CN110263083B - Knowledge graph processing method, device, equipment and medium - Google Patents
Knowledge graph processing method, device, equipment and medium Download PDFInfo
- Publication number
- CN110263083B CN110263083B CN201910537133.5A CN201910537133A CN110263083B CN 110263083 B CN110263083 B CN 110263083B CN 201910537133 A CN201910537133 A CN 201910537133A CN 110263083 B CN110263083 B CN 110263083B
- Authority
- CN
- China
- Prior art keywords
- entity evidence
- knowledge graph
- unit
- attribute
- entity
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/26—Visual data mining; Browsing structured data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Animal Behavior & Ethology (AREA)
- Pathology (AREA)
- Computational Linguistics (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses a method, a device, equipment and a medium for processing a knowledge graph, and relates to the technical field of knowledge graphs. The method comprises the following steps: selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified; determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units; and selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit. According to the embodiment of the invention, at least two candidate entity evidence units are selected from the target field entity evidence units, the matching degree of the candidate entity evidence units and the knowledge graph to be detected is determined, and the target entity evidence units are selected according to the matching degree, so that the human intervention is reduced, and the accuracy and the construction efficiency of the knowledge graph are improved.
Description
Technical Field
The embodiment of the invention relates to the technical field of intelligent logistics, in particular to a method, a device, equipment and a medium for processing a knowledge graph.
Background
The construction of an accurate and comprehensive medical industry knowledge map of authority is the basic data requirement of a plurality of upper-layer applications, and medical knowledge in an authority medical book is a smart crystal summarized and demonstrated by people, so that people can extract a plurality of authority medical facts.
In the prior art, medical facts mined from medical books are accurately added into a medical map, and a general method is to perform preliminary screening through frequency information of the facts and then to perform auditing by medical experts. The method is a manual intervention method, so that the problems of large human input and relatively low efficiency exist. Besides depending on experience, the process of examining and verifying medical facts by medical experts also needs a book inquiry tool to inquire and verify in many cases.
Disclosure of Invention
The embodiment of the invention provides a method, a device, equipment and a medium for processing a knowledge graph, and aims to solve the problems of high labor input and relatively low efficiency in the construction of the knowledge graph in the prior art.
In a first aspect, an embodiment of the present invention provides a method for processing a knowledge graph, where the method includes:
selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified;
determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units;
and selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit.
In a second aspect, an embodiment of the present invention provides an apparatus for processing a knowledge graph, where the apparatus includes:
the candidate entity evidence unit selection module is used for selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified;
the matching degree determining module is used for determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units;
and the target entity evidence unit selection module is used for selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, so that the knowledge graph to be verified is verified based on the target entity evidence unit.
In a third aspect, an embodiment of the present invention provides an apparatus, where the apparatus further includes:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method of knowledge-graph processing as described in any of the embodiments of the invention.
In a fourth aspect, the present invention provides a computer readable medium, on which a computer program is stored, which when executed by a processor implements a method for processing a knowledge-graph according to any one of the embodiments of the present invention.
According to the embodiment of the invention, at least two candidate entity evidence units are selected from the target field entity evidence units, the matching degree of the at least two candidate entity evidence units and the knowledge graph to be verified is determined, and the target entity evidence unit is selected according to the matching degree for verifying the knowledge graph to be verified according to the target entity evidence unit, so that the accuracy rate and the construction efficiency of the knowledge graph are improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
FIG. 1 is a flow chart of a knowledge-graph processing method according to an embodiment of the present invention;
FIG. 2 is a flowchart of a knowledge-graph processing method according to a second embodiment of the present invention;
FIG. 3 is a flowchart of a knowledge-graph processing method according to a third embodiment of the present invention;
FIG. 4 is a schematic structural diagram of a knowledge-graph processing apparatus according to a fourth embodiment of the present invention;
fig. 5 is a schematic structural diagram of an apparatus according to a fifth embodiment of the present invention.
Detailed Description
The embodiments of the present invention will be described in further detail with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the embodiments of the invention and that no limitation of the invention is intended. It should be further noted that, for convenience of description, only some structures, not all structures, relating to the embodiments of the present invention are shown in the drawings.
Example one
Fig. 1 is a flowchart of a method for processing a knowledge graph according to an embodiment of the present invention. The embodiment is suitable for the case of constructing a medical knowledge map, and the method can be executed by a processing device of the knowledge map provided by the embodiment of the invention, and the device can be realized in a software and/or hardware manner. As shown in fig. 1, the method may include:
s101, according to the knowledge graph to be verified, at least two candidate entity evidence units are selected from the entity evidence units in the target field to which the knowledge graph to be verified belongs.
The knowledge graph to be verified belongs to the field which can be selected from medical treatment, education, finance, culture and the like. The entity evidence unit is obtained by mining knowledge facts in a target field to which the knowledge graph to be verified belongs, and the function of the entity evidence unit is embodied in the following two aspects: 1. the method can be used as an evidence of the knowledge graph to be verified to assist related auditors in evaluating the knowledge graph to be verified; 2. the knowledge graph to be verified can be included to increase comprehensiveness of the knowledge graph to be verified.
Optionally, S (entity name), P (attribute name), and O (attribute value) in the knowledge graph to be verified are used as search terms to search in the entity evidence units in the target field, and at least two candidate entity evidence units are selected according to the search result.
The size of the entity evidence unit may be large or small, and when the size of the entity evidence unit is large, such as a complete document, a complete paper or a thick book, since it may include multiple entities and multiple attributes corresponding to the multiple entities, the problem of information interference with single entity multiple attributes or multiple entity multiple attributes may occur.
In order to avoid such information interference problem, before S101, optionally, the method further includes:
A. extracting attribute information and chapter title information to which the attribute information belongs from the target field fact;
the attribute information comprises a word description of multiple attributes of the entity, the word description comprises an attribute name and attribute content, and the chapter header information comprises a word description of the entity associated with the attribute information.
Illustratively, taking a book html (hypertext markup language) page as an example, the page mainly has two main parts, wherein the first part is a content block including information describing a chapter, a section, a title and the like; the second part is a content block including description detail information. Specifically, chapter title information to which attribute information belongs, such as "third larynx science", "chapter four specific pharyngitis", and "third larynx syphilis", is extracted in the first section by including a target keyword search method; in the second part, attribute information such as "clinical manifestation hoarseness", "diagnosis key family history" or "treatment scheme and principle plum driving treatment" is extracted by a search method including a target keyword.
B. Taking the chapter title information as a title of an entity evidence unit in the target field;
illustratively, the extracted chapter title information is "third laryngitis science", "chapter four specific pharyngitis" and "third throat syphilis", and then "third laryngitis science", "chapter four specific pharyngitis" and "third throat syphilis" are taken as the titles of the entity evidence units.
C. Taking the attribute name in the attribute information as an attribute key word of an entity evidence unit in the target field;
the attribute name in the attribute information is a higher-level concept of the attribute content.
Illustratively, for example, if the extracted attribute information includes "family history of diagnostic gist", the attribute name is "diagnostic gist", and then "diagnostic gist" is used as an attribute keyword of the entity evidence unit; for another example, if the extracted attribute information includes "clinical manifestation hoarseness", the attribute name is "clinical manifestation", and "clinical manifestation" is used as an attribute keyword of the entity evidence unit.
D. And taking the attribute content in the attribute information as an attribute key value of an entity evidence unit in the target field.
The attribute content in the attribute information is an attribute name lower expression form.
Illustratively, for example, if the extracted attribute information includes "diagnostic gist family history", the attribute content is "family history", and then "family history" is used as an attribute key value of an entity evidence unit; for another example, if the extracted attribute information includes "clinically manifested hoarseness", the attribute content is "hoarseness", and then "hoarseness" is taken as an attribute key value of the entity evidence unit.
By extracting the attribute information and the chapter title information to which the attribute information belongs from the target field fact and finally determining the title, the attribute key word and the attribute key value of the entity evidence unit, the problem of information interference is avoided, so that the subsequent knowledge graph to be tested is matched with at least two candidate entity evidence units, and the method is simpler, more effective and more visual.
By selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs, the matching times of the subsequent knowledge graph to be verified and the entity evidence units are reduced, and the efficiency is improved.
S102, determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units.
The matching degree reflects the similarity between the knowledge graph to be detected and the at least two candidate entity evidence units, the higher the matching degree is, the higher the similarity between the knowledge graph to be detected and the at least two candidate entity evidence units is, and the lower the matching degree is, the lower the similarity between the knowledge graph to be detected and the at least two candidate entity evidence units is.
Specifically, the matching degree between the knowledge graph to be tested and at least two candidate entity evidence units is determined, that is, the similarity between the information contained in the knowledge graph to be tested and the information contained in the at least two candidate entity evidence units is determined.
S103, selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit.
Optionally, the target entity evidence unit of the knowledge graph to be tested is determined according to the sorting result.
According to the embodiment of the invention, at least two candidate entity evidence units are selected from the target field entity evidence units, the matching degree of the at least two candidate entity evidence units and the knowledge graph to be detected is determined, and the target entity evidence unit is selected according to the matching degree, so that the human intervention is reduced, and the verification efficiency and the construction efficiency of the knowledge graph are improved.
Example two
Fig. 2 is a flowchart of a processing method of a knowledge graph according to a second embodiment of the present invention. The embodiment provides a specific implementation manner for the first embodiment, and as shown in fig. 2, the method may include:
s201, taking an entity name in the knowledge graph to be verified as a search word, and searching chapter header information, attribute keywords and attribute key values of the entity evidence unit in the target field to obtain a first entity evidence unit.
Specifically, in order to avoid missing possible candidate entity evidence units as much as possible, an entity evidence unit which is the same as or similar to the entity name in the knowledge graph to be verified exists in the chapter header information, the attribute key words or the attribute key values and serves as a first entity evidence unit.
Illustratively, the name of an entity in the knowledge graph to be verified is "specific pharyngitis", namely a search word, and the chapter header information of the entity evidence unit 1 contains "specific pharyngitis", so that the entity evidence unit 1 is a first entity evidence unit; the attribute key value of the entity evidence unit 2 comprises specific pharyngitis, and then the entity evidence unit 2 is a first entity evidence unit; if no words identical or similar to the "specific pharyngitis" are contained in the chapter title information, the attribute keywords, or the attribute key values of the entity evidence unit 3, the entity evidence unit 3 is not the first entity evidence unit.
In order to avoid missing possible candidate entity evidence units as much as possible, optionally in S201, the "alias" and the "aka" of the entity name in the knowledge-graph to be verified may also be used as search terms. Wherein, the collecting means of the "alias" and the "aka" comprises: 1) utilizing existing map data; 2) the method comprises the steps of acquiring the alias and the named name appearing in the book or the literature, and performing manual review on the alias and the named name.
S202, searching in attribute key values of entity evidence units in the target field by taking attribute values in the knowledge graph to be verified as search words to obtain second entity evidence units.
Specifically, an entity evidence unit which is the same as or similar to the attribute value in the knowledge graph to be verified exists in the attribute key value and serves as a second entity evidence unit.
Illustratively, the attribute value in the knowledge graph to be verified is "cough", that is, a search term, and if the attribute key value of the entity evidence unit 4 includes "cough", the entity evidence unit 4 is a second entity evidence unit; if the attribute key of the entity evidence unit 5 does not include "cough", the entity evidence unit 5 is not the second entity evidence unit.
S203, selecting at least two candidate entity evidence units from the first entity evidence unit and the second entity evidence unit.
Specifically, at least two candidate entity evidence units are selected from the first entity evidence unit and the second entity evidence unit according to a preset rule.
Optionally, different weight values are given to the chapter title information, the attribute keywords and the attribute key values, and the optional weight size relationship is as follows: and after the search is finished, the obtained first entity evidence units and second entity evidence units are sorted according to the weight-size relationship, and a preset number of first entity evidence units and second entity evidence units are selected as candidate entity evidence units according to the weights from large to small.
Illustratively, the chapter header information of the first entity evidence unit a includes an entity name in the knowledge graph to be verified, the attribute key of the first entity evidence unit B includes the entity name in the knowledge graph to be verified, and the attribute key of the first entity evidence unit C includes the entity name in the knowledge graph to be verified, then according to the relationship of the chapter header information, the attribute key, and the attribute key, the first entity evidence unit a, the first entity evidence unit B, and the first entity evidence unit C are ordered as follows: a first entity evidence unit A, a first entity evidence unit B and a first entity evidence unit C.
Optionally, on the basis of the foregoing embodiment, the entity name is given as a search term, and the attribute value is given as a different weight of the search term, where the selectable weight size relationship is: and the entity names are used as search word weights > attribute values and are used as search word weights, after the search is finished, the obtained first entity evidence units and second entity evidence units are ranked according to the weight magnitude relation, and a preset number of first entity evidence units and second entity evidence units are selected from large to small according to the weights and are used as candidate entity evidence units.
Illustratively, the chapter header information of the first entity evidence unit a includes an entity name in the knowledge graph to be verified, the attribute key of the first entity evidence unit B includes the entity name in the knowledge graph to be verified, and the attribute key of the second entity evidence unit C includes the attribute value in the knowledge graph to be verified, then according to the weight size relationship between the chapter header information, the attribute key, and the weight size relationship between the entity name as a search word and the attribute value as a search word, the first entity evidence unit a, the first entity evidence unit B, and the second entity evidence unit C are ordered as follows: : a first entity evidence unit A, a first entity evidence unit B and a second entity evidence unit C.
S204, selecting the knowledge graph matched with the knowledge graph to be verified from the at least two candidate entity evidence units according to the mapping relation between the attribute names in the knowledge graph and the attribute keywords in the entity evidence units and the knowledge graph to be verified, and filtering out other candidate entity evidence units.
The mapping relation embodies the incidence relation between the attribute names in the candidate entity evidence unit spectrum of the knowledge graph and the attribute keywords in the entity evidence units.
Optionally, the attribute keywords in the entity evidence unit are ranked according to the popularity, a preset number of attribute keywords with higher popularity are selected, and a mapping relationship between the attribute names in the knowledge graph and the selected preset number of attribute keywords is determined by a semantic analysis algorithm, wherein, for example, the attribute names "clinical expression" in the knowledge graph and the attribute keywords "physical expression" in the entity evidence unit have a mapping relationship; and finally, selecting at least two candidate entity evidence units which have a mapping relation with the knowledge graph to be verified through the obtained mapping relation as candidate entity evidence units, and filtering the candidate entity evidence units without the mapping relation.
S205, determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units.
Since S204 has already screened the candidate entity evidence units through the dimension of attribute names. Two dimensions of entity names and attribute values are mainly considered when determining the matching degree. Optionally, S205 includes the following: determining the title similarity between the entity name in the knowledge graph to be tested and the title in the candidate entity evidence unit; determining the key value similarity between the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit; and determining the matching degree of the candidate entity evidence unit according to the title similarity and the key value similarity of the candidate entity evidence unit.
Wherein the similarity may be determined by a method including according to a predictive model, or according to a text coincidence relation. Specifically, the matching degree of the candidate entity evidence unit is determined according to the title similarity and the key value similarity of the candidate entity evidence unit and a preset rule. Optionally, the matching degree of the candidate entity evidence unit is determined according to the sum of the title similarity and the key value similarity of the candidate entity evidence unit. And establishing a data base for subsequently selecting a target entity evidence unit of the knowledge graph to be detected by determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units.
S206, selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit.
The embodiment of the invention searches in the entity evidence unit of the target field by taking the entity name and the attribute value in the knowledge graph to be verified as search words to respectively obtain the first entity evidence unit and the second entity evidence unit, and then selects the final candidate entity evidence unit according to the mapping relation between the attribute name in the knowledge graph and the attribute keywords in the entity evidence unit, thereby realizing the effect of obtaining the candidate entity evidence unit and laying a foundation for determining the matching degree of the knowledge graph to be verified and the at least two candidate entity evidence units.
EXAMPLE III
Fig. 3 is a flowchart of a processing method of a knowledge graph according to a third embodiment of the present invention. The embodiment provides a specific implementation manner for the first embodiment, and as shown in fig. 3, the method may include:
s301, according to the knowledge graph to be verified, at least two candidate entity evidence units are selected from the entity evidence units in the target field to which the knowledge graph to be verified belongs.
S302, determining the title similarity between the entity name in the knowledge graph to be tested and the title in the candidate entity evidence unit.
Optionally, the first topic similarity and the second topic similarity between the entity name in the knowledge graph to be checked and the topic in the candidate entity evidence unit are determined by a method according to the prediction model and a method for determining the text coincidence relation, respectively.
Optionally, S302 includes:
A. and taking the entity name in the knowledge graph to be detected and the title in the candidate entity evidence unit as the input of a prediction model to obtain the similarity of the first title.
The prediction model is a deep learning model, and the optional prediction model comprises a pairwise model.
Optionally, the training process of the prediction model includes: taking S, P and O obtained by book map mining and sentences of S and O as positive examples; and constructing counter examples by using some heuristic rules, inputting the obtained positive examples and counter examples into a bidirectional long-short term memory network model or a deep-structured long-short term memory network model including an attention mechanism, and further training to obtain a prediction model.
B. And determining the similarity of a second title according to the text coincidence relation between the entity name of the knowledge graph to be detected and the title of the candidate entity evidence unit.
The text coincidence relation reflects the similarity degree between texts. Alternatively, the second title similarity may be determined by:
1) the values of the four parameters are set to zero, e.g., F1-0, F2-0, F3-0, and F4-0.
2) If the entity name of the knowledge graph to be tested is identical to the title of the candidate entity evidence unit, F1 ═ 1, F2 ═ 0, F3 ═ 0, and F4 ═ 0 are set.
3) If the entity name of the knowledge graph to be tested is not identical to the title of the candidate entity evidence unit, and the entity name of the knowledge graph to be tested is the suffix of the title of the candidate entity evidence unit, F1 ═ 0, F2 ═ 1, F3 ═ 0, and F4 ═ 0 are set.
4) If the entity name of the knowledge graph to be tested is not identical to the title of the candidate entity evidence unit, the entity name of the knowledge graph to be tested is not the suffix of the title of the candidate entity evidence unit, and the title of the candidate entity evidence unit is the suffix of the entity name of the knowledge graph to be tested, F1 ═ 0, F2 ═ 0, F3 ═ 1, and F4 ═ 0 are set.
5) If the entity name of the knowledge graph to be tested is not identical to the title of the candidate entity evidence unit, the entity name of the knowledge graph to be tested is not the suffix of the title of the candidate entity evidence unit, and the title of the candidate entity evidence unit is not the suffix of the entity name of the knowledge graph to be tested, F1 ═ 0, F2 ═ 0, F3 ═ 0, and F4 ═ 1 are set.
6) The second title similarity is F1 × a1+ F2 × a2+ F3 × a3+ F4 × jaccard × a4
Among them, a1, a2, a3 and a4 are fixed constants which can be adjusted according to actual conditions, and preferred settings are a 1-1, a 2-0.95, a 3-0.92 and a 4-0.9. Jaccard is Jaccard similarity, and the calculation process is as follows: 1) respectively segmenting the entity name of the knowledge graph to be detected and the title of the candidate entity evidence unit; 2) calculating the number of the same words after word segmentation; 3) and taking the ratio of the number of the same words after word segmentation to min (the number of entity name words and the number of title words) as the value of jaccard.
S303, determining the key value similarity between the attribute value in the knowledge graph to be checked and the attribute key value in the candidate entity evidence unit.
Optionally, the first key value similarity and the second key value similarity between the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit are determined by a method according to the prediction model and a method for determining the text coincidence relation, respectively.
Optionally, S303 includes:
A. taking the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit as the input of a prediction model to obtain the similarity of a first key value;
B. and determining the similarity of a second key value according to the text coincidence relation between the attribute value of the knowledge graph to be detected and the attribute key value of the candidate entity evidence unit.
S304, if the title similarity of any candidate entity evidence unit is smaller than the first title similarity threshold, or the key value similarity of the candidate entity evidence unit is smaller than the first key value similarity threshold, filtering the candidate entity evidence unit.
Specifically, a first title similarity and a second title similarity included in the title similarities are respectively compared with corresponding threshold values; and respectively comparing the first key value similarity and the second key value similarity included in the key value similarity with corresponding threshold values, and filtering candidate entity evidence units which do not meet the conditions.
Optionally, if the first title similarity of any candidate entity evidence unit is smaller than the first title first similarity sub-threshold and the second title similarity is smaller than the first title second similarity sub-threshold, or the first key value similarity is smaller than the first key value first similarity sub-threshold and the second key value similarity is smaller than the first key value second similarity sub-threshold, the candidate entity evidence unit is filtered.
Illustratively, for example, if the first topic first similarity sub-threshold is set to 0.9, the first topic second similarity sub-threshold is set to 0.85, the first topic similarity of the candidate entity evidence unit a is 0.95, and the second topic similarity is 0.8, then the candidate entity evidence unit a should be filtered; for another example, if the first key value and the first similarity sub-threshold are set to 0.9, the first key value and the second similarity sub-threshold are set to 0.85, the first key value similarity of the candidate entity evidence unit B is 0.9, and the second key value similarity of the candidate entity evidence unit B is 0.9, then the candidate entity evidence unit B is not filtered.
S305, if the title similarity of the candidate entity evidence unit is greater than a second title similarity threshold and the key value similarity is greater than a second key value similarity threshold, adding a first zone bit to the candidate entity evidence unit; otherwise, adding a second flag bit for the candidate entity evidence unit.
Optionally, comparing the maximum similarity between the first title similarity and the second title similarity included in the title similarities with a second title similarity threshold; comparing the maximum similarity of the first key value similarity and the second key value similarity included in the key value similarities with a second key value similarity threshold; if the maximum similarity between the first title similarity and the second title similarity is greater than a second title similarity threshold value, and the maximum similarity between the first key value similarity and the second key value similarity is greater than a second key value similarity threshold value, adding a first zone bit to the candidate entity evidence unit; otherwise, adding a second flag bit for the candidate entity evidence unit.
For example, the second title similarity threshold and the second key value similarity threshold are both set to be 0.95, the first title similarity of the candidate entity evidence unit B is 0.97, the second title similarity is 0.9, the first key value similarity is 0.96, and the second key value similarity is 0.93, then the maximum similarity between the first title similarity and the second title similarity 0.97 > the second title similarity threshold 0.95, and the maximum similarity between the first key value similarity and the second key value similarity 0.96 > the second key value similarity threshold 0.95, then the first flag is added to the candidate entity evidence unit B, and the preferred flag is "1".
For example, the second title similarity threshold and the second key value similarity threshold are both set to be 0.95, the first title similarity of the candidate entity evidence unit C is 0.94, the second title similarity is 0.9, the first key value similarity is 0.93, and the second key value similarity is 0.93, then the maximum similarity 0.94 of the first title similarity and the second title similarity is less than the second title similarity threshold 0.95, and the maximum similarity 0.93 of the first key value similarity and the second key value similarity is less than the second key value similarity threshold 0.95, then a second flag is added to the candidate entity evidence unit C, and the preferred flag is "0".
S306, sequencing at least two candidate entity evidence units according to the matching degree and the flag bit information of the candidate entity evidence units.
Optionally, the first title similarity, the second title similarity, the first key value similarity and the second key value similarity are summed to obtain the matching degree of the candidate entity evidence unit.
Specifically, the matching degree of the candidate entity evidence unit is obtained by linearly superimposing the first title similarity, the second title similarity, the first key value similarity and the second key value similarity. The higher the matching degree is, the more advanced the sorting is, and the candidate entity evidence units of different types of zone bit information can be sorted according to the matching degree respectively. For example, the candidate entity evidence unit with the flag bit of "1" is assigned to the first candidate entity evidence group, and the candidate entity evidence unit with the flag bit of "0" is assigned to the second candidate entity evidence group. And sorting according to the matching degree of the candidate entity evidence units in the first candidate entity evidence group and/or the second candidate entity evidence group.
And S307, determining a target entity evidence unit of the knowledge graph to be detected according to the sequencing result.
Specifically, the candidate entity evidence units within the preset number threshold are determined as target entity evidence units of the knowledge graph to be tested according to the sequence. Wherein the first number threshold of the first candidate entity evidence group is different from the second number threshold of the second candidate entity evidence group, typically the first number threshold is greater than the second number threshold. The target entity evidence units of the knowledge graph to be detected are respectively determined in the first candidate entity evidence group and the second candidate entity evidence group, so that omission of the entity evidence units close to the knowledge graph to be detected is avoided, and the knowledge graph to be detected is more comprehensive and reliable after the target entity evidence units are subsequently included.
According to the embodiment of the invention, the entity name in the knowledge graph to be detected, the title similarity between the entity name and the title in the candidate entity evidence unit, the attribute value in the knowledge graph to be detected and the key value similarity between the attribute key value in the candidate entity evidence unit are respectively calculated, and the target entity evidence unit is determined according to the obtained title similarity and the obtained key value similarity, so that the construction efficiency of the knowledge graph is improved, and the human intervention is reduced.
Example four
Fig. 4 is a schematic structural diagram of a knowledge graph processing apparatus according to a fourth embodiment of the present invention, which is capable of executing a knowledge graph processing method according to any embodiment of the present invention, and includes functional modules corresponding to the execution method and beneficial effects. As shown in fig. 4, the apparatus may include:
a candidate entity evidence unit selection module 41, configured to select, according to the to-be-verified knowledge graph, at least two candidate entity evidence units from entity evidence units in a target field to which the to-be-verified knowledge graph belongs;
a matching degree determining module 42, configured to determine matching degrees of the knowledge-graph to be tested and the at least two candidate entity evidence units;
and a target entity evidence unit selecting module 43, configured to select a target entity evidence unit of the to-be-verified knowledge graph from the at least two candidate entity evidence units according to the matching degree, so as to verify the to-be-verified knowledge graph based on the target entity evidence unit.
On the basis of the above embodiment, the apparatus further includes an entity evidence unit obtaining module, specifically configured to:
extracting attribute information and chapter title information to which the attribute information belongs from the target field fact;
taking the chapter title information as a title of an entity evidence unit in the target field;
taking the attribute name in the attribute information as an attribute key word of an entity evidence unit in the target field;
and taking the attribute content in the attribute information as an attribute key value of an entity evidence unit in the target field.
On the basis of the foregoing embodiment, the candidate entity evidence unit selecting module 41 is specifically configured to:
taking an entity name in a knowledge graph to be verified as a search word, and searching chapter title information, attribute keywords and attribute key values of entity evidence units in a target field to obtain a first entity evidence unit;
taking the attribute value in the knowledge graph to be verified as a search word, and searching in the attribute key value of the entity evidence unit in the target field to obtain a second entity evidence unit;
selecting at least two candidate entity evidence units from the first entity evidence unit and the second entity evidence unit.
On the basis of the foregoing embodiment, the candidate entity evidence unit selecting module 41 is further specifically configured to:
and selecting a candidate entity evidence unit matched with the knowledge graph to be verified from the at least two candidate entity evidence units according to the mapping relation between the attribute name in the knowledge graph and the attribute key word in the entity evidence unit and the knowledge graph to be verified, and filtering other candidate entity evidence units.
On the basis of the foregoing embodiment, the matching degree determining module 42 is specifically configured to:
determining the title similarity between the entity name in the knowledge graph to be tested and the title in the candidate entity evidence unit;
determining the key value similarity between the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit;
and determining the matching degree of the candidate entity evidence unit according to the title similarity and the key value similarity of the candidate entity evidence unit.
On the basis of the foregoing embodiment, the matching degree determining module 42 is further specifically configured to:
taking the entity name in the knowledge graph to be detected and the title in the candidate entity evidence unit as the input of a prediction model to obtain a first title similarity;
and determining the similarity of a second title according to the text coincidence relation between the entity name of the knowledge graph to be detected and the title of the candidate entity evidence unit.
On the basis of the foregoing embodiment, the matching degree determining module 42 is further specifically configured to:
taking the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit as the input of a prediction model to obtain the similarity of a first key value;
and determining the similarity of a second key value according to the text coincidence relation between the attribute value of the knowledge graph to be detected and the attribute key value of the candidate entity evidence unit.
On the basis of the foregoing embodiment, the matching degree determining module 42 is further specifically configured to:
if the title similarity of any candidate entity evidence unit is smaller than the first title similarity threshold, or the key value similarity of the candidate entity evidence unit is smaller than the first key value similarity threshold, filtering the candidate entity evidence unit.
On the basis of the foregoing embodiment, the target entity evidence unit selecting module 43 is specifically configured to:
if the title similarity of the candidate entity evidence unit is greater than a second title similarity threshold and the key value similarity is greater than a second key value similarity threshold, adding a first zone bit for the candidate entity evidence unit; otherwise, adding a second flag bit for the candidate entity evidence unit;
sorting at least two candidate entity evidence units according to the matching degree and the flag bit information of the candidate entity evidence units;
and determining a target entity evidence unit of the knowledge graph to be detected according to the sequencing result.
The processing device of the knowledge graph provided by the embodiment of the invention can execute the processing method of the knowledge graph provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to a method for processing a knowledge graph according to any embodiment of the present invention.
EXAMPLE five
Fig. 5 is a schematic structural diagram of an apparatus according to a fifth embodiment of the present invention. Fig. 5 illustrates a block diagram of an exemplary device 500 suitable for use in implementing embodiments of the present invention. The device 500 shown in fig. 5 is only an example and should not bring any limitations to the functionality and scope of use of the embodiments of the present invention.
As shown in fig. 5, device 500 is in the form of a general purpose computing device. The components of device 500 may include, but are not limited to: one or more processors or processing units 501, a system memory 502, and a bus 503 that couples the various system components (including the system memory 502 and the processing unit 501).
The system memory 502 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)504 and/or cache memory 505. The device 500 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 506 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 5, commonly referred to as a "hard drive"). Although not shown in FIG. 5, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 503 by one or more data media interfaces. Memory 502 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the invention.
A program/utility 508 having a set (at least one) of program modules 507 may be stored, for instance, in memory 502, such program modules 507 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 507 generally perform the functions and/or methodologies of embodiments of the invention as described herein.
The device 500 may also communicate with one or more external devices 509 (e.g., keyboard, pointing device, display 510, etc.), with one or more devices that enable a user to interact with the device 500, and/or with any devices (e.g., network card, modem, etc.) that enable the device 500 to communicate with one or more other computing devices. Such communication may occur via input/output (I/O) interfaces 511. Also, device 500 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet) via network adapter 512. As shown, the network adapter 512 communicates with the other modules of the device 500 over a bus 503. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
The processing unit 501 executes various functional applications and data processing by running a program stored in the system memory 502, for example, implementing a processing method of a knowledge graph provided by an embodiment of the present invention, including:
selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified;
determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units;
and selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit.
EXAMPLE six
An embodiment of the present invention further provides a computer-readable storage medium, where the computer-executable instructions, when executed by a computer processor, are configured to perform a method for processing a knowledge graph, the method including:
selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified;
determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units;
and selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit.
Of course, the storage medium provided by the embodiment of the present invention contains computer-executable instructions, and the computer-executable instructions are not limited to the method operations described above, and may also perform related operations in a method for processing a knowledge graph provided by any embodiment of the present invention. The computer-readable storage media of embodiments of the invention may take any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (15)
1. A method of knowledge graph processing, the method comprising:
selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified; wherein the entity evidence unit is obtained by mining knowledge facts in the target field;
determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units;
and selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, and verifying the knowledge graph to be verified based on the target entity evidence unit.
2. The method according to claim 1, wherein before selecting at least two candidate entity evidence units from the entity evidence units of the target domain to which the knowledge graph to be verified belongs according to the knowledge graph to be verified, the method further comprises:
extracting attribute information and chapter title information to which the attribute information belongs from the target field fact;
taking the chapter title information as a title of an entity evidence unit in the target field;
taking the attribute name in the attribute information as an attribute key word of an entity evidence unit in the target field;
and taking the attribute content in the attribute information as an attribute key value of an entity evidence unit in the target field.
3. The method according to claim 1, wherein selecting at least two candidate entity evidence units from the entity evidence units of the target domain to which the knowledge graph to be verified belongs according to the knowledge graph to be verified comprises:
taking an entity name in a knowledge graph to be verified as a search word, and searching chapter title information, attribute keywords and attribute key values of entity evidence units in a target field to obtain a first entity evidence unit;
taking the attribute value in the knowledge graph to be verified as a search word, and searching in the attribute key value of the entity evidence unit in the target field to obtain a second entity evidence unit;
selecting at least two candidate entity evidence units from the first entity evidence unit and the second entity evidence unit.
4. The method according to claim 3, wherein after selecting at least two candidate entity evidence units from the first entity evidence unit and the second entity evidence unit, further comprising:
and selecting a candidate entity evidence unit matched with the knowledge graph to be verified from the at least two candidate entity evidence units according to the mapping relation between the attribute name in the knowledge graph and the attribute key word in the entity evidence unit and the knowledge graph to be verified, and filtering other candidate entity evidence units.
5. The method according to claim 1, wherein determining the degree of matching of the knowledge-graph to be tested with the at least two candidate entity evidence units comprises:
determining the title similarity between the entity name in the knowledge graph to be tested and the title in the candidate entity evidence unit;
determining the key value similarity between the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit;
and determining the matching degree of the candidate entity evidence unit according to the title similarity and the key value similarity of the candidate entity evidence unit.
6. The method of claim 5, wherein determining title similarity between an entity name in the knowledge-graph to be examined and a title in the candidate entity evidence unit comprises:
taking the entity name in the knowledge graph to be detected and the title in the candidate entity evidence unit as the input of a prediction model to obtain a first title similarity;
and determining the similarity of a second title according to the text coincidence relation between the entity name of the knowledge graph to be detected and the title of the candidate entity evidence unit.
7. The method of claim 6, wherein determining key value similarity between attribute values in the knowledge-graph to be tested and attribute key values in the candidate entity evidence units comprises:
taking the attribute value in the knowledge graph to be tested and the attribute key value in the candidate entity evidence unit as the input of a prediction model to obtain the similarity of a first key value;
and determining the similarity of a second key value according to the text coincidence relation between the attribute value of the knowledge graph to be detected and the attribute key value of the candidate entity evidence unit.
8. The method according to claim 7, wherein before determining the matching degree of the candidate entity evidence unit according to the title similarity and the key value similarity of the candidate entity evidence unit, further comprising:
if the title similarity of any candidate entity evidence unit is smaller than the first title similarity threshold, or the key value similarity of the candidate entity evidence unit is smaller than the first key value similarity threshold, filtering the candidate entity evidence unit.
9. The method according to claim 7, wherein selecting the target entity evidence unit of the knowledge-graph to be tested from the at least two candidate entity evidence units comprises:
if the title similarity of the candidate entity evidence unit is greater than a second title similarity threshold and the key value similarity is greater than a second key value similarity threshold, adding a first zone bit for the candidate entity evidence unit; otherwise, adding a second flag bit for the candidate entity evidence unit;
sorting at least two candidate entity evidence units according to the matching degree and the flag bit information of the candidate entity evidence units;
and determining a target entity evidence unit of the knowledge graph to be detected according to the sequencing result.
10. An apparatus for processing a knowledge graph, the apparatus comprising:
the candidate entity evidence unit selection module is used for selecting at least two candidate entity evidence units from the entity evidence units in the target field to which the knowledge graph to be verified belongs according to the knowledge graph to be verified; wherein the entity evidence unit is obtained by mining knowledge facts in the target field;
the matching degree determining module is used for determining the matching degree of the knowledge graph to be detected and the at least two candidate entity evidence units;
and the target entity evidence unit selection module is used for selecting a target entity evidence unit of the knowledge graph to be verified from the at least two candidate entity evidence units according to the matching degree, so that the knowledge graph to be verified is verified based on the target entity evidence unit.
11. The apparatus according to claim 10, further comprising an entity evidence unit obtaining module, specifically configured to:
extracting attribute information and chapter title information to which the attribute information belongs from the target field fact;
taking the chapter title information as a title of an entity evidence unit in the target field;
taking the attribute name in the attribute information as an attribute key word of an entity evidence unit in the target field;
and taking the attribute content in the attribute information as an attribute key value of an entity evidence unit in the target field.
12. The apparatus according to claim 10, wherein the candidate entity evidence unit selection module is specifically configured to:
taking an entity name in a knowledge graph to be verified as a search word, and searching chapter title information, attribute keywords and attribute key values of entity evidence units in a target field to obtain a first entity evidence unit;
taking the attribute value in the knowledge graph to be verified as a search word, and searching in the attribute key value of the entity evidence unit in the target field to obtain a second entity evidence unit;
selecting at least two candidate entity evidence units from the first entity evidence unit and the second entity evidence unit.
13. The apparatus according to claim 12, wherein the candidate entity evidence unit selection module is further configured to:
and selecting a candidate entity evidence unit matched with the knowledge graph to be verified from the at least two candidate entity evidence units according to the mapping relation between the attribute name in the knowledge graph and the attribute key word in the entity evidence unit and the knowledge graph to be verified, and filtering other candidate entity evidence units.
14. An apparatus, characterized in that the apparatus further comprises:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method of processing a knowledge-graph as claimed in any one of claims 1 to 9.
15. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out a method of processing a knowledge-graph as claimed in any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910537133.5A CN110263083B (en) | 2019-06-20 | 2019-06-20 | Knowledge graph processing method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910537133.5A CN110263083B (en) | 2019-06-20 | 2019-06-20 | Knowledge graph processing method, device, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110263083A CN110263083A (en) | 2019-09-20 |
| CN110263083B true CN110263083B (en) | 2022-04-05 |
Family
ID=67919850
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910537133.5A Active CN110263083B (en) | 2019-06-20 | 2019-06-20 | Knowledge graph processing method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263083B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110750654A (en) * | 2019-10-28 | 2020-02-04 | 中国建设银行股份有限公司 | Knowledge graph acquisition method, device, equipment and medium |
| CN111640511B (en) * | 2020-05-29 | 2023-08-04 | 北京百度网讯科技有限公司 | Medical fact verification method, device, electronic equipment and storage medium |
Citations (5)
| 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 |
| CN108376160A (en) * | 2018-02-12 | 2018-08-07 | 北京大学 | A kind of Chinese knowledge mapping construction method and system |
| CN108388580A (en) * | 2018-01-24 | 2018-08-10 | 平安医疗健康管理股份有限公司 | Merge the dynamic knowledge collection of illustrative plates update method of medical knowledge and application case |
| CN109086356A (en) * | 2018-07-18 | 2018-12-25 | 哈尔滨工业大学 | The incorrect link relationship diagnosis of extensive knowledge mapping and modification method |
| CN109829056A (en) * | 2018-11-09 | 2019-05-31 | 广东外语外贸大学 | Predicate explains the fact that template-driven Abductive reasoning method |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108268581A (en) * | 2017-07-14 | 2018-07-10 | 广东神马搜索科技有限公司 | The construction method and device of knowledge mapping |
| US11574204B2 (en) * | 2017-12-06 | 2023-02-07 | Accenture Global Solutions Limited | Integrity evaluation of unstructured processes using artificial intelligence (AI) techniques |
| CN109299285A (en) * | 2018-09-11 | 2019-02-01 | 中国医学科学院医学信息研究所 | A kind of pharmacogenomics knowledge mapping construction method and system |
-
2019
- 2019-06-20 CN CN201910537133.5A patent/CN110263083B/en active Active
Patent Citations (5)
| 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 |
| CN108388580A (en) * | 2018-01-24 | 2018-08-10 | 平安医疗健康管理股份有限公司 | Merge the dynamic knowledge collection of illustrative plates update method of medical knowledge and application case |
| CN108376160A (en) * | 2018-02-12 | 2018-08-07 | 北京大学 | A kind of Chinese knowledge mapping construction method and system |
| CN109086356A (en) * | 2018-07-18 | 2018-12-25 | 哈尔滨工业大学 | The incorrect link relationship diagnosis of extensive knowledge mapping and modification method |
| CN109829056A (en) * | 2018-11-09 | 2019-05-31 | 广东外语外贸大学 | Predicate explains the fact that template-driven Abductive reasoning method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110263083A (en) | 2019-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11836440B1 (en) | Web browsing | |
| CN107992596B (en) | Text clustering method, text clustering device, server and storage medium | |
| US9558264B2 (en) | Identifying and displaying relationships between candidate answers | |
| CN111797214A (en) | FAQ database-based problem screening method and device, computer equipment and medium | |
| CN110457439B (en) | One-stop intelligent writing auxiliary method, device and system | |
| US20180341866A1 (en) | Method of building a sorting model, and application method and apparatus based on the model | |
| CN111475603B (en) | Enterprise identification recognition method, enterprise identification recognition device, computer equipment and storage medium | |
| US9483462B2 (en) | Generating training data for disambiguation | |
| US20090007272A1 (en) | Identifying data associated with security issue attributes | |
| CN109033385B (en) | Picture retrieval method, device, server and storage medium | |
| CN110162786B (en) | Method and device for constructing configuration file and extracting structured information | |
| CN113544689A (en) | Generating and providing additional content for a source view of a document | |
| CN111723260B (en) | Recommended content acquisition method and device, electronic equipment and readable storage medium | |
| US20200118022A1 (en) | Determining Comprehensiveness of Question Paper Given Syllabus | |
| KR101541306B1 (en) | Computer enabled method of important keyword extraction, server performing the same and storage media storing the same | |
| US20220309332A1 (en) | Automated contextual processing of unstructured data | |
| US11669556B1 (en) | Method and system for document retrieval and exploration augmented by knowledge graphs | |
| KR102421904B1 (en) | the method to advance the analysis of the causes of disasters | |
| CN110750627A (en) | Material retrieval method and device, electronic equipment and storage medium | |
| CN110263083B (en) | Knowledge graph processing method, device, equipment and medium | |
| CN116402166B (en) | Training method and device of prediction model, electronic equipment and storage medium | |
| CN113314207A (en) | Object recommendation method and device, storage medium and electronic equipment | |
| CN109977197B (en) | Electronic exercise processing method, device, equipment and storage medium | |
| CN108733702B (en) | Method, device, electronic equipment and medium for extracting upper and lower relation of user query | |
| US11663215B2 (en) | Selectively targeting content section for cognitive analytics and search |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |