CN110263051A - Question and answer for question answering system are to update method, device, equipment and storage medium - Google Patents
Question and answer for question answering system are to update method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN110263051A CN110263051A CN201910501631.4A CN201910501631A CN110263051A CN 110263051 A CN110263051 A CN 110263051A CN 201910501631 A CN201910501631 A CN 201910501631A CN 110263051 A CN110263051 A CN 110263051A
- Authority
- CN
- China
- Prior art keywords
- answer
- target
- question
- corrigendum
- answering system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/23—Updating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Present disclose provides a kind of question and answer for question answering system to update method, comprising: the target answer after obtaining the corrigendum of target problem；Editable database is written into target answer after target problem and corrigendum, the index of the target answer after establishing target problem and corrigendum；And based on index, editable database provides the target answer after the corrigendum of target problem to question answering system.The disclosure additionally provides a kind of question and answer for question answering system to updating device, electronic equipment and readable storage medium storing program for executing.
Description
Technical field
This disclosure relates to which a kind of question and answer for question answering system are to update method, updating device, electronic equipment and readable deposit
Storage media.
Background technique
Question answering system (question answer system, abbreviation QA) is for answering user's input or proposing each
The problem of kind various kinds.For example user puts question to " Chinese national territorial area is how many ", QA system answer " 9,600,000 square kilometres "；With
Put question to " how many Chinese population ", QA system answer " 1,300,000,000 " in family.
Question and answer refer to stored (question answer pair or key value pair, abbreviation KV to)
Good problem and answer pair.Such as: Chinese national territorial area is how many --- 9,600,000 square kilometres, how many Chinese population ---
1300000000, here it is two KV pairs.
In the prior art, the answer for answering out in question answering system, be greatly by retrieval question and answer to come
It arrives.As soon as when the semantic similarity between the query of user's input and the K of storage is greater than the threshold value of a setting, K pairs
The V answered regards the answer of the query, returns to user.
In existing retrieval and KV update scheme, at least 10 minutes time of needs is newly indexed, and be manually inserted into every time
Needing to click to update url by hand again after KV can just make to come into force on line.It cannot be guaranteed that heat updates in real time on line.
It is asked when the real-time error Capability Requirement to question answering system is high, such as in specific product line, such as children
It answers, customer service etc., it is expected that the problem of asking for a user, if answering wrong, question answering system can be supplemented at once and correctly be answered
Case.Users other in this way can obtain satisfied answer.So needing a question and answer to the mechanism of real-time update.
Summary of the invention
At least one of in order to solve the above-mentioned technical problem, present disclose provides a kind of question and answer for question answering system
To update method, updating device, electronic equipment and readable storage medium storing program for executing.
According to one aspect of the disclosure, a kind of question and answer for question answering system are to update method, comprising: obtain target
Target answer after the corrigendum of problem；Editable database is written into target answer after target problem and corrigendum, establishes target
The index of target answer after problem and corrigendum；And based on index, editable database provides target problem to question answering system
Corrigendum after target answer.
According at least one embodiment of the disclosure, editable database includes at least the first subdata base and second
Subdata base；Target answer after target problem and corrigendum is written to the first subdata base in real time；For target problem,
First subdata base and the second subdata base use different matching strategies.
According at least one embodiment of the disclosure, the target answer after target problem and corrigendum is write in non real-time
Enter to the second subdata base.
According at least one embodiment of the disclosure, editable database includes at least the first subdata base, the second son
Database and third subdata base；Target problem and corrigendum after target answer be written in real time the first subdata base and
Second subdata base；The matching used for target problem, the first subdata base, the second subdata base and third subdata base
Strategy is all different.
According at least one embodiment of the disclosure, the target answer after target problem and corrigendum is write in non real-time
Enter to third subdata base.
According at least one embodiment of the disclosure, before the target answer after the corrigendum for obtaining target problem, also
It include: the target answer for finding the mistake of target problem.
According at least one embodiment of the disclosure, the target answer of the mistake of target problem is found, comprising: be based on
Semantic similarity between target problem and target answer come judge target answer whether be mistake target answer.
According at least one embodiment of the disclosure, when question answering system is restarted, question answering system is from editable database
Question and answer pair are read to database with the question and answer of question answering system itself.
Target answer according at least one embodiment of the disclosure, after obtaining the corrigendum of target problem, comprising: base
The target answer of target problem is corrected in semantic similarity.
According to another aspect of the present disclosure, a kind of question and answer for question answering system are to updating device, which is characterized in that packet
Include: acquisition device, acquisition device are used to obtain the target answer after the corrigendum of target problem；Processing unit, processing unit is by mesh
Editable database is written in target answer after mark problem and corrigendum, the rope of the target answer after establishing target problem and corrigendum
Draw；And editable database, editable database provide the mesh after the corrigendum of target problem based on index to question answering system
Mark answer.
According to the another aspect of the disclosure, a kind of electronic equipment, comprising: memory, memory storage computer execution refer to
It enables；And processor, processor executes the computer executed instructions of memory storage, so that processor executes above-mentioned method.
According to the another further aspect of the disclosure, a kind of readable storage medium storing program for executing is stored with computer execution in readable storage medium storing program for executing
Instruction, for realizing above-mentioned method when computer executed instructions are 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
A part of this specification.
Fig. 1 is the schematic stream according to the question and answer for question answering system of one embodiment of the disclosure to update method
Cheng Tu.
Fig. 2 is the question and answer for question answering system according to disclosure another embodiment to the schematic of update method
Flow chart.
Fig. 3 is the question and answer for question answering system according to disclosure another embodiment to the schematic of update method
Flow chart.
Fig. 4 is the question and answer for question answering system according to disclosure another embodiment to the schematic of update method
Flow chart.
Fig. 5 is the schematic frame according to the question and answer for question answering system of one embodiment of the disclosure to updating device
Figure.
Fig. 6 is the question and answer for question answering system according to disclosure another embodiment to the schematic of updating device
Block diagram.
Fig. 7 is the question and answer for question answering system according to disclosure another embodiment to the schematic of updating device
Block diagram.
Fig. 8 is the question and answer for question answering system according to disclosure another embodiment to the schematic of updating device
Block diagram.
Fig. 9 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
For ease of 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.
Fig. 1 is the question and answer for question answering system of one embodiment of the disclosure to update method, comprising: S11, acquisition
Target answer after the corrigendum of target problem；S12, editable database is written into the target answer after target problem and corrigendum,
The index of target answer after establishing target problem and corrigendum；And S13, based on index, editable database is to question answering system
Target answer after the corrigendum of target problem is provided.
Such as user proposes a problem " Chinese national territorial area is how many " using question answering system, question answering system provides
Target answer be " 1,300,000,000 ", then for the question and answer of present embodiment to update method, after the corrigendum for needing to obtain target problem
Target answer be " 9,600,000 square kilometres ".Target after target problem " Chinese national territorial area is how many " and corrigendum is answered
" 9,600,000 square kilometres " write-in editable databases of case establish " Chinese national territorial area is how many " and " 9,600,000 squares of public affairs
In " index, when other users reuse problem system propose " Chinese national territorial area is how many " this target problem
When, it is " 9,600,000 square kilometres " namely question answering system that editable database, which provides the target answer after corrigendum to question answering system,
The target answer after corrigendum is read from editable database.Wherein, the preferred mySQL database of editable database.Preferably,
Target problem and the index of target answer have global uniqueness.Preferably, the target after obtaining the corrigendum of target problem is answered
Case can realize by obtaining target answer edit by personnel, can also be by trained by machine learning algorithm
Semantic similarity model obtain correct target answer to realize.
Semantic similarity model in the prior art, which generally passes through, calculates the distance between word, is reconverted into similarity.
And the distance between semanteme, such as can be counted by a large amount of corpus, this method is mainly by contextual information
Reference foundation of the probability distribution as Similarity of Words utilizes Large Scale Corpus, and the contextual information of word is made
For the reference foundation of Semantic Similarity Measurement.Quantitative analysis method based on statistics can between the Semantic Similarity vocabulary into
The more accurate and effective measurement of row, certain those skilled in the art can also calculate word language by ontology or classification relation
Adopted distance, these belong to the prior art, repeat no more.
Question answering system (Question Answering System, QA) is a kind of advanced form of information retrieval system,
It can answer the problem of user is proposed with natural language with accurate, succinct natural language.Question answering system is with natural language understanding
Technology is core, enables a computer to the conversation content for understanding user, realizes effective communication gap between people and computer
It is logical.Such as the chat robots system generallyd use in computer customer service system is exactly a kind of question answering system, it is a kind of
By natural language technology, can automatically with the artificial intelligence system of user session.Multiple types are had been developed that in the prior art
The question answering system of type, details are not described herein.Present embodiment does not do special limit to the question answering system that update method is applied to question and answer
It is fixed.
MySQL is a kind of Relational DBMS, and relational database saves the data in different tables, and
It is not to be placed on all data in one big warehouse, which adds speed and improves flexibility.The disclosure is not to can
The type of editor's database is particularly limited to, and those skilled in the art can select other kinds of under the enlightenment of the disclosure
Editable database.
Fig. 2 is the question and answer for question answering system of one embodiment of the disclosure to update method, comprising: S21, discovery
The target answer of the mistake of target problem；S22, obtain target problem corrigendum after target answer；S23, by target problem and
Editable database is written in target answer after corrigendum, the index of the target answer after establishing target problem and corrigendum；And
S24, it is based on index, editable database provides the target answer after the corrigendum of target problem to question answering system.
It was found that the target answer of the mistake of target problem, is preferably based on target problem in present embodiment and target is answered
Semantic similarity between case come judge target answer whether be mistake target answer.Preferably, using by machine learning
Algorithm training semantic similarity model come judge target answer whether be mistake target answer.Machine learning algorithm is can
To learn from data and therefrom modified hydrothermal process, without manual intervention.Learning tasks may include input is mapped to it is defeated
Out, learn hiding structure, or " instance-based learning " in unlabelled data, wherein by by new example and come
It is compared to generate class label for new example from the example of training data stored in memory.
Machine learning algorithm is broadly divided into three categories: supervised learning algorithm, unsupervised learning algorithm and intensified learning are calculated
Method.Wherein, supervised learning algorithm learns the mapping from input variable (X) to output variable (Y) using the training data of label
Function, typical supervised learning algorithm such as linear regression algorithm, Logistic regression algorithm, CART algorithm, naive Bayesian
Algorithm etc.；Unsupervised learning algorithm only has input variable (X), but without corresponding output variable, it uses the instruction without label
Practice data and carry out the basic structure of analogue data, typical unsupervised learning algorithm for example Apriori algorithm, K-means algorithm,
PCA algorithm etc.；Nitrification enhancement can be divided into pattern-based nitrification enhancement and non-mode is strengthened according to specified criteria
Learning algorithm and active nitrification enhancement and passive facilitation learning algorithm.
It to the machine learning algorithm type used and is not specially limited in present embodiment, as long as those skilled in the art
It can be used to train semantic similarity model by member, and semantic similarity model is enabled to judge whether target answer is wrong
Target answer accidentally.
Fig. 3 is the question and answer for question answering system of one embodiment of the disclosure to update method, comprising: S31, acquisition
Target answer after the corrigendum of target problem；S32, editable number is written into the target answer after target problem and corrigendum in real time
According to first subdata base in library, the index of the target answer after establishing target problem and corrigendum；And S33, based on index, can
Database is edited to the target answer after the corrigendum of question answering system offer target problem.
Wherein, editable database includes the first subdata base and the second subdata base；Mesh after target problem and corrigendum
Mark answer is written to the first subdata base in real time；For target problem, the first subdata base and the second subdata base are used
Different matching strategies.Target answer after target problem and corrigendum is written to the second subdata base in non real-time.This implementation
In mode, it is preferable that editable database provides the target answer after the corrigendum of target problem to question answering system in real time.Example
Such as, mySQL database is configured with exact matching library and fuzzy matching library, it will be appreciated by those skilled in the art that asking for target
Topic, the matching strategy that exact matching library and fuzzy matching library use are different.In present embodiment, exact matching only does accurate
The problem of matching, i.e., storing in target problem and exact matching library has a word is different can not match, but correspondingly, complete
Full matching library can increase KV newly to finishing to increase speed of the new question and answer to (i.e. KV to) when especially fast within general 1ms.
Generally, exact matching library can be matched to about 20% target problem.Although fuzzy matching can match about 80% mesh
Mark problem, but newly-increased KV pairs of the speed in fuzzy matching library will be slower than exact matching library.
Fig. 4 is the question and answer for question answering system of one embodiment of the disclosure to update method, comprising: S41, acquisition
Target answer after the corrigendum of target problem；S42, editable number is written into the target answer after target problem and corrigendum in real time
According to first subdata base and the second subdata base in library, the index of the target answer after establishing target problem and corrigendum；And
S43, it is based on index, editable database provides the target answer after the corrigendum of target problem to question answering system.
Wherein, editable database includes the first subdata base, the second subdata base and third subdata base；Target is asked
Target answer after topic and corrigendum is written to the first subdata base and the second subdata base in real time；For target problem,
One subdata base, the second subdata base and third subdata base use different matching strategies.Mesh after target problem and corrigendum
Mark answer is written to third subdata base in non real-time.In present embodiment, it is preferable that editable database is in real time to asking
Target answer after the corrigendum of system offer target problem is provided.For example, mySQL database is configured with exact matching library, simple mould
Matching library and complicated fuzzy matching library are pasted, it is exact matching library, simple it will be appreciated by those skilled in the art that for target problem
The matching strategy that fuzzy matching library, complicated fuzzy matching library use is different.In present embodiment, exact matching only does accurate
The problem of matching, i.e., storing in target problem and exact matching library has a word is different can not match, with the reality of front one
The matching strategy for applying mode is identical, repeats no more.
For simple fuzzy control matching library, increases KV clock synchronization newly, the computer representation of target problem, such as target problem need to be calculated
Phonetic, matching when calculate target problem phonetic and stored problem phonetic editing distance；For another example new
Increase KV clock synchronization, calculate the embedding vector of target problem, in matching, calculates the embedding and of target problem
The COS distance etc. of the embedding of the problem of storage.Generally, simple fuzzy control matching library increases KV newly to can be in 500ms
Within complete, and about 80% target problem can be matched to.
Matching library fuzzy for complexity, increases KV clock synchronization newly, needs to need overall situation KV number such as being established inverted index
According to the operation that library calculates, generally require 10 minutes or so.The target that the fuzzy matching library of complexity can be matched to about 80% is asked
Topic.
Wherein, global KV database includes the question and answer of above-mentioned editable database and question answering system itself to database.When
When question answering system is restarted, question answering system reads question and answer pair to database from the question and answer of editable database and question answering system itself.
According to another aspect of the present disclosure, the disclosure additionally provides a kind of question and answer for question answering system to updating device,
Fig. 5 shows the question and answer for question answering system of one embodiment of the disclosure to updating device 100, and question and answer are to updating device
100 include: acquisition device 101, and acquisition device 101 is used to obtain the target answer after the corrigendum of target problem；Processing unit
102, editable database 103 is written in the target answer after target problem and corrigendum by processing unit 102, establishes target problem
With the index of the target answer after corrigendum；And editable database 103, editable database 103 are based on index to question and answer system
System provides the target answer after the corrigendum of target problem.
Preferably, the target answer after obtaining the corrigendum of target problem, acquisition device 101 can be by obtaining through remarkable
The target answer of editor member realizes that acquisition device 101 can also be by similar to the semanteme by machine learning algorithm training
Degree model obtains correct target answer to realize.
For in present embodiment acquisition device 101 and processing unit 102 can be included in question answering system simultaneously
In；It is also possible to one in acquisition device 101 and processing unit 102 to be included in question answering system；It can also be acquisition dress
Set 101 and processing unit 102 be not included in question answering system, but acquisition device 101 and processing unit 102 can be with
Question answering system communication connection.Preferably embodiment, editable database 103 are not included in question answering system, ask
The system of answering can read data from editable database 103.
Fig. 6 shows the question and answer for question answering system of one embodiment of the disclosure to updating device 200, question and answer pair
Updating device 200 includes: judgment means 204, and judgment means 204 are used to find the target answer of the mistake of target problem；It obtains
Device 201, acquisition device 201 are used to obtain the target answer after the corrigendum of target problem；Processing unit 202, processing unit
Editable database 203 is written in target answer after target problem and corrigendum by 202, the mesh after establishing target problem and corrigendum
Mark the index of answer；And editable database 203, editable database 203 are based on index and ask to question answering system offer target
Target answer after the corrigendum of topic.Preferably, judgment means 204 are similar based on the semanteme between target problem and target answer
It spends to judge whether target answer is wrong target answer.Judgment means 204 are instructed for example, by using by machine learning algorithm
Experienced semantic similarity model come judge target answer whether be mistake target answer.
Preferably, the target answer after obtaining the corrigendum of target problem, acquisition device 201 can be by obtaining through remarkable
The target answer of editor member realizes that acquisition device 201 can also be by similar to the semanteme by machine learning algorithm training
Degree model obtains correct target answer to realize.
Can be included simultaneously for judgment means 204, acquisition device 201 and the processing unit 202 in present embodiment
In question answering system；It is also possible to one or two of judgment means 204, acquisition device 201 and processing unit 202 to be wrapped
It includes in question answering system；It can also be that judgment means 204, acquisition device 201 and processing unit 202 are not included in question and answer
In system, but judgment means 204, acquisition device 201 and processing unit 202 can be communicated to connect with question answering system.As
It is preferably carried out mode, editable database 203 is not included in question answering system, and question answering system can be from editable data
Data are read in library 203.
Fig. 7 shows the question and answer for question answering system of one embodiment of the disclosure to updating device 300, question and answer pair
Updating device 300 includes: acquisition device 301, and acquisition device 301 is used to obtain the target answer after the corrigendum of target problem；Place
Device 302 is managed, the target answer after target problem and corrigendum is written editable database 303, establishes mesh by processing unit 302
The index of target answer after mark problem and corrigendum；And editable database 303, editable database 303 are based on index
Target answer after the corrigendum of target problem is provided to question answering system.
Wherein, editable database 303 includes the first subdata base 3031 and the second subdata base 3032；Target problem
It is written to the first subdata base 3031 in real time with the target answer after corrigendum；For target problem, the first subdata base
3031 and second subdata base 3032 use different matching strategies.Target answer after target problem and corrigendum is non real-time
Ground is written to the second subdata base 3032.In present embodiment, it is preferable that editable database 303 is in real time to question answering system
Target answer after the corrigendum of target problem is provided.For example, mySQL database is configured with exact matching library (the first subdata base
3031 example) and fuzzy matching library (example of the second subdata base 3032), it will be appreciated by those skilled in the art that for
Target problem, the matching strategy that exact matching library and fuzzy matching library use are different.In present embodiment, exact matching is only done
The problem of storing in accurate matching, i.e. target problem and exact matching library has a word is different can not match, but corresponding
, exact matching library can increase KV pairs newly to increase the speed of new question and answer to (i.e. KV to) when especially fast within general 1ms
It finishes.Generally, exact matching library can be matched to about 20% target problem.Although fuzzy matching can match about
80% target problem, but newly-increased KV pairs of the speed in fuzzy matching library will be slower than exact matching library.
Preferably, editable database 303 and question answering system configure on a different server, correspondingly, editable number
Database is configured on a different server according to the question and answer of library 303 and question answering system itself, such as editable database 303
It can be only fitted on Cloud Server.
For in present embodiment acquisition device 301 and processing unit 302 can be included in question answering system simultaneously
In；It is also possible to one in acquisition device 301 and processing unit 302 to be included in question answering system；It can also be acquisition dress
Set 301 and processing unit 302 be not included in question answering system, but acquisition device 301 and processing unit 302 can be with
Question answering system communication connection.Preferably embodiment, editable database 303 are not included in question answering system, ask
The system of answering can read data from editable database 303.
Fig. 8 shows the question and answer for question answering system of one embodiment of the disclosure to updating device 400, question and answer pair
Updating device 400 includes: acquisition device 401, and acquisition device 401 is used to obtain the target answer after the corrigendum of target problem；Place
Device 402 is managed, the target answer after target problem and corrigendum is written editable database 403, establishes mesh by processing unit 402
The index of target answer after mark problem and corrigendum；And editable database 403, editable database 403 are based on index
Target answer after the corrigendum of target problem is provided to question answering system.
Wherein, editable database 403 includes the first subdata base 4031, the second subdata base 4032 and third subnumber
According to library 4033；Target answer after target problem and corrigendum is written to the first subdata base 4031 and the second subnumber in real time
According to library 4032；For target problem, the first subdata base 4031, the second subdata base 4032 and third subdata base 4033 are adopted
With different matching strategies.Target answer after target problem and corrigendum is written to third subdata base in non real-time
4033。
In present embodiment, it is preferable that editable database 403 provides the corrigendum of target problem to question answering system in real time
Target answer afterwards.For example, mySQL database is configured with exact matching library (example of the first subdata base 4031), simple mould
It pastes matching library (example of the second subdata base 4032) and complexity is fuzzy matching library (example of third subdata base 4033), this
It will be appreciated by the skilled person that for target problem, exact matching library, simple fuzzy control matching library, complexity obscure matching library in field
The matching strategy of use is different.In present embodiment, accurate matching, i.e. target problem and exact matching library are only done in exact matching
The problem of middle storage, has a word is different can not match, identical as the matching strategy of one embodiment in front, no longer
It repeats.
For simple fuzzy control matching library, increases KV clock synchronization newly, the computer representation of target problem, such as target problem need to be calculated
Phonetic, matching when calculate target problem phonetic and stored problem phonetic editing distance；For another example new
Increase KV clock synchronization, calculate the embedding vector of target problem, in matching, calculates the embedding and of target problem
The COS distance etc. of the embedding of the problem of storage.Generally, simple fuzzy control matching library increases KV newly to can be in 500ms
Within complete, and about 80% target problem can be matched to.
Matching library fuzzy for complexity, increases KV clock synchronization newly, needs to need overall situation KV number such as being established inverted index
According to the operation that library calculates, generally require 10 minutes or so.The target that the fuzzy matching library of complexity can be matched to about 80% is asked
Topic.
Wherein, global KV database includes the question and answer of above-mentioned editable database 403 and question answering system itself to data
Library.When question answering system is restarted, question answering system reads database from the question and answer of editable database 403 and question answering system itself
Take question and answer pair.
For in present embodiment acquisition device 401 and processing unit 402 can be included in question answering system simultaneously
In；It is also possible to one in acquisition device 401 and processing unit 402 to be included in question answering system；It can also be acquisition dress
Set 401 and processing unit 402 be not included in question answering system, but acquisition device 401 and processing unit 402 can be with
Question answering system communication connection.Preferably embodiment, editable database 403 are not included in question answering system, ask
The system of answering can read data from editable database 403.
The disclosure also provides a kind of electronic equipment, as shown in figure 9, the equipment includes: communication interface 1000, memory
2000 and processor 3000.Communication interface 1000 carries out data interaction for being communicated with external device.Memory
The computer program that can be run on processor 3000 is stored in 2000.When processor 3000 executes the computer program
Realize method in above 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, storage
Device 2000 and processor 3000 can be connected with each other by bus and complete mutual communication.The bus can be industry
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 just
It is only indicated with a thick line in expression, the figure, 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 be completed each other by internal interface
Communication.
Any process described otherwise above or method description are construed as in flow chart or herein, indicate packet
Include it is one or more for realizing specific logical function or process the step of executable instruction code module, segment
Or part, and the range of the preferred embodiment of the disclosure includes other realization, wherein can not be by shown or discussion
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
Embodiment of the present disclosure person of ordinary skill in the field is understood.Processor executes each method as described above
And processing.For example, the method implementation in the disclosure may be implemented as software program, it is tangibly embodied in machine
Readable medium, such as memory.In some embodiments, software program it is some or all of can via memory and/
Or communication interface and be loaded into and/or install.When software program is loaded into memory and is executed by processor, can execute
One or more steps in method as described above.Alternatively, in other embodiments, processor can pass through other
Any mode appropriate (for example, by means of firmware) and be 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 readable storage medium storing program for executing, so that (such as computer based system is including processor for instruction execution system, device or equipment
Unite or other can be from instruction execution system, device or equipment instruction fetch and the system executed instruction) it uses, or combine these
Instruction execution system, device or equipment and use.
For the purpose of this specification, " readable storage medium storing program for executing " can be it is any may include, store, communicate, propagate, or transport
Program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment
It sets.The more specific example (non-exhaustive list) of readable storage medium storing program for executing include the following: there is being electrically connected for one or more wirings
Socket part (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, readable storage medium storing program for executing can even is that the paper that can print described program on it or other suitable media,
Because can then be edited, be interpreted or when necessary with other conjunctions for example by carrying out optical scanner to paper or other media
Suitable mode is handled electronically to obtain described program, is then stored in memory.
It should be appreciated that each section of the disclosure can be realized with hardware, software or their combination.In above-mentioned implementation
In mode, multiple steps or method can with storage in memory and by software that suitable instruction execution system executes come
It realizes.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 realize: have for realizing the discrete of the logic gates of logic function to data-signal
Logic circuit, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), scene can compile
Journey gate array (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 readable storage medium storing program for executing, should
Program 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 can at one
It reads in storage medium.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 tool for combining the embodiment/mode or example to describe
Body characteristics, structure, material or feature are contained at least one embodiment/mode or example of the application.In this explanation
In book, 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 with suitable
Mode combines.In addition, without conflicting with each other, those skilled in the art can by it is described in this specification not
Same embodiment/mode or example and different embodiments/mode or exemplary feature are combined.
In addition, term " first ", " second " are used for description purposes only, it is not understood to indicate or imply relatively important
Property or implicitly indicate the quantity of indicated technical characteristic." first " is defined 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
It is a, three 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 question and answer for question answering system are to update method characterized by comprising
Target answer after obtaining the corrigendum of target problem；
Editable database is written into target answer after the target problem and the corrigendum, establishes the target problem and institute
The index of target answer after stating corrigendum；And
Based on the index, after the editable database provides the corrigendum of the target problem to the question answering system
Target answer.
2. question and answer according to claim 1 are to update method, which is characterized in that the editable database includes at least the
One subdata base and the second subdata base；Target answer after the target problem and the corrigendum is written to described in real time
First subdata base；For target problem, first subdata base and the second subdata base use different matching strategies.
3. question and answer according to claim 2 are to update method, which is characterized in that after the target problem and the corrigendum
Target answer is written to second subdata base in non real-time.
4. question and answer according to any one of claim 1 to 3 are to update method, which is characterized in that obtaining target problem
Corrigendum after target answer before, further includes: find the target problem mistake target answer.
5. question and answer according to claim 4 are to update method, which is characterized in that find the mesh of the mistake of the target problem
Mark answer, comprising: the target answer is judged based on the semantic similarity between the target problem and target answer whether
For the target answer of mistake.
6. question and answer according to any one of claim 1 to 3 are to update method, which is characterized in that when the question answering system
When restarting, the question answering system reads question and answer pair to database from the question and answer of the editable database and question answering system itself.
7. question and answer according to any one of claim 1 to 6 are to update method, which is characterized in that obtain target problem
Target answer after corrigendum, comprising: the target answer of the target problem is corrected based on semantic similarity.
8. a kind of question and answer for question answering system are to updating device characterized by comprising
Acquisition device, the acquisition device are used to obtain the target answer after the corrigendum of target problem；
Editable data are written in target answer after the target problem and the corrigendum by processing unit, the processing unit
Library, the index of the target answer after establishing the target problem and the corrigendum；And
Editable database, the editable database provide the target problem based on described index to the question answering system
Target answer after the corrigendum.
9. a kind of electronic equipment characterized by comprising
Memory, the memory storage execute instruction；And
Processor, the processor execute executing instruction for the memory storage, so that the processor is executed as right is wanted
Method described in asking any one of 1 to 7.
10. a kind of readable storage medium storing program for executing, which is characterized in that it is stored with and executes instruction in the readable storage medium storing program for executing, the execution
For realizing the method as described in any one of claims 1 to 7 when instruction is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910501631.4A CN110263051A (en) | 2019-06-11 | 2019-06-11 | Question and answer for question answering system are to update method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910501631.4A CN110263051A (en) | 2019-06-11 | 2019-06-11 | Question and answer for question answering system are to update method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110263051A true CN110263051A (en) | 2019-09-20 |
Family
ID=67917571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910501631.4A Pending CN110263051A (en) | 2019-06-11 | 2019-06-11 | Question and answer for question answering system are to update method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263051A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101093509A (en) * | 2007-07-18 | 2007-12-26 | 中国科学院计算技术研究所 | Interactive querying system and method |
| CN101373532A (en) * | 2008-07-10 | 2009-02-25 | 昆明理工大学 | FAQ Chinese request-answering system implementing method in tourism field |
| CN103902652A (en) * | 2014-02-27 | 2014-07-02 | 深圳市智搜信息技术有限公司 | Automatic question-answering system |
| CN105917334A (en) * | 2014-01-14 | 2016-08-31 | 微软技术许可有限责任公司 | Coherent question answering in search results |
| CN105912697A (en) * | 2016-04-25 | 2016-08-31 | 北京光年无限科技有限公司 | Optimization method and device of dialog system knowledge base |
| US20170103069A1 (en) * | 2015-10-13 | 2017-04-13 | International Business Machines Corporation | Supplementing candidate answers |
| CN106874441A (en) * | 2017-02-07 | 2017-06-20 | 腾讯科技（上海）有限公司 | Intelligent answer method and apparatus |
| CN108108426A (en) * | 2017-12-15 | 2018-06-01 | 杭州网蛙科技有限公司 | Understanding method, device and the electronic equipment that natural language is putd question to |
| US10019434B1 (en) * | 2012-06-01 | 2018-07-10 | Google Llc | Resolving pronoun ambiguity in voice queries |
| CN109145092A (en) * | 2017-06-15 | 2019-01-04 | 阿里巴巴集团控股有限公司 | A kind of database update, intelligent answer management method, device and its equipment |
| CN109815321A (en) * | 2018-12-26 | 2019-05-28 | 出门问问信息科技有限公司 | Question answering method, device, equipment and storage medium |
-
2019
- 2019-06-11 CN CN201910501631.4A patent/CN110263051A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101093509A (en) * | 2007-07-18 | 2007-12-26 | 中国科学院计算技术研究所 | Interactive querying system and method |
| CN101373532A (en) * | 2008-07-10 | 2009-02-25 | 昆明理工大学 | FAQ Chinese request-answering system implementing method in tourism field |
| US10019434B1 (en) * | 2012-06-01 | 2018-07-10 | Google Llc | Resolving pronoun ambiguity in voice queries |
| CN105917334A (en) * | 2014-01-14 | 2016-08-31 | 微软技术许可有限责任公司 | Coherent question answering in search results |
| CN103902652A (en) * | 2014-02-27 | 2014-07-02 | 深圳市智搜信息技术有限公司 | Automatic question-answering system |
| US20170103069A1 (en) * | 2015-10-13 | 2017-04-13 | International Business Machines Corporation | Supplementing candidate answers |
| CN105912697A (en) * | 2016-04-25 | 2016-08-31 | 北京光年无限科技有限公司 | Optimization method and device of dialog system knowledge base |
| CN106874441A (en) * | 2017-02-07 | 2017-06-20 | 腾讯科技（上海）有限公司 | Intelligent answer method and apparatus |
| CN109145092A (en) * | 2017-06-15 | 2019-01-04 | 阿里巴巴集团控股有限公司 | A kind of database update, intelligent answer management method, device and its equipment |
| CN108108426A (en) * | 2017-12-15 | 2018-06-01 | 杭州网蛙科技有限公司 | Understanding method, device and the electronic equipment that natural language is putd question to |
| CN109815321A (en) * | 2018-12-26 | 2019-05-28 | 出门问问信息科技有限公司 | Question answering method, device, equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 陈阳: "互联网"人肉搜索"在公安工作中的创新运用", 《江苏警官学院学报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11816438B2 (en) | Context saliency-based deictic parser for natural language processing | |
| CN108446286B (en) | Method, device and server for generating natural language question answers | |
| Stefik | Introduction to knowledge systems | |
| Patel et al. | Systems, design, and entrepreneurial thinking: Comparative frameworks | |
| Norton et al. | Introductory concepts in information science | |
| CN109919316A (en) | The method, apparatus and equipment and storage medium of acquisition network representation study vector | |
| JP2020071869A (en) | Video-based job provider and job seeker matching server and method | |
| CN105989088A (en) | Learning device under digital environment | |
| Cowen et al. | How to learn and teach economics with large language models, including GPT | |
| CN109783624A (en) | Answer generation method, device and the intelligent conversational system in knowledge based library | |
| CN113871020B (en) | Health management method and system after critical illness diagnosis based on AI machine learning | |
| CN112015896B (en) | Emotion classification method and device based on artificial intelligence | |
| CN111339285B (en) | BP neural network-based enterprise resume screening method and system | |
| Laurell et al. | Disruption and Social Media—Entrant firms as institutional entrepreneurs | |
| CN112463986A (en) | Information storage method and device | |
| CN111552796A (en) | Volume assembling method, electronic device and computer readable medium | |
| CN117556802B (en) | User portrait method, device, equipment and medium based on large language model | |
| Banasiewicz | Organizational learning in the Age of Data | |
| CN110263051A (en) | Question and answer for question answering system are to update method, device, equipment and storage medium | |
| Pilkington et al. | A conceptual model of the research methodology domain-with a focus on computing fields of study | |
| Al-Hmouz et al. | A machine learning based framework for adaptive mobile learning | |
| CN110222186A (en) | Reduplicated word class question processing method, processing unit, equipment and storage medium | |
| Olsson | Schools of systems thinking-Development trends in systems methodology | |
| Shedlock et al. | Indigeneity and technology: Assigning Indigenous properties to the agent-based IT artefact ensemble | |
| CN112102304A (en) | Image processing method, image processing device, computer equipment and computer readable 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 |
Effective date of registration: 20211118 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. |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190920 |
|
| RJ01 | Rejection of invention patent application after publication |