and labeling information thereof to calculate the matching degree score of the question and the answer.
And step 208, outputting the query result.
It should be noted that, if the query result includes multiple clauses, the multiple clauses may be output at the same time; or sequentially outputting each clause, and outputting one clause each time, which is not limited in the embodiment of the present invention.
Furthermore, the query result can be processed first, so that the finally output query result has the characteristic of natural conversation. Specifically, the query result may be converted into a natural language text, which is then output.
Converting the query result into natural language text may be performed in any of the following ways:
1) filling the query result into a preset template to obtain a natural language text;
2) and inputting the query result into a pre-established text conversion model, and obtaining a natural language text according to the output of the text conversion model.
By the above processing, not only the flexibility of the output contents can be improved, but also the repeated output of some words can be avoided by the output rewriting.
According to the medical map query method provided by the embodiment of the invention, the patient characteristic information is extracted from the query information input by the user, each node related to the patient characteristic information in the medical map is determined, the node with the highest similarity is used as the query node, the query result is determined according to the text data of the child nodes of the query node, and the query result is output to the user. Because the medical map is established based on the inquiry dialogue data, the naturalness and the accuracy of the query result are ensured. And each child node not only records the characteristic data of the child node, but also records the characteristic data of all father nodes of the child node, thereby effectively avoiding the interference of similar conversations caused by different conversation sequences.
Correspondingly, the embodiment of the invention also provides a medical atlas setting device, which is a structural block diagram of the device as shown in fig. 3.
In this embodiment, the medical atlas setting apparatus includes the following modules:
a data collection module 301 for collecting a large amount of complete inquiry session data;
the preprocessing module 302 is configured to process each piece of session data into a data format in which a user and a doctor interact with each other in turn, so as to obtain an interactive data segment corresponding to each piece of session data;
a feature extraction module 303, configured to extract a node feature for each segment of data in the interactive data segment;
the graph generation module 304 is configured to use a first segment of data starting from an interactive data segment corresponding to each session data as a root node, use a next segment of data as child nodes, perform unidirectional linking according to a sequence to construct a medical graph, merge nodes with the same or similar node characteristics, and record characteristic data of each child node and characteristic data of all father nodes thereof, where the characteristic data includes text data and node characteristics.
In the embodiment of the present invention, the preprocessing module 302 needs to format the inquiry dialogue data of the data collecting module 301, specifically, data of each complete inquiry is used as a dialogue data, and several sentences spoken by the patient or the doctor continuously are integrated into one segment to form a data format for the patient and the doctor to interact alternately. Further, in the medical map, each piece of data in the interactive data piece of each piece of dialogue data serves as one node.
The node characteristics may be any one or more of the following in combination:
a text-based node feature, the text-based node feature being the dialog text itself;
node features based on text features, wherein the node features based on the text features are vectors of dialog texts;
the node characteristics based on the entity information are the entity information extracted from the dialogue text;
and the node feature based on the entity feature is a vector of entity information extracted from the dialog text.
Wherein the vector of the dialog text or the vector of the entity information can be obtained by a pre-trained neural network model. The training process of the neural network model may employ the prior art and is not described in detail herein. The extraction manner of the entity information is similar to the extraction manner of the keywords in the prior art, and is not described in detail here.
Thus, the obtained medical map can have a plurality of trees, each tree has a root node, and related nodes exist among different trees, that is, one child node can trace to different root nodes. In addition, since the nodes having the same node characteristics are combined when the medical map is constructed, there are a plurality of texts corresponding to such nodes. It can be seen that each node in the medical map may have one or more of the above-mentioned combined node features, and each node may correspond to one or more texts. In addition, each child node records not only its own text data but also the text data of all its parents.
The medical map establishing device provided by the embodiment of the invention is based on a large amount of complete inquiry dialogue data, arranges the inquiry dialogue data into a data format in which a user and a doctor interact in turn to obtain an interactive data segment corresponding to each dialogue data, extracts the node characteristics of each segment of data, then takes the first segment of data starting in the interactive data segment corresponding to each dialogue data as a root node, takes the next segment of data as a child node, constructs a medical map by unidirectional linking according to the sequence, merges nodes with the same or similar node characteristics, and records the characteristic data of each child node and the characteristic data of all father nodes thereof, thereby ensuring the richness of the content of the medical map and having higher practicability and universality.
Correspondingly, the embodiment of the invention also provides a medical map query device, which is used for extracting the patient characteristic information from the query information input by the user aiming at the medical map constructed based on the inquiry dialogue data, determining each node related to the patient characteristic information in the medical map, taking the node with the highest similarity as a query node, determining a query result according to the text data of the child nodes of the query node, and outputting the query result to the user.
Fig. 4 is a block diagram of a medical atlas query apparatus according to an embodiment of the invention.
In this embodiment, the medical atlas query apparatus includes the following modules:
a receiving module 401, configured to receive query information input by a user;
an information extraction module 402, configured to extract patient characteristic information from the query information;
a query module 403 for determining nodes of the medical atlas 300 related to the patient characteristic information;
a judging module 404, configured to judge whether there is only one relevant node;
a query node determining module 405, configured to use, when there is only one relevant node, the relevant node as a query node; otherwise, respectively calculating the similarity between the text data of each node and the query information, and taking the node with the highest similarity as a query node;
a query result determining module 406, configured to determine a query result according to the text data of the child node of the query node;
and the output module 407 is configured to output the query result.
In an embodiment of the present invention, the patient characteristic information includes: keyword information and entity information. Accordingly, the query module 403 may specifically search for nodes including the patient feature information in the medical map, and use these nodes as nodes related to the patient feature information.
As mentioned above, the node characteristics of each node in the medical map may be a plurality of different characteristics, and accordingly, when searching each node containing the patient characteristic information, the searching manner may be different according to the different node characteristics. Moreover, when searching for each node containing the patient characteristic information, a precise matching or partial matching mode may be adopted, and specifically, the searching may be performed in the order from high priority to low priority of the node characteristics.
In addition, if there are multiple related nodes, the query node determining module 405 needs to determine the most related node, that is, the query node, according to the similarity between the text data of each node and the query information. It should be noted that, when the medical map is created, nodes with the same or similar node characteristics are merged, so that a plurality of texts corresponding to the merged nodes exist. Correspondingly, in the case that a node related to the patient feature information, that is, a related node, corresponds to a plurality of self texts, the query node determining module 405 may respectively calculate the similarity between each text data of the node and the current input information, and select the maximum value thereof as the similarity between the text data of the node and the query information.
Further, if the query node has a plurality of child nodes, the query result determination module 406 may determine a final query result according to a matching degree of the text data of each child node and the query information. Accordingly, the query result determination module 406 may include a matching degree calculation unit and a selection unit; wherein:
the matching degree calculating unit is used for calculating the matching degree of the text of each child node and the query information when the query node has a plurality of child nodes;
the selection unit is used for taking the text of the child node with the highest matching degree as a query result.
Similarly, if the child node corresponds to a plurality of own texts, the matching degree calculation unit may calculate the matching degree between each text data of the node and the current input information, and select the maximum value as the matching degree between the text data of the node and the query information.
The calculation of the matching degree has been described in detail in the foregoing embodiments of the method of the present invention, and is not described herein again.
It should be noted that, if the query result includes multiple clauses, the output module 407 may output the multiple clauses at the same time; or sequentially outputting each clause, and outputting one clause each time, which is not limited in the embodiment of the present invention.
Further, the output module 407 may also process the query result first, so that the finally output query result has a natural conversation characteristic. For example, a specific structure of the output module 407 includes the following units:
the text conversion unit is used for converting the query result into a natural language text;
and a text output unit which outputs the natural language text.
Wherein the text conversion unit may include: a template filling unit or a text generating unit;
the template filling unit is used for filling the inquiry information or the inquiry result into a preset template to obtain a natural language text;
the text generation unit is used for inputting the inquiry information or the inquiry result into a pre-established inquiry model and obtaining a natural language text according to the output of the inquiry model.
The medical map query device provided by the embodiment of the invention extracts the patient characteristic information from the query information input by the user, determines each node related to the patient characteristic information in the medical map, takes the node with the highest similarity as the query node, determines the query result according to the text data of the child nodes of the query node, and outputs the query result to the user. Because the medical map is established based on the inquiry dialogue data, the naturalness and the accuracy of the query result are ensured. And each child node not only records the characteristic data of the child node, but also records the characteristic data of all father nodes of the child node, thereby effectively avoiding the interference of similar conversations caused by different conversation sequences.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
It should be noted that the method and apparatus in the embodiments of the present invention may be applied to various terminal devices, such as a mobile phone, a computer, and a notebook.
Fig. 5 is a block diagram illustrating an apparatus 800 for a medical atlas query method, according to an example embodiment. For example, the apparatus 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, an exercise device, a personal digital assistant, and the like.
Referring to fig. 5, the apparatus 800 may include one or more of the following components: processing component 802, memory 804, power component 806, multimedia component 808, audio component 810, input/output (I/O) interface 812, sensor component 814, and communication component 816.
The processing component 802 generally controls overall operation of the device 800, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing elements 802 may include one or more processors 820 to execute instructions to perform all or a portion of the steps of the methods described above. Further, the processing component 802 can include one or more modules that facilitate interaction between the processing component 802 and other components. For example, the processing component 802 can include a multimedia module to facilitate interaction between the multimedia component 808 and the processing component 802.
The memory 804 is configured to store various types of data to support operation at the device 800. Examples of such data include instructions for any application or method operating on device 800, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 804 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The multimedia component 808 includes a screen that provides an output interface between the device 800 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front facing camera and/or a rear facing camera. The front-facing camera and/or the rear-facing camera may receive external multimedia data when the device 800 is in an operating mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
The audio component 810 is configured to output and/or input audio signals. For example, the audio component 810 includes a Microphone (MIC) configured to receive external audio signals when the apparatus 800 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in the memory 804 or transmitted via the communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.
The I/O interface 812 provides an interface between the processing component 802 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The sensor assembly 814 includes one or more sensors for providing various aspects of state assessment for the device 800. For example, the sensor assembly 814 may detect the open/closed state of the device 800, the relative positioning of the components, such as a display and keypad of the apparatus 800, the sensor assembly 814 may also detect a change in position of the apparatus 800 or a component of the apparatus 800, the presence or absence of user contact with the apparatus 800, orientation or acceleration/deceleration of the apparatus 800, and a change in temperature of the apparatus 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
The communication component 816 is configured to facilitate communications between the apparatus 800 and other devices in a wired or wireless manner. The device 800 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast associated information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communications component 816 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
In an exemplary embodiment, the apparatus 800 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.
In an exemplary embodiment, a non-transitory computer-readable storage medium comprising instructions, such as the memory 804 comprising instructions, executable by the processor 820 of the device 800 to perform the key press false touch correction method described above is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
A non-transitory computer readable storage medium having instructions therein which, when executed by a processor of a mobile terminal, enable the mobile terminal to perform a key press mis-touch error correction method, the method comprising: in the input process of a user, acquiring pressing information when each key is triggered; determining a false triggering key according to the acquired pressing information; correcting error of the false triggering key; and determining each candidate word corresponding to the corrected complete input string.
Fig. 6 is a schematic structural diagram of a server in an embodiment of the present invention. The server 1900, which may vary widely in configuration or performance, may include one or more Central Processing Units (CPUs) 1922 (e.g., one or more processors) and memory 1932, one or more storage media 1930 (e.g., one or more mass storage devices) that store applications 1942 or data 1944. Memory 1932 and storage medium 1930 can be, among other things, transient or persistent storage. The program stored in the storage medium 1930 may include one or more modules (not shown), each of which may include a series of instructions operating on a server. Still further, a central processor 1922 may be provided in communication with the storage medium 1930 to execute a series of instruction operations in the storage medium 1930 on the server 1900.
The server 1900 may also include one or more power supplies 1926, one or more wired or wireless network interfaces 1950, one or more input-output interfaces 1958, one or more keyboards 1956, and/or one or more operating systems 1941, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.
A non-transitory computer readable storage medium having instructions therein which, when executed by a processor of a device, enable the device to perform the key press mis-touch correction method described above.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is only limited by the appended claims
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A method of medical atlas creation, the method comprising:
collecting a large amount of complete inquiry dialogue data;
processing each dialogue data into a data format of alternate interaction between a user and a doctor to obtain an interactive data segment corresponding to each dialogue data;
extracting node characteristics of each section of data in the interactive data section;
the method comprises the steps of taking the first section of data which begins in an interactive data section corresponding to each conversation data as a root node, taking the next section of data as child nodes, carrying out one-way linkage according to the sequence to construct a medical map, merging the nodes with the same or similar node characteristics, recording the characteristic data of each child node and the characteristic data of all father nodes of the child node, wherein the characteristic data comprises text data and node characteristics.
2. The method of claim 1, wherein the node characteristics comprise any one or more of the following in combination:
a text-based node feature, the text-based node feature being the dialog text itself;
node features based on text features, wherein the node features based on the text features are vectors of dialog texts;
the node characteristics based on the entity information are the entity information extracted from the dialogue text;
and the node feature based on the entity feature is a vector of entity information extracted from the dialog text.
3. A medical atlas query method, the method comprising:
receiving query information input by a user;
extracting patient characteristic information from the query information;
determining nodes in a medical map that are related to the patient characteristic information;
if only one relevant node exists, taking the relevant node as a query node;
otherwise, respectively calculating the similarity between the text data of each node and the query information, and taking the node with the highest similarity as a query node;
determining a query result according to the text data of the child nodes of the query node;
and outputting the query result.
4. The method of claim 3, wherein the patient characteristic information comprises: keyword information and entity information.
5. The method of claim 3, wherein determining query results from text data of children of the query node comprises:
if the query node has a plurality of child nodes, respectively calculating the matching degree of the text of each child node and the query information;
and taking the text of the child node with the highest matching degree as a query result.
6. The method of claim 3, wherein outputting the query result comprises:
and if the query result comprises a plurality of clauses, sequentially outputting each clause, and outputting one clause each time.
7. A medical atlas creation apparatus, the apparatus comprising:
the data collection module is used for collecting a large amount of complete inquiry dialogue data;
the preprocessing module is used for processing each piece of dialogue data into a data format in which a user and a doctor interact in turn to obtain an interactive data segment corresponding to each piece of dialogue data;
the feature extraction module is used for extracting node features from each segment of data in the interactive data segment;
the graph generation module is used for taking the first section of data which begins in the interactive data section corresponding to each piece of session data as a root node, taking the next section of data as child nodes, performing one-way link according to the sequence to construct a medical graph, merging the nodes with the same or similar node characteristics, and recording the characteristic data of each child node and the characteristic data of all father nodes thereof, wherein the characteristic data comprises text data and node characteristics.
8. A medical atlas query apparatus, the apparatus comprising:
the receiving module is used for receiving query information input by a user;
the information extraction module is used for extracting patient characteristic information from the query information;
the query module is used for determining each node related to the patient characteristic information in the medical map;
the judging module is used for judging whether only one related node exists or not;
the query node determining module is used for taking the related node as a query node when only one related node exists; otherwise, respectively calculating the similarity between the text data of each node and the query information, and taking the node with the highest similarity as a query node;
the query result determining module is used for determining a query result according to the text data of the child nodes of the query node;
and the output module is used for outputting the query result.
9. A computer device, comprising: one or more processors, memory;
the memory is for storing computer-executable instructions, and the processor is for executing the computer-executable instructions to implement the method of any one of claims 3 to 6.
10. A readable storage medium having stored thereon instructions that are executed to implement the method of any of claims 3 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911164404.3A CN112836059A (en) | 2019-11-25 | 2019-11-25 | Medical map establishing method and device and medical map inquiring method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911164404.3A CN112836059A (en) | 2019-11-25 | 2019-11-25 | Medical map establishing method and device and medical map inquiring method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112836059A true CN112836059A (en) | 2021-05-25 |
Family
ID=75922185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911164404.3A Pending CN112836059A (en) | 2019-11-25 | 2019-11-25 | Medical map establishing method and device and medical map inquiring method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112836059A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107729481A (en) * | 2017-10-16 | 2018-02-23 | 北京神州泰岳软件股份有限公司 | The Text Information Extraction result screening technique and device of a kind of custom rule |
| CN108763523A (en) * | 2018-05-31 | 2018-11-06 | 康键信息技术（深圳）有限公司 | Automatic session implementation method, server and storage medium |
| US20190018839A1 (en) * | 2017-07-14 | 2019-01-17 | Guangzhou Shenma Mobile Information Technology Co., Ltd. | Knowledge map-based question-answer method, device, and storage medium |
| CN109792402A (en) * | 2016-07-08 | 2019-05-21 | 艾赛普公司 | Automated to respond to the request of user |
| CN109829036A (en) * | 2019-02-12 | 2019-05-31 | 浙江核新同花顺网络信息股份有限公司 | A kind of dialogue management method and relevant apparatus |
-
2019
- 2019-11-25 CN CN201911164404.3A patent/CN112836059A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109792402A (en) * | 2016-07-08 | 2019-05-21 | 艾赛普公司 | Automated to respond to the request of user |
| US20190018839A1 (en) * | 2017-07-14 | 2019-01-17 | Guangzhou Shenma Mobile Information Technology Co., Ltd. | Knowledge map-based question-answer method, device, and storage medium |
| CN107729481A (en) * | 2017-10-16 | 2018-02-23 | 北京神州泰岳软件股份有限公司 | The Text Information Extraction result screening technique and device of a kind of custom rule |
| CN108763523A (en) * | 2018-05-31 | 2018-11-06 | 康键信息技术（深圳）有限公司 | Automatic session implementation method, server and storage medium |
| CN109829036A (en) * | 2019-02-12 | 2019-05-31 | 浙江核新同花顺网络信息股份有限公司 | A kind of dialogue management method and relevant apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10977452B2 (en) | Multi-lingual virtual personal assistant | |
| CN110288985B (en) | Voice data processing method and device, electronic equipment and storage medium | |
| CN107797984B (en) | Intelligent interaction method, equipment and storage medium | |
| CN108399914B (en) | Voice recognition method and device | |
| KR102544453B1 (en) | Method and device for processing information, and storage medium | |
| CN107515900B (en) | Intelligent robot and event memo system and method thereof | |
| CN111128183B (en) | Speech recognition method, apparatus and medium | |
| CN108257676B (en) | Medical case information processing method, device and equipment | |
| CN111368541A (en) | Named entity identification method and device | |
| RU2733816C1 (en) | Method of processing voice information, apparatus and storage medium | |
| CN111898382A (en) | Named entity recognition method and device for named entity recognition | |
| CN111435595A (en) | Text normalization method and device | |
| CN112836058A (en) | Medical knowledge map establishing method and device and medical knowledge map inquiring method and device | |
| CN112036174B (en) | Punctuation marking method and device | |
| CN110674246A (en) | Question-answering model training method, automatic question-answering method and device | |
| CN114880480A (en) | Question-answering method and device based on knowledge graph | |
| WO2018214663A1 (en) | Voice-based data processing method and apparatus, and electronic device | |
| CN112133295B (en) | Speech recognition method, device and storage medium | |
| WO2021179703A1 (en) | Sign language interpretation method and apparatus, computer device, and storage medium | |
| CN112836026B (en) | Dialogue-based inquiry method and device | |
| CN116166843A (en) | Text video cross-modal retrieval method and device based on fine granularity perception | |
| CN110968246A (en) | Intelligent Chinese handwriting input recognition method and device | |
| CN113539235B (en) | Text analysis and speech synthesis method, device, system and storage medium | |
| CN112836059A (en) | Medical map establishing method and device and medical map inquiring method and device | |
| JP7208968B2 (en) | Information processing method, device 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 |