CN109271371A - A kind of Distributed-tier big data analysis processing model based on Spark - Google Patents
A kind of Distributed-tier big data analysis processing model based on Spark Download PDFInfo
- Publication number
- CN109271371A CN109271371A CN201810956427.7A CN201810956427A CN109271371A CN 109271371 A CN109271371 A CN 109271371A CN 201810956427 A CN201810956427 A CN 201810956427A CN 109271371 A CN109271371 A CN 109271371A
- Authority
- CN
- China
- Prior art keywords
- layer
- distributed
- tier
- big data
- real
- 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
- 238000007405 data analysis Methods 0.000 title claims abstract description 22
- 238000012545 processing Methods 0.000 title abstract description 11
- 230000005540 biological transmission Effects 0.000 claims description 18
- 238000013480 data collection Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000000034 method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of, and the Distributed-tier big data analysis based on Spark handles model, including expression layer (PT), front end switching layer (FST), rear end switching layer (BST), real time business logical layer (RBLT), non-real-time service logical layer (NRBLT) and data access layer (DAT).The invention proposes a kind of, and the Distributed-tier big data analysis based on Spark handles model, can effectively reduce the analysis speed of mass data, and the Heterogeneous Information in support system between each subsystem is linked up and stored with data.It is sufficient for the short-term trend forecast demand of high frequency trade market.The application value with higher in high frequency, big data processing system.
Description
Technical field
The present invention relates to big data analysis process fields, more particularly, to a kind of Distributed-tier based on Spark
Big data analysis handles model.
Background technique
Big data can help user to improve insight, be promoted in higher level, wider array of visual angle, bigger range
Decision edge.But some values having often are hidden in big data, show value density it is extremely low, distribution extremely not
Rule, Information hiding are in the highest degree, discovery is useful is worth extremely difficult distinct characteristic.As the high frequency of stock market is traded
(HFT), because of short-term market trend and quickly quotation, people are difficult to determine when buy or sell in time, to big data
Accuracy, the rapidity of analysis have high requirement.
Summary of the invention
Present invention aim to address said one or multiple defects, propose that a kind of Distributed-tier based on Spark is big
Data Analysis Services model.
To realize the above goal of the invention, the technical solution adopted is that:
A kind of Distributed-tier big data analysis processing model based on Spark, including expression layer (PT), front end exchange
Layer (FST), rear end switching layer (BST), real time business logical layer (RBLT), non-real-time service logical layer (NRBLT) and data are visited
Ask layer (DAT);Wherein expression layer (PT) carries out data transmission with front end switching layer (FST), the output of front end switching layer (FST)
End is connect with the input terminal of medium;Medium carries out data transmission with rear end switching layer (BST);The output of rear end switching layer (BST)
End is connect with the input terminal of the input terminal of real time business logical layer (RBLT) and non-real-time service logical layer (NRBLT);Real-time industry
Be engaged in logical layer (RBLT) output end and non-real-time service logical layer (NRBLT) output end with data access layer (DAT)
Input terminal connection.
Preferably, the expression layer (PT) is obtained data and is serviced using Facade and handled from user to rear from BLT
Hold all requests of cluster.
Preferably, the front end switching layer (FST) further includes the front-end server being deployed on node, the front end
Switching layer (FST) is responsible for receiving web request, and web request is transferred to Facade by Kafka message system.
Preferably, the front-end server is the front-end server for deploying MongoDB, before the MongoDB passes through
End switching layer (FST) is sent to Kafka to avoid enter into rear end cluster.
Preferably, the rear end switching layer (BST) obtains message from Kafka, carries out front end by BST ingress interface
Server and rear end switching layer carry out information transmission.
Preferably, the real time business logical layer (RBLT) further includes indicating node and docking center;The expression section
Point is carried out data transmission by spout and medium;The docking center is carried out data transmission by bolt and medium.
Preferably, the non-real-time service logical layer (NRBLT) is for storing decision strategy;The wherein decision plan
It is slightly stored in MongoDB, can be obtained the interface of quickly access large data collection using R program and Spark RDD.
Preferably, the data access layer (DAT) includes real time data resources bank, switching centre, baseline and data bins
Library;Wherein real time data resources bank carries out real-time data access to switching centre.
Compared with prior art, the beneficial effects of the present invention are:
The invention proposes a kind of, and the Distributed-tier big data analysis based on Spark handles model, can effectively reduce sea
The analysis speed of data is measured, and the Heterogeneous Information in support system between each subsystem is linked up and stored with data.It is sufficient for high frequency
The short-term trend forecast demand of trade market.The application value with higher in high frequency, big data processing system.
Detailed description of the invention
Fig. 1 is the distributed architecture figure of this system;
Fig. 2 is real time business logical layer structure figure;
Fig. 3 is status center topology diagram;
Fig. 4 is original design figure;
Fig. 5 is HFT topology diagram;
Fig. 6 is the average calculation times figure that state of market calculates;
Fig. 7 is computing market each second status number figure;
Specific embodiment
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent;
Below in conjunction with drawings and examples, the present invention is further elaborated.
Embodiment 1
A kind of Distributed-tier big data analysis processing model based on Spark, referring to FIG. 1, including expression layer
(PT), front end switching layer (FST), rear end switching layer (BST), real time business logical layer (RBLT), non-real-time service logical layer
(NRBLT) and data access layer (DAT);Wherein expression layer (PT) carries out data transmission with front end switching layer (FST), and front end is handed over
The input terminal of the output end and medium that change layer (FST) connects;Medium carries out data transmission with rear end switching layer (BST);It hands over rear end
Change the defeated of the output end of layer (BST) and the input terminal of real time business logical layer (RBLT) and non-real-time service logical layer (NRBLT)
Enter end connection;The output end of the output end of real time business logical layer (RBLT) and non-real-time service logical layer (NRBLT) with
The input terminal of data access layer (DAT) connects.This framework is from Triple distribution architectural evolution.Finally, we will
Business Logic is separated into real time business logical layer and non-real-time service logical layer.In addition, we use the message of two ranks
Middleware is transmitted to solve the high frequency requirements in whole system.
Expression layer (PT) this layer obtains data from BLT, and prepares the user that web page is presented to online browse.In order to
Accelerate loading velocity, reduces the delay of access time, the present embodiment services to handle from user to rear end cluster using Facade
All requests.Architecture is set to have more loose couplings.
Front end switching layer (FST) is responsible for receiving web request, and is passed them to by Kafka message system
Facade.This layer includes the front-end server being deployed on node.In view of operation efficiency, the present embodiment disposes MongoDB
It in front-end server, is run through front end switching layer and is sent to Kafka, it is not necessary to enter rear end cluster and carry out data processing.
In the present embodiment, the front-end server is the front-end server for deploying MongoDB, and the MongoDB passes through
Front end switching layer (FST) is sent to Kafka to avoid enter into rear end cluster.
In the present embodiment, the rear end switching layer (BST) obtains message from Kafka, before being carried out by BST ingress interface
Server and rear end switching layer is held to carry out information transmission.
In the present embodiment, the real time business logical layer (RBLT) further includes indicating node and docking center;The expression
Node is carried out data transmission by spout and medium;The docking center is carried out data transmission by bolt and medium.Real-time industry
Business logical layer (RBLT) is the key component of radio frequency system, is mainly responsible for the processing and calculating of real time data.It includes two weights
The service wanted, data analysis and decision.Such as a stock trade price forecasting system, it is necessary to a storm topology, with quick
Real-time price quotations stream is handled, and is stored into HBase.Rket state is calculated for HDFS.That is: the signal bought in or sold is calculated.
As shown in Fig. 2, if user terminal and transaction platform are divided into two topological networks.Pass through Kafka message system computing market
State simultaneously passes it to user.In order to improve efficiency with higher speed, we incorporate the two topology, and will
Kafka messaging middleware replaces with Netty, realizes high-frequency therapeutic treatment and the transmission of information.In Storm topology, Netty
Speed be about 10 times of Kafka.
In the present embodiment, the function of non-real-time service logical layer (NRBLT) calculates user according to big data
Information result carries out decision strategy.Decision strategy is stored in MongoDB, is quickly accessed convenient for user from front end node.It utilizes
R program and Spark RDD, so that it may obtain the interface of quickly access large data collection.
In the present embodiment, the data access layer (DAT) includes real time data resources bank, switching centre, baseline and data
Warehouse;Wherein real time data resources bank carries out real-time data access to switching centre.Data access layer (DAT) comes for accessing
From all data of database or external data source.As DAT provides an order interface, and big data information can be combined
At a K-Bar, middleware is transmitted to user's immediate feedback external data information by Kafka unified message.
By above-mentioned model framework, our one stock exchange big data analysis decision calculated examples of virtual development, to calculation
Method process is analyzed.Stock trade price provides real-time price quotations and marketing state by network trading platform.Due to needing
The requirement for meeting machine learning and quickly calculating.We are first using network trading center as a topology, to realize algorithm
Low latency.See the most entire status center topology of Fig. 3.
In topology, KafkaSpout is serviced from external RealtimeDataPublisher and is received real-time price quotations, and is led to
It crosses distributed information system Kafka and constantly sends market real-time deal price.Then KafkaSpout by Price pass-through give with
18 ComputeStateBolt afterwards.Each ComputeStateBolt has different computer logics, and is come using it
Calculate state of market defined in specific TA logic.Then, result state of market is sent to spy by 18 ComputeStateBolt
Fixed TA WriteDataBolt.HBase is written in corresponding TA data by each WriteDataBolt.For example,
State of market is sent to MAWriteDataBolt by ComputeStateBolt, special to store MA state of market.In topology
Outside, all black lines all indicate that Kafka, Netty distributed messaging system transmit.
The purpose of high frequency trading market data analysis is to acquire marketing and price status for user.Therefore, Wo Menxu
Machine learning algorithm is used, historic market data are learnt, then according to historical trend changing rule, help constructs investment plan
Slightly.In order to solve the problems, such as large data sets Fast Learning, herein using the Plan Center operation in Apache Spark frame
Machine learning algorithm loads large-scale history data set from HBase, and learns in a short time and analyze.Plan Center branch
Hold vector machine (SVM), logistic regression (LR) and classification.By Spark RDD, Plan Center can be by the city of hundreds of gb
Field status data is loaded into memory, and multiple nodes in the cluster calculate analysis.User is helped to provide trading strategies.It hands over
After easy strategy generating, user can choose the investment decision used on web page.Large data sets handle model framework such as Fig. 4
It is shown.
In order to reduce big data analysis, processing and the overhead time of transmission, we are by status center topology and trade us
Above topology is merged into one large-scale topology, forms a large size HFT system such as Fig. 5.
HFT after integration extracts data from Kafka queue, and writes data into HBase and MongoDB.Fig. 5 is shown
Entire HFT topological structure.Pass through the integration to network trading center and user terminal, so that it may the cost time of information transmission
Shorten to several milliseconds.But due to the complexity of large-scale transaction system architecture, it is necessary to carry out efficient cluster resource pipe
Reason, can just effectively improve the calculating speed of algorithm.Therefore, we are started most of services using yarn and managed on cluster
All resources.And each node and Hadoop service status on the configuration monitoring Cloudera by customizing Hadoop service,
Realize the cluster service of large data sets.
Due to high frequency and real time data processing requirement, trade center needs calculate millions of a markets in one second
State.Therefore, simulated experimental environments have been built herein, compare the algorithm performance processing result of different number futures exchange.We
8 computers are prepared as cluster, wherein 6 run Storm topology as manager.Experimental situation is as shown in table 1.
The details of 1 cluster of table
In order to test the extreme efficiency of the architecture and find out the configuration of most suitable cluster, we are to each experiment
The average calculation times of all state of market compare.
For check algorithm performance, we be added in original topological structure one it is entitled
The new bolt of ExpStateReceiverBolt is flat by calculating quickly to collect all calculating metric datas of state of market
Mean testing algorithm performance.Fig. 6 shows results of property, and Fig. 7 shows the state of market number of N number of stock.
Obviously, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be pair
The restriction of embodiments of the present invention.For those of ordinary skill in the art, may be used also on the basis of the above description
To make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.It is all this
Made any modifications, equivalent replacements, and improvements etc., should be included in the claims in the present invention within the spirit and principle of invention
Protection scope within.
Claims (8)
1. a kind of Distributed-tier big data analysis based on Spark handles model, which is characterized in that including expression layer (PT),
Front end switching layer (FST), rear end switching layer (BST), real time business logical layer (RBLT), non-real-time service logical layer (NRBLT)
With data access layer (DAT);Wherein expression layer (PT) carries out data transmission with front end switching layer (FST), front end switching layer (FST)
Output end and medium input terminal connect;Medium carries out data transmission with rear end switching layer (BST);Rear end switching layer (BST)
Output end connect with the input terminal of the input terminal of real time business logical layer (RBLT) and non-real-time service logical layer (NRBLT);
The output end of the output end of real time business logical layer (RBLT) and non-real-time service logical layer (NRBLT) is and data access layer
(DAT) input terminal connection.
2. a kind of Distributed-tier big data analysis based on Spark according to claim 1 handles model, feature exists
Ask the visitor in for the institute for obtaining data from BLT in, the expression layer (PT) and servicing using Facade to handle from user to rear end cluster
It asks.
3. a kind of Distributed-tier big data analysis based on Spark according to claim 1 handles model, feature exists
In the front end switching layer (FST) further includes the front-end server being deployed on node, and the front end switching layer (FST) is responsible for
Web request is received, and is transferred to web request by Kafka message system
4. a kind of Distributed-tier big data analysis based on Spark according to claim 3 handles model, feature exists
In the front-end server is the front-end server for deploying MongoDB, and the MongoDB is sent out by front end switching layer (FST)
It send to Kafka to avoid enter into rear end cluster.
5. a kind of Distributed-tier big data analysis based on Spark according to claim 1 handles model, feature exists
In the rear end switching layer (BST) obtains message from Kafka, carries out front-end server by BST ingress interface and exchanges with rear end
Layer carries out information transmission.
6. a kind of Distributed-tier big data analysis based on Spark according to claim 1 handles model, feature exists
In the real time business logical layer (RBLT) further includes indicating node and docking center;The expression node passes through spout and matchmaker
Jie carries out data transmission;The docking center is carried out data transmission by bolt and medium.
7. a kind of Distributed-tier big data analysis based on Spark according to claim 1 handles model, feature exists
In the non-real-time service logical layer (NRBLT) is for storing decision strategy;Wherein the decision strategy is stored in MongoDB
In, it can be obtained the interface of quickly access large data collection using R program and Spark RDD.
8. a kind of distributed big data analysis based on Spark according to claim 1 handles model, which is characterized in that
The data access layer (DAT) includes real time data resources bank, switching centre, baseline and data warehouse;Wherein real time data provides
Source library carries out real-time data access to switching centre.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810956427.7A CN109271371B (en) | 2018-08-21 | 2018-08-21 | Spark-based distributed multi-layer big data analysis processing model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810956427.7A CN109271371B (en) | 2018-08-21 | 2018-08-21 | Spark-based distributed multi-layer big data analysis processing model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109271371A true CN109271371A (en) | 2019-01-25 |
CN109271371B CN109271371B (en) | 2022-02-11 |
Family
ID=65154176
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810956427.7A Expired - Fee Related CN109271371B (en) | 2018-08-21 | 2018-08-21 | Spark-based distributed multi-layer big data analysis processing model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109271371B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110502509A (en) * | 2019-08-27 | 2019-11-26 | 广东工业大学 | A kind of traffic big data cleaning method and relevant apparatus based on Hadoop Yu Spark frame |
CN111177765A (en) * | 2020-01-06 | 2020-05-19 | 广州知弘科技有限公司 | Financial big data processing method, storage medium and system |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1997020258A1 (en) * | 1995-11-29 | 1997-06-05 | Hybrithms Corp. | Multiple-agent hybrid control architecture |
US6112183A (en) * | 1997-02-11 | 2000-08-29 | United Healthcare Corporation | Method and apparatus for processing health care transactions through a common interface in a distributed computing environment |
US20040143602A1 (en) * | 2002-10-18 | 2004-07-22 | Antonio Ruiz | Apparatus, system and method for automated and adaptive digital image/video surveillance for events and configurations using a rich multimedia relational database |
CN102063306A (en) * | 2011-01-06 | 2011-05-18 | 夏春秋 | Technical implementation method for application development through electronic form |
CN102364523A (en) * | 2011-05-11 | 2012-02-29 | 武汉理工大学 | Method for realizing three-dimensional virtual city system based on RIA (rich Internet application) architecture |
CN102385739A (en) * | 2011-11-15 | 2012-03-21 | 中国电力科学研究院 | Integrated information management platform for county-level power supply enterprises |
CN102523246A (en) * | 2011-11-23 | 2012-06-27 | 陈刚 | Cloud computation treating system and method |
CN105162826A (en) * | 2015-07-15 | 2015-12-16 | 中山大学 | Cloud computing multilayer cloud architecture |
CN107274062A (en) * | 2017-05-11 | 2017-10-20 | 王嫣然 | Share books management system and the sharing method using the system in a kind of campus based on school's LAN |
CN107292473A (en) * | 2016-04-10 | 2017-10-24 | 国网山东省电力公司经济技术研究院 | The online estimating and examining system of planning feasibility study business and method based on process optimization |
CN107657569A (en) * | 2016-07-25 | 2018-02-02 | 湖南移商动力网络技术有限公司 | J2EE and cloud computing design a kind of intelligence community system |
-
2018
- 2018-08-21 CN CN201810956427.7A patent/CN109271371B/en not_active Expired - Fee Related
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1997020258A1 (en) * | 1995-11-29 | 1997-06-05 | Hybrithms Corp. | Multiple-agent hybrid control architecture |
US6112183A (en) * | 1997-02-11 | 2000-08-29 | United Healthcare Corporation | Method and apparatus for processing health care transactions through a common interface in a distributed computing environment |
US20040143602A1 (en) * | 2002-10-18 | 2004-07-22 | Antonio Ruiz | Apparatus, system and method for automated and adaptive digital image/video surveillance for events and configurations using a rich multimedia relational database |
CN102063306A (en) * | 2011-01-06 | 2011-05-18 | 夏春秋 | Technical implementation method for application development through electronic form |
CN102364523A (en) * | 2011-05-11 | 2012-02-29 | 武汉理工大学 | Method for realizing three-dimensional virtual city system based on RIA (rich Internet application) architecture |
CN102385739A (en) * | 2011-11-15 | 2012-03-21 | 中国电力科学研究院 | Integrated information management platform for county-level power supply enterprises |
CN102523246A (en) * | 2011-11-23 | 2012-06-27 | 陈刚 | Cloud computation treating system and method |
CN105162826A (en) * | 2015-07-15 | 2015-12-16 | 中山大学 | Cloud computing multilayer cloud architecture |
CN107292473A (en) * | 2016-04-10 | 2017-10-24 | 国网山东省电力公司经济技术研究院 | The online estimating and examining system of planning feasibility study business and method based on process optimization |
CN107657569A (en) * | 2016-07-25 | 2018-02-02 | 湖南移商动力网络技术有限公司 | J2EE and cloud computing design a kind of intelligence community system |
CN107274062A (en) * | 2017-05-11 | 2017-10-20 | 王嫣然 | Share books management system and the sharing method using the system in a kind of campus based on school's LAN |
Non-Patent Citations (2)
Title |
---|
吴冕冠: "基于Spark的大数据应用开发支持环境研究开发", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
王宇轲: "基于BA-BP算法的汽车配件需求预测系统研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110502509A (en) * | 2019-08-27 | 2019-11-26 | 广东工业大学 | A kind of traffic big data cleaning method and relevant apparatus based on Hadoop Yu Spark frame |
CN110502509B (en) * | 2019-08-27 | 2023-04-18 | 广东工业大学 | Traffic big data cleaning method based on Hadoop and Spark framework and related device |
CN111177765A (en) * | 2020-01-06 | 2020-05-19 | 广州知弘科技有限公司 | Financial big data processing method, storage medium and system |
Also Published As
Publication number | Publication date |
---|---|
CN109271371B (en) | 2022-02-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8473422B2 (en) | Method and system for social network analysis | |
CN108038145A (en) | Distributed Services tracking, system, storage medium and electronic equipment | |
Sun et al. | The cost-efficient deployment of replica servers in virtual content distribution networks for data fusion | |
Li et al. | Topology-aware neural model for highly accurate QoS prediction | |
CN110245178A (en) | Marketing automation management platform system and its management method | |
CN111130842B (en) | Dynamic network map database construction method reflecting network multidimensional resources | |
WO2023185090A1 (en) | Scheduling method and apparatus based on microservice link analysis and reinforcement learning | |
CN1956454B (en) | Method and system for bundling and sending work units to a server based on a weighted cost | |
CN104410699A (en) | Resource management method and system of open type cloud computing | |
WO2023217127A1 (en) | Causation determination method and related device | |
CN113392150A (en) | Data table display method, device, equipment and medium based on service domain | |
US8341263B2 (en) | Peer to peer monitoring framework for transaction tracking | |
CN109271371A (en) | A kind of Distributed-tier big data analysis processing model based on Spark | |
Akay et al. | Predicting the performance measures of an optical distributed shared memory multiprocessor by using support vector regression | |
CN115373888A (en) | Fault positioning method and device, electronic equipment and storage medium | |
Zhu et al. | Analysis of stock market based on visibility graph and structure entropy | |
Wu et al. | Blender: A container placement strategy by leveraging zipf-like distribution within containerized data centers | |
CN110380890A (en) | A kind of CDN system service quality detection method and system | |
Yue et al. | Desis: Efficient Window Aggregation in Decentralized Networks. | |
Deng | The Informatization of Small and Medium‐Sized Enterprises Accounting System Based on Sensor Monitoring and Cloud Computing | |
CN114579311B (en) | Method, device, equipment and storage medium for executing distributed computing task | |
Liu et al. | Towards dynamic reconfiguration of composite services via failure estimation of general and domain quality of services | |
Zhu et al. | An Influence Maximization Algorithm Based on Improved K-Shell in Temporal Social Networks. | |
Sudhakar et al. | Path based optimization of mpi collective communication operation in cloud | |
CN111522662B (en) | Node system for financial analysis and implementation method thereof |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220211 |