CN109543047A - A kind of knowledge mapping construction method based on medical field website - Google Patents
A kind of knowledge mapping construction method based on medical field website Download PDFInfo
- Publication number
- CN109543047A CN109543047A CN201811388329.4A CN201811388329A CN109543047A CN 109543047 A CN109543047 A CN 109543047A CN 201811388329 A CN201811388329 A CN 201811388329A CN 109543047 A CN109543047 A CN 109543047A
- Authority
- CN
- China
- Prior art keywords
- entity
- relationship
- website
- attribute
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of knowledge mapping construction methods based on medical field website, which is characterized in that including step 1: acquiring entity, entity attribute and corpus according to the data source of default medical field;Step 2: the Relation extraction between the entity based on BiLSTM (two-way shot and long term memory network) model;Step 3: knowledge fusion and map generate.Reach and provides effective, complete, reliable disease knowledge in knowledge level;The beneficial effect that semantic search and inquiry understand in intelligent answer and field in medical assistance field.
Description
Technical field
The present invention relates to artificial intelligence and data mining technology field, more particularly to a kind of based on medical field website
Knowledge mapping construction method.
Background technique
The fast development of the technologies such as big data, Internet of Things and deep learning, so that artificial intelligence was obtaining very in recent years
Great development.In medical field, the problems such as medical worker's diagnosis and treatment pressure is big, common people's the difficulty of getting medical service, doctor-patient dispute take place frequently, makes medical treatment
Field becomes one of most active and widest field of artificial intelligence application.Artificial intelligence is related in the application of medical field auxiliary
Diagnosis and treatment, health control, information system management, medical image etc. are helped, wherein semantic search, inquiry understanding and automatic question answering all need
The knowledge mapping of medical field is relied on, field question is understood and solved with secondary computer.
Knowledge mapping is a kind of data structure based on figure, is made of the relationship (side) between entity (node) and entity, this
Matter is a kind of semantic network.Knowledge mapping different types of data connection into a relational network, and then have from " close
System " angle goes the ability of problem analysis.Knowledge mapping can be divided into world knowledge map and domain knowledge map according to covering surface, lead to
The range that knowledge is emphasized with knowledge mapping is mainly used in search engine, i.e. " semantic search ", the standard retrieved in single field
True rate is not high.Domain knowledge map emphasizes the depth of knowledge, has stronger specific aim and professional, is generally used in field
Data mining or decision support etc..By domain knowledge map, user can effectively obtain rapidly relevant knowledge and know
Logical relation between knowledge, and then more fully understand realm information.
Summary of the invention
It is a kind of based on medical field website the technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide
Knowledge mapping construction method.
In order to solve the above technical problems, the present invention provides a kind of knowledge mapping construction method based on medical field website,
Characterized by comprising the following steps:
Step 1: entity, entity attribute and corpus are acquired according to the data source of default medical field;The industry in selection field
Website as data source, the entity and entity attributes that medical field is extracted from structuring webpage as structural data,
The long text of medical field is extracted from non-structured web page as unstructured data, i.e. corpus;
Step 2: the Relation extraction entity between the entity based on BiLSTM (two-way shot and long term memory network) model;Specifically
To select word2vec model to be trained, obtaining single word from the corpus acquired in step 1 characterized by Chinese individual character
The insertion of symbol indicates (char embedding), characterized by char embedding, by the text comprising entity pair to be identified
Digitized representations are carried out, and is input in BiLSTM model and is trained identification, export the result of entity relationship pair;
Step 3: knowledge fusion and map generate;Conflict verification is carried out to the entity and entity attributes that obtain in step 1,
Website world rankings and frequency of occurrence according to its source web carry out knowledge fusion;The knowledge fusion refers to, to conflict
Entity, to B, compares the popularity of its source web, retains the higher website data of popularity, when popularity can not area to A and entity
Point both, or both website popularity it is close when, compare the frequency of the entity to A and entity to B in corpus, it is higher to retain the frequency
Entity pair, entity pair identical to the frequency or close, carry out desk checking;
In step 2 based on the result of entity relationship pair, the knowledge fusion of entity relationship is carried out, comprising: merge and repeat
Entity relationship pair, the entity relationship of conflict is verified, according to the website world rankings and frequency of occurrence of its source web,
Retain the entity relationship pair identified in the highest data source of website world rankings;By Redis database with Key-Value's
Form saves entity and entity attributes, and entity is saved in the form of Node-Relation-Node chart database Neo4j
Between relationship, two databases form the knowledge mapping in the field by key (Key) association in Redis.
Knowledge mapping is disease knowledge map in the step 1;In the conceptual model of the disease knowledge map, define
Four entity classifications, the classification of the entity include: disease class, symptom class, organ class and routine inspection class;The disease class
Attribute includes: title, alias, and definition (describes), department, crowd, cause, Symptoms, infectiousness, heredity, prevention and
Health care;The disease class includes " complication ", " sequelae ", the entity relationship of " whole-part ";The attribute packet of the symptom class
It includes: title, alias, prevention and health care, emergency treatment;The attribute of the organ class includes: title, and also known as, brief introduction prevents and protects
It is strong;The attribute of the routine inspection class includes: title, alias, department, points for attention;Relationship between the different classifications is
Bidirectional relationship, including " disease-causes-symptom ", " symptom-embodiment-disease ", " disease-association-organ ", " organ-association-disease
Disease " relationship;The structural data includes the infobox in website, list etc., using in web page tag directly extraction field
Entity, attribute-name and attribute value are saved entity as candidate entity, and attribute-name and attribute value are protected as candidate entity attributes
It deposits, obtains entity-attribute pair, i.e., entity-attribute is saved to according to (entity, attribute-name, attribute value) triple form;It is described non-
In structural data, the result of entity relationship pair is saved according to the triple form of (entity, relationship, entity), wherein BiLSTM
Model exports a series of relationships and its probability, format according to relationship probability sorting are as follows:
Noi: relationshipi, probabilityi,Noi+1: relationshipi+1, probabilityi+1... (i indicates serial number);
Relationship of the highest relationship 1 of select probability as entity 1 and entity 2 forms triple (entity 1, relationship 1, entity
2);In the step 3 conflict verification is carried out to triple to conflict verification is carried out to entity attribute pair and entity relationship, and
Duplicate removal is carried out in triple of the character layer after verification.
For BiLSTM model in the step 2 by training in advance, training set is at least 5000 marks manually marked
Corpus.
The structural data and unstructured data are acquired using network directional crawler, and the medical field website is should
The first two ten website of the world website ranking in field.
Training set in the step 2 is the mark corpus that artificial mark reaches 7500.
The conflict of the entity relationship pair, which is examined, to be turned by manually verifying.
Advantageous effects of the invention:
A) effective, complete, reliable disease knowledge is provided in knowledge level.Cross-cutting user is helped to understand rapidly simultaneously
Understand business scenario, such as the developer of medical treatment APP.
B) intelligent answer in medical assistance field.Intelligent answer achieves certain progress in electricity pin customer service field, but
The high industry of field threshold, such as medical treatment, health care field do not have related application also.
C) semantic search and inquiry understand in field.The semantic search of knowledge based map is no longer retrieved from literal,
But from the term of physical layer foliation solution user, to capture the intention of user input query, more accurately returns and meet use
The search result of family demand, further optimization recommendation mechanisms.
Detailed description of the invention
Fig. 1 is method flow schematic diagram of the invention;
Fig. 2 is the disease knowledge map conceptual model figure based on medical field website in exemplary embodiment of the present invention.
Specific embodiment
The present invention is further illustrated with exemplary embodiment with reference to the accompanying drawing:
The present invention constructs disease knowledge map, is designed by domain-conceptual model using medical field website as data source,
The feature for basing oneself upon disease in medical domain, the multirelation between disease and symptom design the conceptual model towards disease, and fixed
The association attributes of adopted disease class and symptom class.
Building based on disease knowledge map is divided into mode layer design and data layer building:
As shown in Fig. 2, the mode layer design of knowledge mapping should surround its application scenarios, to the feature of things, thing in field
Relationship between object and things is abstract, obtains conceptual model.In the conceptual model of disease knowledge map, four major class: disease are defined
Sick class, symptom class, organ class and routine inspection class, and the relationship its different characteristic and class is abstracted.With disease
For class, disease class is to " hypertension ", and the objective disease such as " diabetes " is abstracted, and attribute includes: title, alias, definition
(description), department, crowd, cause, Symptoms, infectiousness, heredity, prevention and health care.In terms of relationship, " diabetes ",
There is " concurrent " between illnesss such as " hypertension ", the accompanying relationships such as " rear to lose ".It is abstract with this, exist between disease class " complication ", " after
Something lost disease ", " whole-part " relationship, the relationship between class and class is bidirectional relationship, therefore is referred in the example shown with class name, such as
What " disease-state " relationship referred to is " disease-causes-symptom ", " symptom-embodiment-disease " two class relationships, " disease-organ "
What relationship referred to is " disease-association-organ ", " organ-association-disease " two class relationships, and so on.
As shown in Figure 1, conceptual model based on above-mentioned definition, the building of data Layer is carried out using medical web site.
Step 1: domain entities, entity attribute and corpus are acquired according to preset data source.It is high to choose medical field popularity
Industrial sustainability as data source, entity and entity attributes data are extracted from structuring webpage, from non-structured web page
Middle extraction field correlation long text.Wherein entity and entity attributes will be saved in knowledge by data fusion in the next steps
In map, and Relation extraction of the relevant long text in field by the corpus as the training of field term vector, for next step.
Step 11, select the medical field website higher website of world rankings as medical field data source, wherein website
The traffic ranking that popularity is provided according to the website Alexa, including " 39 healthy net ", " 120ask.com ", " China Public health
20 websites such as net ", " good doctor is online ".
Step 12, structural data and unstructured data are acquired from above-mentioned data source using web crawlers.
Step 13, the format of structural data is neat, systematicness is strong, including the infobox in website, list etc..
Step 14, the data volume of unstructured data is big, and format is changeable, in above-mentioned website mostly in the form of the information of field
It presents.
It step 15,, can entity, attribute-name and category directly in extraction field using web page tag from structural data
Property value, will extract entity saves as candidate entity, and attribute-name and attribute value are as candidate entity attributes preservation.Wherein, real
The attribute of body is saved according to the triple form of (entity, attribute-name, attribute value).
Step 16, after non-structured field long text obtains, it is organized into long text data set (field corpus), is
Basis is made in subsequent character incorporation model training.Arranging step includes removal repeated text, deletes blank and spcial character, described
Spcial character includes emoticon, web page tag symbol etc..
Step 2: Relation extraction between the entity based on BiLSTM model.The entity and its attribute extracted in data source, which is in, to be divided
Bulk state identifies discovery by BiLSTM (two-way shot and long term memory network) deep learning model from the field corpus of acquisition
Relationship classification between set entity pair, wherein relationship classification comes from conceptual model defined in step 1, and field corpus is from step
The field long text obtained in rapid 2, set entity is to the entity obtained in step 2, finally, after Relation extraction,
Entity and relationship are formed into entity relationship pair, the i.e. form of triple (entity, relationship, entity), knowledge mapping will by relationship
The solid tissue of " free " plays its effect in terms of semantic search, semantic query at semantic network.
Step 21, entity 1 and entity 2 are entity pair to be identified.
Step 22, text to be identified refers to the text comprising entity 1 and entity 2.
Step 23, long text data set, i.e., the data set obtained in step 16, for training the insertion table of character in field
Show, i.e. char embedding.
Step 24, the character incorporation model based on character is trained, based on the field corpus in step 23, with Chinese single
Word is characterized, and selects the training common word2vec model of term vector, and the insertion that training obtains single character indicates
(charembedding)。
Step 25, characterized by the insertion of character indicates, the texts digitization in step 22 is indicated, and is input to
In BiLSTM model.
Further, the BiLSTM model in step 25 has already passed through training, and training set is 5000 marks manually marked
Corpus is infused, especially up to 7500 mark corpus make its accuracy rate reach 70% or more, especially up to 80%.
Step 26, BiLSTM model exports a series of relationships and its probability, format No according to relationship probability sortingi: it closes
Systemi, probabilityi,Noi+1: relationshipi+1, probabilityi+1... highest 1 (the i.e. pass of sequence first of relationship of (i indicates serial number) select probability
System) relationship as entity 1 and entity 2 in step 21, formation triple (entity 1, relationship 1, entity 2).
Step 3: knowledge fusion and map generate.Based on the triple of acquisition, the knowledge fusion of entity relationship is carried out,
Include: to merge duplicate entity relationship pair, the entity relationship of conflict is verified.Source number of the verification according to entity relationship pair
According to and frequency of occurrence, i.e. the relationship pair that identifies in the reservation higher data source of world website ranking.Complicated to relationship simultaneously
Entity pair realizes verification by artificial.Entity and its attribute finally are saved by Redis, saves the pass between entity by Neo4j
System generates disease knowledge map.
Step 31, entity relationship is to the entity relationship pair for referring to obtaining from step 2, i.e. triple (entity 1, relationship
1, entity 2).Entity attribute to refer to step 15 directly extracted from website obtained by entity attribute pair, i.e., triple (entity,
Attribute, attribute value).
Step 32, conflict verification is carried out to the triple of previous step.Triple conflict includes entity attribute conflict and entity
Conflict of relationships, wherein entity attribute conflict is melted according to the source web of entity attribute and its frequency of occurrence in corpus
It closes, specifically, comparing the popularity of its source web to B to A and entity to the entity of conflict, retaining the higher net of popularity
Stand data, when popularity both cannot be distinguished, or both website popularity it is close when, compare entity to A and entity to B in corpus
In the frequency, retain the higher entity pair of the frequency, entity pair identical to the frequency or close, carry out desk checking.Entity is closed
System conflict the case where because the relationship in field between disease entity be not it is single, and model training identification relationship accuracy rate
Limited, old friend's work verifies.
Step 33, the triple duplicate removal in character layer after examining.
Step 34, entity and its attribute are stored in Redis database in field, and the relationship between entity is stored in Neo4j
In database, the two forms the disease knowledge map of medical field by key (KEY) association in Redis.
Present invention is mainly used for provide a kind of knowledge mapping construction method based on medical field website, its advantages
It is:
A) effective, complete, reliable disease knowledge is provided in knowledge level.Cross-cutting user is helped to understand rapidly simultaneously
Understand business scenario, such as the developer of medical treatment APP.
B) intelligent answer in medical assistance field.Intelligent answer achieves certain progress in electricity pin customer service field, but
The high industry of field threshold, such as medical treatment, health care field do not have related application also.
C) semantic search and inquiry understand in field.The semantic search of knowledge based map is no longer retrieved from literal,
But from the term of physical layer foliation solution user, to capture the intention of user input query, more accurately returns and meet use
The search result of family demand, further optimization recommendation mechanisms.
Above embodiments do not limit the present invention in any way, all to be made in a manner of equivalent transformation to above embodiments
Other improvement and application, belong to protection scope of the present invention.
Claims (6)
1. a kind of knowledge mapping construction method based on medical field website, which comprises the following steps:
Step 1: entity, entity attribute and corpus are acquired according to the data source of default medical field;The industrial sustainability in selection field
As data source, the entity and entity attributes that medical field is extracted from structuring webpage are as structural data, from non-
The long text of medical field is extracted in structuring webpage as unstructured data, i.e. corpus;
Step 2: the Relation extraction entity between the entity based on BiLSTM (two-way shot and long term memory network) model;Specifically, from
In the corpus acquired in step 1 characterized by Chinese individual character, selects word2vec model to be trained, obtain the embedding of single character
Entering indicates that (char embedding) is counted the text comprising entity pair to be identified characterized by char embedding
Wordization indicates, and is input in BiLSTM model and is trained identification, exports the result of entity relationship pair;
Step 3: knowledge fusion and map generate;Conflict verification, foundation are carried out to the entity and entity attributes that obtain in step 1
The website world rankings and frequency of occurrence of its source web carry out knowledge fusion;The knowledge fusion refers to, to the entity of conflict
To A and entity to B, compares the popularity of its source web, retain the higher website data of popularity, when popularity cannot be distinguished two
Person, or both website popularity it is close when, compare the frequency of the entity to A and entity to B in corpus, retain the higher reality of the frequency
Body pair, entity pair identical to the frequency or close carry out desk checking;
In step 2 based on the result of entity relationship pair, the knowledge fusion of entity relationship is carried out, comprising: merge duplicate reality
Body relationship pair verifies the entity relationship of conflict, according to the website world rankings and frequency of occurrence of its source web, retains
The entity relationship pair identified in the highest data source of website world rankings;Through Redis database in the form of Key-Value
Entity and entity attributes are saved, between chart database Neo4j saves entity in the form of Node-Relation-Node
Relationship, two databases form the knowledge mapping in the field by key (Key) association in Redis.
2. a kind of knowledge mapping construction method based on medical field website as described in claim 1, it is characterised in that: described
Knowledge mapping is disease knowledge map in step 1;In the conceptual model of the disease knowledge map, four entities point are defined
Class, the classification of the entity include: disease class, symptom class, organ class and routine inspection class;The attribute of the disease class includes:
Title, alias, definition (describe), department, crowd, cause, Symptoms, infectiousness, heredity, prevention and health care;It is described
Disease class includes " complication ", " sequelae ", the entity relationship of " whole-part ";The attribute of the symptom class includes: title,
Alias, prevention and health care, emergency treatment;The attribute of the organ class includes: title, nickname, brief introduction, prevention and health care;It is described
The attribute of routine inspection class includes: title, alias, department, points for attention;Relationship between the different classifications is two-way pass
System, including " disease-causes-symptom ", " symptom-embodiment-disease ", " disease-association-organ ", " organ-association-disease " close
System;The structural data includes the infobox in website, list etc., using in web page tag directly extraction field entity,
Attribute-name and attribute value are saved entity as candidate entity, and attribute-name and attribute value are saved as candidate entity attributes, are obtained
To entity-attribute pair, i.e. entity-attribute is saved to according to (entity, attribute-name, attribute value) triple form;It is described non-structural
Change in data, the result of entity relationship pair is saved according to the triple form of (entity, relationship, entity), wherein BiLSTM model
According to relationship probability sorting, a series of relationships and its probability, format are exported are as follows:
Noi: relationshipi, probabilityi,Noi+1: relationshipi+1, probabilityi+1... (i indicates serial number);
Relationship of the highest relationship 1 of select probability as entity 1 and entity 2 is formed triple (entity 1, relationship 1, entity 2);
In the step 3 conflict verification is carried out to triple, and in word to conflict verification is carried out to entity attribute pair and entity relationship
Symbol level carries out duplicate removal to the triple after verification.
3. a kind of knowledge mapping construction method based on medical field website as claimed in claim 2, it is characterised in that: described
For BiLSTM model in step 2 by training in advance, training set is at least 5000 mark corpus manually marked.
4. a kind of knowledge mapping construction method based on medical field website as claimed in claim 3, it is characterised in that: described
Structural data and unstructured data are acquired using network directional crawler, and the medical field website is the WorldNet in the field
The first two ten website of ranking of standing.
5. a kind of knowledge mapping construction method based on medical field website as claimed in claim 3, it is characterised in that: described
Training set in step 2 is the mark corpus that artificial mark reaches 7500.
6. a kind of knowledge mapping construction method based on medical field website as claimed in claim 4, it is characterised in that: described
The conflict of entity relationship pair, which is examined, to be turned by manually verifying.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811388329.4A CN109543047A (en) | 2018-11-21 | 2018-11-21 | A kind of knowledge mapping construction method based on medical field website |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811388329.4A CN109543047A (en) | 2018-11-21 | 2018-11-21 | A kind of knowledge mapping construction method based on medical field website |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109543047A true CN109543047A (en) | 2019-03-29 |
Family
ID=65848655
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811388329.4A Pending CN109543047A (en) | 2018-11-21 | 2018-11-21 | A kind of knowledge mapping construction method based on medical field website |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109543047A (en) |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110222319A (en) * | 2019-06-19 | 2019-09-10 | 北京百度网讯科技有限公司 | Method and apparatus for mining data |
CN110275894A (en) * | 2019-06-24 | 2019-09-24 | 恒生电子股份有限公司 | A kind of update method of knowledge mapping, device, electronic equipment and storage medium |
CN110289101A (en) * | 2019-07-02 | 2019-09-27 | 京东方科技集团股份有限公司 | A kind of computer equipment, system and readable storage medium storing program for executing |
CN110413798A (en) * | 2019-07-24 | 2019-11-05 | 厦门快商通科技股份有限公司 | A kind of medical and beauty treatment knowledge mapping method for auto constructing, system and storage medium |
CN110609910A (en) * | 2019-09-18 | 2019-12-24 | 金色熊猫有限公司 | Medical knowledge graph construction method and device, storage medium and electronic equipment |
CN110795528A (en) * | 2019-09-05 | 2020-02-14 | 腾讯科技(深圳)有限公司 | Data query method and device, electronic equipment and storage medium |
CN110957042A (en) * | 2020-01-17 | 2020-04-03 | 广州慧视医疗科技有限公司 | Prediction and simulation method of eye diseases under different conditions based on domain knowledge migration |
CN111125309A (en) * | 2019-12-23 | 2020-05-08 | 中电云脑(天津)科技有限公司 | Natural language processing method and device, computing equipment and storage medium |
CN111177411A (en) * | 2019-12-27 | 2020-05-19 | 赣州市智能产业创新研究院 | Knowledge graph construction method based on NLP |
CN111191048A (en) * | 2020-01-02 | 2020-05-22 | 南京邮电大学 | Emergency call question-answering system construction method based on knowledge graph |
CN111339252A (en) * | 2020-02-25 | 2020-06-26 | 腾讯科技(深圳)有限公司 | Searching method, searching device and storage medium |
CN111341456A (en) * | 2020-02-21 | 2020-06-26 | 中南大学湘雅医院 | Method and device for generating diabetic foot knowledge map and readable storage medium |
CN111475641A (en) * | 2019-08-26 | 2020-07-31 | 北京国双科技有限公司 | Data extraction method and device, storage medium and equipment |
CN111950285A (en) * | 2020-07-31 | 2020-11-17 | 合肥工业大学 | Intelligent automatic construction system and method of medical knowledge map based on multi-modal data fusion |
CN112163092A (en) * | 2020-10-10 | 2021-01-01 | 成都数之联科技有限公司 | Entity and relation extraction method, system, device and medium |
CN112233803A (en) * | 2020-09-11 | 2021-01-15 | 北京欧应信息技术有限公司 | Data mining device for assisting doctor in optimizing diagnosis and treatment |
CN112309521A (en) * | 2020-10-31 | 2021-02-02 | 海南大学 | Cross-modal user medical data analysis method for essential computing |
CN112347204A (en) * | 2021-01-08 | 2021-02-09 | 药渡经纬信息科技(北京)有限公司 | Method and device for constructing drug research and development knowledge base |
CN112463973A (en) * | 2019-09-06 | 2021-03-09 | 医渡云(北京)技术有限公司 | Construction method, device and medium of medical knowledge graph and electronic equipment |
CN113012803A (en) * | 2019-12-19 | 2021-06-22 | 京东方科技集团股份有限公司 | Computer device, system, readable storage medium and medical data analysis method |
CN113032469A (en) * | 2019-12-24 | 2021-06-25 | 医渡云(北京)技术有限公司 | Text structured model training and medical text structured method and device |
WO2021139282A1 (en) * | 2020-06-24 | 2021-07-15 | 平安科技(深圳)有限公司 | Medical field knowledge graph construction method and apparatus, device and storage medium |
CN113449118A (en) * | 2021-06-29 | 2021-09-28 | 华南理工大学 | Standard document conflict detection method and system based on standard knowledge graph |
CN113505243A (en) * | 2021-07-29 | 2021-10-15 | 深圳万海思数字医疗有限公司 | Intelligent question-answering method and device based on medical knowledge graph |
CN113569055A (en) * | 2021-07-26 | 2021-10-29 | 东北大学 | Strip mine knowledge map construction method based on genetic algorithm optimization neural network |
CN113641784A (en) * | 2021-06-25 | 2021-11-12 | 合肥工业大学 | Medical knowledge recommendation method and system integrating medical teaching and research |
WO2022126799A1 (en) * | 2020-12-14 | 2022-06-23 | 谈斯聪 | Intelligent multi-data fusion disease-identifying method |
CN117198547A (en) * | 2023-09-01 | 2023-12-08 | 湖北中医药大学 | Intelligent extraction method and system for traditional Chinese medicine clinical scientific research knowledge |
Citations (4)
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 |
US20180232443A1 (en) * | 2017-02-16 | 2018-08-16 | Globality, Inc. | Intelligent matching system with ontology-aided relation extraction |
CN108492887A (en) * | 2018-04-13 | 2018-09-04 | 合肥工业大学 | medical knowledge map construction method and device |
CN108829722A (en) * | 2018-05-08 | 2018-11-16 | 国家计算机网络与信息安全管理中心 | A kind of Dual-Attention relationship classification method and system of remote supervisory |
-
2018
- 2018-11-21 CN CN201811388329.4A patent/CN109543047A/en active Pending
Patent Citations (4)
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 |
US20180232443A1 (en) * | 2017-02-16 | 2018-08-16 | Globality, Inc. | Intelligent matching system with ontology-aided relation extraction |
CN108492887A (en) * | 2018-04-13 | 2018-09-04 | 合肥工业大学 | medical knowledge map construction method and device |
CN108829722A (en) * | 2018-05-08 | 2018-11-16 | 国家计算机网络与信息安全管理中心 | A kind of Dual-Attention relationship classification method and system of remote supervisory |
Non-Patent Citations (2)
Title |
---|
袁凯琦 等: "医学知识图谱构建技术与研究进展", 《计算机应用研究》 * |
邹玉薇: "中文医疗知识图谱半自动化构建研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (38)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110222319A (en) * | 2019-06-19 | 2019-09-10 | 北京百度网讯科技有限公司 | Method and apparatus for mining data |
CN110275894A (en) * | 2019-06-24 | 2019-09-24 | 恒生电子股份有限公司 | A kind of update method of knowledge mapping, device, electronic equipment and storage medium |
CN110275894B (en) * | 2019-06-24 | 2021-12-14 | 恒生电子股份有限公司 | Knowledge graph updating method and device, electronic equipment and storage medium |
CN110289101A (en) * | 2019-07-02 | 2019-09-27 | 京东方科技集团股份有限公司 | A kind of computer equipment, system and readable storage medium storing program for executing |
CN110413798A (en) * | 2019-07-24 | 2019-11-05 | 厦门快商通科技股份有限公司 | A kind of medical and beauty treatment knowledge mapping method for auto constructing, system and storage medium |
CN111475641A (en) * | 2019-08-26 | 2020-07-31 | 北京国双科技有限公司 | Data extraction method and device, storage medium and equipment |
CN110795528B (en) * | 2019-09-05 | 2023-10-13 | 腾讯科技(深圳)有限公司 | Data query method and device, electronic equipment and storage medium |
CN110795528A (en) * | 2019-09-05 | 2020-02-14 | 腾讯科技(深圳)有限公司 | Data query method and device, electronic equipment and storage medium |
CN112463973A (en) * | 2019-09-06 | 2021-03-09 | 医渡云(北京)技术有限公司 | Construction method, device and medium of medical knowledge graph and electronic equipment |
CN110609910A (en) * | 2019-09-18 | 2019-12-24 | 金色熊猫有限公司 | Medical knowledge graph construction method and device, storage medium and electronic equipment |
CN113012803A (en) * | 2019-12-19 | 2021-06-22 | 京东方科技集团股份有限公司 | Computer device, system, readable storage medium and medical data analysis method |
CN111125309A (en) * | 2019-12-23 | 2020-05-08 | 中电云脑(天津)科技有限公司 | Natural language processing method and device, computing equipment and storage medium |
CN113032469B (en) * | 2019-12-24 | 2024-02-20 | 医渡云(北京)技术有限公司 | Text structured model training and medical text structuring method and device |
CN113032469A (en) * | 2019-12-24 | 2021-06-25 | 医渡云(北京)技术有限公司 | Text structured model training and medical text structured method and device |
CN111177411A (en) * | 2019-12-27 | 2020-05-19 | 赣州市智能产业创新研究院 | Knowledge graph construction method based on NLP |
CN111191048A (en) * | 2020-01-02 | 2020-05-22 | 南京邮电大学 | Emergency call question-answering system construction method based on knowledge graph |
CN110957042A (en) * | 2020-01-17 | 2020-04-03 | 广州慧视医疗科技有限公司 | Prediction and simulation method of eye diseases under different conditions based on domain knowledge migration |
CN111341456B (en) * | 2020-02-21 | 2024-02-23 | 中南大学湘雅医院 | Method and device for generating diabetic foot knowledge graph and readable storage medium |
CN111341456A (en) * | 2020-02-21 | 2020-06-26 | 中南大学湘雅医院 | Method and device for generating diabetic foot knowledge map and readable storage medium |
CN111339252A (en) * | 2020-02-25 | 2020-06-26 | 腾讯科技(深圳)有限公司 | Searching method, searching device and storage medium |
WO2021139282A1 (en) * | 2020-06-24 | 2021-07-15 | 平安科技(深圳)有限公司 | Medical field knowledge graph construction method and apparatus, device and storage medium |
CN111950285A (en) * | 2020-07-31 | 2020-11-17 | 合肥工业大学 | Intelligent automatic construction system and method of medical knowledge map based on multi-modal data fusion |
CN111950285B (en) * | 2020-07-31 | 2024-01-23 | 合肥工业大学 | Medical knowledge graph intelligent automatic construction system and method with multi-mode data fusion |
CN112233803A (en) * | 2020-09-11 | 2021-01-15 | 北京欧应信息技术有限公司 | Data mining device for assisting doctor in optimizing diagnosis and treatment |
CN112163092A (en) * | 2020-10-10 | 2021-01-01 | 成都数之联科技有限公司 | Entity and relation extraction method, system, device and medium |
CN112163092B (en) * | 2020-10-10 | 2022-07-12 | 成都数之联科技股份有限公司 | Entity and relation extraction method, system, device and medium |
CN112309521B (en) * | 2020-10-31 | 2021-09-10 | 海南大学 | Cross-modal user medical data analysis method for essential computing |
CN112309521A (en) * | 2020-10-31 | 2021-02-02 | 海南大学 | Cross-modal user medical data analysis method for essential computing |
WO2022126799A1 (en) * | 2020-12-14 | 2022-06-23 | 谈斯聪 | Intelligent multi-data fusion disease-identifying method |
CN112347204A (en) * | 2021-01-08 | 2021-02-09 | 药渡经纬信息科技(北京)有限公司 | Method and device for constructing drug research and development knowledge base |
CN113641784A (en) * | 2021-06-25 | 2021-11-12 | 合肥工业大学 | Medical knowledge recommendation method and system integrating medical teaching and research |
CN113449118A (en) * | 2021-06-29 | 2021-09-28 | 华南理工大学 | Standard document conflict detection method and system based on standard knowledge graph |
CN113449118B (en) * | 2021-06-29 | 2022-09-20 | 华南理工大学 | Standard document conflict detection method and system based on standard knowledge graph |
CN113569055B (en) * | 2021-07-26 | 2023-09-22 | 东北大学 | Strip mine knowledge graph construction method based on genetic algorithm optimization neural network |
CN113569055A (en) * | 2021-07-26 | 2021-10-29 | 东北大学 | Strip mine knowledge map construction method based on genetic algorithm optimization neural network |
CN113505243A (en) * | 2021-07-29 | 2021-10-15 | 深圳万海思数字医疗有限公司 | Intelligent question-answering method and device based on medical knowledge graph |
CN117198547A (en) * | 2023-09-01 | 2023-12-08 | 湖北中医药大学 | Intelligent extraction method and system for traditional Chinese medicine clinical scientific research knowledge |
CN117198547B (en) * | 2023-09-01 | 2024-04-30 | 湖北中医药大学 | Intelligent extraction method and system for traditional Chinese medicine clinical scientific research knowledge |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109543047A (en) | A kind of knowledge mapping construction method based on medical field website | |
CN109460473B (en) | Electronic medical record multi-label classification method based on symptom extraction and feature representation | |
CN108182262B (en) | Intelligent question-answering system construction method and system based on deep learning and knowledge graph | |
CN108073569B (en) | Law cognition method, device and medium based on multi-level multi-dimensional semantic understanding | |
CN106250412B (en) | Knowledge mapping construction method based on the fusion of multi-source entity | |
JP5904559B2 (en) | Scenario generation device and computer program therefor | |
CN103646032B (en) | A kind of based on body with the data base query method of limited natural language processing | |
CN104391942B (en) | Short essay eigen extended method based on semantic collection of illustrative plates | |
CN110222201A (en) | A kind of disease that calls for specialized treatment knowledge mapping construction method and device | |
CN111639190A (en) | Medical knowledge map construction method | |
CN100416569C (en) | Web page metadata based formalized description method for user access behaviors | |
CN105760439B (en) | A kind of personage's cooccurrence relation map construction method based on specific behavior co-occurrence network | |
CN103646112B (en) | Dependency parsing field self-adaption method based on web search | |
Wisselgren et al. | Testing methods of record linkage on Swedish censuses | |
JP5907393B2 (en) | Complex predicate template collection device and computer program therefor | |
CN109657062A (en) | A kind of electronic health record text resolution closed-loop policy based on big data technology | |
CN104298714B (en) | A kind of mass text automatic marking method based on abnormality processing | |
CN106446162A (en) | Orient field self body intelligence library article search method | |
JP2015121896A (en) | Phrase pair gathering device, and computer program for the same | |
CN107943514A (en) | The method for digging and system of core code element in a kind of software document | |
CN112420212B (en) | Method for constructing brain stroke traditional Chinese medicine knowledge graph | |
CN110377751A (en) | Courseware intelligent generation method, device, computer equipment and storage medium | |
CN114580418B (en) | Police physical training knowledge graph system | |
CN110600121A (en) | Knowledge graph-based primary etiology diagnosis method | |
CN104820681A (en) | Response method and system for online Q&A service |
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: 20190329 |