CN113779090B - Intelligent prediction caching method for Handle identification analysis - Google Patents

Intelligent prediction caching method for Handle identification analysis Download PDF

Info

Publication number
CN113779090B
CN113779090B CN202111073296.6A CN202111073296A CN113779090B CN 113779090 B CN113779090 B CN 113779090B CN 202111073296 A CN202111073296 A CN 202111073296A CN 113779090 B CN113779090 B CN 113779090B
Authority
CN
China
Prior art keywords
handle
cache
identifier
association rule
code
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.)
Active
Application number
CN202111073296.6A
Other languages
Chinese (zh)
Other versions
CN113779090A (en
Inventor
白宏钢
霍健
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mako Workshop Industrial Technology Beijing Co ltd
Original Assignee
Mako Workshop Industrial Technology Beijing Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Mako Workshop Industrial Technology Beijing Co ltd filed Critical Mako Workshop Industrial Technology Beijing Co ltd
Priority to CN202111073296.6A priority Critical patent/CN113779090B/en
Publication of CN113779090A publication Critical patent/CN113779090A/en
Application granted granted Critical
Publication of CN113779090B publication Critical patent/CN113779090B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2455—Query execution
    • G06F16/24552—Database cache management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/242—Query formulation
    • G06F16/2433—Query languages

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses an intelligent prediction caching method for Handle identification analysis, which comprises the following steps: dividing the analysis whole according to the Handle identifier into an identifier prefix coding query and an identifier complete coding query, and caching analysis results of the Handle identifier prefix and the Handle complete identifier originally, wherein the association rule trainer respectively carries out association rule mining on the Handle identifier prefix and the Handle identifier complete code by utilizing Aprior algorithm according to logs of the prior analysis process, and finally stores the association rules into an association rule trainer in the cache. And then in the process of analyzing the identifier, the intelligent prediction cache can combine the stored identifier codes in the cache, predict the upcoming Handle identifier prefix and the Handle identifier complete codes in real time according to the association rule in the association rule trainer, and then respectively obtain the analysis results corresponding to the prefix and the Handle identifier complete codes in advance and store the analysis results in the cache.

Description

Intelligent prediction caching method for Handle identification analysis
Technical Field
The invention belongs to the technical field of Handle identification analysis of industrial Internet, relates to a caching method in Handle identification analysis, and particularly relates to a method for intelligently predicting, analyzing and encoding and intelligently caching Handle identification.
Background
With the continuous development of industrial Internet identification analysis platforms, especially the continuous popularization and application of Handle identification analysis in the industrial Internet identification analysis field of China, the whole Handle identification analysis system will receive more and more Handle identification inquiry requests in the future, the communication traffic of the system will be larger and larger, and the communication pressure of the system is larger and larger, so that the speed and the efficiency of Handle identification analysis can be subjected to larger test.
The Handle identifier resolution system is a hierarchical identifier resolution system, the upper layer is GHR (GLobal HANDLE REGISTRY ), and the lower layer is LHS (Local HANDLE SERVICE ). The Handle client firstly sends a query request coded by the identification prefix to the GHR through the recursion node, then the GHR returns information of the corresponding LHS service site, then the Handle client sends a query analysis request coded by the whole identification code to the corresponding LHS service site through the recursion node, and finally the LHS returns a final analysis result to the Handle client.
Although, in order to improve the analysis speed and shorten the analysis time delay, the Handle identifier analysis system is designed with corresponding cache service. There are two types of caching services in the Handle identity resolution system: (1) The method is a cache service of a request result of the Handle identification prefix, namely when the Handle client sends an identification prefix query request to the GHR through a recursion node, the GHR returns corresponding LHS site information, and then the Handle identification analysis system stores the queried Handle identification prefix and the corresponding LHS site information thereof in a cache, and when the same identification prefix request exists, the corresponding LHS site information can be directly acquired from the cache for subsequent query without passing through the GHR. (2) The other type is a cache service of the analysis result of the whole Handle identifier, namely the Handle identifier analysis system can directly store the Handle identifier which is inquired to be successful by the Handle client and the analysis result thereof in a cache, and when the Handle client inquires the same identifier again, a series of inquiry processes are not needed, and the corresponding analysis result can be directly obtained from the cache.
However, the local cache is always stored based on the queried identifier code no matter the prefix of the Handle identifier or the complete Handle identifier code, after each time of the whole query request process, the corresponding identifier code successfully queried at the time can be cached and stored, and the storage efficiency is low, so that the identifier analysis efficiency is reduced correspondingly, and the analysis time delay is prolonged.
In fact, certain relevance exists in some inquiry requests of identification codes, for example, after inquiring certain types of toothpastes of a certain brand, other types of toothpastes of the brand are likely to be inquired continuously, or other brands of toothpastes of the same type are likely to be inquired, and the like, so that certain trend exists in the inquiry of the identification codes in the Handle identification analysis process, and a brand-new intelligent prediction caching method is designed aiming at the defects of the current cache: according to relevance and trend of identification code inquiry, the method utilizes Aprior algorithm to intelligently predict the Handle identification code to be analyzed, and caches the corresponding analysis result in advance, so that the inquired identification code analysis result and the predicted identification code analysis result exist in the cache, the intelligent prediction cache replaces the original local cache, communication pressure when a large number of analysis requests exist can be further relieved, analysis time delay is reduced, and analysis efficiency is improved.
At present, related intelligent prediction caching methods are briefly mentioned in intelligent caching method patents (an internet of things identifier analysis method with an intelligent caching module, CN 201210328283.3) related to the internet of things identifier analysis field published by researchers, but the specific method is specially aimed at the internet of things identifier analysis field, and is different from the concept that a new cache is built to replace the original cache on the basis of the original common cache, and in addition, the internet of things identifier is greatly different from an industrial internet Handle identifier coding method and an industrial internet Handle identifier analysis method, so that the method cannot be fully applied to the industrial internet Handle identifier analysis field; in addition, many researchers also put forward related intelligent prediction caching methods in other fields of the internet, for example, patents (a time-series database caching management method based on access trend prediction, CN 201510733108.6) issued by researchers on caching management methods of time-series databases also put forward methods for data prediction and loading in advance according to data access trends, but the methods are specially put forward for data access characteristics in the time-series databases, and the characteristics of the methods are not in accordance with Handle identification analysis in practice, so that various methods related to the internet cannot be fully applied to Handle identification analysis.
Disclosure of Invention
The invention aims to overcome the problems in the prior art and provides an intelligent prediction caching method based on Aprior algorithm, which is applicable to the field of Handle identification analysis.
In order to achieve the purpose of the invention, the technical scheme adopted is as follows:
The invention discloses an intelligent prediction caching method oriented to Handle identification analysis, which designs a new intelligent prediction caching, and the intelligent prediction caching comprises an association rule trainer, and comprises the following steps:
step 1, an association rule trainer in an intelligent prediction cache performs association rule mining;
In the association rule trainer, according to the characteristic that the whole analysis of the Handle identifier is divided into two parts of identifier prefix coding inquiry and identifier complete coding inquiry, and the Handle identifier analysis system originally comprises two types of caching services of analysis result caching of the Handle identifier prefix and analysis result caching of the Handle complete identifier, the association rule is mined on the Handle identifier prefix and the Handle identifier complete coding respectively by utilizing Aprior algorithm according to logs of the prior analysis process, and finally the association rules are stored in the association rule trainer in the cache.
Step 2, constructing and applying an intelligent prediction cache in Handle identification analysis;
In the process of identification analysis, the intelligent prediction cache can combine the stored identification codes in the cache, predict the upcoming Handle identification prefix and Handle identification complete codes in real time according to the association rules in the association rule trainer, then respectively obtain the analysis results corresponding to the respective Handle identification prefix and Handle identification complete codes in advance and store the analysis results in the cache, thereby constructing a new intelligent prediction cache on the basis of the original cache function, replacing the original common cache which only stores the queried identification codes, and further improving the analysis efficiency.
(1) The step 1 specifically comprises the following steps:
1) Firstly, setting a periodic updating time T (T should be a smaller time range, and is within a range of 1S-5S) of the content of an association rule trainer in an intelligent prediction cache, and setting a minimum support degree S and a minimum confidence degree C of an association rule;
2) Respectively establishing a transaction table A of a Handle identifier complete code and a transaction table B of a Handle identifier prefix code according to a log file in an analysis process, wherein the transaction table A comprises Handle identifier complete code records inquired at regular intervals, and the transaction table B comprises Handle identifier prefix code records inquired at regular intervals;
3) Excavating a frequent item set a of a Handle identifier complete code meeting the minimum support degree S according to the transaction table A, and excavating a frequent item set B of a Handle identifier prefix meeting the minimum support degree S according to the transaction table B;
4) Calculating a correlation rule R a of the Handle identifier complete code meeting the minimum confidence coefficient C according to the frequent item set a, and similarly, calculating a correlation rule R b of the Handle identifier prefix code meeting the minimum confidence coefficient C according to the frequent item set b;
5) Finally, the mined association rule R a and the association rule R b are stored in an association rule trainer, so that mining and storing of the identification code association rule is completed once.
6) The invention can also be designed: and (3) updating the association rule in the association rule trainer once every a set time T, namely repeating the steps 1) to 5), so as to realize dynamic updating of the association rule and ensure the accuracy and the effectiveness of the intelligent prediction cache storage to a certain extent.
Further, let i= { I 1,i2,...,im } be a set of m different items, each I k (k=1, 2.,; m) is referred to as an item, assuming that the association rule is: wherein/> And/>The calculation formulas of the minimum support S and the minimum confidence C of the association rule in the step 1) are as follows:
further, in the step 6), when the association rule in the "association rule trainer" is updated periodically, the memory of the trainer is limited due to the limited memory of the intelligent prediction cache, and the present invention designs to replace the content in the trainer by a method similar to LRU (LEAST RECENTLY Used ) algorithm, which comprises the following steps:
a. The invention is designed as follows: the time when each association rule was successfully matched last is stored in the association rule trainer, so that the heat degree of the association rule used is represented (namely, the closer the time is, the higher the heat degree is; the farther the time is, the lower the heat degree is);
b. When the update time T is reached, the present invention is designed to: the association rules obtained by the re-prediction in the steps 1) to 5) are replaced in sequence, namely the association rules which are not recorded with the matching time in the trainer, namely the association rules which are not matched are replaced in sequence;
c. and then, replacing from far to near in sequence according to the recorded last successful matching time until all association rules obtained by new training are successfully stored or the memory of the trainer is full.
(2) The step 2 specifically comprises the following procedures:
1) Firstly, the intelligent prediction cache judges whether the prediction cache combining the Handle identifier complete code and the Handle identifier prefix stored in the current cache can be performed in advance according to the association rule in the association rule trainer:
a. if predictive caching can be performed in advance, obtaining a Handle identifier complete code and a Handle identifier prefix which can be predicted at present and are about to be provided with an analysis request, and storing the corresponding analysis result in the cache in advance;
b. if the prediction storage of any Handle identifier complete code or Handle identifier prefix cannot be realized, skipping the step, and directly starting from the step 2);
2) Whenever a Handle client initiates a Handle identifier complete code query analysis request, the Handle client will first query in an intelligent prediction cache whether there is a current Handle identifier complete code analysis result:
a. If the query is successful, the identification code may be queried before the query or the intelligent prediction cache successfully predicts the Handle identification code, and then the analysis result is directly returned, namely the query is finished;
b. If the inquiry fails, continuing to step 3);
3) When the Handle client side does not directly inquire the analysis result of the complete code of the current Handle identifier in the intelligent prediction cache, further inquiry analysis is needed, and at this time, the analysis system firstly carries out matching inquiry on the prefix code of the Handle identifier of the current request and the prefix of the Handle identifier stored in the intelligent prediction cache:
a. If the query is successful, the fact that the identification prefix code is queried or the intelligent prediction cache successfully predicts the identification prefix code possibly exists before is indicated, corresponding LHS service site information can be directly obtained without prefix query to GHR, then a complete identification code request is continuously sent to the LHS service site, an analysis result is directly obtained, and correspondingly, the system also stores the complete identification code analysis result into the intelligent prediction cache and continues to carry out the step b;
b. Step a is carried out, wherein a new Handle identifier complete code is stored in the cache at this time, the intelligent prediction cache can judge whether the Handle identifier complete code which is about to be provided with an analysis request and is not stored in the cache can be predicted according to an association rule R a in an association rule trainer in real time, if the new Handle identifier complete code can be predicted, the predicted Handle identifier complete code and an analysis result thereof are stored in the cache in advance, the query is finally ended, and otherwise, the query is carried out to the step a;
c. If the inquiry fails, continuing to step 4);
4) When the Handle client cannot inquire the analysis result of the prefix code of the current Handle identifier in the intelligent prediction cache, the Handle identifier analysis needs to finish inquiry according to the original complete analysis flow, and the method comprises the following steps:
firstly, a Handle client sends an identification prefix query request to a GHR through a recursion node to acquire information of a corresponding LHS service site, and synchronously stores a Handle identification prefix coding analysis result into an intelligent prediction cache;
b. Step a is performed, wherein a new Handle identifier prefix code is stored in the cache, the intelligent prediction cache can judge whether the Handle identifier prefix code which is about to be requested for analysis and is not stored in the cache can be predicted according to the association rule R b in the association rule trainer in real time, and if the brand new Handle identifier prefix code can be predicted, the predicted Handle identifier prefix code and the analysis result thereof are stored in the cache in advance;
c. meanwhile, after the step a, the Handle client continues to send a Handle complete identification query request to the corresponding LHS service site through the recursion node, and an analysis result is obtained and stored in the intelligent prediction cache;
d. and c, after the step of storing a new Handle identifier complete code in the cache, judging whether the Handle identifier complete code which is about to be submitted to an analysis request and is not stored in the cache can be predicted in real time according to the association rule R a in the association rule trainer by the intelligent prediction cache, storing the predicted Handle identifier complete code and the analysis result thereof in the cache in advance if the brand new Handle identifier complete code can be predicted, and finally ending the query, otherwise, ending the query in the step of c.
Compared with the prior art, the invention has the advantages that:
1. according to the method, the trend and the relevance existing in the Handle identification code analysis request are utilized, the upcoming Handle identification code is predicted according to the association rule of the mined Handle identification code, the analysis result is put into the cache in advance, a new intelligent prediction cache is constructed on the basis of keeping the original cache function, the cache for only storing the queried identification code analysis result is replaced, and the cache storage efficiency is further improved.
2. The intelligent prediction cache designed by the invention not only can directly obtain the queried identification code analysis result, but also can directly obtain the queried but predicted identification code analysis result, thereby greatly improving the Handle identification analysis efficiency, further relieving the communication pressure when a large number of Handle identification requests and accelerating the analysis speed.
3. Aiming at the characteristic of analysis and query of the Handle identifier, the method not only predicts and caches the Handle identifier code, but also predicts and caches the prefix of the Handle identifier. The analysis result of the Handle identification complete code can not be directly obtained in the cache, when analysis query is still needed to be carried out on each level of nodes, the current identification prefix code can be more efficiently matched and queried in the prediction cache, LHS service site information is obtained by matching in advance, and the analysis efficiency of the whole Handle identification system is greatly improved.
4. The intelligent prediction cache designed by the invention is implemented by a real-time prediction cache mechanism, and particularly means that the intelligent prediction cache can conduct possible prediction in real time according to the mined association rule after the content in the cache changes every time the cache stores a new Handle identification code, so that the timeliness of the cache content is ensured to a certain extent, and the association rule stored in the association rule trainer is periodically and dynamically updated, so that the accuracy and the effectiveness of the intelligent prediction cache storage are ensured to a certain extent.
Drawings
FIG. 1 shows a Handle identity resolution overview flowchart of the present invention.
FIG. 2 shows a Handle identity resolution architecture diagram of the present invention.
FIG. 3 shows a flow chart of the association rule trainer mining association rules.
FIG. 4 shows an internal block diagram of the intelligent prediction cache.
Detailed Description
The present invention will be described in detail below by way of specific examples with reference to the accompanying drawings, but is not limited thereto.
The embodiment provides a Handle identification analysis example adopting the intelligent prediction caching method.
An intelligent prediction caching method for Handle identification analysis comprises the following steps:
step 1, firstly, an association rule trainer in an intelligent prediction cache performs association rule mining, wherein the association rule mining process of the association rule trainer is as follows:
(1) In the present embodiment, it is assumed that the minimum support S is 75% and the minimum confidence C is 80%.
(2) According to the log file in the analysis process, a transaction table A1 of a Handle identifier complete code and a transaction table B1 of a Handle identifier prefix code are respectively established, and are shown in the following table (assuming that C represents the Handle identifier complete code and C h represents the Handle identifier prefix code):
table 1 transaction table A1
Table 2 transaction table B1
(3) The process of mining association rules for the Handle identifier complete code is as follows:
a. According to the transaction table A1, the frequent item set a 1 of the Handle identifier complete code is mined as follows: { {86.1000.12/0011111, 86.1000.16/0033333}, wherein the {86.1000.12/0011111, 86.1000.16/0033333} item set appears 5 times in total, and the total number of transactions is 6, the minimum support S is calculated according to the calculation formula (1) in the summary of the invention:
Meeting the condition that the minimum support is 75% or more, proving that the frequent item set is correctly mined;
b. The association rule R a1 is calculated according to the frequent item set a 1: Association rule/>
Wherein for association rulesThe minimum confidence coefficient C is calculated according to the calculation formula (2) in the invention content, wherein the total occurrence of the {86.1000.12/0011111, 86.1000.16/0033333} item sets is 5 times, and the total occurrence of 86.1000.12/0011111 codes is 6 times:
Meets the condition that the minimum confidence coefficient is more than 80 percent, and proves the association rule The calculation is correct;
wherein for association rules The minimum confidence coefficient C is calculated according to the calculation formula (2) in the summary of the invention, wherein the total occurrence of the {86.1000.12/0011111, 86.1000.16/0033333} item set is 5 times, and the total occurrence of 86.1000.16/0033333 codes is 5 times:
Meets the condition that the minimum confidence coefficient is more than 80 percent, and proves the association rule The calculation is correct;
c. Association rules to be mined And association rules/>Stored in the association rule trainer.
(4) The association rule process of mining Handle identification prefix codes is as follows:
a. According to the transaction table B1, the frequent item set B 1 of the Handle identification prefix code is mined as follows: { {86.1000.12, 86.1001.23}, {86.1000.12, 86.1000.16}, wherein the {86.1000.12, 86.1001.23} item sets occur 5 times together, and the total number of transactions is 6, the minimum support S is calculated according to the calculation formula (1) in the summary of the invention:
Meeting the condition that the minimum support is 75 percent, and proving that the frequent item set is correctly mined;
Wherein {86.1000.12, 86.1000.16} term sets appear 5 times in total, and the total number of transactions is 6, the minimum support S is calculated according to the calculation formula (1) in the summary of the invention:
Meeting the condition that the minimum support is 75 percent, and proving that the frequent item set is correctly mined;
b. Calculating the association rule according to the frequent item set b 1 Association rulesAssociation rule/> Association rules
Wherein for association rulesThe total occurrence of the {86.1000.12, 86.1001.23} item set is 5 times, the total occurrence of the 86.1000.12 prefix code is 6 times, and the minimum confidence coefficient C is calculated according to the calculation formula (2) in the summary of the invention:
Meets the condition that the minimum confidence coefficient is more than 80 percent, and proves the association rule The calculation is correct;
wherein for association rules The {86.1000.12, 86.1001.23} term set appears 5 times altogether, and the 86.1001.23 prefix code appears 5 times altogether, so that the minimum confidence coefficient C is calculated according to the calculation formula (2) in the summary of the invention:
Meets the condition that the minimum confidence coefficient is more than 80 percent, and proves the association rule The calculation is correct;
wherein for association rules The total occurrence of the {86.1000.12, 86.1000.16} item set is 5 times, the total occurrence of the 86.1000.12 prefix code is 6 times, and the minimum confidence coefficient C is calculated according to the calculation formula (2) in the summary of the invention:
Meets the condition that the minimum confidence coefficient is more than 80 percent, and proves the association rule The calculation is correct;
wherein for association rules The {86.1000.12, 86.1000.16} term set appears 5 times altogether, and the 86.1000.16 prefix code appears 5 times altogether, so that the minimum confidence coefficient C is calculated according to the calculation formula (2) in the summary of the invention:
Meets the condition that the minimum confidence coefficient is more than 80 percent, and proves the association rule The calculation is correct;
(5) Association rules to be mined And/>Stored in the association rule trainer.
Step 2, assuming that the stored Handle identifier complete code and Handle identifier prefix in the intelligent prediction cache do not include all the Handle identifier codes and prefix codes mentioned in the above process, the intelligent prediction cache cannot perform any one of the Handle identifier codes or the Handle identifier prefix prediction cache in advance according to the association rule of the mining, and in this embodiment, the following processes are continuously performed in the following 3 cases (where case 2 and case 3 are performed when case 1 ends):
Case 1, when a Handle client sends a "86.1000.12/0011111" id code query request:
C, if the handle client does not inquire the analysis result of the identification code in the intelligent prediction cache, continuing to carry out the step b;
c, if the handle client side does not inquire the analysis result of the prefix code 86.1000.12 in the intelligent prediction cache, continuing to carry out the step c;
c, sending a query request of prefix code 86.1000.12 to the GHR through a recursion node by the handle client, acquiring corresponding LHS service site information, and storing the prefix code and an analysis result thereof into an intelligent prediction cache;
d. At this time, according to the association rule in the association rule trainer And association rules/>And prefix code "86.1000.12" stored in the cache, predicting the Handle identification prefix code to be used for making the query request as: "86.1000.16" and "86.1001.23", and obtain the analysis results of these two prefix codes in advance, namely the corresponding LHS service site information, then store these two prefix codes and their analysis results in the intellectual prediction cache;
e. Meanwhile, the Handle client continues to send '86.1000.12/0011111' identification code query requests to the corresponding LHS service sites, and finally obtains corresponding analysis results, and simultaneously stores the Handle identification codes and the analysis results thereof into an intelligent prediction cache;
f. at this time, according to the association rule in the association rule trainer And the stored identification code '86.1000.12/0011111' in the cache predicts that the Handle identification code of the query request is about to be made as: "86.1000.16/0033333", and obtaining the analysis result of the identification code in advance, then storing the identification code and the analysis result thereof into the intelligent prediction cache, and finally ending the round of inquiry.
In case 2, when a Handle client sends a '86.1000.16/0033333' identifier code query request, according to case 1, the Handle client can directly query the analysis result of the identifier code in the intelligent prediction cache, and then directly returns the analysis result, and the query is finished.
Case 3, when a Handle client sends a "86.1001.23/0031422" identifier code query request:
C, if the handle client does not inquire the analysis result of the identification code in the intelligent prediction cache, continuing to carry out the step b;
b. according to the condition 1, the handle client can query the analysis result of the identification prefix code 86.1001.23 in the intelligent prediction cache, and then directly acquire corresponding LHS service site information without sending a query request to the GHR;
D, the handle client directly sends a '86.1001.23/0031422' identification code query request to the corresponding LHS service site through the recursion node, and finally obtains an analysis result of the identification code, and meanwhile, the identification code and the analysis result thereof are stored in an intelligent prediction cache to carry out the step d;
d. Through step c, the "86.1001.23/0031422" identifier code and its analysis result are newly stored in the cache, but at this time, according to the association rule And association rules/>And c, if the new Handle identification complete code cannot be predicted, the round of inquiry is finished after the step c.
Finally, it should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention and not for limiting the same, and although the detailed description is given with reference to the embodiments of the present invention, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, and it should be covered by the scope of the claims of the present invention.

Claims (4)

1. An intelligent prediction caching method for Handle identification analysis is characterized in that: comprising the following steps:
(1) The intelligent prediction cache comprises an association rule trainer, according to the characteristic that the analysis whole of the Handle identifier is divided into two parts of identifier prefix coding inquiry and identifier complete coding inquiry, and the Handle identifier analysis system originally comprises two types of cache services of analysis result cache of the Handle identifier prefix and analysis result cache of the Handle complete identifier, the association rule trainer respectively carries out association rule mining on the Handle identifier prefix and the Handle identifier complete coding according to logs of the prior analysis process by utilizing Aprior algorithm, and finally stores the association rules into the association rule trainer in the cache;
the specific flow is as follows:
1) Firstly, setting minimum support degree S and minimum confidence degree C of an association rule in an intelligent prediction cache;
let i= { I 1,i2,...,im } be a set of m different items, each I k (k=1, 2,., m) being referred to as an item, assuming that the association rule is: wherein/> And/>The calculation formula of the minimum support S and the minimum confidence C of the association rule is as follows:
2) Respectively establishing a transaction table A of a Handle identifier complete code and a transaction table B of a Handle identifier prefix code according to a log file in an analysis process, wherein the transaction table A comprises Handle identifier complete code records inquired at regular intervals, and the transaction table B comprises Handle identifier prefix code records inquired at regular intervals;
3) Excavating a frequent item set a of a Handle identifier complete code meeting the minimum support degree S according to the transaction table A, and excavating a frequent item set B of a Handle identifier prefix meeting the minimum support degree S according to the transaction table B;
4) Calculating a correlation rule R a of the Handle identifier complete code meeting the minimum confidence coefficient C according to the frequent item set a, and similarly, calculating a correlation rule R b of the Handle identifier prefix code meeting the minimum confidence coefficient C according to the frequent item set b;
5) Finally, the mined association rule R a and the association rule R b are stored in an association rule trainer, so that mining and storing of the identification code association rule are completed once;
(2) Then in the process of analyzing the marks, the intelligent prediction cache can combine the mark codes stored in the cache, predict the upcoming Handle mark prefix and the Handle mark complete codes in real time according to the association rules in the association rule trainer, and then respectively obtain the analysis results corresponding to the prefix and the Handle mark complete codes in advance and store the analysis results in the cache;
the specific flow is as follows:
1) Firstly, the intelligent prediction cache judges whether the prediction cache combining the Handle identifier complete code and the Handle identifier prefix stored in the current cache can be performed in advance according to the association rule in the association rule trainer:
a. if predictive caching can be performed in advance, obtaining a Handle identifier complete code and a Handle identifier prefix which can be predicted at present and are about to be provided with an analysis request, and storing the corresponding analysis result in the cache in advance;
b. if the prediction storage of any Handle identifier complete code or Handle identifier prefix cannot be realized, skipping the step, and directly starting from the step 2);
2) Whenever a Handle client initiates a Handle identifier complete code query analysis request, the Handle client will first query in an intelligent prediction cache whether there is a current Handle identifier complete code analysis result:
a. If the query is successful, the identification code may be queried before the query or the intelligent prediction cache successfully predicts the Handle identification code, and then the analysis result is directly returned, namely the query is finished;
b. If the inquiry fails, continuing to step 3);
3) When the Handle client side does not directly inquire the analysis result of the complete code of the current Handle identifier in the intelligent prediction cache, further inquiry analysis is needed, and at this moment, the analysis system can firstly carry out matching inquiry on the prefix code of the Handle identifier of the current request and the prefix of the Handle identifier stored in the intelligent prediction cache:
a. If the query is successful, the fact that the identification prefix code is queried or the intelligent prediction cache successfully predicts the identification prefix code possibly exists before is indicated, corresponding LHS service site information can be directly obtained without prefix query to GHR, then a complete identification code request is continuously sent to the LHS service site, an analysis result is directly obtained, and correspondingly, the system also stores the complete identification code analysis result into the intelligent prediction cache and continues to carry out the step b;
b. Step a is carried out, wherein a new Handle identifier complete code is stored in the cache at this time, the intelligent prediction cache can judge whether the Handle identifier complete code which is about to be provided with an analysis request and is not stored in the cache can be predicted according to an association rule R a in an association rule trainer in real time, if the new Handle identifier complete code can be predicted, the predicted Handle identifier complete code and an analysis result thereof are stored in the cache in advance, the query is finally ended, and otherwise, the query is carried out to the step a;
c. If the inquiry fails, continuing to step 4);
4) When the Handle client cannot inquire the analysis result of the prefix code of the current Handle identifier in the intelligent prediction cache, the Handle identifier analysis needs to finish inquiry according to the original complete analysis flow, and the method comprises the following steps:
firstly, a Handle client sends an identification prefix query request to a GHR through a recursion node to acquire information of a corresponding LHS service site, and synchronously stores a Handle identification prefix coding analysis result into an intelligent prediction cache;
b. Step a is performed, wherein a new Handle identifier prefix code is stored in the cache, the intelligent prediction cache can judge whether the Handle identifier prefix code which is about to be requested for analysis and is not stored in the cache can be predicted according to the association rule R b in the association rule trainer in real time, and if the brand new Handle identifier prefix code can be predicted, the predicted Handle identifier prefix code and the analysis result thereof are stored in the cache in advance;
c. meanwhile, after the step a, the Handle client continues to send a Handle complete identification query request to the corresponding LHS service site through the recursion node, and an analysis result is obtained and stored in the intelligent prediction cache;
d. and c, after the step of storing a new Handle identifier complete code in the cache, judging whether the Handle identifier complete code which is about to be submitted to an analysis request and is not stored in the cache can be predicted in real time according to the association rule R a in the association rule trainer by the intelligent prediction cache, storing the predicted Handle identifier complete code and the analysis result thereof in the cache in advance if the brand new Handle identifier complete code can be predicted, and finally ending the query, otherwise, ending the query in the step of c.
2. The intelligent prediction caching method for Handle identification resolution according to claim 1, wherein the method comprises the following steps: in the step (1), the association rule in the association rule trainer is updated once every a set time T, namely, the steps 1) to 5) are repeated, so that the dynamic update of the association rule is realized.
3. The intelligent prediction caching method for Handle identification resolution according to claim 2, wherein the method is characterized in that: the updating operation is specifically as follows:
a. the time when each association rule is successfully matched last time is stored and recorded in the association rule trainer, so that the heat degree of the association rule used is represented;
b. When the updating time T is reached, the association rule which is obtained through the re-prediction in the steps 1) to 5) is replaced in sequence, namely the association rule which is not recorded with the matching time in the trainer, namely the association rule which is not matched is replaced in sequence;
c. and then, replacing from far to near in sequence according to the recorded last successful matching time until all association rules obtained by new training are successfully stored or the memory of the trainer is full.
4. The intelligent prediction caching method for Handle identification resolution according to claim 2, wherein the method is characterized in that: the periodic updating time T is 1 s-5 s.
CN202111073296.6A 2021-09-14 2021-09-14 Intelligent prediction caching method for Handle identification analysis Active CN113779090B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111073296.6A CN113779090B (en) 2021-09-14 2021-09-14 Intelligent prediction caching method for Handle identification analysis

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111073296.6A CN113779090B (en) 2021-09-14 2021-09-14 Intelligent prediction caching method for Handle identification analysis

Publications (2)

Publication Number Publication Date
CN113779090A CN113779090A (en) 2021-12-10
CN113779090B true CN113779090B (en) 2024-05-03

Family

ID=78843491

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111073296.6A Active CN113779090B (en) 2021-09-14 2021-09-14 Intelligent prediction caching method for Handle identification analysis

Country Status (1)

Country Link
CN (1) CN113779090B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5819275A (en) * 1995-06-07 1998-10-06 Trusted Information Systems, Inc. System and method for superimposing attributes on hierarchically organized file systems
WO2008055908A1 (en) * 2006-11-07 2008-05-15 Nokia Siemens Networks Gmbh & Co. Kg Method for forwarding address prefix data in a network and device
CN103678338A (en) * 2012-09-07 2014-03-26 中国科学院计算机网络信息中心 Internet of things identification analytic method with intelligent cache module
CN111291043A (en) * 2020-01-09 2020-06-16 中国信息通信研究院 Identification value query method, identification resolution server and storage medium
CN112016015A (en) * 2020-08-31 2020-12-01 中国信息通信研究院 Data processing method and device based on distributed identification analysis

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9319307B2 (en) * 2013-09-06 2016-04-19 At&T Intellectual Property I, L.P. Providing differentiated service to traffic flows obscured by content distribution systems

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5819275A (en) * 1995-06-07 1998-10-06 Trusted Information Systems, Inc. System and method for superimposing attributes on hierarchically organized file systems
WO2008055908A1 (en) * 2006-11-07 2008-05-15 Nokia Siemens Networks Gmbh & Co. Kg Method for forwarding address prefix data in a network and device
CN103678338A (en) * 2012-09-07 2014-03-26 中国科学院计算机网络信息中心 Internet of things identification analytic method with intelligent cache module
CN111291043A (en) * 2020-01-09 2020-06-16 中国信息通信研究院 Identification value query method, identification resolution server and storage medium
CN112016015A (en) * 2020-08-31 2020-12-01 中国信息通信研究院 Data processing method and device based on distributed identification analysis

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
CONCURRENT FREQUENT ITEMSETS MINING IN A SHARED PREFIX TREE USING THE APRIORI ALGORITHM;Puscian, M;《 PHOTONICS APPLICATIONS IN ASTRONOMY, COMMUNICATIONS, INDUSTRY, AND HIGH-ENERGY PHYSICS EXPERIMENTS 2018》;全文 *
电力装备制造业工业互联网标识解析体系应用发展研究;李晋航等;《电力系统保护与控制》;第 48 卷(第 12 期);正文第1-3节 *

Also Published As

Publication number Publication date
CN113779090A (en) 2021-12-10

Similar Documents

Publication Publication Date Title
JP4856627B2 (en) Partial query caching
Zhao et al. Rest: A reference-based framework for spatio-temporal trajectory compression
CN102810118B (en) A kind of change weighs net K nearest neighbor search method
Thomsen et al. Effective caching of shortest paths for location-based services
CN110119928B (en) Vehicle matching recommendation method based on driver characteristics
CN111666305B (en) Method and system for realizing correlation between redis cache and database
Wong et al. Mining fuzzy association rules for web access case adaptation
CN105740445A (en) Database query method and device
CN105740383A (en) Method and system for realizing data consistency of database and cache in big data platform
CN106547828A (en) A kind of database caches system and method based on neutral net
CN107122437A (en) A kind of big data processing method supported many condition retrieval and analyzed in real time
CN113254630A (en) Domain knowledge map recommendation method for global comprehensive observation results
CN113779090B (en) Intelligent prediction caching method for Handle identification analysis
CN112751948B (en) DNS cache recommendation method based on collaborative filtering
CN104750806A (en) Large data query method and system
CN117194502A (en) Database content cache replacement method based on long-term and short-term memory network
CN116166895A (en) Materialized view-based space-time range query method, system, equipment and medium
CN111723092A (en) Data processing method and device
CN115114289A (en) Data query method and device and electronic equipment
CN114116702A (en) Key value database-based quick query method for adjacent points
CN101936741B (en) Reverse base graph based vehicle navigation road network incremental data updating system for correcting spatial-temporal data model
CN112989255A (en) Identification analysis system
CN117076467B (en) Smart city identification method, system, equipment and medium based on MA system
CN117076523B (en) Local data time sequence storage method
US7552137B2 (en) Method for generating a choose tree for a range partitioned database table

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