CN108062411A - A kind of system and method for searching electronic component data message - Google Patents
A kind of system and method for searching electronic component data message Download PDFInfo
- Publication number
- CN108062411A CN108062411A CN201711482590.6A CN201711482590A CN108062411A CN 108062411 A CN108062411 A CN 108062411A CN 201711482590 A CN201711482590 A CN 201711482590A CN 108062411 A CN108062411 A CN 108062411A
- Authority
- CN
- China
- Prior art keywords
- electronic component
- data
- component data
- reptile
- neutral net
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Abstract
The invention discloses a kind of system and method for searching electronic component data message, which includes:Data collection module, for collecting the electronic component data of separate sources, different-format;Training unit is trained for the electronic component being collected into data to be input to neutral net;Correction unit, for being corrected to the data that there is mistake in the electronic component data after training;Query unit, for user to the information of neural network inputs inquiry electronic component products, to obtain query result.User of the present invention is by inputting the product information of some electronic component to query unit, it can precisely, quickly find the electronic component, it is easy to use, it searches precisely, quickly, solves and inaccurate technical barrier is difficult to search or searched due to electronic component wide variety.
Description
Technical field
The present invention relates to the system and method for searching electronic component data message, more specifically a kind of lookup electronics
The system and method for component data message.
Background technology
It is according to some mostly using traditional database or search engine in traditional electronic component lookup system
Characteristic value searches corresponding electronic component data.It, can only be by inputting its item number, if not if necessary to search certain component
Know item number, then need accurately to know that certain several limited characteristic value can just find some component.At present, electronic component
Species is of all kinds, and data format is intricate, and similar title and parameter have very much, therefore in the feelings for not knowing specific item number
With traditional mode it is difficult to inquire some component product wanted to look up under condition.
The content of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide a kind of electronic component data messages searched to be
System and method.
To achieve the above object, the present invention uses following technical scheme:A kind of electronic component data message searched is
System, the system comprises:
Data collection module, for collecting the electronic component data of separate sources, different-format;
Training unit is trained for the electronic component being collected into data to be input to neutral net;
Correction unit, for being corrected to the data that there is mistake in the electronic component data after training;
Query unit, for user to the information of neural network inputs inquiry electronic component products, to obtain inquiry knot
Fruit.
Its further technical solution is:The data collection module includes:
Website reptile module is made of multiple reptile terminals, for obtaining the electronic component data of targeted website;
Task allocating module, for carrying out task distribution and deployment to each reptile terminal;
Data summarization module, the electronic component data for all reptile terminals to be got are parsed, summarized;
Data memory module, the electronic component data after parsing, summarize for preservation.
Its further technical solution is:The training unit includes:
Receiving module, for receiving, handling the electronic component data of input neutral net;
Memory module, for electronic component data to be remembered to treated;
Output module, for exporting the electronic component data result after remembering.
Its further technical solution is:The correction unit includes:
Input module, for inquiring about data to neural one electronic component of network inputs;
Feedback module, neutral net inquire about data according to the electronic component of input and provide feedback result;
Judgment module, for judging whether feedback result is correct;
Module is encouraged, if feedback result is correct, gives neutral net positive incentive, if feedback result mistake, to nerve
Network reverse energization.
A kind of method for searching electronic component data message, the described method includes:
Collect separate sources, the electronic component data of different-format;
Neutral net is input to being collected into electronic component data to be trained;
The data that there is mistake in electronic component data after training are corrected;
User inquires about the information of electronic component products to neural network inputs, to obtain query result.
Its further technical solution is:It is described collect separate sources, different-format electronic component data the step of, tool
Body comprises the following steps:
It is obtained by the electronic component data of reptile terminal-pair targeted website;
The electronic component data that all reptile terminals are got are parsed, are summarized;
Preserve parsing, summarize after electronic component data.
Its further technical solution is:The electronic component data by reptile terminal-pair targeted website are obtained
The step of before, further include, task distribution and deployment carried out to each reptile terminal.
Its further technical solution is:The step that the electronic component data input neutral net of described pair of collection is trained
Suddenly, following steps are specifically included:
It receives, the electronic component data of processing input neutral net;
To treated, electronic component data are remembered;
Electronic component data result after output memory.
Its further technical solution is:The data that there is mistake in electronic component data after described pair of training are entangled
Positive step, specifically includes following steps:
To neural one electronic component inquiry data of network inputs;
Neutral net inquires about data according to the electronic component of input and provides feedback result;
Judge whether feedback result is correct;
If feedback result is correct, neutral net positive incentive is given, if feedback result mistake, is reversely swashed to neutral net
It encourages.
Compared with the prior art, the invention has the advantages that:A kind of electronic component data message searched of the present invention is
System collects the electronic component data of separate sources, different-format by data collection module, will be collected by training unit
Electronic component data, which are input in neutral net, to be trained, and unit of rectifying a deviation can be to existing in electronic component after training
The data of mistake are corrected, and neutral net is enable comprehensively to grasp electronic component data, and by rectifying a deviation, unit constantly entangles
The accuracy that neutral net grasps data is continuously improved in the data of lookup error, and user to query unit by inputting some electricity
The product information of sub- component, you can precisely, the electronic component is quickly found, it is easy to use, it searches precisely, quickly,
It solves and inaccurate technical barrier is difficult to search or searched due to electronic component wide variety.
Above description is only the general introduction of technical solution of the present invention, can in order to better understand technical measure
Content according to specification is practiced, and in order to make above and other objects of the present invention, feature and advantage brighter
Aobvious understandable, special below to lift preferred embodiment, detailed description are as follows.
Description of the drawings
Fig. 1 is a kind of structure chart for the system specific embodiment for searching electronic component data message of the present invention;
Fig. 2 is data collection module in a kind of system specific embodiment for searching electronic component data message of the present invention
Structure chart;
Fig. 3 is the structure of training unit in a kind of system specific embodiment for searching electronic component data message of the present invention
Figure;
Fig. 4 is the structure of correction unit in a kind of system specific embodiment for searching electronic component data message of the present invention
Figure;
Fig. 5 is a kind of flow chart for the method specific embodiment for searching electronic component data message of the present invention;
Fig. 6 is to collect electronic component in a kind of method specific embodiment for searching electronic component data message of the present invention
The flow chart of step;
Fig. 7 is that neutral net is instructed in a kind of method specific embodiment for searching electronic component data message of the present invention
Practice the flow chart of step;
Fig. 8 is that wrong data corrects step in a kind of method specific embodiment for searching electronic component data message of the present invention
Rapid flow chart;
Fig. 9 is a kind of exemplary plot for the system and method specific embodiment for searching electronic component data message of the present invention.
Specific embodiment
In order to more fully understand the present invention technology contents, with reference to specific embodiment to technical scheme into
One step introduction and explanation, but not limited to this.
The specific embodiment shown in -4 is please referred to Fig.1, this implementation provides a kind of lookup electronic component data message
System, the system include:
Data collection module 1, for collecting the electronic component data of separate sources, different-format;
Training unit 2 is trained for the electronic component being collected into data to be input to neutral net;
Correction unit 3, for being corrected to the data that there is mistake in the electronic component data after training;
Query unit 4, for user to the information of neural network inputs inquiry electronic component products, to obtain inquiry knot
Fruit.
Specifically, data collection module 1, different electronic component data, these electronics member devices are collected from each website
Number of packages is according to the data for including different electronic component products or the number of the different expression-forms of same electronic component products
According to.Training unit 2 is trained for the electronic component being collected into data to be input to neutral net, is to allow nerve
Network can remember these electronic component data, and neutral net can constantly be allowed to remember various electronic component numbers
According to.Due to there is the data in the presence of mistake from the electronic component that different web sites are collected, then, it is necessary to right by unit 3 of rectifying a deviation
The data of mistake are corrected.Query unit 4 is mainly used for user and inquires about electronic component, and user is in inquiry electronic component
When, this electronic component can be found out with the relevant parameter of the electronic component by simply entering some, but in order to improve
Precision is searched, user can be several with the relevant parameter of this electronic component with multi input, then the result inquired is just more
Precisely.
Further, since the electronic component data of neutral net memory are when correcting the data of mistake by unit 3 of rectifying a deviation
It is not corrected to, then whether correct, so as to further right if can collect the data that user is inquired by query unit 4
Electronic component data in neutral net are corrected, by persistently being entangled to the electronic component data in neutral net
Just, search user more efficiently and accurate during electronic component.
Further, data collection module 1 includes:
Website reptile module 11 is made of multiple reptile terminals, for obtaining the electronic component data of targeted website;
Task allocating module 12, for carrying out task distribution and deployment to each reptile terminal;
Data summarization module 13, the electronic component data for all reptile terminals to be got are parsed, summarized;
Data memory module 14, the electronic component data after parsing, summarize for preservation.
Website reptile module 11 is made of multiple reptile terminals, and is parsed for the page elements of website, and is taken out
Data message useful in website is taken, such as, it would be desirable to reptile terminal collects electronic component data, then, reptile terminal
Just only it can collect the relevant data of electronic component.It, can be right by task allocating module 12 before reptile terminal crawls data
The situation of each reptile terminal distribution and deployment task, distribution and deployment task can be according to each website data resource or reality
Demand determine.Data summarization module 13 is that the electronic component data that reptile terminal is got are parsed and summarized,
And pass through data memory module 14 and preserve.In addition, during the collection of electronic component data, it is also necessary to whole to each reptile
End working condition is monitored, and judges whether reptile terminal works are normal, and whether the data crawled are correct etc..It is also required to climbing
Worm terminal carries out the parameter configuration of related work, website reptile is upgraded, the life cycle of reptile terminal and work carry out pipe
Reason.In addition, it is also necessary to which management is scheduled to reptile terminal operating mode, time, stopping.
Further, training unit 2 includes:
Receiving module 21, for receiving, handling the electronic component data of input neutral net;
Memory module 22, for electronic component data to be remembered to treated;
Output module 23, for exporting the electronic component data result after remembering.
Wherein, memory module 22 is for remembering electronic component data, the ratio that this Memory Process can be popular
Make people and remembering some thing, it is notable that people in Memory Process it is possible that can forget, but neutral net memory thing
It will not forget, it is only necessary to which once training can be remembered.Output module 23 is the electronic component data result after output memory,
Can be regarded as man memory some thing the result is that kind of.
Further, correction unit 3 includes:
Input module 31, for inquiring about data to neural one electronic component of network inputs;
Feedback module 32, neutral net inquire about data according to the electronic component of input and provide feedback result;
Judgment module 33, for judging whether feedback result is correct;
Module 34 is encouraged, if feedback result is correct, gives neutral net positive incentive, if feedback result mistake, to god
Through network reverse energization.
Specifically, judgment module 33, for judging whether feedback result is correct, the foundation of judgement can be by manually sentencing
Disconnected, since reptile terminal is swashed from network, the data got are variant or mistake when being possible to, and at this moment just need manually to sentence
Disconnected and verification, if correct, then by the way that module 34 is encouraged to tell neutral net to one positive incentive of neutral net
This data is correct, if artificial judgment is with verification, this data is wrong, then by encouraging module 34 to nerve net
One reverse energization of network tells that this data of neutral net are wrong.
Referring to Fig. 9, when the neutral net that training is used to complete searches for electronic component, for example, input:Input voltage
370 VDC Input Voltage=85 VAC to of=85VAC to 264 VAC, 120 VDC to 264 VAC, 120
370 VDC of VDC to 370 VDC or 85 VAC to 264 VAC, 120 VDC to etc. is various can to express the similar meaning
Natural language, corresponding electronic component products can be successfully searched.Multiple parameters can be further inputted simultaneously
It more precisely filters, parameter and the respective value of any combination of Chinese and English can be inputted to search electronic component products.
In addition, present embodiments providing the code on reptile terminal, code is as follows:
CrawlClient.java
package com.giiso.ai.crawl;
importjava.util.List;
/**
* reptile client, comprising a series of index, server-side index, decision how many task will be returned to according to these values.
*@author giiso
*
*/
public class CrawlClient{
//cpu indexs
private int cpu;
// indicator memory
private int memo;
// network index
private int net;
The task quantity that // client receives every time
private int size;
// comprehensive scores
private int score;
// blacklist, the data that server-side will not distribute in blacklist are given to client private List<String>
blackList;
public int getCpu(){
return cpu;
}
public void setCpu(int cpu){
This.cpu=cpu;
}
public int getMeno(){
return memo;
}
public void setMeno(int meno){
This.memo=meno;
}
public int getNet(){
return net;
}
public void setNet(int net){
This.net=net;
}
public int getSize(){
return size;
}
public void setSize(int size){
This.size=size;
}
public int getScore(){
return score;
}
public void setScore(int score){
This.score=score;
}
public List<String>getBlackList(){
return blackList;
}
public void setBlackList(List<String>blackList){
This.blackList=blackList;
}
}
CrawlManager.java
package com.giiso.ai.crawl;
/**
* reptile management class, manages all reptile clients
*@author giiso
*
*/
public class CrawlManager{
/**
* reptile client holder
*/
private CrawlClientHolder crawlclientHolder;
/**
* reptile client registers
*@param client
*/
public void register(CrawlClient client){
clientHolder.add(client);
}
/**
* all reptile clients are returned
*@return
*/
public List<CrawlClient>listAll(){
clientHolder.listAll();
}
/**
* paging returns to reptile client
*@return
*/
public List<CrawlClient>lis(int pageSize,int pageSize){
clientHolder.listAll(int pageSize,int pageSize);
}
}
TaskDistribute.java
package com.giiso.ai.crawl;
/**
* task is distributed
*@author giiso
*
*/
public class TaskDistribute{
/**
* reptile task pool
*/
private CrawlTaskPool crawltaskPool;
/**
* reptile task,
*@param crawlclient reptiles task client object
*@return
*/
public CrawlTask obtainTask(CrawlClient crawlclient){
taskPool.obtainTask(crawlclient);
}
}
The present embodiment additionally provides the code of neural metwork training, and code is as follows:
TrainManager.java
package com.giiso.ai.train;
/**
* data training management device
*@author giiso
*
*/
public class TrainManager{
Private Neural neural=NeuralHolder.getNeural ();public void train
(TrainData trainData){
neural.train(trainData);
}
public ResultData getData(String keyword){
return neural.getData(keyword);
}
public void reward(String resultId){
reward(resultId,1);
}
public void reward(String resultId,int score){
neural.reward(resultId,score);
}
public void punish(String resultId){
punish(resultId,1);
}
public void punish(String resultId,int score){
neural.punish(resultId,score);
}
}
The specific embodiment shown in Fig. 5-8 is referred to, present embodiments provides and electronics is searched using the quick of above system
The method of component data message, this method include:
S10, separate sources, the electronic component data of different-format are collected;
S20, neutral net is input to being collected into electronic component data is trained;
S30, the data that there is mistake in the electronic component data after training are corrected;
S40, user inquire about the information of electronic component products to neural network inputs, to obtain query result.
Specifically, collecting different electronic component data from each website, these electronic component data include difference
The data of the data of the electronic component products either different expression-forms of same electronic component products.For that will be collected into
Electronic component data be input to neutral net and be trained, be to allow neutral net that can remember these electronic component numbers
According to neutral net can constantly be allowed to remember various electronic component data.Due to the electronics collected from different web sites
There are the data in the presence of mistake in component, then, it is necessary to be corrected to the data of mistake.User is in inquiry electronic component
When, this electronic component can be found out with the relevant parameter of the electronic component by simply entering some, but in order to improve
Precision is searched, user can be several with the relevant parameter of this electronic component with multi input, then the result inquired is just more
Precisely.
Further, since the electronic component data of neutral net memory are not corrected to when correcting the data of mistake, that
Can collect whether the data that user inquires are correct, so as to further to the electronic component data in neutral net
It is corrected, by persistently being corrected to the electronic component data in neutral net, when making user's lookup electronic component
It is more efficiently and accurate.
Further, in certain embodiments, step S10 specifically includes following steps:It is S11, right
Each reptile terminal carries out task distribution and deployment.
S12, obtained by the electronic component data of reptile terminal-pair targeted website;
S13, the electronic component data that all reptile terminals are got are parsed, are summarized;
S14, preserve parsing, summarize after electronic component data.
Specifically, reptile terminal is parsed for the page elements of website, and data message useful in website is extracted,
Such as, it would be desirable to reptile terminal looks for electronic component data, then, it is relevant that reptile terminal just can only collect electronic component
Data., can be by each reptile terminal distribution and deployment task before reptile terminal crawls data, distribution and deployment task
Situation can be determined according to the demand of each website data resource or reality.The electronic component that reptile terminal is got
Data are parsed and summarized, and are preserved.In addition, during the collection of electronic component data, it is also necessary to each reptile
Terminal works situation is monitored, and judges whether reptile terminal works are normal, and whether the data crawled are correct etc..It is also required to pair
Reptile terminal carries out the parameter configuration of related work, and website reptile is upgraded, and the life cycle of reptile terminal and work carry out
Management.In addition, it is also necessary to which management is scheduled to reptile terminal operating mode, time, stopping.
Further, in certain embodiments, step S20 specifically includes following steps:
S21, reception, the electronic component data of processing input neutral net;
S22, to treated, electronic component data are remembered;
Electronic component data result after S23, output memory.
Wherein, electronic component data are remembered, what this Memory Process can be popular be compared to, and people is remembering some
Thing, it is notable that people is it is possible that can forget in Memory Process, but neutral net memory thing will not be forgotten, only need
Training it can once remember.Output module 23 is the electronic component data result after output memory, you can is understood for people
Memory some thing the result is that kind of.
Further, in certain embodiments, step S30 specifically includes following steps:
S31, data are inquired about to neural one electronic component of network inputs;
S32, neutral net inquire about data according to the electronic component of input and provide feedback result;
S33, judge whether feedback result is correct;
If S34, feedback result are correct, neutral net positive incentive is given, if feedback result mistake, anti-to neutral net
To excitation.
Specifically, judging whether feedback result is correct, the foundation of judgement can be by artificial judgment, due to reptile terminal
Mistake, at this moment just needs artificial judgment and verification when the data got of swashing from network are variant or are possible to, if
It is correct, then give neutral net one positive incentive, tell this data of neutral net be it is correct, if artificial judgment and
It is wrong to verify this data, then gives neutral net one reverse energization, tells that this data of neutral net are wrong.
Referring to Fig. 9, when the neutral net that training is used to complete searches for electronic component, for example, input:Input voltage
370 VDC Input Voltage=85 VAC to of=85 VAC to 264 VAC, 120 VDC to, 264 VAC,
370 VDC of 120 VDC to 370 VDC or 85 VAC to 264 VAC, 120 VDC to etc. is various can express it is similar
The natural language of the meaning can be successfully searched corresponding electronic component products.It can further input simultaneously multiple
Parameter more precisely filters, parameter and the respective value of any combination of Chinese and English can be inputted to search electronic component products.
In addition, present embodiments providing the code on reptile terminal, code is as follows:
CrawlClient.java
package com.giiso.ai.crawl;
importjava.util.List;
/**
* reptile client, comprising a series of index, server-side index, decision how many task will be returned to according to these values.
*@author giiso
*
*/
public class CrawlClient{
//cpu indexs
private int cpu;
// indicator memory
private int memo;
// network index
private int net;
The task quantity that // client receives every time
private int size;
// comprehensive scores
private int score;
// blacklist, the data that server-side will not distribute in blacklist are given to client
private List<String>blackList;
public int getCpu(){
return cpu;
}
public void setCpu(int cpu){
This.cpu=cpu;
}
public int getMeno(){
return memo;
}
Public void setMeno (int meno) { this.memo=meno;
}
public int getNet(){
return net;
}
public void setNet(int net){
This.net=net;
}
public int getSize(){
return size;
}
public void setSize(int size){
This.size=size;
}
public int getScore(){
return score;
}
public void setScore(int score){
This.score=score;
}
public List<String>getBlackList(){
return blackList;
}
public void setBlackList(List<String>blackList){
This.blackList=blackList;
}
}
CrawlManager.java
package com.giiso.ai.crawl;
/**
* reptile management class, manages all reptile clients
*@author giiso
*
*/
public class CrawlManager{
/**
* reptile client holder
*/
private CrawlClientHolder crawlclientHolder;
/**
* reptile client registers
*@param client
*/
public void register(CrawlClient client){
clientHolder.add(client);
}
/**
* all reptile clients are returned
*@return
*/
public List<CrawlClient>listAll(){
clientHolder.listAll();
}
/**
* paging returns to reptile client
*@return
*/
public List<CrawlClient>lis(int pageSize,int pageSize){
clientHolder.listAll(int pageSize,int pageSize);
}
}
TaskDistribute.java
package com.giiso.ai.crawl;
/**
* task is distributed
*@author giiso
*
*/
public class TaskDistribute{
/**
* reptile task pool
*/
private CrawlTaskPool crawltaskPool;
/**
* reptile task,
*@param crawlclient reptiles task client object
*@return
*/
public CrawlTask obtainTask(CrawlClient crawlclient){
taskPool.obtainTask(crawlclient);
}
}
The present embodiment additionally provides the code of neural metwork training, and code is as follows:
TrainManager.java
package com.giiso.ai.train;
/**
* data training management device
*@author giiso
*
*/
public class TrainManager{
Private Neural neural=NeuralHolder.getNeural ();
public void train(TrainData trainData){
neural.train(trainData);
}
public ResultData getData(String keyword){
return neural.getData(keyword);
}
public void reward(String resultId){
reward(resultId,1);
}
public void reward(String resultId,int score){
neural.reward(resultId,score);
}
public void punish(String resultId){
punish(resultId,1);
}
public void punish(String resultId,int score){
neural.punish(resultId,score);
}
}
It is above-mentioned only with embodiment come the technology contents that further illustrate the present invention, in order to which reader is easier to understand, but not
It represents embodiments of the present invention and is only limitted to this, any technology done according to the present invention extends or recreation, by the present invention's
Protection.Protection scope of the present invention is subject to claims.
Claims (9)
1. a kind of system for searching electronic component data message, which is characterized in that the system comprises:
Data collection module, for collecting the electronic component data of separate sources, different-format;
Training unit is trained for the electronic component being collected into data to be input to neutral net;
Correction unit, for being corrected to the data that there is mistake in the electronic component data after training;
Query unit, for user to the information of neural network inputs inquiry electronic component products, to obtain query result.
A kind of 2. system for searching electronic component data message according to claim 1, which is characterized in that the data
Collector unit includes:
Website reptile module is made of multiple reptile terminals, for obtaining the electronic component data of targeted website;
Task allocating module, for carrying out task distribution and deployment to each reptile terminal;
Data summarization module, the electronic component data for all reptile terminals to be got are parsed, summarized;
Data memory module, the electronic component data after parsing, summarize for preservation.
A kind of 3. system for searching electronic component data message according to claim 1, which is characterized in that the training
Unit includes:
Receiving module, for receiving, handling the electronic component data of input neutral net;
Memory module, for electronic component data to be remembered to treated;
Output module, for exporting the electronic component data result after remembering.
A kind of 4. system for searching electronic component data message according to claim 1, which is characterized in that the correction
Unit includes:
Input module, for inquiring about data to neural one electronic component of network inputs;
Feedback module, neutral net inquire about data according to the electronic component of input and provide feedback result;
Judgment module, for judging whether feedback result is correct;
Module is encouraged, if feedback result is correct, neutral net positive incentive is given, if feedback result mistake, to neutral net
Reverse energization.
A kind of 5. method for searching electronic component data message, which is characterized in that the described method includes:
Collect separate sources, the electronic component data of different-format;
Neutral net is input to being collected into electronic component data to be trained;
The data that there is mistake in electronic component data after training are corrected;
User inquires about the information of electronic component products to neural network inputs, to obtain query result.
A kind of 6. method for searching electronic component data message according to claim 5, which is characterized in that the collection
Separate sources, different-format electronic component data the step of, specifically include following steps:
It is obtained by the electronic component data of reptile terminal-pair targeted website;
The electronic component data that all reptile terminals are got are parsed, are summarized;
Preserve parsing, summarize after electronic component data.
7. a kind of method for searching electronic component data message according to claim 6, which is characterized in that described to pass through
Before the step of electronic component data of reptile terminal-pair targeted website are obtained, further include, to each reptile terminal into
Row task is distributed and deployment.
A kind of 8. method for searching electronic component data message according to claim 5, which is characterized in that described pair of receipts
The step of electronic component data input neutral net of collection is trained, specifically includes following steps:
It receives, the electronic component data of processing input neutral net;
To treated, electronic component data are remembered;
Electronic component data result after output memory.
A kind of 9. method for searching electronic component data message according to claim 5, which is characterized in that described pair of instruction
The step of data that there is mistake in electronic component data after white silk are corrected, specifically includes following steps:
To neural one electronic component inquiry data of network inputs;
Neutral net inquires about data according to the electronic component of input and provides feedback result;
Judge whether feedback result is correct;
If feedback result is correct, neutral net positive incentive is given, if feedback result mistake, gives neutral net reverse energization.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711482590.6A CN108062411A (en) | 2017-12-29 | 2017-12-29 | A kind of system and method for searching electronic component data message |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711482590.6A CN108062411A (en) | 2017-12-29 | 2017-12-29 | A kind of system and method for searching electronic component data message |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108062411A true CN108062411A (en) | 2018-05-22 |
Family
ID=62140897
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711482590.6A Pending CN108062411A (en) | 2017-12-29 | 2017-12-29 | A kind of system and method for searching electronic component data message |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108062411A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110674384A (en) * | 2019-09-27 | 2020-01-10 | 厦门晶欣电子有限公司 | Component model matching method |
CN111274158A (en) * | 2020-02-27 | 2020-06-12 | 北京首汽智行科技有限公司 | Data verification method |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1182592A1 (en) * | 2000-08-18 | 2002-02-27 | SER Systeme AG Produkte und Anwendungen der Datenverarbeitung | Method of process modelling |
CN101196905A (en) * | 2007-12-05 | 2008-06-11 | 覃征 | Intelligent pattern searching method |
CN103593478A (en) * | 2013-11-28 | 2014-02-19 | 广东工业大学 | Resource retrieval method and device of electronic component |
CN104572965A (en) * | 2014-12-31 | 2015-04-29 | 南京理工大学 | Search-by-image system based on convolutional neural network |
CN106202010A (en) * | 2016-07-12 | 2016-12-07 | 重庆兆光科技股份有限公司 | The method and apparatus building Law Text syntax tree based on deep neural network |
CN106339756A (en) * | 2016-08-25 | 2017-01-18 | 北京百度网讯科技有限公司 | Training data generation method and device and searching method and device |
CN107220296A (en) * | 2017-04-28 | 2017-09-29 | 北京拓尔思信息技术股份有限公司 | The generation method of question and answer knowledge base, the training method of neutral net and equipment |
-
2017
- 2017-12-29 CN CN201711482590.6A patent/CN108062411A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1182592A1 (en) * | 2000-08-18 | 2002-02-27 | SER Systeme AG Produkte und Anwendungen der Datenverarbeitung | Method of process modelling |
CN101196905A (en) * | 2007-12-05 | 2008-06-11 | 覃征 | Intelligent pattern searching method |
CN103593478A (en) * | 2013-11-28 | 2014-02-19 | 广东工业大学 | Resource retrieval method and device of electronic component |
CN104572965A (en) * | 2014-12-31 | 2015-04-29 | 南京理工大学 | Search-by-image system based on convolutional neural network |
CN106202010A (en) * | 2016-07-12 | 2016-12-07 | 重庆兆光科技股份有限公司 | The method and apparatus building Law Text syntax tree based on deep neural network |
CN106339756A (en) * | 2016-08-25 | 2017-01-18 | 北京百度网讯科技有限公司 | Training data generation method and device and searching method and device |
CN107220296A (en) * | 2017-04-28 | 2017-09-29 | 北京拓尔思信息技术股份有限公司 | The generation method of question and answer knowledge base, the training method of neutral net and equipment |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110674384A (en) * | 2019-09-27 | 2020-01-10 | 厦门晶欣电子有限公司 | Component model matching method |
CN111274158A (en) * | 2020-02-27 | 2020-06-12 | 北京首汽智行科技有限公司 | Data verification method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109145216A (en) | Network public-opinion monitoring method, device and storage medium | |
CN102710646B (en) | Method and system for collecting phishing websites | |
CN109408821B (en) | Corpus generation method and device, computing equipment and storage medium | |
EA201300375A1 (en) | THE METHOD OF ORGANIZING A SEARCH DATABASE USING FUZZY CRITERIA | |
CN101902470A (en) | Form feature-based Web security vulnerability dynamic testing method | |
CN105302815B (en) | The filter method and device of the uniform resource position mark URL of webpage | |
CN106776567A (en) | A kind of internet big data analyzes extracting method and system | |
CN112989348B (en) | Attack detection method, model training method, device, server and storage medium | |
CN112989157A (en) | Method and device for detecting crawler request | |
CN103294732A (en) | Web page crawling method and spider | |
CN106547803B (en) | Method and device for crawling incremental resources of website | |
CN103399872A (en) | Method and device for optimizing webpage capture | |
CN108062411A (en) | A kind of system and method for searching electronic component data message | |
CN111737494A (en) | Knowledge graph generation method of intelligent learning system | |
CN112559923A (en) | Website resource recommendation method and device, electronic equipment and computer storage medium | |
CN105630937A (en) | Method and device for searching answers to exam questions | |
CN109302418A (en) | A kind of malice domain name detection method and device based on deep learning | |
CN103778156A (en) | Method and device for searching for data and server for data search | |
CN109698798A (en) | A kind of recognition methods of application, device, server and storage medium | |
CN103593344B (en) | A kind of information collecting method and device | |
CN113918794B (en) | Enterprise network public opinion benefit analysis method, system, electronic equipment and storage medium | |
CN105183843B (en) | list page identification system and method | |
CN106484913A (en) | Method and server that a kind of Target Photo determines | |
CN111158973B (en) | Web application dynamic evolution monitoring method | |
CN106547780A (en) | Article reprints statistics of variables method and device |
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 |