CN112037920A - Medical knowledge map construction method, device, equipment and storage medium - Google Patents
Medical knowledge map construction method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN112037920A CN112037920A CN202010899991.7A CN202010899991A CN112037920A CN 112037920 A CN112037920 A CN 112037920A CN 202010899991 A CN202010899991 A CN 202010899991A CN 112037920 A CN112037920 A CN 112037920A
- Authority
- CN
- China
- Prior art keywords
- data
- medical
- knowledge
- entities
- medical knowledge
- 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
- 238000010276 construction Methods 0.000 title claims abstract description 71
- 230000004927 fusion Effects 0.000 claims abstract description 116
- 238000000605 extraction Methods 0.000 claims abstract description 62
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 44
- 238000005516 engineering process Methods 0.000 claims abstract description 20
- 230000014509 gene expression Effects 0.000 claims description 22
- 230000011218 segmentation Effects 0.000 claims description 17
- 230000009193 crawling Effects 0.000 claims description 14
- 238000012216 screening Methods 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 238000012544 monitoring process Methods 0.000 claims description 8
- 201000010099 disease Diseases 0.000 claims description 7
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 7
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 208000024891 symptom Diseases 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 3
- 238000000034 method Methods 0.000 abstract description 31
- 238000003745 diagnosis Methods 0.000 abstract description 10
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 8
- 239000003814 drug Substances 0.000 description 6
- 238000011282 treatment Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 4
- 238000005065 mining Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000012552 review Methods 0.000 description 3
- 229960004099 azithromycin Drugs 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 229940079593 drug Drugs 0.000 description 2
- 238000007499 fusion processing Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 241000239290 Araneae Species 0.000 description 1
- 235000004494 Sorbus aria Nutrition 0.000 description 1
- 244000158209 Sorbus aria Species 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- MQTOSJVFKKJCRP-BICOPXKESA-N azithromycin Chemical compound O([C@@H]1[C@@H](C)C(=O)O[C@@H]([C@@]([C@H](O)[C@@H](C)N(C)C[C@H](C)C[C@@](C)(O)[C@H](O[C@H]2[C@@H]([C@H](C[C@@H](C)O2)N(C)C)O)[C@H]1C)(C)O)CC)[C@H]1C[C@@](C)(OC)[C@@H](O)[C@H](C)O1 MQTOSJVFKKJCRP-BICOPXKESA-N 0.000 description 1
- 230000003115 biocidal effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000012795 verification 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
- 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
-
- 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
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- 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
-
- 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
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Public Health (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention relates to the field of artificial intelligence and discloses a medical knowledge map construction method, a medical knowledge map construction device, medical knowledge map construction equipment and a medical knowledge map storage medium. The method determines a knowledge feature extraction mode based on a data acquisition source to analyze medical data in a targeted manner, so that accurate entity names are acquired, entities with different entity names but the same body are fused by using a fusion algorithm, the entities are unified, and finally, a knowledge graph is constructed based on the fused entities, so that the problem of low precision of the knowledge graph caused by non-unified related medical knowledge entities is solved, meanwhile, powerful support and guarantee are provided for diagnosis means and methods in the medical field, and knowledge sharing and management can be completed more reasonably and efficiently. In addition, the invention also relates to a blockchain technology, and the medical knowledge map and the entity can be stored in the blockchain.
Description
Technical Field
The application relates to the field of artificial intelligence, in particular to a medical knowledge graph construction method, device, equipment and storage medium.
Background
The knowledge map is a knowledge base with a map structure, belongs to the field of knowledge engineering, and is an important basic measure for realizing artificial intelligence at present. The knowledge graph is applied, so that the connotation of the original scientific knowledge graph is expanded, and the application scene of the knowledge graph is extended.
However, the application of the current knowledge graph is still limited to the aspects of a search engine, a question-answering system and the like, and the application of other aspects is less. In the medical field, there are usually complex and intricate relationships among diseases, illnesses and diagnosis and treatment means, and the data storage mode of the existing relationship model is not convenient for the expansion of the content of the medical knowledge system, so that the recognition accuracy is not high in the process of recognizing medical entities based on the existing medical knowledge graph, and the work efficiency is affected.
Disclosure of Invention
The invention mainly aims to solve the technical problem that the medical knowledge map constructed by the existing method is low in accuracy.
The invention provides a medical knowledge graph construction method in a first aspect, which comprises the following steps:
crawling medical knowledge from a plurality of data sources using a web crawler tool;
determining a corresponding knowledge feature extraction mode according to the data source of the medical knowledge;
performing knowledge extraction from the medical knowledge according to the knowledge feature extraction mode to form medical knowledge data, wherein the medical knowledge data comprises data of at least two data types;
extracting entities in the medical knowledge data, attribute information of the entities and relationship information among the entities;
calling a preset fusion algorithm to fuse the entities, the attribute information of the entities and the relationship information among the entities to obtain map data;
and constructing a medical knowledge map according to the map data.
Optionally, in a first implementation of the first aspect of the present invention, the medical knowledge comprises structured data, semi-structured data, and unstructured data;
the extracting knowledge from the medical knowledge according to the knowledge feature extraction manner to form medical knowledge data includes:
if the medical knowledge is structured data, extracting the knowledge of the medical data by constructing a regular expression, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the structured data comprises medical professional books;
if the medical knowledge is semi-structured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by using a regular expression and a data index, constructing triple data by using extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the semi-structured data is inquiry data;
if the medical knowledge is unstructured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by utilizing a semantic annotation-based POS-CBOW correlation model algorithm, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the unstructured data is network medical data.
Optionally, in a second implementation manner of the first aspect of the present invention, the extracting knowledge of the medical data through a regular expression and a data index includes:
carrying out word segmentation processing on sentences in the inquiry data according to a regular expression to obtain a word segmentation sequence;
and extracting medical entity characteristic words in the word segmentation sequence by using a TextRank keyword extraction algorithm, wherein the entity characteristic words comprise characteristics of diseases, symptoms, causes and the like.
Optionally, in a third implementation manner of the first aspect of the present invention, the invoking a preset fusion algorithm to fuse the entities, the attribute information of the entities, and the relationship information between the entities to obtain the map data includes:
grouping the extracted entities to obtain a plurality of entity groups;
respectively calculating the similarity between the entities in each entity group through a dynamic programming algorithm;
and merging and removing the duplication of the entities in the same group according to the similarity to obtain the map data.
Optionally, in a fourth implementation manner of the first aspect of the present invention, the invoking a preset fusion algorithm to fuse the entities, the attribute information of the entities, and the relationship information between the entities to obtain the map data includes:
performing ontology abstraction on each entity through a concept abstraction technology to obtain a corresponding entity ontology;
fusing each entity body by adopting a fusion model constructed based on a graph neural network to obtain a preliminary fusion result;
calculating a fusion degree value of the preliminary fusion result, and comparing the fusion degree value with a preset fusion degree;
if the fusion degree is greater than the preset fusion degree, outputting map data;
and if the fusion degree is not greater than the preset fusion degree, re-executing the fusion step until the fusion degree is greater than the preset fusion degree and outputting the map data.
Optionally, in a fifth implementation manner of the first aspect of the present invention, the constructing a medical knowledge-map from the map data includes:
self-learning and reasoning are carried out on the map data by adopting a tensor decomposition algorithm, and multidimensional medical entities are mined from the map data;
and constructing the medical knowledge graph according to the graph data and the medical entity.
Optionally, in a sixth implementation manner of the first aspect of the present invention, after the constructing of the medical knowledge-map according to the map data, the method further includes:
and calling a data monitoring interface to monitor the medical data in the medical data webpage in real time through the Internet, and extracting entities in the medical data and corresponding attribute information to update the entities in the medical data to the medical knowledge map.
A second aspect of the present invention provides a medical knowledge map construction apparatus including:
a crawling module for crawling medical knowledge from a plurality of data sources using a web crawler tool;
the matching module is used for determining a corresponding knowledge feature extraction mode according to the data source of the medical knowledge;
the knowledge extraction module is used for extracting knowledge from the medical knowledge according to the knowledge characteristic extraction mode to form medical knowledge data, wherein the medical knowledge data comprises data of at least two data types;
the entity extraction module is used for extracting entities in the medical knowledge data, attribute information of the entities and relationship information among the entities;
the fusion module is used for calling a preset fusion algorithm to fuse the entities, the attribute information of the entities and the relationship information among the entities to obtain map data;
and the construction module is used for constructing the medical knowledge map according to the map data.
Optionally, in a first implementation of the second aspect of the invention, the medical knowledge comprises structured data, semi-structured data, and unstructured data; the matching module is specifically configured to:
if the medical knowledge is structured data, extracting the knowledge of the medical data by constructing a regular expression, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the structured data comprises medical professional books;
if the medical knowledge is semi-structured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by using a regular expression and a data index, constructing triple data by using extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the semi-structured data is inquiry data;
if the medical knowledge is unstructured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by utilizing a semantic annotation-based POS-CBOW correlation model algorithm, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the unstructured data is network medical data.
Optionally, in a second implementation manner of the second aspect of the present invention, the knowledge extraction module includes:
the word segmentation unit is used for carrying out word segmentation processing on the sentences in the inquiry data according to the regular expression to obtain a word segmentation sequence;
and the entity extraction unit is used for extracting medical entity characteristic words in the word segmentation sequence by using a TextRank keyword extraction algorithm, wherein the entity characteristic words comprise characteristics such as diseases, symptoms and causes.
Optionally, in a third implementation manner of the second aspect of the present invention, the fusion module includes:
the grouping unit is used for grouping the extracted entities to obtain a plurality of entity groups;
the calculating unit is used for respectively calculating the similarity between the entities in each entity group through a dynamic programming algorithm;
and the merging unit is used for merging and removing the duplication of the entities in the same group according to the similarity to obtain the map construction data.
Optionally, in a fourth implementation manner of the second aspect of the present invention, the fusion module includes:
the abstraction unit is used for performing ontology abstraction on each entity through a concept abstraction technology to obtain a corresponding entity ontology;
the fusion unit is used for carrying out fusion processing on each entity body by adopting a fusion model constructed based on a graph neural network to obtain a preliminary fusion result;
the comparison unit is used for calculating the fusion degree value of the preliminary fusion result and comparing the fusion degree value with a preset fusion degree;
the output unit is used for outputting map construction data when the fusion degree value is larger than the preset fusion degree; and when the fusion degree value is not greater than the preset fusion degree, re-executing the fusion step until the fusion degree value is greater than the preset fusion degree and outputting map construction data.
Optionally, in a fifth implementation manner of the second aspect of the present invention, the building module includes:
the mining unit is used for carrying out self-learning and reasoning on the map construction data by adopting a tensor decomposition algorithm and mining multi-dimensional medical entities from the map construction data;
and the construction unit is used for constructing the medical knowledge map according to the map construction data and the medical entity.
Optionally, in a sixth implementation manner of the second aspect of the present invention, the medical knowledge-map constructing apparatus further includes a monitoring module, which is specifically configured to:
and calling a data monitoring interface to monitor the medical data in the medical data webpage in real time through the Internet, and extracting entities in the medical data and corresponding attribute information to update the entities in the medical data to the medical knowledge map.
A third aspect of the present invention provides a medical knowledge map construction apparatus comprising: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line;
the at least one processor invokes the instructions in the memory to cause the medical knowledge-map construction apparatus to perform the medical knowledge-map construction method described above.
A fourth aspect of the present invention provides a computer-readable storage medium having stored therein a computer program which, when run on a computer, causes the computer to execute the medical knowledge map construction method described above.
In the technical scheme provided by the invention, the medical data is pertinently analyzed by determining the knowledge characteristic extraction mode based on the acquired data source, so that accurate entity names are obtained, entities with different entity names but the same body are fused by using a fusion algorithm, the entities are unified, and finally, the knowledge graph is constructed based on the fused entities, so that the problem of low precision of the knowledge graph caused by non-uniform related medical knowledge entities is solved, meanwhile, powerful support and guarantee are provided for diagnosis means and methods in the medical field, and knowledge sharing and management can be completed more reasonably and efficiently.
Drawings
FIG. 1 is a schematic diagram of a first embodiment of a medical knowledge map construction method in an embodiment of the invention;
FIG. 2 is a schematic diagram of a second embodiment of a medical knowledge map construction method according to an embodiment of the invention;
FIG. 3 is a diagram of a third embodiment of a medical knowledge map construction method according to an embodiment of the invention;
FIG. 4 is a schematic diagram of one embodiment of a medical knowledge map construction apparatus in an embodiment of the invention;
FIG. 5 is a schematic diagram of another embodiment of a medical knowledge map construction apparatus in an embodiment of the invention;
fig. 6 is a schematic diagram of an embodiment of the medical knowledge map construction apparatus in the embodiment of the invention.
Detailed Description
Aiming at the problem of low accuracy of the existing medical entity identification and knowledge graph construction, the application provides a method for solving the problems of information acquisition and knowledge fusion of the knowledge graph by using the technologies of a regular expression, a hidden Markov model, dependency syntax analysis, a graph neural network and the like. The problems of low precision and unreasonable utilization of structural information are solved by adopting a means of combining manual review and a computer. The completeness of the medical knowledge base is further ensured by the dynamic knowledge map updating and complementing mechanism.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
For understanding, a specific flow of an embodiment of the present invention is described below, and referring to fig. 1, a first embodiment of a medical knowledge graph construction method according to an embodiment of the present invention includes:
101. crawling medical knowledge from a plurality of data sources using a web crawler tool;
in this embodiment, the plurality of data sources at least include a network data source, a diagnosis and treatment data source, and a medical professional book data source, and corresponding data, such as medical textbooks, documents, monographs, and on-line inquiry data, in practical applications, different data sources can be acquired by the web crawler tool in a unified manner, and different tools can be selected according to different data sources.
Specifically, for a web data source, it may use a web crawler (also called a web spider, a web robot, a program for automatically crawling web information or a script program according to certain rules) to crawl a web page of a target website, and then obtain medically-related knowledge in the web page by analyzing a source code of the web page.
For the diagnosis and treatment data source, the diagnosis and treatment data source can be directly analyzed from an electronic medical record in a medical database, and then the analyzed content is formed into a text to obtain medical knowledge.
For a medical professional book data source, reading an electronic text from a webpage library or a database of a thesis website, and then extracting knowledge of the electronic document to obtain medical knowledge; furthermore, for books in non-electronic format, the books need to be converted into texts in electronic format by scanning or photo recognition, and then extracted.
102. Determining a corresponding knowledge feature extraction mode according to a data source of medical knowledge;
when determining a knowledge feature extraction mode, firstly, determining a data type of medical knowledge based on a data source of the medical knowledge, and inquiring a knowledge feature extraction mode corresponding to the medical data of the currently determined type based on the corresponding relation between the data type and the knowledge feature extraction mode, wherein the data type at least comprises a structured medical professional book, unstructured network medical data and semi-structured patient diagnosis data.
103. Performing knowledge extraction from medical knowledge according to a knowledge feature extraction mode to form medical knowledge data;
in this embodiment, the medical knowledge data includes data of at least two data types, the data formats obtained from different data sources are different, that is, the data types are different, the types of the data formats at least include a structured type, a semi-structured type and an unstructured type, in practical applications, both the structured type data (e.g., data managed in a relational data table format) and the semi-structured type data (e.g., log file, XML document, JSON document, etc.) refer to data having a basic fixed structure pattern, and the unstructured type data (e.g., word document, PDF document, PPT document, etc.) refer to data without a fixed structure pattern.
For data with a fixed structure mode, when knowledge extraction is realized, extraction can be specifically performed according to labeling information of rows and columns in text data, for example, diagnostic data, and when the extraction is performed, extraction is performed through item fields in a diagnostic sheet, and data corresponding to the extracted item fields are divided according to a triple division principle to construct triple knowledge data, so that medical knowledge data are obtained.
For data without a fixed structure mode, firstly converting captured data into text data, then extracting sentences based on the text data to obtain a sentence set, then calling a preset knowledge extraction template to extract medical or medical characteristic words and relations among the characteristic words in the sentence set, and finally dividing the extracted data according to a triple data division principle to obtain medical knowledge data.
And finally, combining the data with the fixed structure mode and the data without the fixed structure mode to obtain the final medical knowledge data.
104. Extracting entities, attribute information of the entities and relationship information among the entities in the medical knowledge data;
in this step, the method may specifically include extracting entities, attribute information of the entities, and relationship information between the entities from the medical knowledge data, and storing the extracted entities, attribute information of the entities, and relationship information between the entities in a form of a two-dimensional table, where the extraction of the entities may be implemented by an entity naming recognition model trained based on medical text data with entity labels.
In practical application, the relation between the entity and the attribute can be extracted by analyzing the medical knowledge data through dependency syntax, the dependency syntax mainly explains the syntax structure by analyzing the dependency relation before the components in the language unit, the core verb in the sentence is claimed to be the center component which governs other components, and the core word is obtained based on the analysis to analyze the dependency relation before and after, so that the entity and the attribute and the relation information of the entity are obtained.
105. Calling a preset fusion algorithm to fuse each entity, attribute information of each entity and relationship information among the entities to obtain map data;
in practical application, because the sources of knowledge in the medical database are different, and there are situations of various different formats or structures, the names of the entities of the data from different sources are different, which results in the problems of uneven knowledge quality, repeated meanings, and the like. In this regard, the step integrates, disambiguates and processes the entities through knowledge fusion, enhances the logic and expression ability inside the knowledge base, and updates old knowledge or supplements new knowledge for the medical knowledge map.
In this embodiment, the fusion can be specifically realized by combining a fusion model of a graph neural network with manual verification, or by directly selecting the fusion model to realize alone, but the fusion model needs to be trained and tested, and after the online requirement is met, the model can be used to fuse the entities;
further, before using the fusion model, the method further comprises: the entity is abstracted, the entity cover surface of the entity is abstracted, the entity concept is input into the fusion model, the fusion model calculates the similarity between the entities and the parameters such as editing distance, semantics and the like, and the same entities are combined based on the calculated parameters, so that the diversity of entity naming is reduced, and the unity of the entities is ensured.
106. And constructing the medical knowledge map according to the map data.
In this embodiment, when the medical knowledge graph is constructed based on the graph data, the medical knowledge graph may be constructed by inputting the graph data into a preset graph tree. Here, the map tree refers to a tree structure diagram including a plurality of parent nodes and child nodes, and specifically, includes a plurality of designated concepts in the medical knowledge system, as an example, the plurality of designated concepts at least includes: diseases, causes, symptoms, drugs, examinations, organs, treatments, etc. In the embodiment of the invention, the designated concept is taken as a root node, the lower concept of the designated concept is taken as a middle node, and the entity is taken as a leaf node to construct the graph tree aiming at each designated concept.
For example, the map tree of "medicine" is a sub-concept of "medicine" which belongs to the middle node, and further, the map tree of "medicine" is a sub-concept of "antibiotic" which belongs to the middle node, and the map tree of "medicine" is a specific entity of "azithromycin" which belongs to the leaf node. For convenience of description, the example of the map tree "drugs" may then be denoted as "drugs-antibiotics-erythromycin-antibiotics-azithromycin".
When the medical knowledge graph is constructed based on the graph tree, the similarity between the graph data and entities on nodes in the graph tree is calculated, if the similarity reaches a threshold value, whether the entities in the graph data are the same as the entities on the nodes is judged, the same should be understood that the information represented by the entities is the same or contained, if the information which is not contained in the tree nodes exists in the graph data, the graph data needs to be added to the tree nodes, and if the information does not exist, the graph data is kept unchanged.
Of course, in practical application, besides the above modification, the construction may also be performed according to the skeleton of the atlas tree and according to the entity, attribute and relationship information in the atlas data, so as to perform the combing of node relationship, and form the medical knowledge atlas.
In this embodiment, it may also be to construct a medical knowledge map: self-learning and reasoning are carried out on the map data by adopting a tensor decomposition algorithm, and the medical entity with dimensionality is mined from the map data; and constructing the medical knowledge graph according to the graph data and the medical entity. In practical application, in order to ensure the accuracy of the atlas construction data, the data can be stored in a block chain mode, so that the uniqueness of the data in the process of constructing the medical knowledge atlas is ensured.
Specifically, the disambiguated entities, relationships and attributes are manually checked and evaluated and then stored in a graph database, a medical knowledge map knowledge body is constructed, knowledge fusion of various source data is completed, and a medical knowledge map with strong professional performance and rich content is constructed. The self-learning and reasoning can be carried out by adopting a tensor decomposition algorithm, target information is mined from the current knowledge graph, and the target information is fed back and stored in the knowledge graph as a part of knowledge.
According to the embodiment of the method, the medical data are analyzed in a targeted manner by determining the knowledge feature extraction mode based on the acquired data source, so that accurate entity names are obtained, entities with different entity names but the same body are fused by using a fusion algorithm to unify the entities, and finally, the knowledge graph is constructed based on the fused entities, so that the problem of low precision of the knowledge graph caused by non-uniformity of related medical knowledge entities is solved;
furthermore, the problems of information acquisition and knowledge fusion of the knowledge graph are solved by using the technologies such as a regular expression and a graph neural network. The problems of low precision and unreasonable utilization of structural information are solved by adopting a means of combining manual review and a computer. The dynamic knowledge map updating and complementing mechanism further ensures the completeness of the medical knowledge base, simultaneously provides more powerful support and guarantee for the diagnosis means and method in the medical field, and can complete knowledge sharing and management more reasonably and efficiently.
Referring to fig. 2, a second embodiment of the medical knowledge-map constructing method according to the embodiment of the present invention includes:
201. crawling medical knowledge from a plurality of data sources using a web crawler tool;
202. determining a type of medical knowledge;
in this step, the types of medical knowledge include a structured type, a semi-structured type, and a structured type.
In practical application, for the determination of the type of the medical knowledge, the data source is specifically determined by detecting an interface for acquiring data when the web crawler tool acquires the medical knowledge, for example, in the embodiment, three data sources are summarized, which are respectively a hospital, a web page and a medical book, and a medical system of each data source is provided with an individual access interface, and the medical system accesses the database of the corresponding data source through the authorization information access interface to acquire the corresponding data, thereby determining the corresponding data type.
203. Determining a corresponding knowledge feature extraction mode according to a data source of medical knowledge;
in the step, the knowledge feature extraction mode is obtained by querying through a corresponding relation table, wherein the corresponding relation table is provided with a unique feature extraction mode in advance according to different data types, after the data source is determined, the interface ID is determined based on the corresponding relation between the data source and the interface, the field with the same ID in the corresponding relation is queried based on the interface ID, the line content corresponding to the field is extracted, so that corresponding data type information is obtained, and the knowledge feature extraction mode is determined based on the data type information.
204. If the medical knowledge is structured data, performing knowledge extraction on the medical data by constructing a regular expression, constructing triple data by using extracted knowledge characteristics, and taking the triple data as medical knowledge data;
in this step, the structured data comprises a medical professional book.
205. If the medical knowledge is semi-structured data, screening the medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by using a regular expression and a data index, constructing triple data by using extracted knowledge characteristics, and taking the triple data as medical knowledge data;
in this step, the semi-structured data is interrogation data.
206. If the medical knowledge is unstructured data, screening the medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by utilizing a semantic annotation-based POS-CBOW correlation model algorithm, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data;
in this step, the unstructured data is network medical data.
In practical application, for structured data, the constructing of the regular expression to perform knowledge extraction refers to: extracting useful information from the acquired structured data, converting the information into an XML file, and converting the XML file into an RDF file through a recursive algorithm according to a medical knowledge map model;
for unstructured data, extracting knowledge entities, relations and attributes of the unstructured data by using a semantic annotation-based POS-CBOW association model algorithm sequentially comprises the following steps:
setting a mode of the unstructured phrases in combination with (medical field) knowledge and the mode to segment unstructured data;
removing redundancy of the segmented words by adopting a POS-CBOW association model algorithm to obtain corresponding entities, relations and attributes:
in the above equation, Sim (Vi, Vj) is the cosine similarity of two different entities Vi and Vj, Set (Vi, Vj) is the part-of-speech similarity of Vi and Vj, depth Vi is the level of entity Vi, and Dist (Vi, Vj) is the distance between Vi and Vj in the level tree.
In the present embodiment, for knowledge extraction of medical knowledge of semi-structured data, the implementation steps include:
carrying out word segmentation processing on sentences in the inquiry data according to a regular expression to obtain a word segmentation sequence;
and extracting medical entity characteristic words in the word segmentation sequence by using a TextRank keyword extraction algorithm, wherein the entity characteristic words comprise characteristics of diseases, symptoms, causes and the like.
207. Extracting entities, attribute information of the entities and relationship information among the entities in the medical knowledge data;
208. calling a preset fusion algorithm to fuse each entity, attribute information of each entity and relationship information among the entities to obtain map data;
209. and constructing the medical knowledge map according to the map data.
In practical applications, the step 207-.
In the embodiment, the medical data is analyzed in a targeted manner by determining the knowledge feature extraction mode based on the acquired data source, so that accurate entity names are acquired, entities with different entity names but the same body are fused by using a fusion algorithm to unify the entities, and finally, the knowledge graph is constructed based on the fused entities, so that the problem of low precision of the knowledge graph caused by non-uniformity of medical knowledge entities is solved.
Referring to fig. 3, a third embodiment of the medical knowledge-map constructing method according to the embodiment of the present invention includes:
301. crawling medical knowledge from a plurality of data sources using a web crawler tool;
302. determining a corresponding knowledge feature extraction mode according to a data source of medical knowledge;
303. performing knowledge extraction from medical knowledge according to a knowledge feature extraction mode to form medical knowledge data;
304. extracting entities, attribute information of the entities and relationship information among the entities in the medical knowledge data;
305. performing ontology abstraction on each entity through a concept abstraction technology to obtain a corresponding entity ontology;
306. fusing each entity body by adopting a fusion model constructed based on a graph neural network to obtain a primary fusion result;
in practical application, the process of fusing the entity ontology through the fusion model specifically comprises the following steps:
selecting part of entity bodies as fusion objects, determining the attributes of the fusion objects, and calculating the edit distance of the attributes between every two fusion objects through a dynamic programming algorithm to obtain the attribute similarity, wherein the calculation formula is as follows:
in the above formula, D (i, j) is the minimum edit distance for converting the attribute i to the attribute j, M is the number of characters of the attribute j, N is the number of characters of the attribute i, and +1 represents the cost of the insertion, deletion, and replacement operations;
and calculating the weight of the attributes before setting similar entities based on the similarity, calculating the weighted value between the entities, and performing entity body alignment and relationship information fusion through the weighted value so as to obtain a preliminary fusion result.
307. Calculating a fusion degree value of the preliminary fusion result, and comparing the fusion degree value with a preset fusion degree;
308. if the fusion degree is greater than the preset fusion degree, outputting map data;
309. and if the fusion degree is not greater than the preset fusion degree, re-executing the fusion step until the fusion degree is greater than the preset fusion degree and outputting the map data.
In practical application, for the preliminary fusion result which does not meet the preset fusion degree, the fusion number of the entities can be reduced, so that the preliminary fusion result is recalculated, the fusion degree is calculated, and whether the condition is met or not is judged.
In order to further ensure the accuracy of the medical knowledge graph, the construction method provided by this embodiment further includes monitoring external data in real time, and updating the entity of the medical knowledge graph according to the monitored result, and the specific implementation steps are as follows:
after the medical knowledge map is constructed according to the map data, the method further comprises the following steps:
and calling a data monitoring interface to monitor the medical data in the medical data webpage in real time through the Internet, and extracting entities in the medical data and corresponding attribute information to update the entities in the medical data to the medical knowledge map.
In this embodiment, the step of calling the preset fusion algorithm to fuse the entities, the attribute information of the entities, and the relationship information between the entities to obtain the map data may be specifically implemented by the following method:
grouping the extracted entities to obtain a plurality of entity groups;
respectively calculating the similarity between the entities in each entity group through a dynamic programming algorithm;
and merging and removing the duplication of the entities in the same group according to the similarity to obtain the map data.
In conclusion, the proposal provides a corresponding solution by solving the problem that related medical knowledge entities are not unified in the construction process of the medical knowledge map. The information acquisition and knowledge fusion problems of the knowledge graph are solved by utilizing the technologies of a regular expression, a hidden Markov model, dependency syntax analysis, a graph neural network and the like. The problems of low precision and unreasonable utilization of structural information are solved by adopting a means of combining manual review and a computer. The completeness of the medical knowledge base is further ensured by the dynamic knowledge map updating and complementing mechanism.
The effective utilization of medical data and the construction method of the map can provide more powerful support and guarantee for the diagnosis means and method in the medical field, and can complete knowledge sharing and management more reasonably and efficiently.
With reference to fig. 4, the medical knowledge graph constructing apparatus according to an embodiment of the present invention is described above, and a first embodiment of the medical knowledge graph constructing apparatus according to an embodiment of the present invention includes:
a crawling module 401 for crawling medical knowledge from multiple data sources using a web crawler tool;
a matching module 402, configured to determine a corresponding knowledge feature extraction manner according to a data source of the medical knowledge;
a knowledge extraction module 403, configured to perform knowledge extraction from the medical knowledge according to the knowledge feature extraction manner to form medical knowledge data, where the medical knowledge data includes data of at least two data types;
an entity extraction module 404, configured to extract entities in the medical knowledge data, attribute information of the entities, and relationship information between the entities;
a fusion module 405, configured to invoke a preset fusion algorithm to fuse the entities, the attribute information of the entities, and the relationship information between the entities, so as to obtain map data;
a construction module 406 for constructing a medical knowledge graph according to the graph data.
In this embodiment, the medical knowledge graph construction apparatus operates the medical knowledge graph construction method, and the method determines a knowledge feature extraction manner based on a data source to analyze medical data in a targeted manner, so as to obtain precise entity names, and further fuses entities with different entity names but the same entity by using a fusion algorithm to unify the entities, and finally constructs a knowledge graph based on the fused entities, so that the problem of low knowledge graph precision caused by non-unity of related medical knowledge entities is solved, meanwhile, powerful support and guarantee are provided for diagnosis means and methods in the medical field, and knowledge sharing and management can be completed more reasonably and efficiently.
Referring to fig. 5, a second embodiment of the medical knowledge-map constructing apparatus according to the embodiment of the present invention specifically includes:
a crawling module 401 for crawling medical knowledge from multiple data sources using a web crawler tool;
a matching module 402, configured to determine a corresponding knowledge feature extraction manner according to a data source of the medical knowledge;
a knowledge extraction module 403, configured to perform knowledge extraction from the medical knowledge according to the knowledge feature extraction manner to form medical knowledge data, where the medical knowledge data includes data of at least two data types;
an entity extraction module 404, configured to extract entities in the medical knowledge data, attribute information of the entities, and relationship information between the entities;
a fusion module 405, configured to invoke a preset fusion algorithm to fuse the entities, the attribute information of the entities, and the relationship information between the entities, so as to obtain map data;
a construction module 406 for constructing a medical knowledge graph according to the graph data.
Optionally, the medical knowledge comprises structured data, semi-structured data, and unstructured data; the matching module 402 is specifically configured to:
if the medical knowledge is structured data, extracting the knowledge of the medical data by constructing a regular expression, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the structured data comprises medical professional books;
if the medical knowledge is semi-structured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by using a regular expression and a data index, constructing triple data by using extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the semi-structured data is inquiry data;
if the medical knowledge is unstructured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by utilizing a semantic annotation-based POS-CBOW correlation model algorithm, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the unstructured data is network medical data.
Optionally, the knowledge extraction module 403 includes:
a word segmentation unit 4031, configured to perform word segmentation processing on the sentences in the inquiry data according to the regular expression to obtain a word segmentation sequence;
and the entity extraction unit 4032 is used for extracting medical entity feature words in the word segmentation sequence by using a TextRank keyword extraction algorithm, wherein the entity feature words comprise features such as diseases, symptoms and causes.
Optionally, the fusion module 405 includes:
a grouping unit 4051, configured to group the extracted entities to obtain a plurality of entity groups;
a calculating unit 4052, configured to calculate similarity between entities in each entity group through a dynamic programming algorithm;
the merging unit 4053 is configured to merge and deduplicate the entities in the same group according to the similarity, so as to obtain map construction data.
Optionally, the fusion module 405 includes:
an abstraction unit 4054, configured to perform ontology abstraction on each entity through a concept abstraction technology to obtain a corresponding entity ontology;
the fusion unit 4055 is configured to perform fusion processing on each entity body by using a fusion model constructed based on a graph neural network to obtain a preliminary fusion result;
a comparing unit 4056, configured to calculate a fusion degree value of the preliminary fusion result, and compare the fusion degree value with a preset fusion degree;
the output unit 4057 is configured to output map construction data when the fusion degree value is greater than a preset fusion degree; and when the fusion degree value is not greater than the preset fusion degree, re-executing the fusion step until the fusion degree value is greater than the preset fusion degree and outputting map construction data.
Optionally, the building module 406 includes:
the mining unit 4061 is used for self-learning and reasoning the map construction data by adopting a tensor decomposition algorithm, and mining multi-dimensional medical entities from the map construction data;
a constructing unit 4062, configured to construct the medical knowledge graph according to the graph construction data and the medical entity.
The medical knowledge map constructing apparatus further includes a monitoring module 407, which is specifically configured to:
and calling a data monitoring interface to monitor the medical data in the medical data webpage in real time through the Internet, and extracting entities in the medical data and corresponding attribute information to update the entities in the medical data to the medical knowledge map.
The medical knowledge map construction device in the embodiment of the present invention is described in detail in the above fig. 4 and fig. 5 from the perspective of the modular functional entity, and the medical knowledge map construction equipment in the embodiment of the present invention is described in detail in the following from the perspective of hardware processing, and the medical knowledge map construction device can be configured in a plug-in manner to implement identification of medical data with the medical knowledge map construction equipment.
Fig. 6 is a schematic structural diagram of a medical knowledge map constructing apparatus 600 according to an embodiment of the present invention, which may have relatively large differences due to different configurations or performances, and may include one or more processors (CPUs) 610 (e.g., one or more processors) and a memory 620, and one or more storage media 630 (e.g., one or more mass storage devices) for storing applications 633 or data 632. Memory 620 and storage medium 630 may be, among other things, transient or persistent storage. The program stored on the storage medium 630 may include one or more modules (not shown), each of which may include a sequence of instructions operating on the medical knowledge map construction apparatus 600. Still further, the processor 610 may be configured to communicate with the storage medium 630 and execute a series of instruction operations in the storage medium 630 on the medical knowledge map construction apparatus 600 to implement the steps of the medical knowledge map construction method described above.
The medical knowledge map construction apparatus 600 may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input-output interfaces 660, and/or one or more operating systems 631, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, and the like. Those skilled in the art will appreciate that the medical knowledge map construction apparatus configuration shown in FIG. 6 does not constitute a limitation of the medical knowledge map construction apparatus provided herein, and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, or a volatile computer-readable storage medium, and the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to execute the steps of the medical knowledge map construction method provided in the above embodiments.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will 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; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A medical knowledge graph construction method is characterized by comprising the following steps:
crawling medical knowledge from a plurality of data sources using a web crawler tool;
determining a corresponding knowledge feature extraction mode according to the data source of the medical knowledge;
performing knowledge extraction from the medical knowledge according to the knowledge feature extraction mode to form medical knowledge data, wherein the medical knowledge data comprises data of at least two data types;
extracting entities in the medical knowledge data, attribute information of the entities and relationship information among the entities;
calling a preset fusion algorithm to fuse the entities, the attribute information of the entities and the relationship information among the entities to obtain map data;
and constructing a medical knowledge map according to the map data.
2. The medical knowledge map construction method of claim 1, wherein the medical knowledge comprises structured data, semi-structured data, and unstructured data;
the extracting knowledge from the medical knowledge according to the knowledge feature extraction manner to form medical knowledge data includes:
if the medical knowledge is structured data, extracting the knowledge of the medical data by constructing a regular expression, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the structured data comprises medical professional books;
if the medical knowledge is semi-structured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by using a regular expression and a data index, constructing triple data by using extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the semi-structured data is inquiry data;
if the medical knowledge is unstructured data, screening medical data from the medical knowledge by adopting a Hadoop big data technology, extracting the medical data by utilizing a semantic annotation-based POS-CBOW correlation model algorithm, constructing triple data by using the extracted knowledge characteristics, and taking the triple data as medical knowledge data, wherein the unstructured data is network medical data.
3. The medical knowledge graph construction method according to claim 2, wherein the extracting of knowledge of the medical data through regular expressions and data indicators comprises:
carrying out word segmentation processing on sentences in the inquiry data according to a regular expression to obtain a word segmentation sequence;
and extracting medical entity characteristic words in the word segmentation sequence by using a TextRank keyword extraction algorithm, wherein the entity characteristic words comprise characteristics of diseases, symptoms, causes and the like.
4. The medical knowledge graph construction method according to claim 2, wherein the step of calling a preset fusion algorithm to fuse the entities, the attribute information of the entities and the relationship information between the entities to obtain graph data comprises the steps of:
grouping the extracted entities to obtain a plurality of entity groups;
respectively calculating the similarity between the entities in each entity group through a dynamic programming algorithm;
and merging and removing the duplication of the entities in the same group according to the similarity to obtain the map data.
5. The medical knowledge graph construction method according to claim 2, wherein the step of calling a preset fusion algorithm to fuse the entities, the attribute information of the entities and the relationship information between the entities to obtain graph data comprises the steps of:
performing ontology abstraction on each entity through a concept abstraction technology to obtain a corresponding entity ontology;
fusing each entity body by adopting a fusion model constructed based on a graph neural network to obtain a preliminary fusion result;
calculating a fusion degree value of the preliminary fusion result, and comparing the fusion degree value with a preset fusion degree;
if the fusion degree is greater than the preset fusion degree, outputting map data;
and if the fusion degree is not greater than the preset fusion degree, re-executing the fusion step until the fusion degree is greater than the preset fusion degree and outputting the map data.
6. The medical knowledge-map construction method according to any one of claims 1-5, wherein the construction of a medical knowledge-map from the map data comprises:
self-learning and reasoning are carried out on the map data by adopting a tensor decomposition algorithm, and multidimensional medical entities are mined from the map data;
and constructing the medical knowledge graph according to the graph data and the medical entity.
7. The medical knowledge-map construction method according to any one of claims 1-5, further comprising, after said construction of a medical knowledge-map from said map data:
and calling a data monitoring interface to monitor the medical data in the medical data webpage in real time through the Internet, and extracting entities in the medical data and corresponding attribute information to update the entities in the medical data to the medical knowledge map.
8. A medical knowledge map construction apparatus, characterized in that the medical knowledge map construction apparatus comprises:
a crawling module for crawling medical knowledge from a plurality of data sources using a web crawler tool;
the matching module is used for determining a corresponding knowledge feature extraction mode according to the data source of the medical knowledge;
the knowledge extraction module is used for extracting knowledge from the medical knowledge according to the knowledge characteristic extraction mode to form medical knowledge data, wherein the medical knowledge data comprises data of at least two data types;
the entity extraction module is used for extracting entities in the medical knowledge data, attribute information of the entities and relationship information among the entities;
the fusion module is used for calling a preset fusion algorithm to fuse the entities, the attribute information of the entities and the relationship information among the entities to obtain map data;
and the construction module is used for constructing the medical knowledge map according to the map data.
9. A medical knowledge map construction apparatus, characterized in that the medical knowledge map construction apparatus comprises: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line;
the at least one processor invokes the instructions in the memory to cause the medical knowledge graph construction apparatus to perform the medical knowledge graph construction method of any one of claims 1-7.
10. A computer-readable storage medium, having stored thereon a computer program, wherein the computer program, when executed by a processor, implements a medical knowledge map construction method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010899991.7A CN112037920A (en) | 2020-08-31 | 2020-08-31 | Medical knowledge map construction method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010899991.7A CN112037920A (en) | 2020-08-31 | 2020-08-31 | Medical knowledge map construction method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112037920A true CN112037920A (en) | 2020-12-04 |
Family
ID=73586643
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010899991.7A Pending CN112037920A (en) | 2020-08-31 | 2020-08-31 | Medical knowledge map construction method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112037920A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112466463A (en) * | 2020-12-10 | 2021-03-09 | 求臻医学科技（北京）有限公司 | Intelligent answering system based on tumor accurate diagnosis and treatment knowledge graph |
| CN112711626A (en) * | 2020-12-29 | 2021-04-27 | 长威信息科技发展股份有限公司 | Unified convergence method and system for multi-source heterogeneous data |
| CN112732941A (en) * | 2021-01-15 | 2021-04-30 | 医渡云（北京）技术有限公司 | Model-based medical knowledge graph construction method, device, equipment and medium |
| CN112732940A (en) * | 2021-01-15 | 2021-04-30 | 医渡云（北京）技术有限公司 | Model-based medical knowledge graph reasoning method, device, equipment and medium |
| CN112784065A (en) * | 2021-02-01 | 2021-05-11 | 东北大学 | Unsupervised knowledge graph fusion method and unsupervised knowledge graph fusion device based on multi-order neighborhood attention network |
| CN112905746A (en) * | 2021-03-08 | 2021-06-04 | 国能大渡河流域水电开发有限公司 | System archive knowledge mining processing method based on knowledge graph technology |
| CN113065003A (en) * | 2021-04-22 | 2021-07-02 | 国际关系学院 | Knowledge graph generation method based on multiple indexes |
| CN113190689A (en) * | 2021-05-25 | 2021-07-30 | 广东电网有限责任公司广州供电局 | Construction method, device, equipment and medium of electric power safety knowledge graph |
| CN113420161A (en) * | 2021-06-24 | 2021-09-21 | 平安科技（深圳）有限公司 | Node text fusion method and device, computer equipment and storage medium |
| CN113434692A (en) * | 2021-06-22 | 2021-09-24 | 上海交通大学医学院附属仁济医院 | Method, system and equipment for constructing graph neural network model and recommending diagnosis and treatment scheme |
| CN113486187A (en) * | 2021-03-24 | 2021-10-08 | 平安科技（深圳）有限公司 | Buddhism knowledge graph construction method, device, equipment and storage medium |
| CN113643821A (en) * | 2021-10-13 | 2021-11-12 | 浙江大学 | Multi-center knowledge graph joint decision support method and system |
| CN113707297A (en) * | 2021-08-26 | 2021-11-26 | 平安国际智慧城市科技股份有限公司 | Medical data processing method, device, equipment and storage medium |
| CN114400099A (en) * | 2021-12-31 | 2022-04-26 | 北京华彬立成科技有限公司 | Disease information mining and searching method and device, electronic equipment and storage medium |
| CN114969386A (en) * | 2022-08-03 | 2022-08-30 | 神州医疗科技股份有限公司 | Disambiguation method, apparatus, electronic device, and medium applied to medical field |
| CN115062628A (en) * | 2022-06-15 | 2022-09-16 | 北京信息科技大学 | Automatic simulation method for doctor-patient communication conversation based on knowledge graph |
| CN115630697A (en) * | 2022-10-26 | 2023-01-20 | 泸州职业技术学院 | Knowledge graph construction method and system capable of distinguishing single-phase and double-phase affective disorder |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021281A (en) * | 2016-04-29 | 2016-10-12 | 京东方科技集团股份有限公司 | Method for establishing medical knowledge graph, device for same and query method for same |
| CN107609163A (en) * | 2017-09-15 | 2018-01-19 | 南京深数信息科技有限公司 | Generation method, storage medium and the server of medical knowledge collection of illustrative plates |
| CN109271530A (en) * | 2018-10-17 | 2019-01-25 | 长沙瀚云信息科技有限公司 | A kind of disease knowledge map construction method and plateform system, equipment, storage medium |
| CN109471948A (en) * | 2018-11-08 | 2019-03-15 | 威海天鑫现代服务技术研究院有限公司 | A kind of the elder's health domain knowledge question answering system construction method |
| CN109885698A (en) * | 2019-02-13 | 2019-06-14 | 北京航空航天大学 | A kind of knowledge mapping construction method and device, electronic equipment |
| CN110222201A (en) * | 2019-06-26 | 2019-09-10 | 中国医学科学院医学信息研究所 | A kind of disease that calls for specialized treatment knowledge mapping construction method and device |
| CN110674311A (en) * | 2019-09-05 | 2020-01-10 | 国家电网有限公司 | Knowledge graph-based power asset heterogeneous data fusion method |
| CN110704631A (en) * | 2019-08-16 | 2020-01-17 | 北京紫冬认知科技有限公司 | Construction method and device of medical knowledge map |
-
2020
- 2020-08-31 CN CN202010899991.7A patent/CN112037920A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021281A (en) * | 2016-04-29 | 2016-10-12 | 京东方科技集团股份有限公司 | Method for establishing medical knowledge graph, device for same and query method for same |
| CN107609163A (en) * | 2017-09-15 | 2018-01-19 | 南京深数信息科技有限公司 | Generation method, storage medium and the server of medical knowledge collection of illustrative plates |
| CN109271530A (en) * | 2018-10-17 | 2019-01-25 | 长沙瀚云信息科技有限公司 | A kind of disease knowledge map construction method and plateform system, equipment, storage medium |
| CN109471948A (en) * | 2018-11-08 | 2019-03-15 | 威海天鑫现代服务技术研究院有限公司 | A kind of the elder's health domain knowledge question answering system construction method |
| CN109885698A (en) * | 2019-02-13 | 2019-06-14 | 北京航空航天大学 | A kind of knowledge mapping construction method and device, electronic equipment |
| CN110222201A (en) * | 2019-06-26 | 2019-09-10 | 中国医学科学院医学信息研究所 | A kind of disease that calls for specialized treatment knowledge mapping construction method and device |
| CN110704631A (en) * | 2019-08-16 | 2020-01-17 | 北京紫冬认知科技有限公司 | Construction method and device of medical knowledge map |
| CN110674311A (en) * | 2019-09-05 | 2020-01-10 | 国家电网有限公司 | Knowledge graph-based power asset heterogeneous data fusion method |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112466463A (en) * | 2020-12-10 | 2021-03-09 | 求臻医学科技（北京）有限公司 | Intelligent answering system based on tumor accurate diagnosis and treatment knowledge graph |
| CN112466463B (en) * | 2020-12-10 | 2023-08-18 | 求臻医学科技(浙江)有限公司 | Intelligent answering system based on tumor accurate diagnosis and treatment knowledge graph |
| CN112711626A (en) * | 2020-12-29 | 2021-04-27 | 长威信息科技发展股份有限公司 | Unified convergence method and system for multi-source heterogeneous data |
| CN112732941A (en) * | 2021-01-15 | 2021-04-30 | 医渡云（北京）技术有限公司 | Model-based medical knowledge graph construction method, device, equipment and medium |
| CN112732940A (en) * | 2021-01-15 | 2021-04-30 | 医渡云（北京）技术有限公司 | Model-based medical knowledge graph reasoning method, device, equipment and medium |
| CN112732940B (en) * | 2021-01-15 | 2023-07-14 | 医渡云（北京）技术有限公司 | Reasoning method, device, equipment and medium of medical knowledge graph based on model |
| CN112732941B (en) * | 2021-01-15 | 2023-07-07 | 医渡云（北京）技术有限公司 | Method, device, equipment and medium for constructing medical knowledge graph based on model |
| CN112784065A (en) * | 2021-02-01 | 2021-05-11 | 东北大学 | Unsupervised knowledge graph fusion method and unsupervised knowledge graph fusion device based on multi-order neighborhood attention network |
| CN112784065B (en) * | 2021-02-01 | 2023-07-14 | 东北大学 | Unsupervised knowledge graph fusion method and device based on multi-order neighborhood attention network |
| CN112905746A (en) * | 2021-03-08 | 2021-06-04 | 国能大渡河流域水电开发有限公司 | System archive knowledge mining processing method based on knowledge graph technology |
| CN113486187A (en) * | 2021-03-24 | 2021-10-08 | 平安科技（深圳）有限公司 | Buddhism knowledge graph construction method, device, equipment and storage medium |
| CN113065003B (en) * | 2021-04-22 | 2023-05-26 | 国际关系学院 | Knowledge graph generation method based on multiple indexes |
| CN113065003A (en) * | 2021-04-22 | 2021-07-02 | 国际关系学院 | Knowledge graph generation method based on multiple indexes |
| CN113190689A (en) * | 2021-05-25 | 2021-07-30 | 广东电网有限责任公司广州供电局 | Construction method, device, equipment and medium of electric power safety knowledge graph |
| CN113434692A (en) * | 2021-06-22 | 2021-09-24 | 上海交通大学医学院附属仁济医院 | Method, system and equipment for constructing graph neural network model and recommending diagnosis and treatment scheme |
| CN113434692B (en) * | 2021-06-22 | 2023-08-01 | 上海交通大学医学院附属仁济医院 | Method, system and equipment for constructing graphic neural network model and recommending diagnosis and treatment scheme |
| CN113420161A (en) * | 2021-06-24 | 2021-09-21 | 平安科技（深圳）有限公司 | Node text fusion method and device, computer equipment and storage medium |
| CN113707297B (en) * | 2021-08-26 | 2024-04-05 | 深圳平安智慧医健科技有限公司 | Medical data processing method, device, equipment and storage medium |
| CN113707297A (en) * | 2021-08-26 | 2021-11-26 | 平安国际智慧城市科技股份有限公司 | Medical data processing method, device, equipment and storage medium |
| CN113643821A (en) * | 2021-10-13 | 2021-11-12 | 浙江大学 | Multi-center knowledge graph joint decision support method and system |
| WO2023061377A1 (en) * | 2021-10-13 | 2023-04-20 | 浙江大学 | Multi-center knowledge graph joint decision support method and system |
| CN114400099A (en) * | 2021-12-31 | 2022-04-26 | 北京华彬立成科技有限公司 | Disease information mining and searching method and device, electronic equipment and storage medium |
| CN115062628A (en) * | 2022-06-15 | 2022-09-16 | 北京信息科技大学 | Automatic simulation method for doctor-patient communication conversation based on knowledge graph |
| CN114969386A (en) * | 2022-08-03 | 2022-08-30 | 神州医疗科技股份有限公司 | Disambiguation method, apparatus, electronic device, and medium applied to medical field |
| CN115630697A (en) * | 2022-10-26 | 2023-01-20 | 泸州职业技术学院 | Knowledge graph construction method and system capable of distinguishing single-phase and double-phase affective disorder |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112037920A (en) | Medical knowledge map construction method, device, equipment and storage medium | |
| CN111767410B (en) | Method, device, equipment and storage medium for constructing clinical medical knowledge graph | |
| Su et al. | Information resources processing using linguistic analysis of textual content | |
| CN109886294B (en) | Knowledge fusion method, apparatus, computer device and storage medium | |
| Pivk et al. | Transforming arbitrary tables into logical form with TARTAR | |
| US20110302168A1 (en) | Graphical models for representing text documents for computer analysis | |
| EP3127020A1 (en) | Ontology mapping method and apparatus | |
| WO2015161338A1 (en) | Ontology aligner method, semantic matching method and apparatus | |
| Prudhomme et al. | Interpretation and automatic integration of geospatial data into the Semantic Web: Towards a process of automatic geospatial data interpretation, classification and integration using semantic technologies | |
| Sleeman et al. | Entity type recognition for heterogeneous semantic graphs | |
| WO2015161340A1 (en) | Ontology browser and grouping method and apparatus | |
| Li et al. | Natural language data management and interfaces: Recent development and open challenges | |
| CN113486187A (en) | Buddhism knowledge graph construction method, device, equipment and storage medium | |
| Tekli et al. | Minimizing user effort in XML grammar matching | |
| CN111831624A (en) | Data table creating method and device, computer equipment and storage medium | |
| Albuquerque et al. | UlyssesNER-Br: a corpus of Brazilian legislative documents for named entity recognition | |
| Thang et al. | XML schema automatic matching solution | |
| KR101783791B1 (en) | Compression apparatus and method for managing provenance | |
| CN117573880A (en) | Rolling process data element model and data space construction method and system | |
| Phillips et al. | Using Metadata Record Graphs to understand controlled vocabulary and keyword usage for subject representation in the UNT theses and dissertations collection. | |
| Ren et al. | Effectively processing medical term queries on the UMLS Metathesaurus by layered dynamic programming | |
| D'Souza | Parser extraction of triples in unstructured text | |
| KR102132142B1 (en) | Method and apparatus for recommending vocabulary from data dictionary based on natural language processing technique | |
| Fudholi et al. | Code (common ontology development): A knowledge integration approach from multiple ontologies | |
| Thakare et al. | Extraction of template using clustering from heterogeneous web documents |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201204 |