CN109902161A - Answer processing method, device, equipment and the storage medium of question answering system - Google Patents
Answer processing method, device, equipment and the storage medium of question answering system Download PDFInfo
- Publication number
- CN109902161A CN109902161A CN201910102807.9A CN201910102807A CN109902161A CN 109902161 A CN109902161 A CN 109902161A CN 201910102807 A CN201910102807 A CN 201910102807A CN 109902161 A CN109902161 A CN 109902161A
- Authority
- CN
- China
- Prior art keywords
- attribute
- answer
- subject entity
- same name
- 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.)
- Granted
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Present disclose provides a kind of answer processing methods of question answering system, comprising: the problem of identifying problem subject entity obtains the asked attribute of problem and/or the synonym of asked attribute；It is retrieved according to knowledge mapping, obtains m corresponding with problem subject entity answer subject entities of the same name, wherein m > 1；And selection answer subject entity is carried out from m answer subject entities of the same name according to the distribution of the temperature of m answer subject entities of the same name.The disclosure additionally provides answer processing unit, electronic equipment and the computer readable storage medium of a kind of question answering system.
Description
Technical field
This disclosure relates to a kind of answer processing method of question answering system, the answer processing unit of question answering system, electronic equipment
And computer readable storage medium.
Background technique
In existing question answering system, for there is ambiguity subject, question answering system can be from multiple possible entities
One entity (such as temperature is highest) of middle selection is answered as the subject of the problem.In this way, can not be according to problem
And other information judge user want inquiry the problem of subject entity when, be unable to satisfy the demand of user.
And in existing system, the case where answer of the same name should export simultaneously cannot be covered, for example, must be temperature
Non- famous entity less than 10,000 can be just removed, such as be not required to be that different types of entity can just trigger.
Summary of the invention
At least one of in order to solve the above-mentioned technical problem, present disclose provides a kind of answer processing sides of question answering system
Answer processing unit, electronic equipment and the readable storage medium storing program for executing of method, question answering system.
According to one aspect of the disclosure, a kind of answer processing method of question answering system, comprising: the problem of identifying problem master
Language entity obtains the asked attribute of problem and/or the synonym of asked attribute；It is retrieved according to knowledge mapping, obtains and ask
The corresponding m answer subject entity of the same name of subject entity is inscribed, wherein m > 1；And the heat according to m answer subject entities of the same name
Degree is distributed to carry out selection answer subject entity from m answer subject entities of the same name.
According at least one embodiment of the disclosure, selected from m answer subject entities of the same name when being distributed according to temperature
When selecting answer subject entity, when the temperature of the highest answer subject entity of middle temperature of m answer subject entities of the same name is low, heat
Spend the appearance of the temperature gap in highest answer subject entity and m answer subject entities of the same name between remaining answer subject entity
It is big to bear value；And when the temperature of the highest answer subject entity of middle temperature of m answer subject entities of the same name is high, temperature highest
Answer subject entity and m answer subject entities of the same name in temperature gap between remaining answer subject entity tolerance value
It is small.
According at least one embodiment of the disclosure, carry out selecting answer master when from a answer subject entities of the same name of m
After language entity, further includes: distinguish selected answer subject entity.
According at least one embodiment of the disclosure, distinguish selected answer subject entity include: seek it is selected
Answer subject entity attributes intersection；The identical attribute of attribute value is deleted in dependence intersection；And from delete attribute value phase
With selective discrimination attribute in the attribute intersection of attribute, to use differentiation attribute to distinguish answer subject of the same name.
According at least one embodiment of the disclosure, when seeking selected answer subject entity attributes intersection,
When the different same attribute of attribute value is not present between answer subject entity, then delete in selected answer subject entity
The minimum answer subject entity of temperature, and selected answering after the minimum answer subject entity of temperature is deleted is sought again
Case subject entity attributes intersection.
According at least one embodiment of the disclosure, in selective discrimination attribute, the attribute for surely belonging to sexual intercourse collection is endowed
When priority, according to the priority of attribute come selective discrimination attribute, when surely belonging to the attribute of sexual intercourse collection and being not endowed with priority, according to
Text similarity between the attribute value of attribute intersection carrys out selective discrimination attribute.
According at least one embodiment of the disclosure, by the editing distance between the attribute value of computation attribute intersection come
It determines text similarity, selects the attribute of the smallest attribute value of text similarity as attribute is distinguished.
According to another aspect of the present disclosure, the answer processing unit of a kind of question answering system, comprising: obtain module, identification is asked
The problem of topic subject entity, obtain the asked attribute of problem and/or the synonym of asked attribute；Retrieval module, according to knowledge graph
Spectrum is retrieved, and obtains m corresponding with problem subject entity answer subject entities of the same name, wherein m > 1；And selecting module,
It is real that selection answer subject is carried out from m answer subject entities of the same name according to the distribution of the temperature of m answer subject entities of the same name
Body.
According to the another further aspect of the disclosure, a kind of electronic equipment, comprising: memory, memory store executable instruction；With
And processor, processor executes the executable instruction of memory storage, so that processor executes above-mentioned method.
According to the another aspect of the disclosure, a kind of readable storage medium storing program for executing is stored with executable instruction in readable storage medium storing program for executing,
For realizing above-mentioned method when executable instruction is executed by processor.
Detailed description of the invention
Attached drawing shows the illustrative embodiments of the disclosure, and it is bright together for explaining the principles of this disclosure,
Which includes these attached drawings to provide further understanding of the disclosure, and attached drawing is included in the description and constitutes this
Part of specification.
Fig. 1 is the schematic flow chart according to the answer processing method of the question answering system of one embodiment of the disclosure.
Fig. 2 is the schematic flow chart according to the answer processing method of the question answering system of one embodiment of the disclosure.
Fig. 3 is according to the schematic of the differentiation step of the answer processing method of the question answering system of one embodiment of the disclosure
Flow chart.
Fig. 4 is the schematic flow chart according to the answer processing unit of the question answering system of one embodiment of the disclosure.
Fig. 5 is the explanatory view according to the electronic equipment of one embodiment of the disclosure.
Specific embodiment
The disclosure is described in further detail with embodiment with reference to the accompanying drawing.It is understood that this place
The specific embodiment of description is only used for explaining related content, rather than the restriction to the disclosure.It also should be noted that being
Convenient for description, part relevant to the disclosure is illustrated only in attached drawing.
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the disclosure can
To be combined with each other.The disclosure is described in detail below with reference to the accompanying drawings and in conjunction with embodiment.
Question answering system (Question Answering System, abbreviation QA system) is for answering user's input or mentioning
Various problems out.Such as user puts question to " whom the wife of Zhou Jielun is ", QA system then answers " elder brother's icepro "；User's enquirement " in
How many population of state ", QA system answer " 1,300,000,000 ".
Knowledge mapping (Knowledge Graph, abbreviation KG), target will cover the letter of all entities and entity in the world
Breath.The generic representation method of the information of entity is triple form, such as: Zhou Jielun, wife, elder brother insult.
Problem subject entity: QA system receive the problem of in, problem subject entity refer to it is being asked in this problem,
As that entity of subject in question sentence.Such as: " Zhou Jielun " is problem subject entity in " whom the wife of Zhou Jielun is ", and
" China " in " how many Chinese population " is problem subject entity.
Answer subject entity of the same name: in the problem of QA system receives, problem subject entity may correspond to multiple same
Name answer subject entity, and which answer subject entity can not judge that the problem inquired by problem is.Such as
When user inquires " apple ", " apple " may refer to the answer subject entity " apple " as fruit, it is also possible to refer to conduct
The answer subject entity " apple " of Apple Inc..In this way will likely be there are two answer subject entity of the same name, one is as fruit
Apple, the other is the apple as company.For another example, when user inquires " whom the author of The Orchid Pavilion sequence is ", in the problem
The problem of subject entity be " The Orchid Pavilion sequence ", and the answer subject entity of The Orchid Pavilion sequence have it is multiple, for example, The Orchid Pavilion sequence may both refer to the king legendary ruler of great antiquity, the first of the Three August Ones
The calligraphy work " The Orchid Pavilion sequence " write；It may also refer to the song " The Orchid Pavilion sequence " that Zhou Jielun is sung, it in this way will be there are two of the same name
Answer subject entity " the The Orchid Pavilion sequence of article " and " the The Orchid Pavilion sequence of song ".
In accordance with one embodiment of the present disclosure, a kind of answer processing method of question answering system is provided.
As shown in Figure 1, the answer processing method 10 includes: step S11, the problem of identifying problem subject entity, asked
The asked attribute of topic and/or the synonym of asked attribute；Step S12, is retrieved according to knowledge mapping, is obtained and problem subject
Corresponding m answer subject entities of the same name of entity, wherein m > 1；And step S13, according to a answer subject entity of the same name of m
Temperature is distributed to carry out selection answer subject entity from m answer subject entities of the same name.
In step s 11, according to the problem of user come subject entity the problem of identifying problem, the asked attribute of parsing problem,
And extend the synonym of asked attribute.
As an example, the problem of user puts question to " The Orchid Pavilion sequence is who writes? ", the subject entity that goes wrong can be identified at this time
For " The Orchid Pavilion sequence ", and attribute is " author " the problem of can be parsed out the problem, then extends the synonymous of the attribute " author "
Word, such as " writing words " (merely illustrative herein, can also according to circumstances to expand more synonyms) can be expanded, finally will
The problem of obtaining problem subject entity is " The Orchid Pavilion sequence ", and the attribute which is asked is " author " and " writing words " etc..
In step s 12, it is retrieved according to knowledge mapping, obtains m corresponding with problem subject entity answers of the same name
Subject entity, wherein m > 1.
As an example, when retrieving in knowledge mapping, the entity of those hit alias can be used.Such as work as user
Inquire " sun is much? ", because the alias of South Korea artist Dong Yongpei is also the sun, although his physical name in knowledge mapping
It is not the sun, but according to alias, which is also hit.At this moment can be used as can choose as multiple answer subjects realities
One in body.As another example, for example, when user put question to " when rewinding this song issues? ", it is assumed that knowing
Know in map and exists: the rewinding of rewinding, the rewinding that Cai Yilin is sung and Xiao Jingteng performance that Zhou Jielun is sung, at this moment by knowing
Know to be retrieved according to problem subject entity " rewinding " in map and obtains three answer subject entities of the same name " Zhou Jielun is sung
Rewinding ", " Cai Yilin sing rewinding " and " rewinding that Xiao Jingteng is sung ", at this moment by can be according to problem in knowledge mapping
Subject entity " rewinding " and retrieve obtain.Here, quantity " three " is merely to example acts on, there may be more in practice
It is a.Be re-used as an example, if user inquire " when rewinding this song issues? ", at this moment pass through retrieval knowledge
Map obtains " rewinding that Zhou Jielun is sung ", " rewinding that Cai Yilin is sung " and " rewinding that Xiao Jingteng is sung ".But for " week
The rewinding that Jie Lun is sung " not " issuing date " this attribute, because can be rejected, not as by option.And " Cai Yi
The rewinding that woods is sung " and " rewinding that Xiao Jingteng is sung " have the attribute of " issuing date ", by this two as by option.Again
One example is provided, with the problem of user " The Orchid Pavilion sequence is who writes? " for, it is false by being retrieved in knowledge mapping
Be located at retrieved in knowledge mapping The Orchid Pavilion preface chapter author be the king legendary ruler of great antiquity, the first of the Three August Ones's and The Orchid Pavilion sequence song write words the side's of being mountain of papers the case where
Under, at this moment can be retrieved in knowledge mapping obtain two answer subject entities " the The Orchid Pavilion sequence as article " of the same name and " as
The The Orchid Pavilion sequence of song ".
In step s 13, in the function that entity of the same name exports simultaneously, according to the temperature of multiple answer subject entities of the same name
Distribution selects the answer subject entity of the same name it is more likely that user's inquiry, rather than just selection that highest entity of temperature.
In an embodiment of the disclosure, select to answer from m answer subject entities of the same name when according to temperature distribution
When case subject entity, when the temperature of the highest answer subject entity of middle temperature of m answer subject entities of the same name is low, temperature is most
The tolerance value of temperature gap in high answer subject entity and m answer subject entities of the same name between remaining answer subject entity
Greatly；And when the temperature of the highest answer subject entity of middle temperature of m answer subject entities of the same name is high, temperature is highest to be answered
The tolerance value of temperature gap in case subject entity and m answer subject entities of the same name between remaining answer subject entity is small.
It is illustrated combined with specific embodiments below to according to temperature distribution selection answer subject entity.
In question answering system, when retrieving the highest answer of temperature according to knowledge mapping, it at this moment can't stop retrieving,
And it is to continue with whether the remaining answer subject entity of the same name of retrieval includes the attribute that can answer customer problem.
When temperature gap " excessive " of remaining answer subject entity of the same name and the highest answer subject entity of temperature, then stop
Only retrieve.For example, " China " this subject, in answer subject entity of the same name existing " People's Republic of China (PRC) ", referred to as China,
There is the China in " reed former China ", refer to the human world world in Japanese mythology, it is also possible to refer to " Luoyang, henan ", the Lip river when Zhou Dynasty
Sun is once referred to as China.And in these entities, the temperature of the People's Republic of China (PRC) is much higher than both other, and quizmaster is true
Be only concerned association attributes under the People's Republic of China's entity in fact, therefore, question answering system do not need to subsequent two entities into
Row retrieval, the needs of user can be met by directly returning to the answer under the People's Republic of China's entity.For another example asking as user
Entitled " Zhao Benshan is much ", it is exactly that Zhao Benshan as performer that temperature is highest, at this time if in knowledge mapping
There are multiple Zhao Benshans (such as multiple people of duplication of name), and other than performer Zhao Benshan, remaining people to bear the same name is obscure same
Name other people (temperature is very low), at this moment other people can be rejected.
Above-mentioned temperature gap " excessive " can be fitted actual conditions by piecewise function.Such as when answer subject entity
Temperature it is lower when, such as in the case where temperature only position, ten or hundreds, although the temperature of these entities is in numerical value
The multiple of upper gap is very big, and for the people of real life, although the poor multiple of institute is very big, but temperature is very low, therefore this
There is no too many differences for entities are from temperature a bit, therefore larger for the tolerance of the temperature gap of these entities value, are arranged
Such as it when indicating by piecewise function, can treat them as being not different.When being gradually increased with the temperature of entity, for
The tolerance value of gap between entity temperature gradually becomes smaller.For example, by there are for two entities, when the temperature of two entities is equal
When very low, one is units and another is hundreds, although the poor multiple of the two institute is hundred times or so, very due to temperature
Low, because tolerance value is larger at this time (such as hundred times or so of gap is not considered gap " excessive ") therefore two entities are considered as heat
Indistinction on degree may export simultaneously two entities；But when an entity temperature in two entities is higher, for example, it is hot
Degree is myriabit number or more, and the temperature of another entity is lower, such as hundreds, and the poor multiple of institute is also hundred times of left sides therebetween
The right side, because tolerance value is smaller at this time, hundred times of left and right differences may only input high that of temperature away from being considered as gap " excessive "
Entity.
According to the disclosure another, embodiment there is provided a kind of answer processing methods of question answering system.As shown in Fig. 2,
The answer processing method 20 further includes other than including the steps that step S21-S23 identical as 10 S11-S13 of method respectively
Step S24 distinguishes selected answer subject entity.
In step s 24, according to the differentiation attribute of selected multiple answer subject entities come to multiple answer subject entities
It distinguishes.In this step, because multiple possible answer subject entities have been found, it is therefore desirable to these entities into
Row is distinguished, otherwise can not judge actually which/which answer is only for desired answer.In the disclosure, using multiple same
The differentiation attribute of name answer subject entity distinguishes, such as finds differentiation attribute, thus will multiple answer subjects realities of the same name
Body distinguishes.For example, the sun as celestial body is owned by age attribute with the sun as people's nickname, when user is apprised of certain
When the age of a sun is tens years old, then user is obviously known that this sun is the people that nickname is the sun, not as
The sun of celestial body.
As shown in figure 3, step S24 further comprises: S241: seeking selected answer subject entity attributes intersection；
S242: the identical attribute of attribute value is deleted in dependence intersection；And S243: it is handed over from the attribute for deleting the identical attribute of attribute value
Selective discrimination attribute is concentrated, to use the differentiation attribute to distinguish answer subject of the same name.
Specifically, in step S241, multiple answer subject entities is compared, the intersection of their attribute is found.
When seeking attribute intersection, if all gathered around between these answer subject entities of the same name there is no one other than answer attributes
The minimum entity of temperature is then rejected, seeks handing over again to remaining answer subject entity by attribute have but different attribute value
Collection has been look for one or more attributes current that all entities are owned by and that attribute value is different.
In step S242, from this intersection set, the identical attribute of those attribute values can be deleted, to find out that
A little attributes with discrimination, that is to say, that the attribute that the respective attribute value of multiple answer subject entities of the same name is different from.This
When this intersection set in attribute, these answer subject entities can be distinguished, this is because each answer subject is real
Body is owned by these attributes and respective attribute value is all different.
In step S243, when surely belonging to the attribute of sexual intercourse collection and being endowed priority, area is selected according to the priority of attribute
Adhering to separately property, when surely belonging to the attribute of sexual intercourse collection and being not endowed with priority, according to the text similarity between the attribute value of attribute intersection
Carry out selective discrimination attribute.It specifically, then can be directly preferential according to what is assigned in advance when attribute is assigned priority in advance
Grade carrys out selective discrimination attribute.In the disclosure, higher priority can be assigned for/mono- attribute, and other attributes then have
There is lower priority, in this way, the attribute of highest priority can be used as attribute is distinguished.Preferably, it also can be set
Attribute blacklist is distinguished, such as some attributes with discrimination are not used in as the differentiation attribute for distinguishing entity.For example, such as
When fruit user inquires the case where host " Xie Na ", " Na Na " is referred to, but there are many people for being named as " Na Na ", it is assumed that can distinguish
The attribute of these " Na Na " includes birthplace, height, age, full name.But the discrimination of full name may be apparently higher than birthplace,
Height, age.It therefore can be by the differentiation attribute of full name alternatively.If select other attributes, such as tell user one
A Na Na is the Na Na being born in Sichuan, as soon as a Na Na is the Na Na being born in Xi'an, at this moment needs user to have and knows and thank
The knowledge quantity in Na birthplace, if it is not, can not judge to locate that the Na Na which Na Na is only its care.It at this moment if will
Full name then can directly be told, Na Na is Xie Na, and user can clearly judge that oneself is of concern as attribute is distinguished
That thanks to Na.
In another case, for example, it is many kinds of due to attribute, it is corresponding preferential therefore, it is difficult to assign one by one in advance
Grade, at this moment in the attribute intersection between entity, when not comprising the attribute for being endowed priority, can be handed over by computation attribute
The text similarity between the attribute value of attribute is concentrated to select from selective discrimination attribute in these attributes, for example, when selection text
That the smallest attribute of this similarity is as differentiation attribute.Such as can to calculate text by cosine similarity algorithm similar
Degree.In the disclosure, can optionally calculate between the attribute value of these attributes editing distance (such as Levenshtein away from
From) carry out selective discrimination attribute, when editing distance is bigger, text similarity is smaller, and it is maximum to can choose editing distance at this time
That attribute is as differentiation attribute.
As an example, the software of " credulity " is named as there are three for example, in their attribute intersection, attribute " software category "
Attribute value is respectively " Chat communication software ", " communication mobile office software " and " communication/Android "；Attribute " developer "
Attribute value is respectively " Netease and telecommunications ", " sitsch " and " Jin Cheng Homeway.com ".It is compiled by the calculating between the attribute value of each attribute
Distance is collected, the editing distance of the attribute value of attribute " developer " is greater than the editing distance of the attribute value of attribute " software category ", because
This can choose attribute " developer " as differentiation attribute.
According to the disclosure optional embodiment, the answer processing method of the disclosure further includes that the progress of answer entity is whole
The step of conjunction.At this point, integrate using natural language to multiple answers, integrate out that one complete, user is understandable
Language is returned to user.Pre-selection was arranged the attribute of priority, the sentence after integration can be made using specific template is more
Add smoothness, for not presetting the attribute of priority, can be integrated etc. according to the title of attribute and attribute value.
As an example, for example, when selected answer subject entity be it is multiple when, by multiple answer subject entities and phase
It closes information to be integrated, by the content after integration as answer processing.It include: that selected answer subject is real in the answer of output
Differentiation attribute, the correspondence attribute of selected answer subject entity and the attribute value of the corresponding attribute of body, also may include selected
The quantity of answer subject entity.
For example, for problem " The Orchid Pavilion sequence is who writes? ", retrieval has obtained two answer subjects of the same name from knowledge mapping
Entity " the The Orchid Pavilion sequence of article " and " the The Orchid Pavilion sequence of song "." song " and " article " is that can distinguish two answer subjects realities of the same name
The differentiation attribute of body, " writing words " and " author " is the correspondence attribute of two answer subject entities of the same name, and " Fang Wenshan " and " the king legendary ruler of great antiquity, the first of the Three August Ones
It " then be respectively two corresponding attributes attribute value.Obtained information can be integrated to provide answer at this time, such as answer is
" the writing words for The Orchid Pavilion sequence as song is Fang Wenshan, the author of the The Orchid Pavilion sequence as article be the king legendary ruler of great antiquity, the first of the Three August Ones it ".
For example, for problem " when rewinding this song issues? ", retrieval has obtained two from knowledge mapping
Answer subject entity " rewinding that Zhou Jielun is sung ", " rewinding that Cai Yilin is sung " and " rewinding that Xiao Jingteng is sung " of the same name, because
There is no this attribute of issuing date for " rewinding that Zhou Jielun is sung ", therefore only selects " rewinding that Cai Yilin is sung " and " desolate
Respect the rewinding for rising performance ".The differentiation attribute " singer " of two answer subject entities of the same name is as the differentiation attribute used at this time
To distinguish.At this moment it can be generated that " issuing date of rewinding that Cai Yilin is sung is 2 months 2004；What Xiao Jingteng was sung falls
The issuing date of band is in November, 2009 ".
In the disclosure, it is preferable that answer can be provided a user according to lower frame.It is to be appreciated that the frame
Frame is only example effect, and those skilled in the art can come according to the actual situation using suitable frame.
With the quantity of selected answer subject entity be two as an example, continue with problem " The Orchid Pavilion sequence is who writes? "
For, frame can for " AAA that I knows has BBB, (as/be named as) DDD of the AAA of CCC is EEE, (as/name
Cry) GGG of the AAA of FFF is HHH ... ... ".According to the answer of frame processing for " the The Orchid Pavilion sequence that I knows has 2, as song
Bent writing words for The Orchid Pavilion sequence is Fang Wenshan, the author of the The Orchid Pavilion sequence as article be the king legendary ruler of great antiquity, the first of the Three August Ones it ".
In the frame, " AAA " indicate such as " The Orchid Pavilion sequence " the problem of subject entity." BBB " indicates obtaining for such as " 2 "
Multiple answer subject entities of the same name quantity." AAA of CCC " and " AAA of FFF " expression obtains such as " the The Orchid Pavilion sequence of song "
And multiple answer subject entities of the same name of " the The Orchid Pavilion sequence of article ", wherein " CCC " and " FFF " indicates that multiple answer subjects of the same name are real
The differentiation attribute of such as " song " and " article " of body, the differentiation attribute can carry out area to multiple answer subject entities of the same name
Point, for another example may have the difference (distinguishing attribute) of singer by taking song " rewinding " as an example, at this moment be exactly " Cai Yilin performance "
And " Xiao Jingteng performance "." DDD " and " GGG " indicates the correspondence attribute of multiple answer subject entities of the same name, in the example of The Orchid Pavilion sequence
In, " DDD " is corresponding attribute " writing words " and " GGG " is corresponding attribute " author ", and two corresponding attributes are different here.But
Two corresponding attributes are also possible to identical, such as in the example of song rewinding, " DDD " and " GGG " is corresponding attribute " hair
The row time "." EEE " and " HHH " indicates the attribute value of the correspondence attribute of each answer subject entity of the same name, such as " as song
The Orchid Pavilion sequence is write words " attribute value be " Fang Wenshan ", the attribute value of " author of the The Orchid Pavilion sequence as article " is " the king legendary ruler of great antiquity, the first of the Three August Ones it ",
The attribute value of " issuing date for the rewinding that Cai Yilin is sung " is " 2 months 2004 ", " when the distribution for the rewinding that Xiao Jingteng is sung
Between " attribute value be " in November, 2009 ".
In said frame bracket " as/be named as " etc. words depend on " CCC " and " FFF " be " AAA " what attribute
Attribute value, for showing that " CCC " and " FFF " is the attribute of " AAA ".
Finally, obtained text is exported as answer, later can according to technology commonly used in the art by its
Voice is converted into be exported.
The accuracy rate of question answering system can be improved according to the above method of the disclosure, thus by providing more information
The experience of user is improved to user.
According to a further embodiment of the disclosure, a kind of answer processing unit of question answering system is provided.As shown in figure 4,
The answer processing unit 40 of the question answering system includes obtaining module 41, retrieval module 42 and selecting module 43.
Obtain module 41, identify problem the problem of subject entity, obtain problem asked attribute and/or asked attribute it is same
Adopted word；Retrieval module 42, is retrieved according to knowledge mapping, obtains m corresponding with problem subject entity answer subjects of the same name
Entity, wherein m > 1；And selecting module 43, according to the distribution of the temperature of m answer subject entity of the same name come from m answers of the same name
Selection answer subject entity is carried out in subject entity.
And the treatment process executed in above-mentioned each module is opposite with the respective process specifically described in the above method respectively
It answers.
As shown in figure 4, the answer processing unit 40 of the question answering system can also include discriminating module 44 and integrate module 45.
Discriminating module 44 integrates module 45 for integrating to answer entity for distinguishing selected answer subject entity.It should
Treatment process performed by two modules is corresponding with integration step with differentiation step in the above method respectively.
The disclosure also provides a kind of electronic equipment, as shown in figure 5, the equipment includes: communication interface 1000, memory 2000
With processor 3000.Communication interface 1000 carries out data interaction for being communicated with external device.In memory 2000
It is stored with the computer program that can be run on processor 3000.Processor 3000 is realized above-mentioned when executing the computer program
Method in embodiment.The quantity of the memory 2000 and processor 3000 can be one or more.
Memory 2000 may include high speed RAM memory, can also further include nonvolatile memory (non-
Volatile memory), a for example, at least magnetic disk storage.
If communication interface 1000, memory 2000 and the independent realization of processor 3000, communication interface 1000, memory
2000 and processor 3000 can be connected with each other by bus and complete mutual communication.The bus can be industrial standard
Architecture (ISA, Industry Standard Architecture) bus, external equipment interconnection (PCI, Peripheral
Component) bus or extended industry-standard architecture (EISA, Extended Industry Standard
Component) bus etc..The bus can be divided into address bus, data/address bus, control bus etc..For convenient for expression, the figure
In only indicated with a thick line, it is not intended that an only bus or a type of bus.
Optionally, in specific implementation, if communication interface 1000, memory 2000 and processor 3000 are integrated in one
On block chip, then communication interface 1000, memory 2000 and processor 3000 can complete mutual lead to by internal interface
Letter.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing specific logical function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the disclosure includes other realization, wherein can not press shown or discussed suitable
Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, Lai Zhihang function, this should be by the disclosure
Embodiment person of ordinary skill in the field understood.Processor executes each method as described above and processing.
For example, the method implementation in the disclosure may be implemented as computer software programs, being tangibly embodied in machine can
Read medium, such as memory.In some embodiments, some or all of of computer software programs can be via memory
And/or communication interface and be loaded into and/or install.When computer software programs are loaded into memory and are executed by processor,
One or more steps in method as described above can be executed.Alternatively, in other embodiments, processor can lead to
It crosses other any modes (for example, by means of firmware) appropriate and is configured as executing one of above method.
Expression or logic and/or step described otherwise above herein in flow charts, may be embodied in any
In computer-readable medium, for instruction execution system, device or equipment (such as computer based system, including processor
System or other can be from instruction execution system, device or equipment instruction fetch and the system executed instruction) use, or combine these
Instruction execution system, device or equipment and use.
For the purpose of this specification, " computer readable storage medium ", which can be, any may include, store, communicating, propagating
Or transfer program uses for instruction execution system, device or equipment or in conjunction with these instruction execution systems, device or equipment
Device.The more specific example (non-exhaustive list) of computer-readable medium include the following: there are one or more wirings
Electrical connection section (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable read-only memory
(CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other suitable Jie
Matter, because can then be edited, be interpreted or when necessary with other for example by carrying out optical scanner to paper or other media
Suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of the disclosure can be realized with hardware, software or their combination.In above-mentioned embodiment party
In formula, multiple steps or method can carry out reality in memory and by the software that suitable instruction execution system executes with storage
It is existing.It, and in another embodiment, can be in following technology well known in the art for example, if realized with hardware
Any one or their combination are realized: having a discrete logic for realizing the logic gates of logic function to data-signal
Circuit, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), field-programmable gate array
Arrange (FPGA) etc..
Those skilled in the art are understood that realize all or part of the steps of above embodiment method
It is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer readable storage medium
In, which when being executed, includes the steps that one or a combination set of method implementation.
In addition, can integrate in a processing module in each functional unit in each embodiment of the disclosure, it can also
To be that each unit physically exists alone, can also be integrated in two or more units in a module.It is above-mentioned integrated
Module both can take the form of hardware realization, can also be realized in the form of software function module.The integrated module
If in the form of software function module realize and when sold or used as an independent product, also can store one calculating
In machine readable storage medium storing program for executing.The storage medium can be read-only memory, disk or CD etc..
In the description of this specification, reference term " an embodiment/mode ", " some embodiment/modes ",
The description of " example ", " specific example " or " some examples " etc. means the embodiment/mode or example is combined to describe specific
Feature, structure, material or feature are contained at least one embodiment/mode or example of the application.In this specification
In, schematic expression of the above terms are necessarily directed to identical embodiment/mode or example.Moreover, description
Particular features, structures, materials, or characteristics can be in any one or more embodiment/modes or example in an appropriate manner
In conjunction with.In addition, without conflicting with each other, those skilled in the art can be by different implementations described in this specification
Mode/mode or example and different embodiments/mode or exemplary feature are combined.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present application, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
It will be understood by those of skill in the art that above embodiment is used for the purpose of clearly demonstrating the disclosure, and simultaneously
Non- be defined to the scope of the present disclosure.For those skilled in the art, may be used also on the basis of disclosed above
To make other variations or modification, and these variations or modification are still in the scope of the present disclosure.
Claims (10)
1. a kind of answer processing method of question answering system characterized by comprising
The problem of identification problem subject entity, obtain the asked attribute of described problem and/or the synonym of asked attribute；
It is retrieved according to knowledge mapping, obtains m corresponding with described problem subject entity answer subject entities of the same name, wherein
M > 1；And
It is answered according to the distribution of the temperature of the m answer subject entities of the same name to carry out selection from m answer subject entities of the same name
Case subject entity.
2. the method as described in claim 1, which is characterized in that when real from m answer subjects of the same name according to temperature distribution
When selecting answer subject entity in body,
When the temperature of the highest answer subject entity of middle temperature of m answer subject entities of the same name is low, the temperature is highest
The tolerance value of temperature gap in answer subject entity and m answer subject entities of the same name between remaining answer subject entity is big；
And
When the temperature of the highest answer subject entity of middle temperature of m answer subject entities of the same name is high, the temperature is highest
The tolerance value of temperature gap in answer subject entity and m answer subject entities of the same name between remaining answer subject entity is small.
3. method according to claim 1 or 2, which is characterized in that answered when carrying out selection from m answer subject entities of the same name
After case subject entity, further includes: distinguish selected answer subject entity.
4. method as claimed in claim 3, which is characterized in that distinguishing selected answer subject entity includes:
Seek selected answer subject entity attributes intersection；
The identical attribute of attribute value is deleted from the attribute intersection；And
The selective discrimination attribute from the attribute intersection for deleting attribute value same alike result, to use the differentiation attribute to answer of the same name
Subject distinguishes.
5. method as claimed in claim 4, which is characterized in that seeking selected answer subject entity attributes intersection
When, when the different same attribute of attribute value is not present between answer subject entity, then delete selected answer subject entity
In the minimum answer subject entity of temperature, and seek again selected after the minimum answer subject entity of temperature is deleted
Answer subject entity attributes intersection.
6. method as described in claim 4 or 5, which is characterized in that in selective discrimination attribute,
When the attribute of the attribute intersection is endowed priority, according to the priority of attribute come selective discrimination attribute,
When the attribute of the attribute intersection is not endowed with priority, according to the text phase between the attribute value of the attribute intersection
Carry out selective discrimination attribute like degree.
7. method as claimed in claim 6, which is characterized in that the editor between attribute value by calculating the attribute intersection
Distance determines the text similarity, selects the attribute of the smallest attribute value of text similarity as the differentiation attribute.
8. a kind of answer processing unit of question answering system characterized by comprising
Obtain module, identify problem the problem of subject entity, obtain described problem asked attribute and/or asked attribute it is synonymous
Word；
Retrieval module is retrieved according to knowledge mapping, obtains m corresponding with described problem subject entity answer subjects of the same name
Entity, wherein m > 1；And
Selecting module, according to the distribution of the temperature of the m answer subject entity of the same name come from m answer subject entities of the same name into
Row selection answer subject entity.
9. a kind of electronic equipment characterized by comprising
Memory, the memory store executable instruction；And
Processor, the processor executes the executable instruction of the memory storage, so that the processor executes such as right
It is required that method described in any one of 1 to 7.
10. a kind of readable storage medium storing program for executing, which is characterized in that be stored with executable instruction in the readable storage medium storing program for executing, it is described can
It executes instruction when being executed by processor for realizing the method as described in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910102807.9A CN109902161B (en) | 2019-02-01 | 2019-02-01 | Answer processing method, device, equipment and storage medium of question-answering system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910102807.9A CN109902161B (en) | 2019-02-01 | 2019-02-01 | Answer processing method, device, equipment and storage medium of question-answering system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109902161A true CN109902161A (en) | 2019-06-18 |
| CN109902161B CN109902161B (en) | 2023-10-20 |
Family
ID=66944621
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910102807.9A Active CN109902161B (en) | 2019-02-01 | 2019-02-01 | Answer processing method, device, equipment and storage medium of question-answering system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109902161B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111309863A (en) * | 2020-02-10 | 2020-06-19 | 北京声智科技有限公司 | Natural language question-answering method and device based on knowledge graph |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106844603A (en) * | 2017-01-16 | 2017-06-13 | 竹间智能科技（上海）有限公司 | The computational methods and device, application process and device of entity hot topic degree |
| CN106919655A (en) * | 2017-01-24 | 2017-07-04 | 网易（杭州）网络有限公司 | A kind of answer provides method and apparatus |
| CN107247736A (en) * | 2017-05-08 | 2017-10-13 | 广州索答信息科技有限公司 | The kitchen field answering method and system of a kind of knowledge based collection of illustrative plates |
| US20180268300A1 (en) * | 2017-03-15 | 2018-09-20 | International Business Machines Corporation | Generating natural language answers automatically |
| CN109284363A (en) * | 2018-12-03 | 2019-01-29 | 北京羽扇智信息科技有限公司 | A kind of answering method, device, electronic equipment and storage medium |
-
2019
- 2019-02-01 CN CN201910102807.9A patent/CN109902161B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106844603A (en) * | 2017-01-16 | 2017-06-13 | 竹间智能科技（上海）有限公司 | The computational methods and device, application process and device of entity hot topic degree |
| CN106919655A (en) * | 2017-01-24 | 2017-07-04 | 网易（杭州）网络有限公司 | A kind of answer provides method and apparatus |
| US20180268300A1 (en) * | 2017-03-15 | 2018-09-20 | International Business Machines Corporation | Generating natural language answers automatically |
| CN107247736A (en) * | 2017-05-08 | 2017-10-13 | 广州索答信息科技有限公司 | The kitchen field answering method and system of a kind of knowledge based collection of illustrative plates |
| CN109284363A (en) * | 2018-12-03 | 2019-01-29 | 北京羽扇智信息科技有限公司 | A kind of answering method, device, electronic equipment and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111309863A (en) * | 2020-02-10 | 2020-06-19 | 北京声智科技有限公司 | Natural language question-answering method and device based on knowledge graph |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109902161B (en) | 2023-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10853580B1 (en) | Generation of text classifier training data | |
| CN111680173A (en) | CMR model for uniformly retrieving cross-media information | |
| CN107678561A (en) | Phonetic entry error correction method and device based on artificial intelligence | |
| CN1606004B (en) | Method and apparatus for identifying semantic structures from text | |
| CN109726274B (en) | Question generation method, device and storage medium | |
| CN108304375A (en) | A kind of information identifying method and its equipment, storage medium, terminal | |
| US9135396B1 (en) | Method and system for determining sets of variant items | |
| JP6756079B2 (en) | Artificial intelligence-based ternary check method, equipment and computer program | |
| CN108563655A (en) | Text based event recognition method and device | |
| CN107679032A (en) | Voice changes error correction method and device | |
| CN104298690A (en) | Method and device for building index structure for relational database table and method and device for conducting inquiring | |
| CN108228571B (en) | Method and device for generating couplet, storage medium and terminal equipment | |
| CN107203265B (en) | Information interaction method and device | |
| KR102546895B1 (en) | method and apparatus for providing academic information based on the summary and text matching of thesis | |
| CN108614851A (en) | Notes content display methods in tutoring system and device | |
| CN102567455A (en) | Method and system of managing documents using weighted prevalence data for statements | |
| CN107885888A (en) | Information processing method and device, terminal device and computer-readable recording medium | |
| KR101607468B1 (en) | Keyword tagging method and system for contents | |
| CN109933803A (en) | A kind of Chinese idiom information displaying method shows device, electronic equipment and storage medium | |
| CN108153875B (en) | Corpus processing method and device, intelligent sound box and storage medium | |
| JP2018112853A (en) | Topic classification apparatus and program therefor | |
| CN101770291A (en) | Semantic analysis data hashing storage and analysis methods for input system | |
| CN109902161A (en) | Answer processing method, device, equipment and the storage medium of question answering system | |
| CN109753557A (en) | Answer output method, device, equipment and the storage medium of question answering system | |
| CN110019696A (en) | Query intention mask method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211117 Address after: 210034 floor 8, building D11, Hongfeng Science Park, Nanjing Economic and Technological Development Zone, Jiangsu Province Applicant after: New Technology Co.,Ltd. Applicant after: VOLKSWAGEN (CHINA) INVESTMENT Co.,Ltd. Address before: 100094 1001, 10th floor, office building a, 19 Zhongguancun Street, Haidian District, Beijing Applicant before: MOBVOI INFORMATION TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |