CN106155999A - Semantics comprehension on natural language method and system - Google Patents
Semantics comprehension on natural language method and system Download PDFInfo
- Publication number
- CN106155999A CN106155999A CN201510166742.6A CN201510166742A CN106155999A CN 106155999 A CN106155999 A CN 106155999A CN 201510166742 A CN201510166742 A CN 201510166742A CN 106155999 A CN106155999 A CN 106155999A
- Authority
- CN
- China
- Prior art keywords
- semantic
- node
- tree
- syntax tree
- interdependent syntax
- 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
Landscapes
- Machine Translation (AREA)
Abstract
The invention discloses a kind of semantics comprehension on natural language method and system, the method includes: obtain text to be resolved；Described text to be resolved is carried out participle and part-of-speech tagging；Based on context-related information, the text after participle is carried out syntactic analysis, obtain the interdependent syntax tree of natural word of every words in described pending text；According to the described interdependent syntax tree of natural word and the ontology knowledge base that builds in advance, build semantic tree；Described semantic tree is utilized to obtain semantic understanding result.Utilize the present invention, correctness and the motility of semantics comprehension on natural language can be improved.
Description
Technical field
The present invention relates to natural language processing technique field, be specifically related to a kind of semantics comprehension on natural language method
And system.
Background technology
Along with the fast development of Internet technology, the hope that people are the most urgent can freely be entered with machine
Row exchange.Natural language understanding, as realizing the key link of man-machine communication, has become as research worker at present
Study hotspot.In recent years, ontology knowledge base is owing to having good semantic concept level, it is easy to find out
Implicit contact between resource, effectively organizes resource and is widely used in natural language understanding neck
In territory.Existing semantics comprehension on natural language method based on body is mainly extraction from natural language text
Natural word, is then mapped to this pronouns, general term for nouns, numerals and measure words by nature word, obtains corresponding tlv triple in ontology knowledge base, by described
Triplet information is combined to obtain semantic understanding result, and this method typically can only process better simply literary composition
This, when text clause to be resolved is more complicated, semantic understanding is easily made mistakes, and can not get correct semantic understanding
Result.Therefore, how to utilize ontology knowledge base flexibly and accurately to carry out semantics comprehension on natural language and become research
Personnel's problem demanding prompt solution.
Summary of the invention
The embodiment of the present invention provides a kind of semantics comprehension on natural language method and system, to improve semantic understanding
Correctness and motility.
To this end, the embodiment of the present invention following technical scheme of offer:
A kind of semantics comprehension on natural language method, including:
Obtain text to be resolved；
Described text to be resolved is carried out participle and part-of-speech tagging；
Based on context-related information, the text after participle is carried out syntactic analysis, obtain described pending text
In every words the interdependent syntax tree of natural word；
According to the described interdependent syntax tree of natural word and the ontology knowledge base that builds in advance, build semantic tree；
Described semantic tree is utilized to obtain semantic understanding result.
Preferably, described text after participle is carried out syntactic analysis, obtain in described pending text every
The interdependent syntax tree of natural word of words includes:
Use maximum spanning tree algorithm that the text after participle is carried out interdependent syntactic analysis, obtain described pending
The interdependent syntax tree of natural word of every words in text.
Preferably, described according to the described interdependent syntax tree of natural word and the ontology knowledge base that builds in advance, build
Semantic tree includes:
The described interdependent syntax tree of natural word is reflected by the semantic vocabulary according to the natural word built in advance to this pronouns, general term for nouns, numerals and measure words
Penetrate as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words；
It is semantic right to determine in the node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words and limit and described ontology knowledge base
Should be related to；
According to described corresponding relation, described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semantic tree.
Preferably, described according to described corresponding relation, described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semantic tree bag
Include:
According to following rule described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words be converted to semantic tree:
Between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node directly
When being connected, obtain the semantic relation between two nodes being connected according to described ontology knowledge base, and by described
Semantic relation is put on the limit connecting described two nodes；
Pass through between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node
When relation node or attribute node are indirectly connected to, a newly-generated limit by described category node, instant node or
Property value node is joined directly together, and is put on newly-generated limit by the semanteme of relation node or attribute node.
Preferably, described utilize described semantic tree obtain semantic understanding result include:
The mode of employing postorder traversal is bottom-up to be traveled through described semantic tree, obtains on described semantic tree
Each node；
Two nodes each edge on described semantic tree and this limit connected are as a subordinate sentence, according to node
Order carries out semantic combination, obtains the statement after the combination of all subordinate sentences as semantic understanding result.
A kind of semantics comprehension on natural language system, including:
Receiver module, is used for obtaining text to be resolved；
Pretreatment module, for carrying out participle and part-of-speech tagging to described text to be resolved；
Syntactic analysis module, for the text after participle being carried out syntactic analysis based on context-related information,
Obtain the interdependent syntax tree of natural word of every words in described pending text；
Semantic tree builds module, for according to the described interdependent syntax tree of natural word and the ontology knowledge that builds in advance
Storehouse, builds semantic tree；
Semantic module, is used for utilizing described semantic tree to obtain semantic understanding result.
Preferably, described syntactic analysis module, specifically for using maximum spanning tree algorithm to the literary composition after participle
Originally carry out interdependent syntactic analysis, obtain the interdependent syntax tree of natural word of every words in described pending text.
Preferably, described semantic tree structure module includes:
Map unit, is used for the semantic vocabulary according to the natural word built in advance to this pronouns, general term for nouns, numerals and measure words by described natural word
Interdependent syntax tree is mapped as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words；
Corresponding relation determines unit, for determining that the node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words and limit are with described
Corresponding relation semantic in ontology knowledge base；
Converting unit, for being converted to semanteme according to described corresponding relation by described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words
Tree.
Preferably, described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semanteme according to following rule by described converting unit
Tree:
Between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node directly
When being connected, obtain the semantic relation between two nodes being connected according to described ontology knowledge base, and by described
Semantic relation is put on the limit connecting described two nodes；
Pass through between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node
When relation node or attribute node are indirectly connected to, a newly-generated limit by described category node, instant node or
Property value node is joined directly together, and the semanteme of point or attribute node is put on newly-generated limit by relation.
Preferably, described semantic module includes:
Traversal Unit, for using, the mode of postorder traversal is bottom-up to be traveled through described semantic tree,
Each node on described semantic tree；
Assembled unit, two nodes being used for connecting each edge on described semantic tree and this limit are as one point
Sentence, carries out semantic combination according to node sequence, obtains the statement after the combination of all subordinate sentences and ties as semantic understanding
Really.
The semantics comprehension on natural language method and system that the embodiment of the present invention provides, by combining ontology knowledge base
And the mode of nature word interdependent syntax tree structure semantic tree carries out semantics comprehension on natural language, the most as much as possible
Remain the semantic information of natural language text, and can fully demonstrate between each body tlv triple mutual
Relation, therefore substantially increases the accuracy of semantic understanding.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present application or technical scheme of the prior art, below will be to enforcement
In example, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is only
Some embodiments described in the present invention, for those of ordinary skill in the art, it is also possible to according to these
Accompanying drawing obtains other accompanying drawing.
Fig. 1 is the flow chart of embodiment of the present invention semantics comprehension on natural language method；
Fig. 2 is natural word interdependent syntax tree example in the embodiment of the present invention；
Fig. 3 is this pronouns, general term for nouns, numerals and measure words interdependent syntax tree example in the embodiment of the present invention；
Fig. 4 is semantic tree example in the embodiment of the present invention；
Fig. 5 is the structural representation of embodiment of the present invention semantics comprehension on natural language system.
Detailed description of the invention
In order to make those skilled in the art be more fully understood that the scheme of the embodiment of the present invention, below in conjunction with the accompanying drawings
With embodiment, the embodiment of the present invention is described in further detail.
As it is shown in figure 1, be the flow chart of embodiment of the present invention semantics comprehension on natural language method, including following
Step:
Step 101, obtains text to be resolved.
Step 102, carries out participle and part-of-speech tagging to described text to be resolved.
Specifically can use method based on condition random field that described pending text is carried out participle and part of speech
Mark.Certainly, it is possible to use other method to carry out participle and part-of-speech tagging, can be with long word such as participle
Joining, part-of-speech tagging can be with based on HMM (Hidden Markov Model, hidden Markov model)
Method etc..
Such as, to the word segmentation result of " the sheet caudal flexure of Swordman pleasing to the ear like that Who Am I and sing " text it is:
Swordman/n /u sheet caudal flexure/n is pleasing to the ear/a /u like I/n is /v who/r sings/v/u
Wherein, the letter representation part of speech in word segmentation result ,/v represents that verb ,/r represent that pronoun ,/n represent name
Word ,/u represent that auxiliary word ,/a represent adjective.
Step 103, carries out syntactic analysis based on context-related information to the text after participle, obtains described
The interdependent syntax tree of natural word of every words in pending text, in the described interdependent syntax tree of natural word, node table
The natural language terms obtained after showing text participle to be resolved, i.e. natural word, while represent two nodes of connection
The natural word represented dependence in text to be resolved.
Such as, can use maximum spanning tree algorithm that the text after participle is carried out interdependent syntactic analysis, obtain
The interdependent syntax tree of natural word of every words in described pending text.
As in figure 2 it is shown, be that text " the sheet caudal flexure of Swordman pleasing to the ear like that Who Am I sing " is carried out sentence
The interdependent syntax tree of natural word obtained after method analysis.Wherein, ROOT node is dummy node, it is possible to regards as and depends on
Depositing the dependence of syntax tree root node, calculate for convenience, ROOT node does not indicates that any word segmentation result.
Letter abbreviations on limit is dependence, and its implication is as shown in table 1 below:
Table 1:
| Dependence |
| Core word HED |
| Attribute head relation ATT |
| Subject-predicate relation SBV |
| Dynamic guest's relation VOB |
| Appositive relation APP |
| " " word structure DE |
Step 104, according to the described interdependent syntax tree of natural word and the ontology knowledge base that builds in advance, builds language
Justice tree.
Described ontology knowledge base includes body construction and according to knowledge defined in body construction, and these are known
Knowing and be generally stored in knowledge base with the form of tlv triple, some concepts of structure sheaf are connected by attribute
Come, form the triple form with subject, predicate and object, the i.e. shape of<subject, predicate, object>
Formula.
Concrete building process is as follows:
(1) according to the semantic vocabulary of the natural word that builds in advance to this pronouns, general term for nouns, numerals and measure words by the described interdependent syntax of natural word
Tree is mapped as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words, and in the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words, node represents oneself of text to be resolved
The word of storage in the ontology knowledge base that so word is corresponding, i.e. this pronouns, general term for nouns, numerals and measure words, while represent what two nodes of connection represented
This pronouns, general term for nouns, numerals and measure words dependence in text to be resolved.Node in the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words can have inhomogeneity
Type, the most in embodiments of the present invention, can have a node of five types, i.e. category node, instant node,
Relation node, attribute node and property value node.
Described semantic vocabulary can carry out structure according to this pronouns, general term for nouns, numerals and measure words of the semanteme in dictionary, ontology knowledge base and correspondence
Building, the semantic vocabulary structure such as " Swordman " is as follows:
Semantic: film
Word: Swordman
NLword: Swordman, Swordsman, State of Divinity
Wherein what Semantic was corresponding is the semanteme in ontology knowledge base, represents movies category at this；Word
Corresponding is the word of storage, i.e. this pronouns, general term for nouns, numerals and measure words in ontology library；What NLword was corresponding is may to go out in natural language
Existing word, i.e. natural word.
According to semantic vocabulary, the natural word in the natural interdependent syntax tree of word is directly mapped with this pronouns, general term for nouns, numerals and measure words, with
Time delete the natural word that cannot map, thus the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words after being mapped.
As Fig. 3 shows that natural language sentences " the sheet caudal flexure of Swordman pleasing to the ear like that Who Am I and sing " are right
The interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words answered.
(2) determine that the node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words and limit are semantic with described ontology knowledge base
Corresponding relation.
According to the semantic field in semantic vocabulary, i.e. Semantic field, determine the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words
The corresponding relation that interior joint and limit are semantic with ontology knowledge base.In embodiments of the present invention, described body is known
The semanteme known in storehouse can have following 5 kinds, i.e. classification, example, relation, attribute and property value.Certainly,
According to the different demands of application, the semanteme in ontology knowledge base can be arranged accordingly, this is sent out
Bright embodiment does not limits.
Interdependent syntax tree interior joint and limit as shown in Figure 3 are divided with the corresponding relation of semanteme in ontology knowledge base
As follows:
" Swordman " node correspondence movies category example, " liking me " node correspondence song class instance,
The property value of " pleasing to the ear " node correspondence attribute song label, attribute is " song label ", and " sheet caudal flexure " saves
The corresponding relation that the corresponding movies category example " Swordman " of point " likes me " with song class instance, "？
Singer " node correspondence singer's classification, " performance " node correspondence singer's classification and song class instance " like me "
Corresponding relation.ROOT node is dummy node, the semanteme in the most corresponding ontology knowledge base.
(3) according to described corresponding relation, described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semantic tree.
Specifically, the bottom-up interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words of scanning, according to described interdependent syntax tree interior joint and
While interdependent syntax tree is changed by the semantic relation in ontology knowledge base, obtain semantic tree.
If Fig. 4 is the semantic tree built according to the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words shown in Fig. 3, when conversion, permissible
According to following transformational rule:
A) between the category node in the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words, instant node or property value node directly
When being connected, by the semantic relation between node described in ontology knowledge library inquiry, described semantic relation is put into
On the limit that two nodes connect.
As shown in Figure 3." pleasing to the ear " property value node is joined directly together with " liking me " song instant node, logical
Crossing the inquiry ontology library semantic relation that obtains between described node is " song label ", therefore, by " song
Label " it is put on the limit of " pleasing to the ear " node and the connection of " liking me " node.
B) pass through between the category node in the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words, instant node or property value node
When relation node or attribute node are indirectly connected to, a newly-generated limit is by category node, instant node or attribute
Value node is joined directly together, and is put on newly-generated limit by the semanteme of relation node or attribute node, described indirectly
Connecting and refer generally to only be spaced between node a node, described interval node does not include that dummy node ROOT saves
Point.
As it is shown on figure 3, " Swordman " movies category instant node saves with " liking me " song class instance
Point is indirectly connected with by relation node " sheet caudal flexure ", then can directly generate a new limit by two instant node
It is directly connected to, and semanteme corresponding for relation node is put on newly-generated limit；After process completes, now
Song class instance node of " liking me " and "？Singer " song category node " sung " by relation node
It is indirectly connected with, then directly generates a new limit and two nodes are connected, and relation node " is sung " be put into
Bian Shang.
Step 105, utilizes described semantic tree to obtain semantic understanding result.
ROOT node on semantic tree is dummy node, does not consider this node during semantic understanding.
Two nodes that each edge on semantic tree and this limit connect are as a subordinate sentence, if semantic tree has many
, then there is multiple subordinate sentence on bar limit.The mode of employing postorder traversal is bottom-up to be traveled through semantic tree, obtains
Each node on semantic tree, carries out the combination of semanteme according to node sequence, and the result of final semantic understanding is
The semanteme expressed by statement after the combination of all subordinate sentences.The mode of described postorder traversal refers to the most first time
Go through the child node of semantic tree, then traverse up the father node of child node successively, until the root node of semantic tree.
If Fig. 4 is the semantic tree built, the node sequence that postorder traversal obtains after terminating is Swordman
-> pleasing to the ear->？Singer-> like me, is combined the subordinate sentence at each node place, obtains the knot of semantic understanding
Fruit is " Swordman sheet caudal flexure likes me, likes that my song label is pleasing to the ear, the love which singer sings I ".
The semantics comprehension on natural language method of the embodiment of the present invention, depends on by combining ontology knowledge base and natural word
The mode depositing syntax tree structure semantic tree carries out semantics comprehension on natural language, remains nature the most as much as possible
The semantic information of language text, and the mutual relation between each body tlv triple can be fully demonstrated, therefore
Substantially increase the accuracy of semantic understanding.
Correspondingly, the embodiment of the present invention also provides for a kind of semantics comprehension on natural language system, as it is shown in figure 5,
It it is a kind of structural representation of this system.
In this embodiment, described system includes:
Receiver module 501, is used for obtaining text to be resolved；
Pretreatment module 502, for carrying out participle and part-of-speech tagging to described text to be resolved；
Syntactic analysis module 503, divides for the text after participle being carried out syntax based on context-related information
Analysis, obtains the interdependent syntax tree of natural word of every words in described pending text；
Semantic tree builds module 504, for according to the described interdependent syntax tree of natural word and the body that builds in advance
Knowledge base 500, builds semantic tree；
Semantic module 505, is used for utilizing described semantic tree to obtain semantic understanding result.
Described syntactic analysis module 503 specifically can use maximum spanning tree calculation to carry out the text after participle
Interdependent syntactic analysis, obtains the interdependent syntax tree of natural word of every words in described pending text.
The natural interdependent syntax tree of word, when building semantic tree, can first be reflected by above-mentioned semantic tree structure module 504
Penetrate as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words, then according to the node in the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words and limit and described body
Corresponding relation semantic in knowledge base, is converted to semantic tree by described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words.
Semantic tree builds a kind of concrete structure of module 504 can include following unit:
Map unit, is used for the semantic vocabulary according to the natural word built in advance to this pronouns, general term for nouns, numerals and measure words by described natural word
Interdependent syntax tree is mapped as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words；
Corresponding relation determines unit, for determining that the node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words and limit are with described
Corresponding relation semantic in ontology knowledge base；
Converting unit, for being converted to semanteme according to described corresponding relation by described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words
Tree.
In actual applications, above-mentioned converting unit can be according to following rule by described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words
Be converted to semantic tree:
Between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node directly
When being connected, obtain the semantic relation between two nodes being connected according to described ontology knowledge base, and by described
Semantic relation is put on the limit connecting described two nodes；
Pass through between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node
When relation node or attribute node are indirectly connected to, a newly-generated limit by described category node, instant node or
Property value node is joined directly together, and is put on newly-generated limit by the semanteme of relation node or attribute node.
Building, based on semantic tree, the semantic tree that module 504 builds, semantic module 505 can use postorder
The mode of traversal is bottom-up to be traveled through semantic tree, obtains each node on semantic tree, according to node
Order carries out the combination of semanteme, and the result of final semantic understanding is expressed by the statement after the combination of all subordinate sentences
Semantic.
Semantic tree builds a kind of concrete structure of module 504 can include Traversal Unit and assembled unit, its
In:
Described Traversal Unit, for using, the mode of postorder traversal is bottom-up to be carried out described semantic tree time
Go through, obtain each node on described semantic tree；
Described assembled unit, for two nodes that each edge on described semantic tree and this limit are connected as
One subordinate sentence, carries out semantic combination according to node sequence, obtains the statement after the combination of all subordinate sentences as semanteme
Understand result.
Above-mentioned each module and unit realize the detailed process of its function and can be found in above the inventive method embodiment
In description, do not repeat them here.
The semantics comprehension on natural language system of the embodiment of the present invention, depends on by combining ontology knowledge base and natural word
The mode depositing syntax tree structure semantic tree carries out semantics comprehension on natural language, remains nature the most as much as possible
The semantic information of language text, and the mutual relation between each body tlv triple can be fully demonstrated, therefore
Substantially increase the accuracy of semantic understanding.
Each embodiment in this specification all uses the mode gone forward one by one to describe, phase homophase between each embodiment
As part see mutually, what each embodiment stressed is different from other embodiments it
Place.For system embodiment, owing to it is substantially similar to embodiment of the method, so describing
Fairly simple, relevant part sees the part of embodiment of the method and illustrates.System described above is implemented
Example is only that schematically the wherein said unit illustrated as separating component can be or may not be
Physically separate, the parts shown as unit can be or may not be physical location, the most permissible
It is positioned at a place, or can also be distributed on multiple NE.Can select according to the actual needs
Some or all of module therein realizes the purpose of the present embodiment scheme.Those of ordinary skill in the art exist
In the case of not paying creative work, i.e. it is appreciated that and implements.
Being described in detail the embodiment of the present invention above, detailed description of the invention used herein is to this
Bright being set forth, the explanation of above example is only intended to help to understand the method and system of the present invention；With
Time, for one of ordinary skill in the art, according to the thought of the present invention, in detailed description of the invention and application
All will change in scope, in sum, this specification content should not be construed as limitation of the present invention.
Claims (10)
1. a semantics comprehension on natural language method, it is characterised in that including:
Obtain text to be resolved；
Described text to be resolved is carried out participle and part-of-speech tagging；
Based on context-related information, the text after participle is carried out syntactic analysis, obtain described pending text
In every words the interdependent syntax tree of natural word；
According to the described interdependent syntax tree of natural word and the ontology knowledge base that builds in advance, build semantic tree；
Described semantic tree is utilized to obtain semantic understanding result.
Method the most according to claim 1, it is characterised in that described text after participle is carried out
Syntactic analysis, obtains the interdependent syntax tree of natural word of every words in described pending text and includes:
Use maximum spanning tree algorithm that the text after participle is carried out interdependent syntactic analysis, obtain described pending
The interdependent syntax tree of natural word of every words in text.
Method the most according to claim 1, it is characterised in that described interdependent according to described natural word
Syntax tree and the ontology knowledge base built in advance, build semantic tree and include:
The described interdependent syntax tree of natural word is reflected by the semantic vocabulary according to the natural word built in advance to this pronouns, general term for nouns, numerals and measure words
Penetrate as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words；
It is semantic right to determine in the node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words and limit and described ontology knowledge base
Should be related to；
According to described corresponding relation, described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semantic tree.
Method the most according to claim 3, it is characterised in that described according to described corresponding relation general
Described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semantic tree and includes:
According to following rule described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words be converted to semantic tree:
Between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node directly
When being connected, obtain the semantic relation between two nodes being connected according to described ontology knowledge base, and by described
Semantic relation is put on the limit connecting described two nodes；
Pass through between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node
When relation node or attribute node are indirectly connected to, a newly-generated limit by described category node, instant node or
Property value node is joined directly together, and is put on newly-generated limit by the semanteme of relation node or attribute node.
5. according to the method described in any one of Claims 1-4, it is characterised in that described in described utilization
Semantic tree obtains semantic understanding result and includes:
The mode of employing postorder traversal is bottom-up to be traveled through described semantic tree, obtains on described semantic tree
Each node；
Two nodes each edge on described semantic tree and this limit connected are as a subordinate sentence, according to node
Order carries out semantic combination, obtains the statement after the combination of all subordinate sentences as semantic understanding result.
6. a semantics comprehension on natural language system, it is characterised in that including:
Receiver module, is used for obtaining text to be resolved；
Pretreatment module, for carrying out participle and part-of-speech tagging to described text to be resolved；
Syntactic analysis module, for the text after participle being carried out syntactic analysis based on context-related information,
Obtain the interdependent syntax tree of natural word of every words in described pending text；
Semantic tree builds module, for according to the described interdependent syntax tree of natural word and the ontology knowledge that builds in advance
Storehouse, builds semantic tree；
Semantic module, is used for utilizing described semantic tree to obtain semantic understanding result.
System the most according to claim 6, it is characterised in that
Described syntactic analysis module, specifically for using maximum spanning tree algorithm to depend on the text after participle
Deposit syntactic analysis, obtain the interdependent syntax tree of natural word of every words in described pending text.
System the most according to claim 6, it is characterised in that described semantic tree builds module and includes:
Map unit, is used for the semantic vocabulary according to the natural word built in advance to this pronouns, general term for nouns, numerals and measure words by described natural word
Interdependent syntax tree is mapped as the interdependent syntax tree of this pronouns, general term for nouns, numerals and measure words；
Corresponding relation determines unit, for determining that the node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words and limit are with described
Corresponding relation semantic in ontology knowledge base；
Converting unit, for being converted to semanteme according to described corresponding relation by described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words
Tree.
System the most according to claim 8, it is characterised in that described converting unit is according to following rule
Then described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words is converted to semantic tree:
Between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node directly
When being connected, obtain the semantic relation between two nodes being connected according to described ontology knowledge base, and by described
Semantic relation is put on the limit connecting described two nodes；
Pass through between the category node in described the interdependent syntax tree of pronouns, general term for nouns, numerals and measure words, instant node or property value node
When relation node or attribute node are indirectly connected to, a newly-generated limit by described category node, instant node or
Property value node is joined directly together, and the semanteme of point or attribute node is put on newly-generated limit by relation.
10. according to the system described in any one of claim 6 to 9, it is characterised in that described semantic analysis
Module includes:
Traversal Unit, for using, the mode of postorder traversal is bottom-up to be traveled through described semantic tree,
Each node on described semantic tree；
Assembled unit, two nodes being used for connecting each edge on described semantic tree and this limit are as one
Subordinate sentence, carries out semantic combination according to node sequence, obtains the statement after the combination of all subordinate sentences as semantic understanding
Result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510166742.6A CN106155999A (en) | 2015-04-09 | 2015-04-09 | Semantics comprehension on natural language method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510166742.6A CN106155999A (en) | 2015-04-09 | 2015-04-09 | Semantics comprehension on natural language method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106155999A true CN106155999A (en) | 2016-11-23 |
Family
ID=57336566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510166742.6A Pending CN106155999A (en) | 2015-04-09 | 2015-04-09 | Semantics comprehension on natural language method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106155999A (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107590130A (en) * | 2017-09-30 | 2018-01-16 | 北京三快在线科技有限公司 | Scene determines method and device, storage medium and electronic equipment |
| CN108363700A (en) * | 2018-03-23 | 2018-08-03 | 北京奇虎科技有限公司 | The method for evaluating quality and device of headline |
| CN108399265A (en) * | 2018-03-23 | 2018-08-14 | 北京奇虎科技有限公司 | Real-time hot news providing method based on search and device |
| CN108470026A (en) * | 2018-03-23 | 2018-08-31 | 北京奇虎科技有限公司 | The sentence trunk method for extracting content and device of headline |
| CN108491512A (en) * | 2018-03-23 | 2018-09-04 | 北京奇虎科技有限公司 | The method of abstracting and device of headline |
| CN108960673A (en) * | 2018-07-24 | 2018-12-07 | 北京天诚同创电气有限公司 | Sewage treatment method for diagnosing faults and device |
| CN108959240A (en) * | 2017-05-26 | 2018-12-07 | 上海醇聚信息科技有限公司 | A kind of proprietary ontology automatic creation system and method |
| CN108959239A (en) * | 2017-05-26 | 2018-12-07 | 上海醇聚信息科技有限公司 | Natural language understanding system and its working method |
| CN109062902A (en) * | 2018-08-17 | 2018-12-21 | 科大讯飞股份有限公司 | A kind of text semantic expression and device |
| CN109815490A (en) * | 2019-01-04 | 2019-05-28 | 平安科技（深圳）有限公司 | Text analyzing method, apparatus, equipment and storage medium |
| CN110222194A (en) * | 2019-05-21 | 2019-09-10 | 深圳壹账通智能科技有限公司 | Data drawing list generation method and relevant apparatus based on natural language processing |
| CN110390085A (en) * | 2019-07-04 | 2019-10-29 | 东软集团股份有限公司 | Analysis method, analytical equipment, storage medium and the electronic equipment of text similarity |
| US10628743B1 (en) | 2019-01-24 | 2020-04-21 | Andrew R. Kalukin | Automated ontology system |
| CN111563385A (en) * | 2020-04-30 | 2020-08-21 | 北京百度网讯科技有限公司 | Semantic processing method, semantic processing device, electronic equipment and media |
| CN111783465A (en) * | 2020-07-03 | 2020-10-16 | 深圳追一科技有限公司 | Named entity normalization method, system and related device |
| CN111950290A (en) * | 2019-04-30 | 2020-11-17 | 广东小天才科技有限公司 | Semantic analysis method and device based on concept graph |
| WO2021000555A1 (en) * | 2019-07-03 | 2021-01-07 | 深圳追一科技有限公司 | Knowledge graph-based question answering method and apparatus, computer device and storage medium |
| CN112269884A (en) * | 2020-11-13 | 2021-01-26 | 北京百度网讯科技有限公司 | Information extraction method, device, equipment and storage medium |
| CN112287663A (en) * | 2020-11-25 | 2021-01-29 | 平安数字信息科技(深圳)有限公司 | Text parsing method, equipment, terminal and storage medium |
| CN112560477A (en) * | 2020-12-09 | 2021-03-26 | 中科讯飞互联(北京)信息科技有限公司 | Text completion method, electronic device and storage device |
| CN116882494A (en) * | 2023-09-07 | 2023-10-13 | 山东山大鸥玛软件股份有限公司 | Method and device for establishing non-supervision knowledge graph oriented to professional text |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1628298A (en) * | 2002-05-28 | 2005-06-15 | 弗拉迪米尔·叶夫根尼耶维奇·涅博利辛 | Method for synthesising self-learning system for knowledge acquistition for retrieval systems |
| CN104252533A (en) * | 2014-09-12 | 2014-12-31 | 百度在线网络技术（北京）有限公司 | Search method and search device |
-
2015
- 2015-04-09 CN CN201510166742.6A patent/CN106155999A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1628298A (en) * | 2002-05-28 | 2005-06-15 | 弗拉迪米尔·叶夫根尼耶维奇·涅博利辛 | Method for synthesising self-learning system for knowledge acquistition for retrieval systems |
| CN104252533A (en) * | 2014-09-12 | 2014-12-31 | 百度在线网络技术（北京）有限公司 | Search method and search device |
Non-Patent Citations (1)
| Title |
|---|
| 丁烨: "基于本体的中文问答系统中问句的语义理解", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108959240A (en) * | 2017-05-26 | 2018-12-07 | 上海醇聚信息科技有限公司 | A kind of proprietary ontology automatic creation system and method |
| CN108959239A (en) * | 2017-05-26 | 2018-12-07 | 上海醇聚信息科技有限公司 | Natural language understanding system and its working method |
| CN107590130B (en) * | 2017-09-30 | 2019-06-14 | 北京三快在线科技有限公司 | Scene determines method and device, storage medium and electronic equipment |
| CN107590130A (en) * | 2017-09-30 | 2018-01-16 | 北京三快在线科技有限公司 | Scene determines method and device, storage medium and electronic equipment |
| CN108363700A (en) * | 2018-03-23 | 2018-08-03 | 北京奇虎科技有限公司 | The method for evaluating quality and device of headline |
| CN108399265A (en) * | 2018-03-23 | 2018-08-14 | 北京奇虎科技有限公司 | Real-time hot news providing method based on search and device |
| CN108470026A (en) * | 2018-03-23 | 2018-08-31 | 北京奇虎科技有限公司 | The sentence trunk method for extracting content and device of headline |
| CN108491512A (en) * | 2018-03-23 | 2018-09-04 | 北京奇虎科技有限公司 | The method of abstracting and device of headline |
| CN108960673A (en) * | 2018-07-24 | 2018-12-07 | 北京天诚同创电气有限公司 | Sewage treatment method for diagnosing faults and device |
| CN109062902A (en) * | 2018-08-17 | 2018-12-21 | 科大讯飞股份有限公司 | A kind of text semantic expression and device |
| CN109062902B (en) * | 2018-08-17 | 2022-12-06 | 科大讯飞股份有限公司 | Text semantic expression method and device |
| CN109815490A (en) * | 2019-01-04 | 2019-05-28 | 平安科技（深圳）有限公司 | Text analyzing method, apparatus, equipment and storage medium |
| CN109815490B (en) * | 2019-01-04 | 2023-11-14 | 平安科技（深圳）有限公司 | Text analysis method, device, equipment and storage medium |
| US10628743B1 (en) | 2019-01-24 | 2020-04-21 | Andrew R. Kalukin | Automated ontology system |
| CN111950290A (en) * | 2019-04-30 | 2020-11-17 | 广东小天才科技有限公司 | Semantic analysis method and device based on concept graph |
| CN110222194A (en) * | 2019-05-21 | 2019-09-10 | 深圳壹账通智能科技有限公司 | Data drawing list generation method and relevant apparatus based on natural language processing |
| WO2020233345A1 (en) * | 2019-05-21 | 2020-11-26 | 深圳壹账通智能科技有限公司 | Natural language processing-based data chart generation method and related device |
| WO2021000555A1 (en) * | 2019-07-03 | 2021-01-07 | 深圳追一科技有限公司 | Knowledge graph-based question answering method and apparatus, computer device and storage medium |
| CN110390085A (en) * | 2019-07-04 | 2019-10-29 | 东软集团股份有限公司 | Analysis method, analytical equipment, storage medium and the electronic equipment of text similarity |
| CN110390085B (en) * | 2019-07-04 | 2023-04-07 | 东软集团股份有限公司 | Text similarity analysis method and device, storage medium and electronic equipment |
| CN111563385A (en) * | 2020-04-30 | 2020-08-21 | 北京百度网讯科技有限公司 | Semantic processing method, semantic processing device, electronic equipment and media |
| CN111563385B (en) * | 2020-04-30 | 2023-12-26 | 北京百度网讯科技有限公司 | Semantic processing method, semantic processing device, electronic equipment and medium |
| CN111783465A (en) * | 2020-07-03 | 2020-10-16 | 深圳追一科技有限公司 | Named entity normalization method, system and related device |
| CN111783465B (en) * | 2020-07-03 | 2024-04-30 | 深圳追一科技有限公司 | Named entity normalization method, named entity normalization system and related device |
| CN112269884A (en) * | 2020-11-13 | 2021-01-26 | 北京百度网讯科技有限公司 | Information extraction method, device, equipment and storage medium |
| CN112269884B (en) * | 2020-11-13 | 2024-03-05 | 北京百度网讯科技有限公司 | Information extraction method, device, equipment and storage medium |
| CN112287663A (en) * | 2020-11-25 | 2021-01-29 | 平安数字信息科技(深圳)有限公司 | Text parsing method, equipment, terminal and storage medium |
| CN112560477A (en) * | 2020-12-09 | 2021-03-26 | 中科讯飞互联(北京)信息科技有限公司 | Text completion method, electronic device and storage device |
| CN112560477B (en) * | 2020-12-09 | 2024-04-16 | 科大讯飞(北京)有限公司 | Text completion method, electronic equipment and storage device |
| CN116882494A (en) * | 2023-09-07 | 2023-10-13 | 山东山大鸥玛软件股份有限公司 | Method and device for establishing non-supervision knowledge graph oriented to professional text |
| CN116882494B (en) * | 2023-09-07 | 2023-11-28 | 山东山大鸥玛软件股份有限公司 | Method and device for establishing non-supervision knowledge graph oriented to professional text |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106155999A (en) | Semantics comprehension on natural language method and system | |
| CN107291687A (en) | It is a kind of based on interdependent semantic Chinese unsupervised open entity relation extraction method | |
| CN108268580A (en) | The answering method and device of knowledge based collection of illustrative plates | |
| CN103116657B (en) | A kind of individuation search method of network teaching resource | |
| CN103309926A (en) | Chinese and English-named entity identification method and system based on conditional random field (CRF) | |
| CN103927360A (en) | Software project semantic information presentation and retrieval method based on graph model | |
| BR112017019015A2 (en) | system that facilitates the use of user-entered keywords to search for related clinical concepts, and method for facilitating the use of user-entered keywords to search for related clinical concepts | |
| CN102982010A (en) | Method and device for abstracting document structure | |
| US20110270882A1 (en) | Resource description framework network construction device and method using an ontology schema having class dictionary and mining rule | |
| KR102100951B1 (en) | System for generating question-answer data for maching learning based on maching reading comprehension | |
| CA2610208A1 (en) | Learning facts from semi-structured text | |
| CN102768681A (en) | Recommending system and method used for search input | |
| CN107992608B (en) | SPARQL query statement automatic generation method based on keyword context | |
| CN103383699A (en) | Character string retrieval method and system | |
| CN106095771A (en) | Writing householder method and device | |
| CN102063493A (en) | Content extraction method based on regular expression group and control logic | |
| CN107193882A (en) | Why not query answer methods based on figure matching on RDF data | |
| CN110532358A (en) | A kind of template automatic generation method towards knowledge base question and answer | |
| CN107656921A (en) | A kind of short text dependency analysis method based on deep learning | |
| CN102968431A (en) | Control device for mining relation between Chinese entities on basis of dependency tree | |
| CN105740235A (en) | Phrase tree to dependency tree transformation method capable of combining Vietnamese grammatical features | |
| CN104142974A (en) | Voice file querying method and device | |
| US20120324346A1 (en) | Method for relational analysis of parsed input for visual mapping of knowledge information | |
| CN104536951A (en) | Microblog text normalizing, word segmenting and part-speech tagging method and system | |
| CN103577398B (en) | A kind of level machine translation method and system based on predicate-argument structure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161123 |