CN108829696A - Towards knowledge mapping node method for auto constructing in metro design code - Google Patents
Towards knowledge mapping node method for auto constructing in metro design code Download PDFInfo
- Publication number
- CN108829696A CN108829696A CN201810349660.9A CN201810349660A CN108829696A CN 108829696 A CN108829696 A CN 108829696A CN 201810349660 A CN201810349660 A CN 201810349660A CN 108829696 A CN108829696 A CN 108829696A
- Authority
- CN
- China
- Prior art keywords
- sentence
- text
- predicate
- node
- word
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
Abstract
The present invention provides one kind towards knowledge mapping node method for auto constructing in metro design code, will《Metro design code》In Subject, Predicate and Object clause specification as input text, around predicate by every specification in subject object mark off come, and generate for stored knowledge map graphic data base Neo4j node create sentence, thus for building the service of construction standards knowledge mapping.The present invention pre-processes predicate dictionary using Hash dictionary, to promote the efficiency for searching label;And after inputting construction standards, can automatically generate knowledge mapping node statement, it is time-saving simultaneously, substantially increase accuracy when creation of knowledge map.
Description
Technical field
The invention belongs to Computer Natural Language Processing technology technical fields, and in particular to towards knowing in metro design code
Know map node method for auto constructing.
Background technique
With the development of computer science and technology, the building of knowledge mapping all trades and professions all have very much realistic meaning and
Application prospect.Knowledge mapping is the description of relationship between entity and entity, releases oneself first edition knowledge graph from Google in 2012
Since spectrum, the whole world has started one upsurge.It the use of the industry field of knowledge mapping include at present search engine, question answering system etc.
Deng.But the informatization of building trade is still in infancy, and the engineering project of a building trade is firstly the need of designing institute
It planned, designed, then designed model and drawing are examined, be provided to unit in charge of construction and construct, most
After come into operation.The construction stage can be reduced to a great extent because the work holdup of doing over again caused by design alteration is existing in accurate drawing and model
As, but it is now very few for the knowledge mapping building of construction specification, therefore traditional architecture industry check of drawings is mostly expert
No matter mode, manual operation, all there is very big problem in time efficiency or accuracy.
Summary of the invention
The object of the present invention is to provide one kind towards knowledge mapping node method for auto constructing in metro design code, has
Metro design code is converted to knowledge mapping node, the spy of automatic compliance inspection is designed convenient for underground railway track traffic engineering
Point.
The technical scheme adopted by the invention is that the knowledge mapping node towards Subject, Predicate and Object clause in metro design code is certainly
Dynamic construction method, which is characterized in that the predicate in construction standards is first stored in Hash dictionary, further according to dictionary, to sentence to be processed
Type structure is that the metro design code of Subject, Predicate and Object form carries out predicate label and extracts the operation of subject object, is ultimately produced
Neo4j database node and its relationship create sentence, include the following steps:
Step 1, predicate Hash dictionary index is constructed using lexicon file;
Step 2, it takes《Metro design code》Subject, Predicate and Object clause specification is as input text S1 to be processed in text;
Step 3, according to predicate dictionary, to text S2 after S1 progress predicate part of speech marking operation output token
Step 4, the Hash dictionary index constructed according to step 1 carries out word to text S2 using reverse maximum matching algorithm
Property label processing, and export result;
Step 5, the division of subject object is carried out to the S2 after label;
Step 6, it generates Neo4j subject node and creates sentence, generate Neo4j object node and create sentence;
Step 7, it generates Neo4j relationship and constructs sentence, and export;
Predicate hash index is constructed in step 1 uses hash_map data structure.
Input text S1 is stored in an array in step 3, S1 [0] is first character;First according to ASCII character value
Space in input text S1, carriage return, line feed are identified, tentatively S1 is divided, and is made with space, carriage return, line feed
Multiple portions are divided into divide node for text S1 is inputted;Then it is encoded according to Chinese character GB2312 and is carried out using height region-position code
It again identifies that, and is divided again, and using Chinese symbol as division node.
Specific step is as follows for reverse maximum matching algorithm in step 4:
Step 4.1, in the text S1 handled through step 3, the text that first time Preliminary division obtains in step 3 is pressed
According to sequence from front to back, a sentence is obtained from first division points;
Step 4.2, if the sentence length obtained in step 4.1 is less than the long n of most major term, using the sentence as matching word
Section w, executes step 4.3;The long length of most major term is taken then since the rightmost side of this if more than or equal to the long n of most major term
Character string executes step 4.3 as matching field w;
Step 4.3, the lexicon file in finding step 1, according in hash index judgment step 4.2 gained w whether
In dictionary, if containing the word in lexicon file, successful match is marked processing to w and exports to S2, and by w from sentence
Middle removal, then remaining sentence is repeated into step 4.2;If it does not exist, 4.4 are thened follow the steps;
Step 4.4, the leftmost side word of matching field w is rejected, uses the field of remaining n-1 word composition as new
With field w, step 4.2 is executed repeatedly, such as rejects to the also non-successful match of single word, then rejects the word from sentence, until
Sentence is sky;
After the completion of step 4.5 handles a sentence, which is rejected from S1, in remaining S1 text, according to
Sequence from front to back obtains a new sentence from first division points;
Step 4.6, step 4.2~4.5 are repeated, until S1 is sky, final output S2.
Specific step is as follows for subject object division part in step 5:
Retrtieval S2 in step 4 is stored in an array by step 5.1, and S2 [0] is first character;According to
ASCII character value identifies the space in input text S2, carriage return, line feed, and input text S2 is divided for multiple portions;Then
Encoded according to Chinese character GB2312 and again identified that using height region-position code, and divided again, and using Chinese symbol as
Divide node.
Treated text according to sequence from front to back, is obtained a sentence by step 5.2 from first division points;
The sentence is from left to right traversed according to ASCII character value, until finding predicate label symbol " { " " } ", " { before " symbol
Ingredient as subject part S3, " } " the subsequent ingredient of symbol is as object part S4, after the completion of processing by the sentence from S2
The step for rejecting, constantly repeating exports S3 and S4 until S2 is sky.
Neo4j subject node is generated in step 6 and creates sentence, generates the specific steps of Neo4j object node creation sentence
For:
Step 6.1 is read through step 5 treated text S3, S4, and S3 and S4 is stored in the very big number of subscript respectively
In group, S3 [0] and S4 [0] are first character;Carriage return, the line feed in input text S3, S4 are carried out according to ASCII character value
Identification will input text S3, and S4 points are multiple portions;
Step 6.2, according to sequence from front to back, divides in the text S3 and S4 handled through step 6.1 from first
A word xx is obtained at point, and is rewritten into " CREATE (xx { name:" xx " }) " form, and output this to S5 and
S6.Then the word is rejected from S3 and S4, constantly executes the step, until S3 and S4 is sky, export S5 and S6.
In step 7:Generating Neo4j relationship building sentence, specific step is as follows:
Step 7.1:For each sentence in step 5, the subject part S3 of sentence each in step 5 is denoted as Z1, and
It is rewritten into the form of " CREATE (Z1) ", and Z1 is rejected from sentence.
Step 7.2:By sentence each in step 5 ":" symbol and " " part between symbol, i.e. predicate part is denoted as V1,
The part is extracted as the relationship between node, be rewritten into "-[:V1] " form, and V1 is rejected from sentence.
Step 7.3:Object part S4 in step 5 is denoted as B1, and be rewritten into "->(B1) " form, and by B1
It is rejected from sentence.
Step 7.4:The ingredient that above-mentioned steps generate is spliced, is exported as relationship node generated statement to S7.
Step 7.5:Judge S2 whether be it is empty, if not empty, then the step of executing step 7.1 to 7.4 repeatedly;If S2 is
Sky then exports knowledge mapping node Neo4j relationship building sentence S7.
The beneficial effects of the invention are as follows:
The present invention carries out the division of guest of honour's language around predicate, has certain versatility for different construction standards, this
Invention, which can automate, generates Neo4j node statement, is the power-assisted agent that knowledge mapping is established, the present invention has effectively evaded artificial
The uncertain factor being likely to occur during check of drawings, false detection rate is low, easy to operate, saves manpower, greatly improves subway and build
Make the completion efficiency of engineering project.
Detailed description of the invention
Fig. 1 be the present invention towards in metro design code the knowledge mapping node method for auto constructing of Subject, Predicate and Object clause it is total
Flow chart;
Fig. 2 is the flow chart of predicate mark part of the invention;
Fig. 3 is the flow chart that subject object of the invention divides part;
Fig. 4 is the flow chart for generating Neo4j node and creating statement part of the invention;
Fig. 5 is the flow chart for generating Neo4j relationship and creating statement part of the invention.
Specific embodiment
The following describes the present invention in detail with reference to the accompanying drawings and specific embodiments.
The present invention is towards knowledge mapping node method for auto constructing in metro design code, first by the predicate in construction standards
It is stored in Hash dictionary, further according to dictionary, predicate mark is carried out to the metro design code that sentence pattern structure to be processed is Subject, Predicate and Object form
The operation for remembering and extracting subject object ultimately produces Neo4j database node and its relationship creation sentence, flow chart such as Fig. 1
Shown, from one specification of input sentence, arrives output relation node sentence, includes the following steps:
Step 1, predicate Hash dictionary index is constructed using lexicon file;
Step 2, it takes《Metro design code》Subject, Predicate and Object clause specification is as input text S1 to be processed in text;
Step 3, according to predicate dictionary, to text S2 after S1 progress predicate part of speech marking operation output token
Step 4, the Hash dictionary index constructed according to step 1 carries out word to text S2 using reverse maximum matching algorithm
Property label processing, and export result;
Step 5, the division of subject object is carried out to the S2 after label;
Step 6, it generates Neo4j subject node and creates sentence, generate Neo4j object node and create sentence;
Step 7, it generates Neo4j relationship and constructs sentence, and export;
Predicate hash index is constructed in step 1 uses hash_map data structure.
In step 3, as shown in Fig. 2, carrying out part of speech label to the predicate in specification, input text S1 is stored in an array
In, S1 [0] is first character;First the space in input text S1, carriage return, line feed are identified according to ASCII character value,
Tentatively S1 is divided, and is divided into multiple portions for text S1 is inputted using space, carriage return, line feed as division node;So
It is encoded according to Chinese character GB2312 and is again identified that using height region-position code afterwards, and divided again, and with Chinese symbol work
To divide node.
Specific step is as follows for reverse maximum matching algorithm in step 4:
Step 4.1, in the text S1 handled through step 3, the text that first time Preliminary division obtains in step 3 is pressed
According to sequence from front to back, a sentence is obtained from first division points;
Step 4.2, if the sentence length obtained in step 4.1 is less than the long n of most major term, using the sentence as matching word
Section w, executes step 4.3;The long length of most major term is taken then since the rightmost side of this if more than or equal to the long n of most major term
Character string executes step 4.3 as matching field w;
Step 4.3, the lexicon file in finding step 1, according in hash index judgment step 4.2 gained w whether
In dictionary, if containing the word in lexicon file, successful match is marked processing to w and exports to S2, and by w from sentence
Middle removal, then remaining sentence is repeated into step 4.2;If it does not exist, 4.4 are thened follow the steps;
Step 4.4, the leftmost side word of matching field w is rejected, uses the field of remaining n-1 word composition as new
With field w, step 4.2 is executed repeatedly, such as rejects to the also non-successful match of single word, then rejects the word from sentence, until
Sentence is sky;
After the completion of step 4.5 handles a sentence, which is rejected from S1, in remaining S1 text, according to
Sequence from front to back obtains a new sentence from first division points;
Step 4.6, step 4.2~4.5 are repeated, until S1 is sky, final output S2.
In step 5, as shown in figure 3, the subject object in sentence is carried out Preliminary division, and subject and object are exported, subject
Object divides part, and specific step is as follows:
Retrtieval S2 in step 4 is stored in an array by step 5.1, and S2 [0] is first character;According to
ASCII character value identifies the space in input text S2, carriage return, line feed, and input text S2 is divided for multiple portions;Then
Encoded according to Chinese character GB2312 and again identified that using height region-position code, and divided again, and using Chinese symbol as
Divide node.
Treated text according to sequence from front to back, is obtained a sentence by step 5.2 from first division points;
The sentence is from left to right traversed according to ASCII character value, until finding predicate label symbol " { " " } ", " { before " symbol
Ingredient as subject part S3, " } " the subsequent ingredient of symbol is as object part S4, after the completion of processing by the sentence from S2
The step for rejecting, constantly repeating exports S3 and S4 until S2 is sky.
In step 6, as shown in figure 4, generating Neo4j subject node by the S3 and S4 that are generated in input step 5 and creating language
Sentence, generate Neo4j object node creation sentence the specific steps are:
Step 6.1 is read through step 5 treated text S3, S4, and S3 and S4 is stored in the very big number of subscript respectively
In group, S3 [0] and S4 [0] are first character;Carriage return, the line feed in input text S3, S4 are carried out according to ASCII character value
Identification will input text S3, and S4 points are multiple portions;
Step 6.2, according to sequence from front to back, divides in the text S3 and S4 handled through step 6.1 from first
A word xx is obtained at point, and is rewritten into " CREATE (xx { name:" xx " }) " form, and output this to S5 and
S6.Then the word is rejected from S3 and S4, constantly executes the step, until S3 and S4 is sky, export S5 and S6.
In step 7, as shown in figure 5, creating sentence by input node, the specific step of Neo4j relationship building sentence is generated
It is rapid as follows:
Step 7.1:For each sentence in step 5, the subject part S3 of sentence each in step 5 is denoted as Z1, and
It is rewritten into the form of " CREATE (Z1) ", and Z1 is rejected from sentence.
Step 7.2:By sentence each in step 5 ":" symbol and " " part between symbol, i.e. predicate part is denoted as V1,
The part is extracted as the relationship between node, be rewritten into "-[:V1] " form, and V1 is rejected from sentence.
Step 7.3:Object part S4 in step 5 is denoted as B1, and be rewritten into "->(B1) " form, and by B1
It is rejected from sentence.
Step 7.4:The ingredient that above-mentioned steps generate is spliced, is exported as relationship node generated statement to S7.
Step 7.5:Judge S2 whether be it is empty, if not empty, then the step of executing step 7.1 to 7.4 repeatedly;If S2 is
Sky then exports knowledge mapping node Neo4j relationship building sentence S7.
Embodiment
《Metro design code》In 28.4.2 entry " smoke removal facility should be arranged in underground station platform " this specification carry out
Processing.
First this is standardized and carries out part of speech marking operation.Specific step is as follows:
Assuming that predicate most major term a length of 6 in dictionary, the predicate in dictionary relevant to this specification is " should be arranged ", output
Text is S2, and label template is " { v:Should be arranged "
(1) S2=" ";S1 is not sky, takes out candidate character strings w=" setting smoke removal facility " from the rightmost side S1;
(2) it consults the dictionary, one word of w Far Left is removed not in dictionary, obtain w=" setting smoke removal facility " by w;
(3) it consults the dictionary, one word of w Far Left is removed not in dictionary, obtain w=" smoke removal facility " by w;
(4) it consults the dictionary, w removes one word of w Far Left not in dictionary, obtains w=" cigarette facility ";
(5) it consults the dictionary, one word of w Far Left is removed not in dictionary, obtain w=" facility " by w;
(6) it consults the dictionary, one word of w Far Left is removed not in dictionary, obtain w=" applying " by w;
(7) w is individual Chinese character at this time, and w then subtracts w not in dictionary from S1, and " smoke evacuation should be arranged in S1=at this time
Facility "
(8) operation for repeating 1-7, until w=" should be arranged ";
(9) it consults the dictionary, operation is marked in dictionary, to w in " should be arranged ", at this time S2=" { v:Should be arranged ribband
Material ";
It repeats the above steps, last S2=" underground station platform { v can be obtained:Should be arranged smoke removal facility "
Table 1 is the test case of part of speech label
It is described in table 1 with " smoke removal facility should be arranged in underground station platform " as experimental subjects, carries out part of speech label
Result after treatment process and label.
After labeled, this is standardized and carries out subject object division operation.Specific step is as follows:
Specification " underground station platform { the v after label:Should be arranged } smoke removal facility " the leftmost side begin stepping through, until finding
" { " symbol, " { " underground station platform " is the subject part of this specification on the left of " symbol, draws it out and is output at this time
In the text document for storing subject.Inverted order is begun stepping through from the rightmost side of the specification again, until find " " symbol, at this time " " symbol
" smoke abatement facility " is the object part of this specification on the right side of number, draws it out and is output in the text document of storage object.
The subject object division of this specification is partially completed at this time.
After division, subject and object to this specification carry out Neo4j node statement and generate operation.Specific steps are such as
Under:
The output file for reading subject and object respectively, with behavior unit, the reading of content of a line a line, and is rewritten
At " CREATE (xx { name:" xx " }) " form, for this specification for, modification result be " (the underground station station CREATE
Platform { name:" underground station platform " }) " and " CREATE (smoke abatement facility { name:" smoke abatement facility " }) "
Text after marking is read again, Neo4j relationship creation sentence is carried out to the sentence and generates operation.Specific steps are such as
Under:
Specification " underground station platform { the v after label:Should be arranged } smoke removal facility " the leftmost side begin stepping through, until finding
" " symbol, will " " symbol left part, i.e. " underground station platform " are rewritten into the form of " CREATE (underground station platform) ",
Be then followed by traversal, until find ":" symbol, then start to read, will ":" symbol and " " between part, i.e., " should be arranged "
Be rewritten into "-[:Should be arranged] " form, finally will " " right part be rewritten into "->The form of (smoke removal facility) ", then by three
Part is spliced, and last relationship creation sentence is generated:" CREATE (underground station platform)-[:Should be arranged]->(smoke evacuation is set
Apply) "
Table 2 is to store test case after generating node statement
One kind of the invention carries out the guest of honour around predicate towards knowledge mapping node method for auto constructing in metro design code
The division of language has certain versatility for different construction standards, and the present invention, which can automate, generates Neo4j node statement,
Be knowledge mapping establish power-assisted agent, the present invention effectively evaded be likely to occur during artificial check of drawings it is uncertain because
Element, false detection rate is low, easy to operate, saves manpower, greatly improves the completion efficiency of subway construction engineering project.
Claims (7)
1. towards knowledge mapping node method for auto constructing in metro design code, which is characterized in that first will be in construction standards
Predicate is stored in Hash dictionary, further according to dictionary, calls to the metro design code that sentence pattern structure to be processed is Subject, Predicate and Object form
Word marks and extracts the operation of subject object, ultimately produces Neo4j database node and its relationship creation sentence, including following step
Suddenly:
Step 1, predicate Hash dictionary index is constructed using lexicon file;
Step 2, it takes《Metro design code》Subject, Predicate and Object clause specification is as input text S1 to be processed in text;
Step 3, according to predicate dictionary, to text S2 after S1 progress predicate part of speech marking operation output token;
Step 4, the Hash dictionary index constructed according to step 1 carries out part of speech mark to text S2 using reverse maximum matching algorithm
Note processing, and export result;
Step 5, the division of subject object is carried out to the S2 after label;
Step 6, it generates Neo4j subject node and creates sentence, generate Neo4j object node and create sentence;
Step 7, it generates Neo4j relationship and constructs sentence, and export.
2. according to claim 1 exist towards knowledge mapping node method for auto constructing, feature in metro design code
In building predicate hash index uses hash_map data structure in step 1.
3. according to claim 1 exist towards knowledge mapping node method for auto constructing, feature in metro design code
In by input text S1 one array of deposit in step 3, S1 [0] is first character;First according to ASCII character value to defeated
Enter the space in text S1, carriage return, line feed are identified, tentatively S1 is divided, and using space, carriage return, line feed as stroke
Node is divided to be divided into multiple portions for text S1 is inputted;Then it is encoded according to Chinese character GB2312 and is carried out again using height region-position code
Identification, and divided again, and using Chinese symbol as division node.
4. according to claim 1 exist towards knowledge mapping node method for auto constructing, feature in metro design code
In specific step is as follows for reverse maximum matching algorithm in step 4:
Step 4.1, in the text S1 handled through step 3, by the text that first time Preliminary division obtains in step 3, according to from
Sequence after going to obtains a sentence from first division points;
Step 4.2, if the sentence length obtained in step 4.1 is less than the long n of most major term, using the sentence as matching field w,
Execute step 4.3;The character of the long length of most major term is taken then since the rightmost side of this if more than or equal to the long n of most major term
String is used as matching field w, executes step 4.3;
Step 4.3, the lexicon file in finding step 1, according in hash index judgment step 4.2 gained w whether in dictionary
In, if containing the word in lexicon file, successful match is marked processing to w and exports to S2, and by w from sentence
It removes, then remaining sentence is repeated into step 4.2;If it does not exist, 4.4 are thened follow the steps;
Step 4.4, the leftmost side word of matching field w is rejected, uses the field of remaining n-1 word composition as new matching word
Section w executes step 4.2, such as rejects to the also non-successful match of single word, then reject the word from sentence, repeatedly until sentence
For sky;
After the completion of step 4.5 handles a sentence, which is rejected from S1, in remaining S1 text, according in the past
Sequence backward obtains a new sentence from first division points;
Step 4.6, step 4.2~4.5 are repeated, until S1 is sky, final output S2.
5. according to claim 1 exist towards knowledge mapping node method for auto constructing, feature in metro design code
In specific step is as follows for subject object division part in step 5:
Retrtieval S2 in step 4 is stored in an array by step 5.1, and S2 [0] is first character;According to ASCII character
Value identifies the space in input text S2, carriage return, line feed, and input text S2 is divided for multiple portions;Then according to the Chinese
Word GB2312 coding is again identified that using height region-position code, and is divided again, and is tied using Chinese symbol as division
Point;
Treated text according to sequence from front to back, is obtained a sentence by step 5.2 from first division points;To this
Sentence is from left to right traversed according to ASCII character value, until finding predicate label symbol " { " " } ", " before " symbol at
It is allocated as subject part S3, " " the subsequent ingredient of symbol as object part S4, picks the sentence after the completion of processing from S2
The step for removing, constantly repeating exports S3 and S4 until S2 is sky.
6. according to claim 1 exist towards knowledge mapping node method for auto constructing, feature in metro design code
In, in step 6 generate Neo4j subject node create sentence, generate Neo4j object node creation sentence the specific steps are:
Step 6.1 is read through step 5 treated text S3, S4, S3 and S4 is stored in respectively in an array, S3 [0] and S4
It [0] is first character;Carriage return, the line feed in input text S3, S4 are identified according to ASCII character value, by input text
This S3, S4 points are multiple portions;
Step 6.2 is in the text S3 and S4 handled through step 6.1, according to sequence from front to back, from first division points
A word xx is obtained, and is rewritten into " CREATE (xx { name:" xx " }) " form, and output this to S5 and S6.So
The word is rejected from S3 and S4 afterwards, constantly executes the step, until S3 and S4 is sky, exports S5 and S6.
7. according to claim 1 exist towards knowledge mapping node method for auto constructing, feature in metro design code
In in step 7:Generating Neo4j relationship building sentence, specific step is as follows:
Step 7.1:For each sentence in step 5, the subject part S3 of sentence each in step 5 is denoted as Z1, and by its
It is rewritten into the form of " CREATE (Z1) ", and Z1 is rejected from sentence;
Step 7.2:By sentence each in step 5 ":" symbol and " " part between symbol, i.e. predicate part is denoted as V1, extracts
The part as the relationship between node, be rewritten into "-[:V1] " form, and V1 is rejected from sentence;
Step 7.3:Object part S4 in step 5 is denoted as B1, and be rewritten into "->(B1) " form, and by B1 subordinate clause
It is rejected in son;
Step 7.4:The ingredient that above-mentioned steps generate is spliced, is exported as relationship node generated statement to S7;
Step 7.5:Judge S2 whether be it is empty, if not empty, then the step of executing step 7.1 to 7.4 repeatedly;If S2 is sky,
It exports knowledge mapping node Neo4j relationship and constructs sentence S7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810349660.9A CN108829696B (en) | 2018-04-18 | 2018-04-18 | Towards knowledge mapping node method for auto constructing in metro design code |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810349660.9A CN108829696B (en) | 2018-04-18 | 2018-04-18 | Towards knowledge mapping node method for auto constructing in metro design code |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108829696A true CN108829696A (en) | 2018-11-16 |
CN108829696B CN108829696B (en) | 2019-10-25 |
Family
ID=64154771
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810349660.9A Active CN108829696B (en) | 2018-04-18 | 2018-04-18 | Towards knowledge mapping node method for auto constructing in metro design code |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108829696B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110046348A (en) * | 2019-03-19 | 2019-07-23 | 西安理工大学 | Main body recognition methods in a kind of rule-based and dictionary metro design code |
CN110347798A (en) * | 2019-07-12 | 2019-10-18 | 之江实验室 | A kind of knowledge mapping auxiliary understanding system based on spatial term technology |
CN111813950A (en) * | 2020-05-20 | 2020-10-23 | 淮阴工学院 | Building field knowledge graph construction method based on neural network self-adaptive optimization and parameter adjustment |
CN112948510A (en) * | 2021-04-21 | 2021-06-11 | 央视国际网络无锡有限公司 | Construction method of knowledge graph in media industry |
CN113032862A (en) * | 2020-07-27 | 2021-06-25 | 深圳市前海数字城市科技有限公司 | Building information model checking method and device and terminal equipment |
CN114004222A (en) * | 2021-11-04 | 2022-02-01 | 中铁第一勘察设计院集团有限公司 | Chinese word segmentation boundary correction method based on frequent items |
CN114004222B (en) * | 2021-11-04 | 2024-04-30 | 中铁第一勘察设计院集团有限公司 | Chinese word segmentation boundary correction method based on frequent items |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103646018A (en) * | 2013-12-20 | 2014-03-19 | 大连大学 | Chinese word segmentation method based on hash table dictionary structure |
US9311302B2 (en) * | 2012-06-19 | 2016-04-12 | City University Of Hong Kong | Method, system and medium for character conversion between different regional versions of a language especially between simplified chinese and traditional chinese |
US20160179901A1 (en) * | 2014-12-23 | 2016-06-23 | Palo Alto Research Center Incorporated | Computer-Implemented System And Method For Providing Selective Contextual Exposure Within Social Network Situations |
CN106815293A (en) * | 2016-12-08 | 2017-06-09 | 中国电子科技集团公司第三十二研究所 | System and method for constructing knowledge graph for information analysis |
CN106919689A (en) * | 2017-03-03 | 2017-07-04 | 中国科学技术信息研究所 | Professional domain knowledge mapping dynamic fixing method based on definitions blocks of knowledge |
CN107590219A (en) * | 2017-09-04 | 2018-01-16 | 电子科技大学 | Webpage personage subject correlation message extracting method |
CN107784088A (en) * | 2017-09-30 | 2018-03-09 | 杭州博世数据网络有限公司 | The knowledge mapping construction method of knowledge based point annexation |
-
2018
- 2018-04-18 CN CN201810349660.9A patent/CN108829696B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9311302B2 (en) * | 2012-06-19 | 2016-04-12 | City University Of Hong Kong | Method, system and medium for character conversion between different regional versions of a language especially between simplified chinese and traditional chinese |
CN103646018A (en) * | 2013-12-20 | 2014-03-19 | 大连大学 | Chinese word segmentation method based on hash table dictionary structure |
US20160179901A1 (en) * | 2014-12-23 | 2016-06-23 | Palo Alto Research Center Incorporated | Computer-Implemented System And Method For Providing Selective Contextual Exposure Within Social Network Situations |
CN106815293A (en) * | 2016-12-08 | 2017-06-09 | 中国电子科技集团公司第三十二研究所 | System and method for constructing knowledge graph for information analysis |
CN106919689A (en) * | 2017-03-03 | 2017-07-04 | 中国科学技术信息研究所 | Professional domain knowledge mapping dynamic fixing method based on definitions blocks of knowledge |
CN107590219A (en) * | 2017-09-04 | 2018-01-16 | 电子科技大学 | Webpage personage subject correlation message extracting method |
CN107784088A (en) * | 2017-09-30 | 2018-03-09 | 杭州博世数据网络有限公司 | The knowledge mapping construction method of knowledge based point annexation |
Non-Patent Citations (2)
Title |
---|
王仁武等: "基于深度学习与图数据库构建中文商业知识图谱的探索研究", 《图书与情报》 * |
甘丽新等: "基于句法语义特征的中文实体关系抽取", 《计算机研究与发展》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110046348A (en) * | 2019-03-19 | 2019-07-23 | 西安理工大学 | Main body recognition methods in a kind of rule-based and dictionary metro design code |
CN110046348B (en) * | 2019-03-19 | 2021-05-25 | 西安理工大学 | Method for recognizing main body in subway design specification based on rules and dictionaries |
CN110347798A (en) * | 2019-07-12 | 2019-10-18 | 之江实验室 | A kind of knowledge mapping auxiliary understanding system based on spatial term technology |
CN110347798B (en) * | 2019-07-12 | 2021-06-01 | 之江实验室 | Knowledge graph auxiliary understanding system based on natural language generation technology |
CN111813950A (en) * | 2020-05-20 | 2020-10-23 | 淮阴工学院 | Building field knowledge graph construction method based on neural network self-adaptive optimization and parameter adjustment |
CN111813950B (en) * | 2020-05-20 | 2024-02-27 | 淮阴工学院 | Building field knowledge graph construction method based on neural network self-adaptive optimization tuning |
CN113032862A (en) * | 2020-07-27 | 2021-06-25 | 深圳市前海数字城市科技有限公司 | Building information model checking method and device and terminal equipment |
CN112948510A (en) * | 2021-04-21 | 2021-06-11 | 央视国际网络无锡有限公司 | Construction method of knowledge graph in media industry |
CN112948510B (en) * | 2021-04-21 | 2024-02-20 | 央视国际网络无锡有限公司 | Construction method of knowledge graph in media industry |
CN114004222A (en) * | 2021-11-04 | 2022-02-01 | 中铁第一勘察设计院集团有限公司 | Chinese word segmentation boundary correction method based on frequent items |
CN114004222B (en) * | 2021-11-04 | 2024-04-30 | 中铁第一勘察设计院集团有限公司 | Chinese word segmentation boundary correction method based on frequent items |
Also Published As
Publication number | Publication date |
---|---|
CN108829696B (en) | 2019-10-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108829696B (en) | Towards knowledge mapping node method for auto constructing in metro design code | |
US10248885B2 (en) | Method for encoding of anatomic curves | |
CN109492077A (en) | The petrochemical field answering method and system of knowledge based map | |
CN105138335B (en) | A kind of function call path extraction method and device based on controlling stream graph | |
CN104503998B (en) | For the kind identification method and device of user query sentence | |
CN110413319B (en) | Code function taste detection method based on deep semantics | |
CN109492106B (en) | Automatic classification method for defect reasons by combining text codes | |
CN105095196B (en) | The method and apparatus of new word discovery in text | |
CN107807910A (en) | A kind of part-of-speech tagging method based on HMM | |
CN110046348A (en) | Main body recognition methods in a kind of rule-based and dictionary metro design code | |
CN110765235A (en) | Training data generation method and device, terminal and readable medium | |
CN109710933A (en) | Acquisition methods, device, computer equipment and the storage medium of training corpus | |
CN115905553A (en) | Construction drawing inspection specification knowledge extraction and knowledge graph construction method and system | |
WO2016112782A1 (en) | Method and system of extracting user living range | |
CN112613321A (en) | Method and system for extracting entity attribute information in text | |
CN113420548A (en) | Entity extraction sampling method based on knowledge distillation and PU learning | |
CN108563637A (en) | A kind of sentence entity complementing method of fusion triple knowledge base | |
CN113656547A (en) | Text matching method, device, equipment and storage medium | |
CN109446704B (en) | Product family finite element model parameterization method based on regular grammar | |
CN107590119A (en) | Character attribute information extraction method and device | |
CN109359701A (en) | A kind of three-dimensional modeling data analytic method of extracted with high accuracy and Fast Classification | |
CN108536724A (en) | Main body recognition methods in a kind of metro design code based on the double-deck hash index | |
CN107220238A (en) | A kind of text object abstracting method based on Mixed Weibull distribution | |
CN106598935A (en) | Method and apparatus for determining emotional tendency of document | |
CN113849621A (en) | Intelligent pushing method for civil colleague case data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |