CN108536720A - Conceive assisting system and design support method - Google Patents
Conceive assisting system and design support method Download PDFInfo
- Publication number
- CN108536720A CN108536720A CN201711235125.2A CN201711235125A CN108536720A CN 108536720 A CN108536720 A CN 108536720A CN 201711235125 A CN201711235125 A CN 201711235125A CN 108536720 A CN108536720 A CN 108536720A
- Authority
- CN
- China
- Prior art keywords
- mentioned
- chart
- relationship
- knowledge
- node
- 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
- G06F40/205—Parsing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Human Computer Interaction (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
Design assisting system and design support method are provided, during user writes a composition or talks, the design of user is supported along the composition of user or the context of speech.Conceive assisting system and keeps following information：Indicate that the information of the construction of the 1st chart comprising node and arrow, node indicate to be formed in the element for the composition proposition for including in document, document specification knowledge, arrow indicates the relationship between the element represented by the proposition；And the information of the appearance sequence of the relationship in expression the paperwork；The device accepts the input of proposition；Extract the relationship between the element for including in the proposition and the element of proposition expression；The 2nd chart is generated, the 2nd chart includes to indicate the node of extracted element and include as arrow using the relationship between the element extracted；In the case where being determined as that the relationship of the 2nd graph representation is contained in the 1st chart, the relationship occurred rearward than the relationship in the paperwork is determined.
Description
Technical field
The present invention relates to design assisting system and design support methods.
Background technology
As the background technology of the art, there are special open 2007-241901 bulletins (patent document 1).In the bulletin
In recorded：" have：Argument extraction mechanism, extraction and above-mentioned theme pass from the theme of input associated opinion text group
Multiple arguments of connection；Intrinsic degree calculates mechanism, by each argument of above-mentioned multiple arguments, calculates intrinsic degree, which indicates
The ratio of the opinion with the position consistent with the argument in the above views text comprising the argument；Importance computer
Structure calculates the importance of the opinion comprising above-mentioned consistent position by each argument of above-mentioned multiple arguments；It is associated with language extractor
Structure, extraction and above-mentioned multiple associated association languages of argument；Opinion selection mechanism is represented, selection is for each of above-mentioned multiple arguments
Multiple in the opinion of the positive or negative of argument represent opinion；And interface agency, it exports from the upper of above-mentioned each mechanism output
State the above-mentioned intrinsic degree of each argument of multiple arguments and above-mentioned importance." (with reference to abstract of description).
Patent document 1：Japanese Unexamined Patent Publication 2007-241901 bulletins
Invention content
For example, it is envisioned that user will write a composition or talk about given subject write.Technology described in patent document 1 passes through pre-
The argument about given theme is first extracted, the composition or speech of user are supported.
But in the case where user has carried out composition or speech, each argument of the technology extraction described in patent document 1
It might not be suitable for the argument of the composition or the context of the session.So the object of the present invention is to provide one kind with
During text or speech are being write in family, the design of user is supported along the composition of user or the context of speech.
In order to solve the above problems, a technical solution of the invention uses structure below.A kind of design assisting system, branch
Quote the design at family, including processor and memory；Above-mentioned memory keeps following information：Chart-information indicates to include node
With the construction of the 1st chart of the 1st kind of arrow, above-mentioned node indicates the element for being formed in the proposition for including in document, above-mentioned document
Knowledge is defined, above-mentioned 1st kind of arrow indicates the relationship between the above-mentioned element represented by above-mentioned proposition；And contextual information, table
Show the appearance sequence of the relationship in above-mentioned document；In above-mentioned processor, the input of proposition is accepted；Extraction has accepted the above-mentioned of input
The element that includes in proposition and the relationship between the element represented by the above-mentioned proposition of input is accepted；The 2nd chart is generated, it should
2nd chart includes the node for the above-mentioned element for indicating to be extracted, and using the relationship between the above-mentioned element extracted as above-mentioned
1st kind of arrow include；It is being determined as that the relationship represented by above-mentioned 2nd chart is contained in the above-mentioned 1st with reference to above-mentioned chart-information
In the case of in chart, with reference to above-mentioned contextual information, the relationship occurred rearward than the relationship in above-mentioned document is determined；Output
The information of above-mentioned relation determined by indicating.
Invention effect
The present invention a technical solution can user writing text or speech during, along user composition or
The context of speech supports the design of user.
Project, structure and effect other than the above can be clearer according to the explanation of the following embodiments and the accompanying drawings.
Description of the drawings
Fig. 1 is the block diagram of the configuration example for the design assisting system for indicating embodiment 1.
Fig. 2 is the flow chart of an example for the processing for indicating that the design assisting system of the present embodiment is implemented.
Fig. 3 A are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 3 B are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 3 C are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 3 D are an examples of the node table of embodiment 1.
Fig. 3 E are an examples of the arrow table of embodiment 1.
Fig. 4 A are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 4 B are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 4 C are an examples of the node table of embodiment 1.
Fig. 4 D are an examples of the arrow table of embodiment 1.
Fig. 5 A are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 5 B are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 5 C are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 5 D are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 5 E are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 6 A are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 6 B are an examples of the node table of embodiment 1.
Fig. 6 C are an examples of the arrow table of embodiment 1.
Fig. 7 A are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 7 B are an examples of the follow-up knowledge table of embodiment 1.
Fig. 8 A are an examples of the knowledge chart of the knowledge in the document for indicate embodiment 1.
Fig. 8 B are an examples of the follow-up knowledge table of embodiment 1.
Fig. 9 A are an examples of the knowledge chart generated during the speech understanding of embodiment 1 is handled.
Fig. 9 B are an examples of the knowledge chart generated during the speech understanding of embodiment 1 is handled.
Fig. 9 C are an examples of the transformation of the picture shown in the display device of embodiment 1.
Label declaration
103 document DB；105 knowledge chart DB；107 context DB；201 design assisting systems；202 input units；203 is aobvious
Showing device；204 communication devices；205 arithmetic units (CPU)；206 memories；207 auxilary units；211 knowledge charts generate
Portion；212 context calculating parts；213 speech understanding portions；214 context identification portions；215 stage identification parts.
Specific implementation mode
Hereinafter, being described with reference to embodiments of the present invention.It should be noted that present embodiment is only for realizing
An example of the present invention does not limit the technical scope of the present invention.
In order to cultivate expressive force and power to think etc., implement the training of (such as argumentative writing) composition in school etc..Composition
From can be different by the knowledge of the grasps such as workbook, it is difficult to learn by oneself, need teacher's to improve the composition technology of student
The situation of individual coaching=tutorial is more.In addition, student, which in the unfamiliar stage of writing a composition, also has, does not know that writes first good asks
Topic needs the situation of the guidance carefully of teacher more in order to solve this problem.In addition, in the training for the session for implementing foreign language
In the case of, also the same generation is not known says that the situation of what good situation is more first.
So present embodiment is by inspiring the composition of user prompt or comment, to support the design for writing text.
In addition same, present embodiment provides the inspiration etc. for next speech in foreign language session, come the auxiliary to conversate.
User can be familiar with foreign language session as a result, and then be easy the session of practice and people.The design assisting system of present embodiment is being made
In text or session the design of user, Jin Erzhi are supported by providing the inspiration of next speech to user, providing comment etc.
Quote the composition writing or session at family.
[embodiment 1]
Fig. 1 is the block diagram of the configuration example for the design assisting system for indicating the present embodiment.The design assisting system of the present embodiment
Such as by having input unit 202, display device 203, communication device 204, arithmetic unit (CPU) 205, memory 206 and auxiliary
The computer of storage device 207 is helped to constitute.
Input unit 202 accepts the input of order from the user etc., e.g. keyboard and mouse etc..Input unit 202
Accept the control of program in order to be executed by arithmetic unit (CPU) 205 and the control for the equipment being connect with design assisting system 201
And the input of order executed etc..Communication device 204 sends design assisting system for example according to the agreement of regulation to external equipment
201 process content receives information from external equipment.
Arithmetic unit (CPU) 205 includes processor, executes the program being stored in memory 206.Memory 206 includes
ROM as the non-volatile memory device and RAM as volatile memory elements.Program that ROM continues to have (such as
BIOS) etc..RAM is the memory element of the such high speeds of DRAM (Dynamic Random Access Memory) and volatibility,
The program executed by processor and the data used in the execution of program are temporarily preserved.
Auxilary unit 207 be, for example, the large capacity of magnetic memory apparatus (HDD), flash memories (SSD) etc. and it is non-easily
The storage device for the property lost preserves the program that arithmetic unit (CPU) 205 executes and the data used in the execution of program.That is, journey
Sequence is read from auxilary unit 207 and is loaded into memory 206, is executed by arithmetic unit (CPU) 205.
The program that arithmetic unit (CPU) 205 executes is via removable medium (CD-ROM, flash memories etc.) or network
It is provided to design assisting system 201, is saved to non-volatile auxilary unit as non-transitory storage medium
In 207.Therefore, design assisting system 201 is preferably with the interface for reading in data from removable medium.
Design assisting system 201 is the multiple calculating physically constituted on a computer or in logic or physically
The computer system constituted on machine, both can on the same computer with individual thread activity, can also be implemented in it is more
It is acted on virtual machine in a physical computer resource.
Arithmetic unit (CPU) 205 include knowledge chart generating unit 211, context calculating part 212, speech understanding portion 213,
Context identification portion 214 and stage identification part 215.For example, arithmetic unit (CPU) 205 is by according to being loaded onto memory 206
In the action of knowledge graph list generation program, functioned as knowledge chart generating unit 211, by according to being loaded onto memory
Context calculation procedure action in 206, functions as context calculating part 212.About in arithmetic unit (CPU) 205
Including other part be also same.
Knowledge chart generating unit 211 generates the chart (knowledge graph of the knowledge for the document expression for showing aftermentioned document DB103
Table).Context calculating part 212 calculates the knowledge represented by the knowledge chart that aftermentioned knowledge chart DB105 is kept in document
Appearance sequence in the document of DB103.Speech understanding portion 213 accepts the input of composition or session from user, make the composition or
The text data of session.
Context identification portion 214 calculates the appearance sequence of the knowledge in the text data made by speech understanding portion 213, will
Appearance sequence and the appearance sequence of 212 calculated knowledge of context calculating part are compared.Stage, identification part 215 was by sentencing
Whether the fixed preset content to composition or Dialog request is insufficient in text input by user, to identify the composition of user
Or the stage of session.
Auxilary unit 207 keeps document DB103, knowledge chart DB105 and context DB107.Document DB103 is kept
Document, particularly argumentative writing for being write with natural language etc..Since the document DB103 documents kept are supported for definition concept
The document of knowledge in device 201, i.e. as the document of knowledge source, so document DB103 preferably keeps recording conduct
The document of the event of the object of discussion or content associated with discussion.
Web documents, news story, report in tissue, scientific paper, administration and international body bulletin, Yi Jizhi
The investigation letter in library etc. is the example for the document that document DB103 is kept.In addition, the past speech resume etc. of user are also
The example for the document that document DB103 is kept.In addition, document DB103 can also also maintain the Bibliographic information etc. of each document.
Knowledge chart DB105 preserves the information of the construction for the knowledge chart for indicating that knowledge chart generating unit 211 generates.Up and down
Literary DB107 preserves the information for the appearance sequence for indicating 212 calculated knowledge of context calculating part.
In addition, in the present embodiment, the information that design assisting system 201 uses is not rely on data configuration, with why
The data configuration performance of sample can.For example, can be by the data that are properly selected from table, list, database or queue
Tectosome preserves information.In addition, in the present embodiment, expression table constructs each number that performance auxilary unit 207 is kept
According to example.
In addition, as long as design assisting system 201 has input unit 202, display device 203 and communication device 204 extremely
It is 1 few.In the case where conceiving assisting system 201 and not having input unit 202, such as communication device 204 is from outside
Equipment accepts the input of order etc..In the case where design assisting system 201 does not have display device 203, such as communication device
204 send the handling result that assisting system 201 generates is conceived to external equipment.
The input that each processing unit can also be executed via memory 206 or auxilary unit 207 to other processing units is defeated
Go out.For example, knowledge chart generating unit 211 actually also may be used in the case where exporting handling result to context calculating part 212
To be that handling result is saved in memory 206 or auxilary unit 207 by knowledge chart generating unit 211, context calculates
Portion 212 obtains the output result being stored in memory 206 or auxilary unit 207 as input.
Fig. 2 is the flow chart of an example for the processing for indicating that the design assisting system 201 of the present embodiment is implemented.Dress is supported in design
The processing for setting 201 execution generally divides, and can be divided into study processing (S110) and with processing this 2 parts (S120).Study
Processing includes the processing of step S111~S112, and operational phase includes the processing of step S121~S125.First, study is handled
(S110) processing illustrates.First, knowledge chart generating unit 211 will be by that will be stored in the letter of the document in document DB103
Breath is transformed to chart type knowledge, generates the knowledge chart (S111) preserved to knowledge chart DB105.Hereinafter, illustrating step S111
Details.
Fig. 3 A~Fig. 3 C are an examples for the knowledge chart for indicating the knowledge in document.In addition, in the present embodiment, using use
The example for the document that Japanese is recorded, even if being that other language such as English also implement same processing.Hereinafter, in chart in figure
The node indicated with the ellipse of substance line, be the reality of the existence or display abstract concept etc. that indicate people, object, tissue etc.
The node of body (entity).In addition, along the relationship between the character representation entity of the arrow record between entity.
Knowledge chart 301 be knowledge chart generating unit 211 according in the document of document DB103 " expert tells about, in school
The use of uniform can improve discipline." sentence generation knowledge chart.
Knowledge chart generating unit 211 extracts entity from sentence first in the generation of knowledge chart.Knowledge chart generates
For example entity is extracted using rote learning in portion 211.It is used in the study for physically carrying label for wanting extraction that is, preparing in advance
Document set, knowledge chart generating unit 211 is constructed for extracting the identifier of entity, and identifier that use production goes out extracts sentence
Entity in son.In addition, it is assumed that the entity in the sentence of Fig. 3 A is " expert ", " uniform " and " discipline ".
In addition, knowledge chart generating unit 211 can also be with reference to pre-prepd word lexicon or the ontology of word
(ontology) entity is extracted.Further, since most of entity is with being equivalent to noun or adjectival phrase (phrase)
Come what is indicated, so knowledge chart generating unit 211 can also carry out syntax parsing, noun or adjective are extracted as entity.
Then, the relationship between 211 computational entity of knowledge chart generating unit." expert tells about, the use of the uniform in school
Discipline can be improved." sentence in, " uniform " and " discipline " is that (" uniform " is subject, " is recorded as the argument of predicate " raising "
Rule " is target langua0).Knowledge chart generating unit 211 is for example extracted by syntax parsing using 2 entities as the predicate of argument in this way,
As the relationship between entity.
Knowledge chart generating unit 211 generates knowledge chart by linking this 2 entities and the relationship arrow between it
301.Knowledge chart generating unit 211 in the generation of knowledge chart, generate using be denoted as subject entity node as starting point,
Using be denoted as predicate entity node as the arrow of terminal.In addition, knowledge chart can also be undirected chart.
In addition, extracting the predicate with entity (subject and target langua0) for argument shown herein as knowledge chart generating unit 211
As the example of relationship, but other relationships between entity can also be extracted.For example, knowledge chart generating unit 211 can also be extracted
Modification between entity can also extract the distance in the sentence between entity as relationship as relationship.In addition, knowledge chart generating unit
211 for example can also extract relationship using the methods of Semantic Role Labelling (semantic character labeling).
Knowledge chart 302 is knowledge chart generating unit 211 according to " the uniform resistance in school in the document of document DB103
Hinder freedom." sentence generation knowledge chart.The method same as the example of Fig. 3 A of knowledge chart generating unit 211, from the sentence
The relationship " obstruction " between 2 entities " uniform " and " freedom " and entity is extracted in son, generates knowledge chart 302.
Knowledge chart generating unit 211 extracts entity from each sentence in the document provided in this way, by the sentence between entity
In relationship described using knowledge chart.
Fig. 3 C are an examples of the knowledge chart after merging 2 knowledge charts.For example, in multiple knowledge charts, there are phases
In the case of same node, knowledge chart generating unit 211 can also be identical by multiple knowledge graph by the way that the node to be considered as
Table merges.Knowledge chart 303 is knowledge chart generating unit 211 by by " uniform " node and knowledge graph in knowledge chart 301
" uniform " node in table 302 is considered as knowledge chart that is identical and generating.
Knowledge chart generating unit 211 for example by surface layer show identical node be considered as it is identical.In addition, using word lexicon
Or in the case that ontology etc. generates node, knowledge chart generating unit 211 for example can also will be identical in word lexicon or ontology
Performance handled as identical node.Specifically, for example, knowledge chart generating unit 211 can also be by the section of synonym
Point is considered as identical.In addition, knowledge chart generating unit 211 can also be by the expression phase in the arrow of 2 nodes of connection on chart
It is considered as with the arrow of relationship identical.
Such as table table shown in Fig. 3 D and Fig. 3 E by the information of the knowledge chart generated of knowledge chart generating unit 211
Show, it will be in the table storage to knowledge chart DB105.Fig. 3 D are an examples for the node table being saved in knowledge chart DB105.
Node table 304 is to maintain the table of the information of the node of knowledge chart.The node table 304 of Fig. 3 D remains knowledge chart 303
The information of node.
Knowledge chart generating unit 211, which assigns the node of the knowledge chart generated, to be used for identifying the section on the knowledge chart
The ID of point, the ID of imparting is saved in node table 304.In addition, knowledge chart generating unit 211 can also by about node its
He is included in node table 304 information.In the example of Fig. 3 D, node table 304 includes the information that the surface layer of each node shows.This
Outside, such as node table 304 can also include the Bibliographic information for being extracted the document that surface layer corresponding with each node shows
Deng.
Fig. 3 E are stored in an example of the arrow table in knowledge chart DB105.Arrow table 305 is the arrow for including knowledge chart
The table of the information of line.In the example of Fig. 3 E, arrow table 305 maintains the information of the arrow of knowledge chart 303.
Knowledge chart generating unit 211, which assigns the arrow of the knowledge chart generated, to be used for identifying the arrow on the knowledge chart
The ID of the ID of each arrow, the ID of starting point node and peripheral node are established corresponding and are saved in arrow table 305 by the ID of line.This
Outside, knowledge chart generating unit 211 can also will be included in about the other information of arrow in arrow table 305.In the example of Fig. 3 E,
Arrow table 305 includes the information that the surface layer of each arrow shows.In addition, for example arrow table 305 can also include be extracted with it is each
The Bibliographic information etc. of the document of the corresponding surface layer performance of arrow.Knowledge chart is defined by node table 304 and arrow table 305
Construction.
Hereinafter, declarative knowledge chart generates another example of processing.Fig. 4 A~Fig. 4 C are the examples of knowledge chart.The sentence of Fig. 4 A
Son is identical as the sentence of Fig. 3 A.Knowledge chart 301 is demonstrated by the information for indicating " uniform improves discipline " shown in the sentence.But
Be, knowledge chart 301 indicated without performance " expert tells about " " uniform improve discipline " shown in the sentence information (that is, about
The information of relationship as " telling about ").
Knowledge chart generating unit 211 is for example, if sentence (statement) as " subduing and improve discipline " is seen and puts into effect
Body can also then generate knowledge chart 402 other than knowledge chart 301.Sentence be indicate constitute sentence multiple elements and
The proposition of relationship between multiple element.In addition, each element is for example either entity, can also be subordinate sentence or sentence.In addition,
Each element can also be sentence.
Since knowledge chart 301 and knowledge chart 402 are independent knowledge charts, so failing the sentence table of performance Fig. 4 A
Relationship between the 2 knowledge charts shown.In addition, related multiple knowledge are used knowledge graphics table independent so respectively
In the case of existing, the quantity of entity becomes huge.
So knowledge chart generating unit 211 generates knowledge chart 403 shown in Fig. 4 B to show such information.
An example of the generation processing of declarative knowledge chart 403.As described above, knowledge chart generating unit 211 is used in the sentence of document
In the case that at least one party of the object of relationship connection is sentence, as knowledge chart 402, regard sentence hypothesis as entity
And generate knowledge chart.
Knowledge chart generating unit 211 is being determined to have the node of knowledge chart and is including whole from identical document
(that is, there are the feelings of the node of nested configuration in the case of the node of the information of the sentence of other knowledge graph representations of generation
Under condition), the node (following also referred to as to represent node) for representing the sentence is generated from the others knowledge chart.
In the example of Fig. 4 A, due to " uniform improve discipline " node of knowledge chart 402 include it is whole from knowledge
The information for the sentence that the knowledge chart 301 that 402 identical document of chart generates indicates, so knowledge chart generating unit 211 generates
Represent node.Knowledge chart generating unit 211 will for example indicate that " raising " arrow of the relationship of knowledge chart 301 is changed to " improve "
Node is determined as representing node.Also, knowledge chart generating unit 211 generate be originally " raising " arrow starting point " uniform "
Node is starting point, the arrow to represent node as terminal, and using represent node as starting point, to be the terminal of " raising " arrow originally
" discipline " node be terminal arrow.The part that the dotted line by rectangle of knowledge chart 403 surrounds is generated as a result,.
As described above, it includes 3 nodes that " raising " node of knowledge chart 403 indicated with doublet, which is representative,
The representative node of 1 sentence of (" uniform ", " raising ", " discipline ").The representative of N number of (N is 1 or more integer) sentence will be represented
Node is referred to as N+1 minor nodes.That is, " raising " node of knowledge chart 403 indicated with doublet is 2 minor nodes.In addition, by table
Show that the node of entity itself is referred to as 1 minor node.That is, in the case that the relationship between n times node is transformed to represent node, the generation
Table node is N+1 minor nodes.
In addition, in knowledge chart 403 with the arrow that doublet indicates be not offered as node between relationship, but be used for table
Show that representing 3 nodes that node on behalf includes in the part of knowledge chart 403 surrounded by the dotted line of rectangle is constituted
The arrow of the case where sentence.Hereinafter, such arrow is referred to as sentence arrow.
In addition, as described above, in knowledge chart 402, by " the system as " expert " node of starting point and as terminal
Clothes improve discipline " the node arrow connection for indicating " telling about " this relationship.Thus, knowledge chart generating unit 211 is by with generation
" raisings " node of table " uniform improve discipline " this sentence is terminal, using " expert " node as starting point, by 2 nodes with indicating
The arrow of " telling about " this relationship connects, and generates knowledge chart 403.
Knowledge chart generating unit 211 can be showed " uniform improves discipline " and " special with 1 knowledge chart 403 as a result,
Family " teaches " uniform improves discipline ".In addition, after the merging treatment for having carried out chart as described above, there remains comprising language
In the case of knowledge chart of the sentence as entity itself, knowledge chart generating unit 211 can also delete the knowledge chart.
Fig. 4 C are an examples of node table 304.Node table 304 can also also keep degree of node.That is, knowledge chart generates
Portion 211 is corresponding with degree of node foundation by node ID and is saved in node table 304.
Fig. 4 D are an examples of arrow table 305.Arrow table 305 can also keep indicate arrow whether be sentence arrow mark
Will.That is, knowledge chart generating unit 211 by arrow ID with indicate whether arrow is that the mark foundation of sentence arrow is corresponding and be saved in
In arrow table 305.In addition, since sentence arrow does not indicate that surface layer shows, so in arrow table 305, with sentence arrow
Null values are preserved in corresponding surface layer performance.
Knowledge chart generating unit 211 makes knowing with the diagram construction comprising 2 minor nodes for indicating sentence in this way
Know chart, the relationship being not only between simple entity, more complicated information can also use knowledge chart to show.
Fig. 5 A~Fig. 5 C are an examples of knowledge chart.Knowledge chart 501 is according to " uniform improves discipline, but hinders certainly
By." sentence make knowledge chart.Knowledge chart generating unit 211 is as using Fig. 3 C explanations, first, according to the sentence,
Make the knowledge chart 303 for indicating " uniform improves discipline " and " uniform hinders freely ".
Knowledge chart generating unit 211 is being determined as having expression anti-in the sentence referring for example to pre-prepd dictionary etc.
In the case of the surface layer performance of opinion, from as " uniform improve discipline " of main subordinate sentence this sentence and as subordinate subordinate sentence
" uniform hinders freely " this sentence, generates represent node respectively.For example, knowledge chart generating unit 211 is same as the above method,
Node " raising " is represented according to the generation of main subordinate sentence, node " obstruction " is represented according to the generation of subordinate subordinate sentence.It is 2 that these, which represent node all,
Minor node.
Also, knowledge chart generating unit 211 generate using be originally " raising " arrow starting point " uniform " node as starting point, with
Node is represented as the arrow of terminal, and using represent node as starting point, be originally " raising " arrow terminal " discipline " node as
The arrow of terminal.Equally, knowledge chart generating unit 211 is generated is with " uniform " node of starting point for being originally " obstruction " arrow
Point is saved to represent arrow of the node as terminal, and to represent node as starting point, with " freedom " of terminal for being originally " obstruction " arrow
Point is the arrow of terminal.These arrows generated are all sentence arrows.
In turn, it is starting point, with subordinate subordinate sentence that knowledge chart generating unit 211, which is generated with the representative node " raising " of main subordinate sentence,
Represent " paradox " arrow that node " obstruction " is terminal, expression paradox relationship.Pass through the processing, knowledge chart generating unit
211 can generate knowledge chart 501.
Hereinafter, explanation is according to the sentence of " according to investigation, uniform improves discipline ", knowledge chart generating unit shown in Fig. 5 B
211 generate the example of knowledge chart 503 shown in Fig. 5 C.Knowledge chart generating unit 211 is same as the example of Fig. 4 A first, generates
Knowledge chart 301.In addition, knowledge chart generating unit 211 generates knowledge chart 502.The generation method of declarative knowledge chart 502
An example.
Knowledge chart generating unit 211 exists in being determined as sentence referring for example to pre-prepd dictionary and indicates information source
Word (" investigation " in Fig. 5 B) in the case of, execute syntax parsing etc., extract the information of the information source-representation (in Fig. 5 B
" uniform improves discipline ").
Knowledge chart generating unit 211 regards " uniform improves discipline " this sentence as entity.Knowledge chart generating unit 211 is given birth to
At using " investigation " node as starting point, using " uniform improve discipline ", node is " information source " arrow of terminal, as between this 2 entities
Relationship.Knowledge chart generating unit 211 can generate knowledge chart 502 shown in Fig. 5 C as a result,.
In turn, knowledge chart generating unit 211 can be by method same as method that Fig. 4 A and Fig. 4 B illustrate is used, will
Knowledge chart 301 merges with knowledge chart 502 and generates knowledge chart 503.
Hereinafter, explanation is according to shown in Fig. 5 D, " expert tells about, and uniform improves discipline, but hinders freely." sentence, knowledge
Chart generating unit 211 generates the example of knowledge chart 505 shown in Fig. 5 E.Knowledge chart generating unit 211 " is made according to the sentence
Clothes improve discipline, but hinder freely " part, the method illustrated by using Fig. 5 A generates knowledge chart 501.In addition, knowledge
Chart generating unit 211 regards " uniform improves discipline, but hinders freely " this sentence as entity, generates knowledge chart 504.
Since " uniform improves discipline, but hinders freely " node of knowledge chart 504 includes whole according to identical sentence
The information for the sentence that the knowledge chart 501 that son generates indicates, so " paradox " arrow for the relationship for indicating knowledge chart 501 is become
It is changed to " paradox " node, is determined as representing node.The number for representing node " paradox " is 3.
Also, knowledge chart generating unit 211 generate using be originally " paradox " arrow starting point " raising " node as starting point, with
" paradox " node be terminal sentence arrow, and using " paradox " node as starting point, be originally " paradox " arrow terminal " resistance
Hinder " node be terminal sentence arrow.
In turn, knowledge chart generating unit 211 passes through with representative " uniform improves discipline, but hinders freely " this sentence
" paradox " node is terminal, using " expert " node as starting point, and 2 nodes are used and indicate that the arrow of " telling about " this relationship connects,
Generate knowledge chart 505.
In this way, the knowledge chart generating unit 211 of the present embodiment by using high order node and linked it is special
Arrow (sentence arrow), by nested information chart type Knowledge representation.Knowledge chart generating unit 211 is not only for letter as a result,
Single information can also show complicated context with less knowledge chart.That is, the design assisting system of the present embodiment
201, although complicated context can be handled, can inhibit the data volume of knowledge chart DB105, carry out to higher speed aftermentioned
Speech understanding processing in knowledge chart control treatment.
Return to the explanation of Fig. 2.Then, the knowledge chart that context calculating part 212 is kept using knowledge chart DB105, holds
Row calculates the context calculation processing (S112) for the sequence that information occurs in article.
As described above, knowledge chart generating unit 211 by the other information extracted from article after carrying out pictorialization, example
Such as node is considered as according to preset benchmark (position in surface layer performance, word lexicon, ontology) identical, thus generated
Knowledge chart shown in Fig. 3 C, Fig. 4 B, Fig. 5 A, Fig. 5 C and Fig. 5 D etc..
Fig. 6 A are an examples of knowledge chart.Article 601 is to oppose that school imports the example of the argumentative writing of uniform.In article
In 601, show the opinion for opposing the importing of uniform, then, while telling about the advantages of improving discipline, or narration hinders to learn
Raw free disadvantage.In turn, in article 601, the fairness in market can be hindered in this way by teaching the use specifically subdued
Viewpoint opinion.
In this way, in composition, session and argumentative writing etc., there is the stream (context) of information alert.The context of the present embodiment
Calculating part 212 calculates the context of the article in document DB103, and result of calculation is saved in context DB107.Up and down
Literary calculating part 212 carries out context calculating by calculating the sequence of the knowledge occurred in article.
The design assisting system 201 of the present embodiment uses knowledge chart DB105 and context DB107, identifies the work of user
Text or the context of session, prompt the user with inspiration appropriate, comment and the query etc. in view of context.Design branch as a result,
Shaking apparatus 201 can deepen the thinking of user and support design.
Hereinafter, being illustrated to the details of context calculation processing (S112).In addition, in context calculation processing
Knowledge, be include in sentence relationship indicate knowledge, i.e., 2 times or more nodes or be not sentence arrow arrow indicate know
Know.The knowledge chart 602 of Fig. 6 A is an example of the knowledge chart generated according to article 601.Fig. 6 B are corresponding with knowledge chart 602
Node table 304 an example.Fig. 6 C are an examples of arrow table corresponding with knowledge chart 602 305.
Context calculating part 212 calculates on knowledge chart in document for example for each document for including in document DB103
The sequence of the knowledge of appearance.Context calculating part 212 for example for the knowledge " uniform improve discipline " in knowledge chart 602,
Can using the ID of node related with the knowledge and arrow as by the tactic vector of the appearance in sentence (N1, E1, N2,
E2, N3) show the knowledge.According to this method, for example, can will " uniform improve discipline, but hinder freely " this knowledge is for example
It shows as (N1, E1, N2, E2, N3, E3, N4, E5, N1, E6, N6, E7, N7).
In addition, context calculating part 212 more compactly shows knowledge preferably with the method for following middle explanations.Tool
For body, context calculating part 212 by it is specified be not sentence arrow arrow and 2 times or more nodes show knowledge.By
This, context calculating part 212 can will be in knowledge chart " uniform improves discipline, but hinders freely " it is such comprising complex sentence,
The complicated knowledge of compound sentence and conjunction etc. is compactly handled.Hereinafter, being illustrated using concrete example.
First, knowledge chart 602 includes 2 minor node N2 and N6 and 3 minor node N4 as 2 times or more nodes.2
Minor node N2 " raising " is linked by sentence arrow with node N1 " uniform " and node N3 " discipline ", indicates that " uniform, which improves, records
This knowledge of rule ".Equally, 2 minor node N6 " obstruction " expressions " uniform hinders freely " this knowledge.3 minor node N4 " paradox " are logical
It crosses sentence arrow and links with the N6 of the N2 of representative " uniform improves discipline " and representative " uniform hinders freely ", so indicating " system
Clothes improve discipline, but hinder freely " this knowledge.
Then, knowledge chart 602 includes E4 and E8 as the arrow for not being sentence arrow.Arrow E4 is " special by node N5
Family " links with node N4 " paradox ", indicates " expert tells about, and uniform improves discipline, but hinders freely " this knowledge.In addition same
Sample, arrow E8 indicate " uniform damage Equity of Market " this knowledge.
Context calculating part 212 by using above method, such as will as " uniform improve discipline, but hinder freely " that
The complicated representation of knowledge of sample is that N4 is such, can use not the ID of ID or 2 time or more node of the arrow for being sentence arrow
Knowledge is compactly indicated.
Context calculating part 212 is in context calculation processing, sequence that calculation knowledge occurs in document.Use Fig. 7 A
And Fig. 7 B illustrate an example of the computational methods.Fig. 7 A are an examples of knowledge chart.Knowledge chart 703 is according in document DB103
Including sentence 701 and sentence 702 generate knowledge chart.The arrow for not being sentence arrow and 2 times are used as described above
In the case of above node performance knowledge, in sentence 701, knowledge occurs with the sequence of N6, E8, N2, E9.In addition, it is same,
In sentence 702, knowledge occurs with the sequence of N6, E9.
Context calculating part 212 is directed to the knowledge occurred in knowledge chart 703, calculates separately in document DB103
The occurrence number of the knowledge (being also referred to as follow-up knowledge below) occurred rearward than the knowledge in document, is saved in context DB107
In the follow-up knowledge table kept.
That is, context calculating part 212 determines and 2 times or more nodes occurring in knowledge chart 703 and be not sentence
The corresponding knowledge of arrow of arrow.Context calculating part 212 is directed to identified knowledge, calculates separately in document DB103
Occur rearward than the knowledge in document and 2 times or more nodes occurring in knowledge chart 703 and be not sentence arrow arrow
The occurrence number of knowledge represented by line.
Fig. 7 B are an examples of follow-up knowledge table.Follow-up knowledge table 704 will node ID corresponding with knowledge or arrow ID,
The follow-up knowledge of the knowledge and the number that the follow-up knowledge occurs after the knowledge are established corresponding and are kept.For example, Fig. 7 B
2nd row of follow-up knowledge table 704 indicates that knowledge E8 occurs 1 time after knowledge N6, and knowledge N2 occurs 1 time, knowledge E9
Occur 2 times.Knowledge E9 due to occurring each 1 time respectively rearward in sentence 701 and sentence 702 than N6, so after Fig. 7 B
In continuous knowledge table 704, the occurrence number of the E9 of the follow-up knowledge as N6 is 2 times.
Context calculating part 212 presses each knowledge by the foundation pair of the occurrence number of follow-up knowledge and follow-up knowledge in this way
It answers and is saved in follow-up knowledge table 704, context DB107 can keep indicating the sequence of the knowledge in document and front and back pass
The information of system etc..
In addition, context calculating part 212 can also be calculated about each knowledge at a distance from follow-up knowledge is on document, protect
It is stored in follow-up knowledge table 704.It is 2 times or more follow-up knowledge about occurrence number, context calculating part 212 is by the knowledge
Average value or minimum value at a distance from the document of the follow-up knowledge etc. are saved in follow-up knowledge table 704.More than, to learning
Habit processing is illustrated.
Then, to being illustrated with processing (S120).In with handling, design 201 use of assisting system is learning
The knowledge chart that is made in processing and follow-up knowledge understand the composition of user or the context of session, in composition or session
Way prompts inspiration appropriate, thus promotes the design of user.
First, speech understanding portion 213 accepts the input of the composition comprising sentence or session from user, is transformed to text
(S121).Specifically, speech understanding portion 213 such as both can from keyboard input interface accept text input, can also will
Text transform is carried out by voice recognition technology from the spoken sounds of the inputs such as microphone.In addition, such as speech understanding portion 213
The article write on paper can be subjected to text transform by OCR etc..In addition, speech understanding portion 213 can also be known using sign language
Not Deng method the movement of user is subjected to text transform.
Then, speech understanding portion 213 implements the speech for understanding the text obtained in step S121 knowledge chart reason
Solution handles (S122).Here, speech understanding portion 213 by the text matches that will input to knowledge chart, to understand saying for user
Words.Hereinafter, illustrating that speech understands the details of processing using Fig. 8 A, Fig. 8 B, Fig. 9 A, Fig. 9 B and Fig. 9 C.
What the node table 304 and arrow table 305 that the knowledge chart 801 of Fig. 8 A is stored in knowledge chart DB105 indicated
A part for knowledge chart.In addition, the follow-up knowledge table of Fig. 8 B is stored in the follow-up knowledge table in context DB107
704 part.
For example, it is assumed that for the theme of " whether school should import uniform ", user is with " agreeing with school to import uniform "
Position writes argumentative writing.And, it is assumed that speech understanding portion 213 accepted shown shown in Fig. 9 C in picture 903 " school should lead
Enter uniform." as sentence input.Speech understanding portion 213 is same using processing is generated with the knowledge chart of step S111
Method, according to the knowledge chart 901 of sentence generation Fig. 9 A.
Then, speech understanding portion 213 compares knowledge chart 901 with knowledge chart 801, is retrieved from knowledge chart 801
The knowledge for including in the knowledge that knowledge chart 901 indicates.If using the performance of the knowledge illustrated in context calculation processing
Method, then since the knowledge E10 of the expression of knowledge chart 801 is (that is, " school " node and " uniform " node " importing " relationship are connected
The knowledge of knot) it is comprised in the knowledge of the expression of knowledge chart 901, speech understanding portion 213 extracts E10.In this way, speech understanding portion
The knowledge that 213 extractions include in the input of user is come by compareing the knowledge extracted with the knowledge chart learnt in advance
Understand the input of user.
In addition, speech understanding portion 213 is for example using technique of expression identical with context calculation processing, by following information to
Context identification portion 214 exports, which indicates the node on the knowledge chart 801 of the knowledge of text representation input by user
Knowledge appearance sequence in ID and arrow ID and the text.In the case where being entered sentence shown in display picture 903, say
Words understanding portion 213 for example exports E10 to context identification part 214.In the case where being entered sentence 702 of Fig. 7 A, say
Words understanding portion 213 is for example exported arrow ID to context identification part 214 with the sequence of N6, E9.
In addition, a part (part being made of N10, E10 and N1) complete one for knowledge chart 901 and knowledge chart 801
It causes.Understanding portion 213 talk in the control of knowledge chart, even if the relationship or entity of the relationship of a side or entity and another party are not
It is completely the same, but it is identical with the another party to be for example considered as a side in the case where similar degree is specified value or more.
For example, it is assumed that speech understanding portion 213 is according to the knowledge chart 902 from text generation input by user Fig. 9 B.Know
The arrow for knowing chart 902 is " adopting ", but understanding portion 213 of talking " can also will adopt " and the E10 of knowledge chart 801 " importing "
Regard as it is synonymous, be determined as knowledge chart 902 indicate knowledge it is consistent with E10.
Understanding portion 213 talk such as according to the nearly adopted relationship computational entity in word lexicon or ontology or the class between relationship
Like degree.Speech understanding portion 213 by being that the relationship of regulation or more or entity are used as synonymous processing by similar degree, can increase with from
Knowledge in user's input in the consistent knowledge chart DB105 of the knowledge extracted, and then inspiration appropriate can be brought to user
Possibility get higher.
Return to the explanation of Fig. 2.Then, context identification portion 214 executes the user obtained in speech understanding processing is defeated
The context identification processing (S123) that the appearance sequence of the knowledge of the text representation entered is compareed with follow-up knowledge table 704.
For example, in the case where user has input the text for being equivalent to knowledge chart 901, exported in step S122 as described above
E10.At this point, context identification portion 214 indicates the 2nd row of the follow-up knowledge table 704 of Fig. 8 B (value of knowledge is the row of E10)
Information to stage identification part 215 export.
Then, in stage identification part 215, the stage identifying processing in the composition of identification user or the stage of session is executed
(S124).For example, for some theme (in the example of Fig. 9 A~Fig. 9 C, " whether school should import uniform "), in user
In the case of being described with the position agreed with or opposed, there is the content to argumentative writing requirement.
Showing whether agreeing with, enumerate agree with or the argument opposed and the reasons why whether agree with and according to prompt etc.,
It is the example to the content of the argumentative writing requirement.Stage identification part 215 presets composition or Dialog request by judgement
Content it is whether insufficient in text input by user, to identify the composition of user or the stage of session.
Fig. 9 C indicate an example of the transformation of the picture shown in display device 203.For example, being shown on display picture 903
Although the stage of the sentence shown be show agree with uniform importing position, but without enumerate argument, without prompt reason
Or according to stage.Stage identification part 215 determines that there are no chatted in text input by user in preset content
The content stated.
In addition, the content of composition or Dialog request is predefined, such as design assisting system 201 is kept in advance
The document of study.For example, design assisting system 201 makes the mark of place corresponding to the content before starting with processing
Label, by using the rote learning of the label, produce the identifier that can identify the content.Stage identification part 215 was using should
Identifier differentiates in text input by user whether include the content.In the example of Fig. 9 C, it is assumed that needed in composition pair
In importing from the uniform to school whether agreeing with show and argument whether agreeing with for showing.
In addition, design assisting system 201 for example can also it is corresponding to the content to document DB103 in place of (for example, comprising
The document of the above or text comprising the above) assign indicate be it is corresponding to the content in place of annotation.
Also, it includes to use comment section that knowledge chart generating unit 211 generates expression in knowledge chart generation processing
Knowledge knowledge chart in the case of, content this information that can also indicate the representation of knowledge comment section is as attached
Information is added to be added on knowledge chart.That is, for example knowledge chart generating unit 211 adds node table 304 and arrow table 305 again
Row, which are saved in additional row.
Return to the explanation of Fig. 2.Then, knot of the stage identification part 215 based on context identification processing and stage identifying processing
Fruit implements to prompt the inspiration prompt inspired to handle (S125) user.Stage, identification part 215 was according to follow-up 704 table of knowledge table
The follow-up knowledge shown generates the inspiration to be prompted.
As described above, in the case where being entered the sentence being shown on display picture 903, at context identification
In reason, the knowledge for exporting follow-up knowledge table 704 is the information of the record of E10.That is, stage identification part 215 has been achieved for table
The follow-up knowledge for advising knowledge E10 is the information of N2 and E9.
Thus, the sentence that the relationship of follow-up knowledge N2 and follow-up knowledge E9 indicate is determined as by stage identification part 215 will be to
The candidate of the inspiration of user's prompt.Stage identification part 215 can also be all defeated to display device 203 as inspiring by the candidate
Go out.
In addition, for example stage identification part 215 can also have the time for being equivalent to following knowledge in stage identifying processing
Include in display device 203 using the candidate preference as inspiring, which is to indicate to be judged as showing in the case of choosing
There is the knowledge of insufficient content in the sentence on display picture 903.
In addition, so-called " some candidate preference is shown ", for example, refer to only by the candidate display in display device 203,
And the candidate is highlighted in display device 203 by whole candidate displays.
In addition, stage identification part 215 can also be by the knowledge of follow-up knowledge table 704 E10 record in go out occurrence
More (in the example of Fig. 8 B, N2 is 10 times, and E9 the is 5 times) knowledge of number is preferentially prompted as inspiration in display device 203
(for example, by occurrence number from big to small sequence prompt specified quantity candidate or by occurrence number be stated number more than time
Choosing all prompts).Stage identification part 215 can will be used as with the stronger knowledge of content relevance input by user and inspire as a result,
Output.
In addition, preserved in follow-up knowledge table 704 knowledge on the document between follow-up knowledge at a distance from the case where
Under, the knowledge of follow-up knowledge table 704 can also be that the smaller knowledge of the distance in the record of E10 is excellent by stage identification part 215
First prompted in display device 203 as inspiration (for example, the candidate of specified quantity is prompted by the sequence of distance from small to large, or
It will all be prompted apart from for specified value candidate below).Stage identification part 215 can will be closed with content input by user as a result,
The stronger knowledge of connection property is exported as inspiring.
For example, it is assumed that being the argument of knowledge N2 and knowledge E9 agreed with or opposed.At this point, stage identification part 215 for example with
Occurrence number is preferential, as shown in the display picture 904 of Fig. 9 C, will " uniform raising discipline " content as opening
Hair prompt.In addition, stage identification part 215 for example comes from user in the case where being speculated as user and needing enlightenment or having accepted
Inspiration prompt in the case of, inspiration is shown in display device 203.Specifically, for example stage identification part 215 both can be
Have passed through from the input of user certain time also without next input in the case of prompt inspire, can also be by user
It prompts to inspire in the case of request from the user by lower button etc..
In addition, stage identification part 215 has only prompted 1 inspiration in showing picture 904 here, but can also prompt multiple
It inspires.In addition, stage identifying processing due to be for will prompt inspiration reduction processing, so can not also execute.
User is reference with the inspiration prompted in showing picture 904, as shown in showing picture 905, is used
The input of text is continued at family, returns to speech understanding processing.In addition, user can also input the text unrelated with suggested inspiration
This.Also, for the text inputted, speech is executed again and understands processing and context identification processing.Have input be shown in it is aobvious
Show that the output in the context identification processing in the case of the sentence on picture 905 is E10 and N2.
At this point, in the follow-up knowledge table 704 of Fig. 8 B, since the common follow-up knowledge of E10 and N2 is E9, so rank
" uniform improves study wish " is used as and is opened in inspiring prompt processing, such as showing picture 906 by section identification part 215
Hair prompt.
The design assisting system 201 of the present embodiment is by above-mentioned processing, when subscriber card is in text input or user
When request inspires, inspiration appropriate can be given to user, additionally it is possible to which the design for supporting user is supported the next of user and said
Words.
In addition, stage identification part 215 can also carry out in inspiring prompt processing using the as shown below of knowledge chart
Processing.As described above, in the case where user has input the sentence being shown on display picture 905, at context identification
Knowledge E10 and knowledge N2 is exported in reason.
In knowledge chart 801, there is the knowledge N6 of the paradox of the knowledge N2 exported in context identification processing.Stage knows
Other portion 215 is including the knowledge of the paradox of the knowledge inputted from context identification portion 214 in this way on knowledge chart and is not having
In the case of the knowledge for inputting the paradox from context identification portion 214, the inspiration for being equivalent to the paradox knowledge can also be prompted.Tool
For body, such as stage identification part 215 can also be by " uniform improves discipline, but also has uniform to hinder shown in display picture 907
Free opinion, how you think" etc. inquiry or comment as inspire prompt.
By the processing, user is able to know that the opinion of the opposite position of itself, and then can realize for deepening opinion
It states or writes a composition, the support of the design of user in session.In addition, in the above example, illustrate that stage identification part 215 will be from upper
The knowledge of the paradox for the knowledge that hereafter identification part 214 inputs is as the example for inspiring prompt, but stage identification part 215 can also incite somebody to action
The knowledge of relationship other than paradox in the knowledge inputted from context identification portion 214 and not by from context identification
The knowledge that portion 214 inputs is prompted as inspiration.
In addition, the present invention is not limited to the above embodiments, including various variations.For example, above-mentioned implementation
Example explains in detail to easily illustrate the present invention, is not limited to the whole knots for centainly having illustrated
Structure.In addition it is also possible to a part for the structure of some embodiment be replaced with to the structure of other embodiment, in addition it is also possible to right
The structure of the structure addition other embodiment of some embodiment.In addition, a part for the structure about each embodiment, can carry out
Addition, deletion, the displacement of other structures.
In addition, above-mentioned each structure, function, processing unit, processing mechanism etc. can also be by their part or all of examples
Such as pass through the hardware realization with IC design.In addition, above-mentioned each structure, function etc. can also be by by processors
By the interpretation of programs for realizing each function, executes and use software realization.Realize the information of program, table, file of each function etc.
The recording device or IC card, SD card, DVD etc. of memory or hard disk, SSD (Solid State Drive) etc. can be placed into
In recording medium.
In addition, about control line or information wire, the part for thinking to need in explanation is shown, it might not on product
Show whole control line or information wire.Actually it is also assumed that most structure is connected to each other.
Claims (10)
1. a kind of design assisting system, supports the design of user, which is characterized in that
Including processor and memory；
Above-mentioned memory keeps following information：
Chart-information indicates that the construction of the 1st chart comprising node and the 1st kind of arrow, above-mentioned node expression are formed in document
Including proposition element, above-mentioned document specification knowledge, above-mentioned 1st kind of arrow indicate between the above-mentioned element represented by above-mentioned proposition
Relationship；And
Contextual information indicates the appearance sequence of the relationship in above-mentioned document；
In above-mentioned processor,
Accept the input of proposition；
Extraction has accepted the element for including in the above-mentioned proposition inputted and has accepted the element represented by the above-mentioned proposition of input
Between relationship；
Generate the 2nd chart, the 2nd chart includes the node for the above-mentioned element for indicating to be extracted, and above-mentioned is wanted what is extracted
Relationship between element includes as above-mentioned 1st kind of arrow；
It is being determined as that the relationship represented by above-mentioned 2nd chart is contained in the situation in above-mentioned 1st chart with reference to above-mentioned chart-information
Under, with reference to above-mentioned contextual information, determine the relationship occurred rearward than the relationship in above-mentioned document；
Output indicates the information of identified above-mentioned relation.
2. design assisting system as described in claim 1, which is characterized in that
Above-mentioned document includes multiple propositions；
The node of above-mentioned 1st chart includes：
1st kind of node indicates the entity for including in above-mentioned multiple propositions；And
2nd kind of node indicates that the relationship of the first proposition, the 1st kind of proposition are the proposition for including in above-mentioned multiple propositions, and
Be other propositions that whole element and relationship are contained among above-mentioned multiple propositions 1 element in proposition；
Above-mentioned 1st chart includes the 2nd kind of arrow, the 2nd kind of arrow will indicate to constitute the node of the element of above-mentioned 1st kind of proposition with
Above-mentioned 2nd kind of node connection；
Above-mentioned processor accepts the input of multiple propositions；
Above-mentioned processor, in the generation of above-mentioned 2nd chart,
Entity in the above-mentioned element extracted is determined as above-mentioned 1st kind of node；
In the case where it includes above-mentioned 1st kind of proposition to be judged to having accepted above-mentioned multiple propositions of input, by the 1st kind of proposition
Relationship is determined as above-mentioned 2nd kind of node；
The node for the element for indicating to constitute the 1st kind of proposition and the above-mentioned 2nd kind of node determined are passed through into above-mentioned 2nd kind of arrow
Connection.
3. design assisting system as described in claim 1, which is characterized in that
Relationship represented by proposition of the above-mentioned contextual information for above-mentioned document, is illustrated respectively in above-mentioned document than the relationship
The occurrence number of the relationship occurred rearward；
The occurrence number of based on represented by above-mentioned contextual information, the identified above-mentioned relation of above-mentioned processor, from determining
Above-mentioned relation in choice relation, and export the information for indicating selected above-mentioned relation.
4. design assisting system as described in claim 1, which is characterized in that
Relationship represented by proposition of the above-mentioned contextual information for above-mentioned document indicates and in above-mentioned document respectively than the pass
It is the distance on above-mentioned document between the relationship occurred rearward；
Above-mentioned processor is based on represented by above-mentioned 2nd chart for including in represented by above-mentioned contextual information, above-mentioned 1st chart
Relationship at a distance from identified above-mentioned relation, the choice relation from identified above-mentioned relation, and exporting selected by expression
Above-mentioned relation information.
5. design assisting system as described in claim 1, which is characterized in that
In the case that above-mentioned processor includes preset relationship in identified above-mentioned relation, output indicates above-mentioned advance
The information of the relationship of setting.
6. a kind of design support method is supported the design of user by design assisting system, which is characterized in that
Above-mentioned design assisting system keeps following information：
Chart-information indicates that the construction of the 1st chart comprising node and the 1st kind of arrow, above-mentioned node expression are formed in document
Including proposition element, above-mentioned document specification knowledge, above-mentioned 1st kind of arrow indicate between the above-mentioned element represented by above-mentioned proposition
Relationship；And
Contextual information indicates the appearance sequence of the relationship in above-mentioned document；
In above-mentioned design support method, above-mentioned design assisting system carries out following processing：
Accept the input of proposition；
Extraction has accepted the element for including in the above-mentioned proposition inputted and has accepted the element represented by the above-mentioned proposition of input
Between relationship；
Generate the 2nd chart, the 2nd chart includes the node for the above-mentioned element for indicating to be extracted, and above-mentioned is wanted what is extracted
Relationship between element includes as above-mentioned 1st kind of arrow；
It is being determined as that the relationship represented by above-mentioned 2nd chart is contained in the situation in above-mentioned 1st chart with reference to above-mentioned chart-information
Under, with reference to above-mentioned contextual information, determine the relationship occurred rearward than the relationship in above-mentioned document；
Output indicates the information of identified above-mentioned relation.
7. design support method as claimed in claim 6, which is characterized in that
Above-mentioned document includes multiple propositions；
The node of above-mentioned 1st chart includes：
1st kind of node indicates the entity for including in above-mentioned multiple propositions；And
2nd kind of node indicates that the relationship of the first proposition, the 1st kind of proposition are the proposition for including in above-mentioned multiple propositions, and
Be other propositions that whole element and relationship are contained among above-mentioned multiple propositions 1 element in proposition；
Above-mentioned 1st chart includes the 2nd kind of arrow, the 2nd kind of arrow will indicate to constitute the node of the element of above-mentioned 1st kind of proposition with
Above-mentioned 2nd kind of node connection；
In above-mentioned design support method, above-mentioned design assisting system carries out following processing：
Accept the input of multiple propositions；
In the generation of above-mentioned 2nd chart,
Entity in the above-mentioned element extracted is determined as above-mentioned 1st kind of node；
In the case where the above-mentioned multiple propositions for being judged to having been accepted input include above-mentioned 1st kind of proposition, by the 1st kind of proposition
Relationship be determined as above-mentioned 2nd kind of node；
The node for the element for indicating to constitute the 1st kind of proposition and the above-mentioned 2nd kind of node determined are passed through into above-mentioned 2nd kind of arrow
Connection.
8. design support method as claimed in claim 6, which is characterized in that
Relationship represented by proposition of the above-mentioned contextual information for above-mentioned document, is illustrated respectively in above-mentioned document than the relationship
The occurrence number of the relationship occurred rearward；
In above-mentioned design support method,
The occurrence number of above-mentioned based on represented by above-mentioned contextual information, the identified above-mentioned relation of design assisting system, from
Choice relation in identified above-mentioned relation, and export the information for indicating selected above-mentioned relation.
9. design support method as claimed in claim 6, which is characterized in that
Relationship represented by proposition of the above-mentioned contextual information for above-mentioned document indicates and in above-mentioned document respectively than the pass
It is the distance on above-mentioned document between the relationship occurred rearward；
In above-mentioned design support method,
Above-mentioned design assisting system is based on above-mentioned 2nd chart for including in represented by above-mentioned contextual information, above-mentioned 1st chart
Represented relationship is at a distance from identified above-mentioned relation, the choice relation from identified above-mentioned relation, and exports expression
The information of selected above-mentioned relation.
10. design support method as claimed in claim 6, which is characterized in that
In the case that above-mentioned design assisting system includes preset relationship in identified above-mentioned relation, in output expression
State the information of preset relationship.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2017-041822 | 2017-03-06 | ||
| JP2017041822A JP6622236B2 (en) | 2017-03-06 | 2017-03-06 | Idea support device and idea support method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108536720A true CN108536720A (en) | 2018-09-14 |
| CN108536720B CN108536720B (en) | 2021-09-21 |
Family
ID=63489748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711235125.2A Active CN108536720B (en) | 2017-03-06 | 2017-11-30 | Concept support device and concept support method |
Country Status (3)
| Country | Link |
|---|---|
| JP (1) | JP6622236B2 (en) |
| KR (1) | KR102039393B1 (en) |
| CN (1) | CN108536720B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP7343820B2 (en) * | 2020-04-10 | 2023-09-13 | 日本電信電話株式会社 | Text data analysis information generation device, text data analysis information generation method, and text data analysis information generation program using ontology |
| CN112148871B (en) * | 2020-09-21 | 2024-04-12 | 北京百度网讯科技有限公司 | Digest generation method, digest generation device, electronic equipment and storage medium |
| JP2022070523A (en) * | 2020-10-27 | 2022-05-13 | 株式会社日立製作所 | Semantic expression analysis system and semantic expression analysis method |
| WO2022091413A1 (en) | 2020-11-02 | 2022-05-05 | 富士通株式会社 | Machine learning program, inference program, device, and method |
| JP2024110172A (en) * | 2023-02-02 | 2024-08-15 | 株式会社日立製作所 | Computer and method for extracting knowledge |
| JP7474536B1 (en) | 2023-08-07 | 2024-04-25 | Ｌｉｎｋｅｄ Ｉｄｅａｌ合同会社 | Information processing system and information processing method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5841900A (en) * | 1996-01-11 | 1998-11-24 | Xerox Corporation | Method for graph-based table recognition |
| WO2002063497A1 (en) * | 2001-02-08 | 2002-08-15 | Soda Technologies Pty Ltd | Proximity for computer represented graphs |
| JP2004342064A (en) * | 2003-05-16 | 2004-12-02 | Seki Tama Yonezawa | Human inference engine |
| WO2007027967A2 (en) * | 2005-08-31 | 2007-03-08 | Eagleforce Associates | System for hypothesis generation |
| KR20110134314A (en) * | 2010-06-08 | 2011-12-14 | 인터내셔널 비지네스 머신즈 코포레이션 | Graphical models for representing text documents for computer analysis |
| CN103124980A (en) * | 2010-09-24 | 2013-05-29 | 国际商业机器公司 | Providing answers to questions including assembling answers from multiple document segments |
| CN103324646A (en) * | 2012-03-19 | 2013-09-25 | 株式会社东芝 | Search supporting device and search supporting method |
| CN104699758A (en) * | 2015-02-04 | 2015-06-10 | 中国人民解放军装甲兵工程学院 | Intelligent generation system and method of graphic library-associated command document |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003006191A (en) * | 2001-06-27 | 2003-01-10 | Ricoh Co Ltd | Device and method for supporting preparation of foreign language document and program recording medium |
| JP2006301967A (en) * | 2005-04-20 | 2006-11-02 | Sony Computer Entertainment Inc | Conversation support device |
| JPWO2007049800A1 (en) * | 2005-10-27 | 2009-04-30 | 株式会社レイテック | Document creation support device |
| WO2007060780A1 (en) * | 2005-11-22 | 2007-05-31 | Nec Corporation | Inspiration support device, inspiration support method, and inspiration support program |
| JP4677563B2 (en) | 2006-03-10 | 2011-04-27 | 国立大学法人 筑波大学 | Decision support system and decision support method |
| KR101353521B1 (en) * | 2012-05-10 | 2014-01-23 | 경북대학교 산학협력단 | A method and an apparatus of keyword extraction and a communication assist device |
-
2017
- 2017-03-06 JP JP2017041822A patent/JP6622236B2/en not_active Expired - Fee Related
- 2017-11-08 KR KR1020170147764A patent/KR102039393B1/en active IP Right Grant
- 2017-11-30 CN CN201711235125.2A patent/CN108536720B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5841900A (en) * | 1996-01-11 | 1998-11-24 | Xerox Corporation | Method for graph-based table recognition |
| WO2002063497A1 (en) * | 2001-02-08 | 2002-08-15 | Soda Technologies Pty Ltd | Proximity for computer represented graphs |
| JP2004342064A (en) * | 2003-05-16 | 2004-12-02 | Seki Tama Yonezawa | Human inference engine |
| WO2007027967A2 (en) * | 2005-08-31 | 2007-03-08 | Eagleforce Associates | System for hypothesis generation |
| KR20110134314A (en) * | 2010-06-08 | 2011-12-14 | 인터내셔널 비지네스 머신즈 코포레이션 | Graphical models for representing text documents for computer analysis |
| CN103124980A (en) * | 2010-09-24 | 2013-05-29 | 国际商业机器公司 | Providing answers to questions including assembling answers from multiple document segments |
| CN103324646A (en) * | 2012-03-19 | 2013-09-25 | 株式会社东芝 | Search supporting device and search supporting method |
| CN104699758A (en) * | 2015-02-04 | 2015-06-10 | 中国人民解放军装甲兵工程学院 | Intelligent generation system and method of graphic library-associated command document |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108536720B (en) | 2021-09-21 |
| JP6622236B2 (en) | 2019-12-18 |
| KR102039393B1 (en) | 2019-11-04 |
| KR20180101991A (en) | 2018-09-14 |
| JP2018147238A (en) | 2018-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Winter | Statistics for linguists: An introduction using R | |
| CN108536720A (en) | Conceive assisting system and design support method | |
| Ahmed et al. | AIML based voice enabled artificial intelligent chatterbot | |
| KR101971582B1 (en) | Method of providing health care guide using chat-bot having user intension analysis function and apparatus for the same | |
| WO2016036851A1 (en) | Method and system for determining edit rules for rewriting phrases | |
| Agarwal et al. | An overview of natural language processing | |
| Kotov et al. | Semantic parser for sentiment analysis and the emotional computer agents | |
| WO2019118257A1 (en) | Assertion-based question answering | |
| JP5726953B2 (en) | Dialog action assignment model learning device, dialog action assignment device, method, and program | |
| Mohapatra et al. | Text classification using NLP based machine learning approach | |
| Mani et al. | Hi, how can I help you?: Automating enterprise IT support help desks | |
| Dong et al. | Revisit input perturbation problems for llms: A unified robustness evaluation framework for noisy slot filling task | |
| Nassiri et al. | Approaches, methods, and resources for assessing the readability of arabic texts | |
| Zimmerman et al. | A blind spot for large language models: Supradiegetic linguistic information | |
| JP7549901B2 (en) | Presentation Evaluation Device | |
| JP7549902B2 (en) | Presentation Evaluation Device | |
| Kumar et al. | Natural language processing | |
| Zhang | Pronunciation ambiguities in Japanese kanji | |
| Pande et al. | Named Entity Recognition for Nepali Using BERT Based Models | |
| Vatian et al. | Comparative analysis of approaches to building medical dialog systems in Russian | |
| Freeman | Parsing using a grammar of word association vectors | |
| JP5690300B2 (en) | Attribute extraction device, attribute extraction method, and attribute extraction program | |
| Shetty et al. | English transliteration of Kannada words with Anusvara and Visarga | |
| Balakrishnan | Appliance of Artificial Intelligence Techniques to Develop a Chatterbot | |
| JP2023130197A (en) | Zero pronoun identifying device, zero pronoun identifying method, and program |
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 |