CN109508420A - A kind of cleaning method and device of knowledge mapping attribute - Google Patents
A kind of cleaning method and device of knowledge mapping attribute Download PDFInfo
- Publication number
- CN109508420A CN109508420A CN201811415629.7A CN201811415629A CN109508420A CN 109508420 A CN109508420 A CN 109508420A CN 201811415629 A CN201811415629 A CN 201811415629A CN 109508420 A CN109508420 A CN 109508420A
- Authority
- CN
- China
- Prior art keywords
- search result
- triple
- attribute value
- percentage
- effective
- 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.)
- Granted
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses the cleaning methods and device of a kind of knowledge mapping attribute, are related to knowledge mapping technical field, using substitution manual operation is automaticly inspected, can directly automatically wash attribute value mistake.Main technical schemes of the embodiment of the present invention are as follows: when receiving a knowledge mapping, triple, the element combinations that the triple is made of entity elements, attribute value element corresponding with the associated property element of the entity elements and the property element are extracted from the knowledge mapping by the parsing to the knowledge mapping；Utilize the corresponding inquiry problem of triple described in default structure of transvers plate；It searches for the inquiry problem on the internet by search engine, obtains corresponding search result；If according to described search result determine attribute value element that the triple includes be it is wrong, delete the attribute value element.The embodiment of the present invention is mainly used in the attribute value for automaticly inspecting and cleaning knowledge mapping.
Description
Technical field
The present embodiments relate to the cleaning method of knowledge mapping technical field more particularly to a kind of knowledge mapping attribute and
Device.
Background technique
Knowledge mapping (Knowledge Graph, KG) is also known as mapping knowledge domains, is known as knowledge domain in books and information group
Visualization or ken map map, are a series of a variety of different figures of explicit knowledge's development process and structural relation,
Knowledge resource and its carrier, excavation, analysis, building, drafting and explicit knowledge and the phase between them are described with visualization technique
Mutually connection.
Currently, the higher external data source of the Reliability ratio of available structuring carrys out direct construction knowledge mapping, such as:
Baidupedia etc., since this kind of external data source is easier to obtain, so but also constructing knowledge mapping all the more simply, just
It is prompt.But due to also including the information of human-edited in external data source, this is existing it is avoided that there are Edit Error
Counter-measure is by manually checking the knowledge mapping of building, this not only labor intensive cost also reduces and checks efficiency.
Summary of the invention
In view of this, the embodiment of the present invention provides the cleaning method and device of a kind of knowledge mapping attribute, main purpose exists
In using substitution manual operation is automaticly inspected, optimizes the checking process of knowledge mapping attribute, improve and check efficiency, it can directly certainly
Dynamic ground washes attribute value mistake.
In order to achieve the above object, the embodiment of the present invention mainly provides the following technical solutions:
In a first aspect, the embodiment of the invention provides a kind of cleaning methods of knowledge mapping attribute, this method comprises:
When receiving a knowledge mapping, three are extracted from the knowledge mapping by the parsing to the knowledge mapping
Tuple, the triple are by entity elements, corresponding with the associated property element of the entity elements and the property element
Attribute value element composition element combinations；
Utilize the corresponding inquiry problem of triple described in default structure of transvers plate；
It searches for the inquiry problem on the internet by search engine, obtains corresponding search result；
If according to described search result determine attribute value element that the triple includes be it is wrong, delete the category
Property value element.
Optionally, the attribute value element for determining that the triple includes according to described search result is wrong, packet
It includes:
Count the total number of the corresponding search result of the inquiry problem；
Crawl the corresponding data information of described search result；
Judge the entity elements for including with the presence or absence of the triple in the corresponding data information of described search result, belong to
Property element and attribute value element；
If so, being effective search result by described search result queue；
Calculate the item number of effective search result and the percentage of the total number；
According to the percentage of the item number of effective search result and the total number, the category that the triple includes is determined
Whether property value element is wrong.
Optionally, described to have according to if searching for the corresponding inquiry problem of the triple by a search engine
The percentage of the item number and the total number of imitating search result determines whether the attribute value element that the triple includes is mistake
Include:
Obtain the item number of the corresponding effective search result of described search engine and the percentage of total number；
Whether the percentage of the item number and the total number that judge effective search result is less than the first preset threshold；
If so, the attribute value element for determining that the triple includes is wrong.
Optionally, described according to institute if searching for the corresponding inquiry problem of the triple respectively by multiple search engines
The percentage of the item number and the total number of stating effective search result determine attribute value element that the triple includes whether be
Mistake includes:
Obtain the item number of the corresponding effective search result of the multiple search engine and the percentage of total number；
According to the weight distributed in advance the multiple search engine, the corresponding percentage of the multiple search engine is held
Row weighting processing obtains weighting treated percentage；
Judge weighting treated the percentage whether less than the first preset threshold；
If so, the attribute value element for determining that the triple includes is wrong.
Optionally, before calculating the percentage of effective searching bar number and the total number, the method also includes:
If finding the property element and attribute that the triple includes in the corresponding data information of described search result
Value element and the entity elements that the triple includes are not found, then extract and wrap in the corresponding data information of described search result
The noun contained；
Judge the noun whether be the entity elements that the triple includes alias；
If so, being effective search result by described search result queue.
Optionally, it is described by described search result queue be effective search result after, the method also includes:
It obtains the answer information for including in the corresponding data information of described search result and the answer information is corresponding
Thumb up information；
Information is thumbed up according to the answer information and the answer information are corresponding, statistics described search result is corresponding
Always thumb up number；
If the number that always thumbs up deletes described search knot less than the second preset threshold in effective search result
Fruit.
Optionally, the method also includes:
Judge the entity elements for including with the presence or absence of the triple in the corresponding heading message of effective search result,
Property element and attribute value element；
If so, being inquiry problem to be selected by the corresponding title mark of the effective search result；
According to the sequencing of web displaying arranged effective search result, determine that the inquiry to be selected is asked
The sequence of topic is successive；
Sequence according to the inquiry problem to be selected is successive, chooses described in the inquiry problem conduct to be selected of preset number
The corresponding newly-increased inquiry problem of triple.
Optionally, described to search for the inquiry problem on the internet by search engine, further comprise:
The corresponding inquiry problem of the triple is searched on the Ask-Answer Community of internet by search engine.
Second aspect, the embodiment of the invention also provides a kind of cleaning device of knowledge mapping attribute, which includes:
Extraction unit, for being known from described by the parsing to the knowledge mapping when receiving a knowledge mapping
Know and extract triple in map, the triple is by entity elements and the associated property element of the entity elements and institute
State the element combinations of the corresponding attribute value element composition of property element；
Structural unit, the corresponding inquiry problem of triple for being extracted using extraction unit described in default structure of transvers plate；
Search unit is obtained for searching for the inquiry problem of the structural unit construction on the internet by search engine
To corresponding search result；
Determination unit, for according to described search unit searches to search result determine attribute that the triple includes
Be worth element whether mistake；
Unit is deleted, is mistake for working as according to the attribute value element that the determination unit determines that the triple includes
, delete the attribute value element.
Optionally, the determination unit includes:
Statistical module, for counting the total number of the corresponding search result of the inquiry problem；
Module is crawled, for crawling the corresponding data information of described search result；
Judgment module whether there is for judging in described crawl in the corresponding data information of search result that module crawls
Entity elements, property element and the attribute value element that the triple includes；
Mark module, for judging to be that there are institutes in the corresponding data information of described search result when the judgment module
Entity elements, property element and attribute value element that triple includes are stated, is that effectively search is tied by described search result queue
Fruit；
Computing module, item number and the statistical module for calculating effective search result of the mark module label are united
The percentage of the total number of meter；
Determining module, the item number of effective search result for being calculated according to the computing module and the hundred of the total number
Divide ratio, determines whether the attribute value element that the triple includes is wrong；
Optionally, the determining module includes:
Acquisition submodule, for obtaining the item number of the corresponding effective search result of described search engine and the percentage of total number
Than；
Judging submodule, the item number and the total number of effective search result for judging the acquisition submodule acquisition
Percentage whether less than the first preset threshold；
Determine submodule, for when the judging submodule judges that the percentage is less than the first preset threshold, really
The attribute value element that the fixed triple includes is wrong.
Optionally, the determining module includes:
The acquisition submodule is also used to obtain the item number of the corresponding effective search result of the multiple search engine
With the percentage of total number；
Submodule is handled, for basis in advance to the weight of the multiple search engine distribution, to the acquisition submodule
The corresponding percentage of multiple search engines obtained executes weighting processing, obtains weighting treated percentage；
The judging submodule, treated for judging the obtained weighting of processing submodule, and whether percentage is less than
First preset threshold；
The determining submodule, for judge described to weight that treated percentage is less than the when the judging submodule
When one preset threshold, the attribute value element for determining that the triple includes is wrong.
Optionally, the determination unit further include:
Extraction module, for before calculating the percentage of effective searching bar number and the total number, if described
The property element and attribute value element that the triple includes are found in the corresponding data information of search result and are not found
The entity elements that the triple includes then extract the noun for including in the corresponding data information of described search result；
The judgment module is also used to judge whether noun that the extraction module extracts is reality that the triple includes
The alias of element of volume；
The mark module is also used to judge that the noun is the entity member that the triple includes when the judgment module
It is effective search result by described search result queue when the alias of element.
Optionally, the determination unit further include:
Module is obtained, for being acquisition described search result after effective search result by described search result queue
The answer information that includes in the corresponding data information and answer information is corresponding thumbs up information；
The statistical module is also used to thumb up information, system according to the answer information and the answer information are corresponding
Described search result is corresponding always thumbs up number for meter；
Removing module, if always thumbing up number less than the second preset threshold, described for the statistical module counts
Described search result is deleted in effective search result.
Optionally, described device further include:
Judging unit, for judging in the corresponding heading message of effective search result with the presence or absence of the triple packet
Entity elements, property element and the attribute value element contained；
Marking unit, for judging it is to exist in the corresponding heading message of effective search result when the judging unit
Entity elements, property element and the attribute value element that the triple includes, by the corresponding title of the effective search result
Labeled as inquiry problem to be selected；
The determination unit is also used to successive suitable according to being arranged effective search result for web displaying
Sequence determines that the sequence of the inquiry problem to be selected is successive；
The determination unit, is also used to successive according to the sequence of the inquiry problem to be selected, chooses the described of preset number
Inquiry problem to be selected is as the corresponding newly-increased inquiry problem of the triple.
The third aspect, the embodiment of the invention also provides a kind of electronic equipment, comprising:
At least one processor；
And at least one processor, the bus being connected to the processor；Wherein,
The processor, memory complete mutual communication by the bus；The processor is described for calling
Program instruction in memory, to execute the cleaning method of knowledge mapping attribute as described above.
Fourth aspect, it is described non-transient the embodiment of the invention also provides a kind of non-transient computer readable storage medium
Computer-readable recording medium storage computer instruction, the computer instruction execute the computer as described above to know
Know the cleaning method of map attribute.
By above-mentioned technical proposal, technical solution provided in an embodiment of the present invention is at least had the advantage that
The cleaning method and device of a kind of knowledge mapping attribute provided in an embodiment of the present invention.The embodiment of the present invention is to utilize
Default template constructs inquiry problem to the triple (that is: entity, attribute and attribute value) of knowledge mapping, is existed by search engine
Search query questions and search result is obtained on internet, and then by the number of the known search result comprising above-mentioned triple
It is believed that whether breath come to judge automatically the attribute value that triple includes be wrong, and then can complete apparent attribute automatically
Value mistake washes.Compared to the prior art, solve causes to expend because of manual inspection knowledge mapping attribute with the presence or absence of mistake
The problem of human cost, low efficiency.The embodiment of the present invention optimizes knowledge mapping attribute using substitution manual operation is automaticly inspected
Checking process improves and checks efficiency, can directly automatically wash attribute value mistake.
Above description is only the general introduction of technical solution of the embodiment of the present invention, in order to better understand the embodiment of the present invention
Technological means, and can be implemented in accordance with the contents of the specification, and in order to allow above and other mesh of the embodiment of the present invention
, feature and advantage can be more clearly understood, the special specific embodiment for lifting the embodiment of the present invention below.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
The limitation of embodiment.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows the cleaning method flow chart of knowledge mapping attribute provided in an embodiment of the present invention；
Fig. 2 shows the cleaning method flow charts of another knowledge mapping attribute provided in an embodiment of the present invention；
Fig. 3 shows a kind of composition block diagram of the cleaning device of knowledge mapping attribute provided in an embodiment of the present invention；
Fig. 4 shows the composition block diagram of the cleaning device of another knowledge mapping attribute provided in an embodiment of the present invention；
Fig. 5 shows a kind of structural representation of the electronic equipment of the cleaning of knowledge mapping attribute provided in an embodiment of the present invention
Figure.
Specific embodiment
The exemplary embodiment for embodiment that the present invention will be described in more detail below with reference to accompanying drawings.Although being shown in attached drawing
The exemplary embodiment of the embodiment of the present invention, it being understood, however, that may be realized in various forms the embodiment of the present invention without answering
It is limited by the embodiments set forth herein.It is to be able to thoroughly understand implementation of the present invention on the contrary, providing these embodiments
Example, and the range of the embodiment of the present invention can be fully disclosed to those skilled in the art.
The embodiment of the invention provides a kind of cleaning methods of knowledge mapping attribute, as shown in Figure 1, this method is that construction is real
Element of volume, property element, attribute value element form the inquiry problem of triple, and search on the internet by means of search engine
The data information of the rope corresponding search result of inquiry problem, the attribute value element to judge that above-mentioned triple includes whether there is
Mistake provides step in detail below to this embodiment of the present invention:
101, when receiving a knowledge mapping, ternary is extracted from knowledge mapping by the parsing to knowledge mapping
Group.
Wherein, triple is by entity elements, category corresponding with the associated property element of entity elements and property element
Property value element composition element combinations, such as: " China-population -1,300,000,000 ".
By parse a knowledge mapping, the triple of available a variety of relationship types, such as: " entity-relationship-
The triples such as entity ", " entity-attribute-attribute value " are by parsing a knowledge mapping simultaneously for the embodiment of the present invention
" entity-attribute-attribute value " triple is therefrom obtained, is examined with the attribute value of the attribute for including to the triple
It looks into.
102, the corresponding inquiry problem of default structure of transvers plate triple is utilized.
Wherein, default template be with above-mentioned " entity-attribute-attribute value " the matched template of triple, it be based on should
Logical relation in triple between entity, attribute and attribute value is come the sentence pattern template that constructs in advance, for example " XXX of XXX is
What is XXX? ", or " XXX of XXX is really XXX? ", and then by the entity for including in triple, attribute and attribute
Value is added to sentence pattern template, so that it may construct inquiry problem.For example, being obtained for triple " China-population -1,300,000,000 "
Corresponding inquiry problem, can be " why Chinese population is 1,300,000,000? " or " Chinese population is really 1,300,000,000? " Deng
Deng in embodiments of the present invention, being not specifically limited to the quantity of the sentence pattern template in default template.
In embodiments of the present invention, it is using the effect of the corresponding inquiry problem of default structure of transvers plate triple: obtains packet
The clause of entity, attribute and property element containing triple, and then pass through search engine on the internet using the clause
Search whether there is corresponding search result, and search engine can be Baidu search, 360 search, Google search etc. here.
It should be noted that for the entity, attribute and attribute that only input triple in the search box of search engine
The keyword of value obtains corresponding search result, and the entity, attribute and attribute value with input comprising triple have logic
Clause and to obtain corresponding search result be different.Since the clause of input has logic, so correspondingly, logical
The search result for crossing search engine feedback generally also has identity logic, such as: " height of Zhang San is really 180 for input
? ", be usually all data information related with the height of Zhang San is discussed correspondingly, obtained search result, compare under,
Clause is inputted in the search box of search engine, more facilitates to obtain preferable search result.
103, pass through search engine search query questions on the internet, obtain corresponding search result.
Wherein, search engine can be Baidu search, 360 search, Google search etc..
In embodiments of the present invention, available in the corresponding inquiry problem of search box input triple of search engine
A plurality of search result, every search result generally comprise title, problem summary, information etc. of answering of online friend.
If 104, according to search result determine attribute value element that triple includes be it is wrong, delete attribute value member
Element.
Wherein, the attribute value element that triple includes is wrong, such as: triple " China-currency-dollar ", again
Alternatively, triple " Liu Dehua-height -175 ", wherein 175 be wrong.
In embodiments of the present invention, since the knowledge mapping to public is much based on directly on the confidence level of structuring
Higher external data source and construct, so can use on network disclosed data information also to verify this knowledge mapping
Attribute attribute value whether there is mistake.For the embodiment of the present invention, be after obtaining the corresponding search result of inquiry problem,
Determine whether the attribute value that triple includes is wrong according to the statistical analysis to search result, if so, directly deletion
The attribute value, and then reach the Rapid Cleaning to the attribute value comprising mistake in triple.Inspection provided in an embodiment of the present invention
The attribute value of triple, can simple, the obvious attribute value that easily will include in triple with the presence or absence of the method for mistake
Mistake is found out, to achieve the purpose that automaticly inspect mistake, be automatically performed cleaning.
A kind of cleaning method of knowledge mapping attribute provided in an embodiment of the present invention.The embodiment of the present invention is to utilize default mould
Plate constructs inquiry problem to the triple (that is: entity, attribute and attribute value) of knowledge mapping, by search engine in internet
Upper search query questions simultaneously obtain search result, and then by the data information of the known search result comprising above-mentioned triple
Whether it is wrong to judge automatically the attribute value that triple includes, and then can completes apparent attribute value mistake automatically
It washes.Compared to the prior art, solve because manual inspection knowledge mapping attribute with the presence or absence of mistake cause labor intensive at
Originally, the problem of low efficiency.The embodiment of the present invention optimizes the inspection of knowledge mapping attribute using substitution manual operation is automaticly inspected
Journey improves and checks efficiency, can directly automatically wash attribute value mistake.
In order to make more detailed explanation to above-described embodiment, the embodiment of the invention also provides another knowledge mappings
The cleaning method of attribute, as shown in Fig. 2, this method is to search for triple in the Ask-Answer Community of internet using different search engines
Corresponding inquiry problem, with the search result quantity that can have more effectively search results by Ask-Answer Community and reduction obtains
But the characteristics of sufficiently achieving search purpose, and then the search result according to Ask-Answer Community feedback is come the attribute for checking triple
No there are mistakes, improve the accuracy checked, efficiency with final, provide step in detail below to this embodiment of the present invention:
201, when receiving a knowledge mapping, ternary is extracted from knowledge mapping by the parsing to knowledge mapping
Group.
Wherein, triple is by entity elements, category corresponding with the associated property element of entity elements and property element
Property value element composition element combinations.
Statement for this step, refers to step 101, and details are not described herein again.
202, the corresponding inquiry problem of default structure of transvers plate triple is utilized.
Statement for this step, refers to step 102, and details are not described herein again.
203, by search engine on the Ask-Answer Community of internet search query questions, obtain corresponding search result.
Wherein, Ask-Answer Community refers to that Baidu knows, 360 question and answer, knows the interacting Question-Answers platforms such as net.
It in embodiments of the present invention, can search inquiry be asked on one or more Ask-Answer Communities by a kind of search engine
Topic, or, by multiple and different search engines on one or more Ask-Answer Community search query questions, this is adapted to
The demand of concrete application scene.
If 204, according to search result determine attribute value element that triple includes be it is wrong, delete attribute value member
Element.
Wherein, determine whether the attribute value element that triple includes is wrong, specific steps according to search result, it can be with
It is as follows:
Firstly, on Ask-Answer Community the corresponding search result of statistical query problem total number；Secondly, crawling search result
Corresponding data information judges the entity elements for including with the presence or absence of triple in the corresponding data information of search result, belongs to
Property element and attribute value element, if so, search result is labeled as effective search result；Finally, calculating effective search result
Item number and total number percentage, if judging, percentage is less than the first preset threshold, it is determined that the attribute that triple includes
Value element is wrong.
For above-mentioned steps, it should be noted that above-mentioned sentenced according only to the search result of Ask-Answer Community feedback
Whether the attribute value element that disconnected triple includes is wrong.Such as: triple " China-alias-U.S. ", corresponding inquiry
Problem is " Chinese alias is really the U.S.? ", a plurality of search result is searched in Ask-Answer Community accordingly, then can
To obtain the data information of each search result by crawling, further successively to search the data of every search result
In information whether include simultaneously triple entity, attribute and attribute value, and will simultaneously comprising the entity of triple, attribute with
And the search result of attribute value is tied as effective search result to count the total search corresponding with inquiry problem of effective search result
The percentage of fruit, and then percentage and a preset threshold are compared, such as 2%, if being less than the preset threshold, then it is assumed that three
The attribute value that tuple includes is wrong, it should be noted that it is above-mentioned that preset threshold is less than according to percentage, indicate that triple
The probability being combined into is very little, thus indirectly assert triple attribute value be it is wrong, still, if there is triple
Attribute value is the situation that can not be verified with the presence or absence of mistake, it may also be said to which this is not apparent attribute error, so for class
Like ambiguous judgement, also it is not considered as the attribute value of triple in the presence of mistake in the embodiment of the present invention, it is possible to keep away
Exempt to judge by accident.
Further, determine whether the attribute value element that triple includes is wrong according to search result as to above-mentioned
The supplement of specific steps can also search for respectively the corresponding inquiry of triple in multiple Ask-Answer Communities in embodiments of the present invention
Problem, the search result more to be enriched, various, finally to improve the attribute for analyzing and determining triple according to search result
Accuracy of the value with the presence or absence of mistake.Specifically, steps are as follows for supplement:
If searching for the corresponding inquiry problem of triple respectively on multiple Ask-Answer Communities, it is right respectively to obtain multiple Ask-Answer Communities
The item number for the effective search result answered and the percentage of total number, then it is right according to the weight distributed in advance multiple Ask-Answer Communities
The corresponding percentage in multiple Ask-Answer Communities executes weighting processing, if judge the percentage obtained after weighting is handled less than first
Preset threshold, it is determined that the attribute value element that triple includes is wrong.
It should be noted that be amount of access, the number of search results etc. according to an Ask-Answer Community in embodiments of the present invention,
It is simply that weight is arranged to it according to the concerned degree of the Ask-Answer Community, confidence level, for measuring to different question and answer
Community obtains the degree of recognition of corresponding above-mentioned percentage, and then is the equal of comprehensive multiple and different question and answer societies based on weighting processing
The data information of area's feedback makes comprehensive inspection result, to improve the accuracy checked.Similarly further analysis, for this
Inventive embodiments can also preset the corresponding weight of different search engines, and preset different Ask-Answer Communities simultaneously
Corresponding weight, and then comprehensive inspection result is obtained according to comprehensive two aspect weights, more preferably to improve the accurate of inspection
Property.
Further, the corresponding effective search result of inquiry problem proposed for the embodiment of the present invention, it is also contemplated that
There may also be corresponding alias for the entity of triple, the triple being made of in this way entity, attribute and attribute value, with entity
The triple of alias, attribute and attribute value composition is equivalent, so in embodiments of the present invention, it should also be in search result
The case where being searched whether in corresponding data information there are the attribute of triple and attribute value but corresponding entity be not present,
Such case if it exists then continues to judge whether the noun for including in the corresponding data information of search result is the triple packet
The alias of the entity elements contained, if so, search result is labeled as effective search result.Such as: triple " China-people
Mouth -1,300,000,000 " is equal with triple " People's Republic of China (PRC)-population -1,300,000,000 ".
It further, can also be by the following method to effective search result obtained above for the embodiment of the present invention
It is screened, to reduce the quantity of effective search result, the effective search result being contracted by here can't be to final inspection three
Whether the attribute value of tuple is that wrong inspection result has influence, but so operation can reduce the search knot of comparison instead
Fruit quantity is to improve to obtain the efficiency of inspection result with final purpose.Specific method may include: to obtain search result pair
It the answer information that includes in the data information answered and answers that information is corresponding to thumb up information, believes according to answering information and answering
Cease it is corresponding thumb up information, search result is corresponding always thumbs up number for statistics, if always thumbing up number less than the second preset threshold,
The Delete Search result in effective search result.It should be noted that above-mentioned be equivalent to be by the corresponding work of search result
Answer mesh and number of answering is corresponding thumbs up number to count the attention rate of a search result, thus high using attention rate,
And effectively search result analyzes and checks the attribute value of triple with the presence or absence of mistake, so that it may reach the embodiment of the present invention
Technical effect, to reduce unnecessary check cost.
In embodiments of the present invention, if checking the triple attribute value that includes, there are mistakes, directly delete attribute value and reach
To the purpose cleaned automatically, it can also be gone out in the presence of mistake with clear indication with the attribute value of marked erroneous, avoid to
Family interferes, and waits manual correction.
205, judge the entity elements for including with the presence or absence of triple in the corresponding heading message of effective search result, attribute
Element and attribute value element.
206, if so, being inquiry problem to be selected by the corresponding title mark of effective search result.
For above-mentioned steps 205-206, in embodiments of the present invention, in addition to using default structure of transvers plate inquiry problem it
Outside, the title of effective search result can also be selected to ask from the corresponding effective search result of inquiry problem as newly-increased inquiry
Topic, to increase the diversity that triple corresponds to inquiry problem.So being to look for effectively searching for knot first for the embodiment of the present invention
Entity, attribute and the attribute value for including with the presence or absence of triple in the heading message of fruit, if so, can be first by such mark
Topic is used as inquiry problem to be selected.
207, according to the sequencing of web displaying arranged effective search result, inquiry problem to be selected is determined
Sequence is successive.
208, the sequence according to inquiry problem to be selected is successive, chooses the inquiry problem to be selected of preset number as triple pair
The newly-increased inquiry problem answered.
For above-mentioned steps 207-208, since there are many obtained inquiry problematic amount to be selected, so needing therefrom preferentially to select
Preset number is selected as newly-increased inquiry problem, the foundation of specific choice is the successive of effective search result on an Ask-Answer Community
Sequence, this is also to complete the preferentially screening that the embodiment of the present invention executes by the intelligent sequencing of Ask-Answer Community, according to Ask-Answer Community
Intelligent sequencing it is found that the first effective search result of sequence is usually and inquiry logic of questions is maximally related, amount of access is high
Either the date is closer preferentially as a result, so the embodiment of the present invention can be selected directly by the intelligent sequencing completion of Ask-Answer Community
It is preferred that selecting the technical need of effective search result, and then optimum selecting inquiry problem to be selected indirectly, and finally determine newly-increased look into
Inquiry topic.The effect that newly-increased triple corresponds to inquiry problem is: so as to the subsequent identical triple to other knowledge mappings into
When row checks, newly-increased inquiry problem can be integrated to execute corresponding search, to increase the diversity of search result, Jin Erli
Analyze to obtain attribute value that triple includes with the presence or absence of the inspection result of mistake, finally to mention with more various search result
The accuracy rate that height checks.
Further, as the realization to method shown in above-mentioned Fig. 1, Fig. 2, the embodiment of the invention provides a kind of knowledge graphs
Compose the cleaning device of attribute.The Installation practice is corresponding with preceding method embodiment, and to be easy to read, present apparatus embodiment is no longer
Detail content in preceding method embodiment is repeated one by one, it should be understood that the device in the present embodiment can correspond to
Realize the full content in preceding method embodiment.Specifically as shown in figure 3, the device includes:
Extraction unit 31, for when receiving a knowledge mapping, by the parsing to the knowledge mapping from described
Extract triple in knowledge mapping, the triple be by entity elements, with the associated property element of the entity elements and
The element combinations of the corresponding attribute value element composition of the property element；
Structural unit 32, the corresponding inquiry of triple for being extracted using extraction unit 31 described in default structure of transvers plate are asked
Topic；
Search unit 33 is asked for searching for the inquiry that the structural unit 32 constructs on the internet by search engine
Topic, obtains corresponding search result；
Determination unit 34, the search result for being searched according to described search unit 33 determine that the triple includes
Attribute value element whether mistake；
Unit 35 is deleted, for being wrong according to the determining triple of the determination unit 34 when the attribute value element for including
Accidentally, delete the attribute value element.
Further, as shown in figure 4, the determination unit 34 includes:
Statistical module 341, for counting the total number of the corresponding search result of the inquiry problem；
Module 342 is crawled, for crawling the corresponding data information of described search result；
Judgment module 343 is in described crawl in the corresponding data information of search result that module 342 crawls for judging
The no entity elements for including there are the triple, property element and attribute value element；
Mark module 344, for when the judgment module 343 judge be in the corresponding data information of described search result
Entity elements, property element and the attribute value element for including there are the triple, are effective by described search result queue
Search result；
Computing module 345, for calculate effective search result that the mark module 344 marks item number and the statistics
The percentage of the total number of module statistics；
Determining module 346, the item number and total item of effective search result for being calculated according to the computing module 345
Several percentage determines whether the attribute value element that the triple includes is wrong；
Further, as shown in figure 4, if searching for the corresponding inquiry problem of the triple, institute by a search engine
State determining module 346 further include:
Acquisition submodule 3461, for obtaining the item number and total number of described search engine corresponding effective search result
Percentage；
Judging submodule 3462, for judging the item number of effective search result that the acquisition submodule obtains and described total
Whether the percentage of item number is less than the first preset threshold；
Submodule 3463 is determined, for judging that the percentage is less than the first preset threshold when the judging submodule
When, the attribute value element for determining that the triple includes is wrong.
Further, as shown in figure 4, being asked if searching for the corresponding inquiry of the triple respectively by multiple search engines
Topic, the determining module 346 further include:
The acquisition submodule 3461 is also used to obtain the corresponding effective search result of the multiple search engine
The percentage of item number and total number；
Submodule 3464 is handled, for basis in advance to the weight of the multiple search engine distribution, to acquisition
The corresponding percentage of multiple search engines that module 3461 obtains executes weighting processing, obtains weighting treated percentage；
The judging submodule 3462 is also used to judge the obtained weighting of processing submodule 3464 treated percentage
Than whether less than the first preset threshold；
The determining submodule 3463 is also used to judge weighting treated the percentage when the judging submodule 3462
When than being less than the first preset threshold, the attribute value element for determining that the triple includes is wrong.
Further, as shown in figure 4, the determination unit 34 further include:
Extraction module 347, for before calculating the percentage of effective searching bar number and the total number, if in institute
It states and finds property element and attribute value element that the triple includes in the corresponding data information of search result and do not search
The entity elements for including to the triple then extract the noun for including in the corresponding data information of described search result；
The judgment module 343, whether the noun for being also used to judge that the extraction module 347 extracts is the triple packet
The alias of the entity elements contained；
The mark module 344 is also used to judge that the noun is that the triple includes when the judgment module 343
It is effective search result by described search result queue when the alias of entity elements.
Further, as shown in figure 4, the determination unit 34 further include:
Module 348 is obtained, for being acquisition described search knot after effective search result by described search result queue
The answer information that includes in the corresponding data information of the fruit and answer information is corresponding thumbs up information；
The statistical module 341 is also used to according to the answer information and the answer information is corresponding thumbs up information,
Described search result is corresponding always thumbs up number for statistics；
Removing module 349, if always thumbing up number less than the second preset threshold for the statistical module 341 statistics,
Described search result is deleted in effective search result.
Further, as shown in figure 4, described device further include:
Judging unit 36, for judging in the corresponding heading message of effective search result with the presence or absence of the triple
Entity elements, property element and the attribute value element for including；
Marking unit 37, for when the judging unit 36 judge be in the corresponding heading message of effective search result
Entity elements, property element and the attribute value element for including there are the triple, effective search result is corresponding
Title mark is inquiry problem to be selected；
The determination unit 34 is also used to successive suitable according to being arranged effective search result for web displaying
Sequence determines that the sequence of the inquiry problem to be selected is successive；
The determination unit 34, is also used to successive according to the sequence of the inquiry problem to be selected, chooses the institute of preset number
Inquiry problem to be selected is stated as the corresponding newly-increased inquiry problem of the triple.
The cleaning device for the knowledge mapping attribute introduced by the present embodiment is that can execute in the embodiment of the present invention
The device of the cleaning method of knowledge mapping attribute, so the cleaning based on knowledge mapping attribute described in the embodiment of the present invention
Method, those skilled in the art can understand the specific embodiment of the cleaning device of the knowledge mapping attribute of the present embodiment
And its various change form, so how the cleaning device at this for the knowledge mapping attribute is realized in the embodiment of the present invention
The cleaning method of knowledge mapping attribute be no longer discussed in detail.As long as those skilled in the art implement in the embodiment of the present invention
Device used by the cleaning method of knowledge mapping attribute belongs to the range to be protected of the application.
The embodiment of the invention also provides a kind of electronic equipment, as shown in Figure 5, comprising: at least one processor
(processor)41；And at least one processor (memory) 42, the bus 43 being connect with the processor 41；Wherein,
The processor 41, memory 42 complete mutual communication by the bus 43；
The processor 41 is used to call the program instruction in the memory 42, to execute in above method embodiment
Step.
The embodiment of the present invention mentions and has also supplied a kind of non-transient computer readable storage medium, and the non-transient computer is readable
Storage medium stores computer instruction, and the computer instruction executes the computer provided by above-mentioned each method embodiment
Method.
In conclusion the cleaning method and device of a kind of knowledge mapping attribute provided in an embodiment of the present invention.The present invention is real
Applying example is to construct inquiry problem using triple (that is: entity, attribute and attribute value) of the default template to knowledge mapping, is passed through
Search engine search query questions and obtains search result on the internet, and then by known searching comprising above-mentioned triple
The data information of hitch fruit can be completed automatically to come whether judge automatically the attribute value that triple includes be wrong
Apparent attribute value mistake washes.Compared to the prior art, it solves because manual inspection knowledge mapping attribute is with the presence or absence of mistake
Mislead the problem of causing labor intensive cost, low efficiency.The embodiment of the present invention optimizes knowledge using substitution manual operation is automaticly inspected
The checking process of map attribute improves and checks efficiency, can directly automatically wash attribute value mistake.In addition, checking
After finishing a knowledge mapping attribute, it can also check that triple is corresponding to be looked into corresponding search result to increase according to this
Inquiry topic, when checking so as to the subsequent identical triple to other knowledge mappings, can integrate newly-increased inquiry problem
Corresponding search is executed, to increase the diversity of search result, and then analyzes to obtain three using more various search result
The attribute value that tuple includes improves the accuracy rate checked with the presence or absence of the inspection result of mistake with final.
It should be understood by those skilled in the art that, embodiments herein can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the application
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the application, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The application is referring to method, the process of equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/
Or the forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable Jie
The example of matter.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates
Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including element
There is also other identical elements in process, method, commodity or equipment.
It will be understood by those skilled in the art that embodiments herein can provide as method, system or computer program product.
Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the application
Form.It is deposited moreover, the application can be used to can be used in the computer that one or more wherein includes computer usable program code
The shape for the computer program product implemented on storage media (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Formula.
The above is only embodiments herein, are not intended to limit this application.To those skilled in the art,
Various changes and changes are possible in this application.It is all within the spirit and principles of the present application made by any modification, equivalent replacement,
Improve etc., it should be included within the scope of the claims of this application.
Claims (17)
1. a kind of cleaning method of knowledge mapping attribute, which is characterized in that the described method includes:
When receiving a knowledge mapping, ternary is extracted from the knowledge mapping by the parsing to the knowledge mapping
Group, the triple are by entity elements, corresponding with the associated property element of the entity elements and the property element
The element combinations of attribute value element composition；
Utilize the corresponding inquiry problem of triple described in default structure of transvers plate；
It searches for the inquiry problem on the internet by search engine, obtains corresponding search result；
If according to described search result determine attribute value element that the triple includes be it is wrong, delete the attribute value
Element.
2. the method according to claim 1, wherein described determine the triple packet according to described search result
The attribute value element contained is wrong, comprising:
Count the total number of the corresponding search result of the inquiry problem；
Crawl the corresponding data information of described search result；
Judge the entity elements for including with the presence or absence of the triple in the corresponding data information of described search result, attribute member
Element and attribute value element；
If so, being effective search result by described search result queue；
Calculate the item number of effective search result and the percentage of the total number；
According to the percentage of the item number of effective search result and the total number, the attribute value that the triple includes is determined
Whether element is wrong.
3. according to the method described in claim 2, it is characterized in that, corresponding if searching for the triple by a search engine
Inquiry problem, then the percentage of the item number according to effective search result and the total number determines the triple
The attribute value element for including whether be mistake include:
Obtain the item number of the corresponding effective search result of described search engine and the percentage of total number；
Whether the percentage of the item number and the total number that judge effective search result is less than the first preset threshold；
If so, the attribute value element for determining that the triple includes is wrong.
4. according to the method described in claim 2, it is characterized in that, if searching for the triple respectively by multiple search engines
Corresponding inquiry problem, then the percentage of the item number according to effective search result and the total number determines described three
The attribute value element that tuple includes whether be mistake include:
Obtain the item number of the corresponding effective search result of the multiple search engine and the percentage of total number；
According to the weight distributed in advance the multiple search engine, the corresponding percentage of the multiple search engine is executed and is added
Power processing obtains weighting treated percentage；
Judge weighting treated the percentage whether less than the first preset threshold；
If so, the attribute value element for determining that the triple includes is wrong.
5. according to the method described in claim 2, it is characterized in that, calculating effective searching bar number and the total number
Before percentage, the method also includes:
If finding the property element and attribute value member that the triple includes in the corresponding data information of described search result
Element and the entity elements that the triple includes are not found, then extract in the corresponding data information of described search result and include
Noun；
Judge the noun whether be the entity elements that the triple includes alias；
If so, being effective search result by described search result queue.
6. according to the method described in claim 2, it is characterized in that, described is being that effectively search is tied by described search result queue
After fruit, the method also includes:
It obtains the answer information for including in the corresponding data information of described search result and the answer information is corresponding thumbs up
Information；
Information, the corresponding total point of statistics described search result are thumbed up according to the answer information and the answer information are corresponding
Praise number；
If the number that always thumbs up deletes described search result less than the second preset threshold in effective search result.
7. according to the method described in claim 2, it is characterized in that, the method also includes:
Judge the entity elements for including with the presence or absence of the triple in the corresponding heading message of effective search result, attribute
Element and attribute value element；
If so, being inquiry problem to be selected by the corresponding title mark of the effective search result；
According to the sequencing of web displaying arranged effective search result, the inquiry problem to be selected is determined
Sequence is successive；
Sequence according to the inquiry problem to be selected is successive, chooses the inquiry problem to be selected of preset number as the ternary
The corresponding newly-increased inquiry problem of group.
8. method according to any one of claim 1 to 7, which is characterized in that it is described by search engine in internet
The upper search inquiry problem further comprises:
The corresponding inquiry problem of the triple is searched on the Ask-Answer Community of internet by search engine.
9. a kind of cleaning device of knowledge mapping attribute, which is characterized in that described device includes:
Extraction unit, for when receiving a knowledge mapping, by the parsing to the knowledge mapping from the knowledge graph
Triple is extracted in spectrum, the triple is by entity elements and the associated property element of the entity elements and the category
Property element corresponding attribute value element composition element combinations；
Structural unit, the corresponding inquiry problem of triple for being extracted using extraction unit described in default structure of transvers plate；
Search unit obtains pair for searching for the inquiry problem of structural unit construction on the internet by search engine
The search result answered；
Determination unit, for according to described search unit searches to search result determine attribute value member that the triple includes
Element whether mistake；
Delete unit, for when according to the determination unit determine attribute value element that the triple includes be it is wrong, delete
Except the attribute value element.
10. device according to claim 9, which is characterized in that the determination unit includes:
Statistical module, for counting the total number of the corresponding search result of the inquiry problem；
Module is crawled, for crawling the corresponding data information of described search result；
Judgment module crawls in the corresponding data information of search result that module crawls described with the presence or absence of described for judging
Entity elements, property element and the attribute value element that triple includes；
Mark module, for judging to be that there are described three in the corresponding data information of described search result when the judgment module
Described search result queue is effective search result by entity elements, property element and the attribute value element that tuple includes；
Computing module, for calculate effective search result of mark module label item number and the statistical module counts
The percentage of total number；
Determining module, the item number of effective search result for being calculated according to the computing module and the percentage of the total number
Than determining whether the attribute value element that the triple includes is wrong.
11. device according to claim 10, which is characterized in that the determining module includes:
Acquisition submodule, for obtaining the item number of the corresponding effective search result of described search engine and the percentage of total number；
Judging submodule, the item number and the hundred of the total number of effective search result for judging the acquisition submodule acquisition
Divide ratio whether less than the first preset threshold；
Submodule is determined, for determining institute when the judging submodule judges that the percentage is less than the first preset threshold
It is wrong for stating the attribute value element that triple includes.
12. device according to claim 10, which is characterized in that the determining module includes:
The acquisition submodule is also used to obtain the item number of the corresponding effective search result of the multiple search engine and total
The percentage of item number；
Submodule is handled, for being obtained to the acquisition submodule according to the weight distributed in advance the multiple search engine
The corresponding percentage of multiple search engines execute weighting processing, obtain weighting treated percentage；
Whether the judging submodule is also used to judge the obtained weighting of processing submodule treated percentage less than the
One preset threshold；
The determining submodule is also used to judge described to weight that treated that percentage is less than first when the judging submodule
When preset threshold, the attribute value element for determining that the triple includes is wrong.
13. device according to claim 10, which is characterized in that the determination unit further include:
Extraction module, for before calculating the percentage of effective searching bar number and the total number, if in described search
As a result the property element and attribute value element that the triple includes are found in corresponding data information and are not found described
The entity elements that triple includes then extract the noun for including in the corresponding data information of described search result；
The judgment module is also used to judge whether the noun that the extraction module extracts is that the entity that the triple includes is first
The alias of element；
The mark module is also used to judge that the noun is the entity elements that the triple includes when the judgment module
It is effective search result by described search result queue when alias.
14. device according to claim 10, which is characterized in that the determination unit further include:
Module is obtained, for it is corresponding to obtain described search result after by described search result queue being effective search result
Data information in include answer information and the answer information is corresponding thumbs up information；
The statistical module is also used to thumb up information according to the answer information and the answer information are corresponding, counts institute
State that search result is corresponding always to thumb up number；
Removing module, if always thumbing up number less than the second preset threshold, described effective for the statistical module counts
Described search result is deleted in search result.
15. device according to claim 10, which is characterized in that described device further include:
Judging unit, for judging that whether there is the triple in the corresponding heading message of effective search result includes
Entity elements, property element and attribute value element；
Marking unit, for judging it is in the presence of described in the corresponding heading message of effective search result when the judging unit
Entity elements, property element and the attribute value element that triple includes, by the corresponding title mark of the effective search result
For inquiry problem to be selected；
The determination unit is also used to the sequencing arranged effective search result according to web displaying, really
The sequence of the fixed inquiry problem to be selected is successive；
The determination unit, is also used to successive according to the sequence of the inquiry problem to be selected, chooses the described to be selected of preset number
Inquiry problem is as the corresponding newly-increased inquiry problem of the triple.
16. a kind of electronic equipment characterized by comprising
At least one processor；
And at least one processor, the bus being connected to the processor；Wherein,
The processor, memory complete mutual communication by the bus；The processor is for calling the storage
Program instruction in device, with the cleaning side of knowledge mapping attribute described in any one of perform claim requirement 1 to claim 8
Method.
17. a kind of non-transient computer readable storage medium, which is characterized in that the non-transient computer readable storage medium is deposited
Store up computer instruction, the computer instruction requires the computer perform claim 1 to described in any one of claim 8
The cleaning method of knowledge mapping attribute.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811415629.7A CN109508420B (en) | 2018-11-26 | 2018-11-26 | Method and device for cleaning attributes of knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811415629.7A CN109508420B (en) | 2018-11-26 | 2018-11-26 | Method and device for cleaning attributes of knowledge graph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109508420A true CN109508420A (en) | 2019-03-22 |
| CN109508420B CN109508420B (en) | 2021-12-07 |
Family
ID=65750508
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811415629.7A Active CN109508420B (en) | 2018-11-26 | 2018-11-26 | Method and device for cleaning attributes of knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109508420B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110287705A (en) * | 2019-06-25 | 2019-09-27 | 北京中科微澜科技有限公司 | A kind of security breaches wrong data modification method based on loophole map |
| CN110909168A (en) * | 2019-09-23 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Knowledge graph updating method and device, storage medium and electronic device |
| CN111026856A (en) * | 2019-12-09 | 2020-04-17 | 出门问问信息科技有限公司 | Intelligent interaction method and device and computer readable storage medium |
| CN111274407A (en) * | 2020-01-15 | 2020-06-12 | 北京百度网讯科技有限公司 | Triple confidence degree calculation method and device in knowledge graph |
| CN111324743A (en) * | 2020-02-14 | 2020-06-23 | 平安科技（深圳）有限公司 | Text relation extraction method and device, computer equipment and storage medium |
| CN111368096A (en) * | 2020-03-09 | 2020-07-03 | 中国平安人寿保险股份有限公司 | Knowledge graph-based information analysis method, device, equipment and storage medium |
| CN111666393A (en) * | 2020-04-29 | 2020-09-15 | 平安科技（深圳）有限公司 | Verification method and device of intelligent question-answering system, computer equipment and storage medium |
| CN111984796A (en) * | 2020-07-31 | 2020-11-24 | 西安理工大学 | Automatic compliance checking method based on standard knowledge graph IFC model |
| CN112417162A (en) * | 2020-11-13 | 2021-02-26 | 中译语通科技股份有限公司 | Method and device for associating entity relationship clue fragments |
| CN117520568A (en) * | 2024-01-04 | 2024-02-06 | 北京奇虎科技有限公司 | Knowledge graph attribute completion method, device, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268216A (en) * | 2014-09-24 | 2015-01-07 | 江苏名通信息科技有限公司 | Data cleaning system based on internet information |
| CN105574098A (en) * | 2015-12-11 | 2016-05-11 | 百度在线网络技术（北京）有限公司 | Knowledge graph generation method and device and entity comparing method and device |
| US10078651B2 (en) * | 2015-04-27 | 2018-09-18 | Rovi Guides, Inc. | Systems and methods for updating a knowledge graph through user input |
| US10102291B1 (en) * | 2015-07-06 | 2018-10-16 | Google Llc | Computerized systems and methods for building knowledge bases using context clouds |
-
2018
- 2018-11-26 CN CN201811415629.7A patent/CN109508420B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268216A (en) * | 2014-09-24 | 2015-01-07 | 江苏名通信息科技有限公司 | Data cleaning system based on internet information |
| US10078651B2 (en) * | 2015-04-27 | 2018-09-18 | Rovi Guides, Inc. | Systems and methods for updating a knowledge graph through user input |
| US10102291B1 (en) * | 2015-07-06 | 2018-10-16 | Google Llc | Computerized systems and methods for building knowledge bases using context clouds |
| CN105574098A (en) * | 2015-12-11 | 2016-05-11 | 百度在线网络技术（北京）有限公司 | Knowledge graph generation method and device and entity comparing method and device |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110287705A (en) * | 2019-06-25 | 2019-09-27 | 北京中科微澜科技有限公司 | A kind of security breaches wrong data modification method based on loophole map |
| CN110287705B (en) * | 2019-06-25 | 2021-03-30 | 北京中科微澜科技有限公司 | Security vulnerability error data correction method based on vulnerability map |
| CN110909168A (en) * | 2019-09-23 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Knowledge graph updating method and device, storage medium and electronic device |
| CN111026856A (en) * | 2019-12-09 | 2020-04-17 | 出门问问信息科技有限公司 | Intelligent interaction method and device and computer readable storage medium |
| CN111274407A (en) * | 2020-01-15 | 2020-06-12 | 北京百度网讯科技有限公司 | Triple confidence degree calculation method and device in knowledge graph |
| CN111324743A (en) * | 2020-02-14 | 2020-06-23 | 平安科技（深圳）有限公司 | Text relation extraction method and device, computer equipment and storage medium |
| CN111368096A (en) * | 2020-03-09 | 2020-07-03 | 中国平安人寿保险股份有限公司 | Knowledge graph-based information analysis method, device, equipment and storage medium |
| CN111666393A (en) * | 2020-04-29 | 2020-09-15 | 平安科技（深圳）有限公司 | Verification method and device of intelligent question-answering system, computer equipment and storage medium |
| CN111984796A (en) * | 2020-07-31 | 2020-11-24 | 西安理工大学 | Automatic compliance checking method based on standard knowledge graph IFC model |
| CN111984796B (en) * | 2020-07-31 | 2022-11-04 | 西安理工大学 | Automatic compliance inspection method based on standard knowledge graph IFC model |
| CN112417162A (en) * | 2020-11-13 | 2021-02-26 | 中译语通科技股份有限公司 | Method and device for associating entity relationship clue fragments |
| CN117520568A (en) * | 2024-01-04 | 2024-02-06 | 北京奇虎科技有限公司 | Knowledge graph attribute completion method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109508420B (en) | 2021-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109508420A (en) | A kind of cleaning method and device of knowledge mapping attribute | |
| US11469963B2 (en) | Cybersecurity incident response and security operation system employing playbook generation through custom machine learning | |
| CN104077396B (en) | Method and device for detecting phishing website | |
| CN109284363A (en) | A kind of answering method, device, electronic equipment and storage medium | |
| CN107545000A (en) | The information-pushing method and device of knowledge based collection of illustrative plates | |
| Marx et al. | Deconstructing Darwin's Naturalization Conundrum in the San Juan Islands using community phylogenetics and functional traits | |
| US8412712B2 (en) | Grouping methods for best-value determination from values for an attribute type of specific entity | |
| CN107077474A (en) | Rapid color is searched for | |
| CN110415107B (en) | Data processing method, data processing device, storage medium and electronic equipment | |
| CN109299258A (en) | A kind of public sentiment event detecting method, device and equipment | |
| CN113098887A (en) | Phishing website detection method based on website joint characteristics | |
| CN109086356A (en) | The incorrect link relationship diagnosis of extensive knowledge mapping and modification method | |
| CN105912712B (en) | Robot dialog control method and system based on big data | |
| CN106934011A (en) | A kind of structuring analysis method and device of JSON data | |
| CN108241867A (en) | A kind of sorting technique and device | |
| CN109213773A (en) | A kind of diagnostic method, device and the electronic equipment of online failure | |
| CN109635089B (en) | Literature work novelty evaluation system and method based on semantic network | |
| WO2019200739A1 (en) | Data fraud identification method, apparatus, computer device, and storage medium | |
| CN111160783A (en) | Method and system for evaluating digital asset value and electronic equipment | |
| CN104767744B (en) | Protocol state machine active estimating method based on protocol knowledge | |
| CN107493275A (en) | The extracted in self-adaptive and analysis method and system of heterogeneous network security log information | |
| CN115203550A (en) | Social recommendation method and system for enhancing neighbor relation | |
| CN107291616A (en) | A kind of online generating platform of project report | |
| Brusco et al. | Deterministic blockmodelling of signed and two‐mode networks: A tutorial with software and psychological examples | |
| CN107566389A (en) | A kind of imitation URL link fishing domain name recognition methods based on C4.5 decision trees |
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 |