CN103177094B - Cleaning method of data of internet of things - Google Patents
Cleaning method of data of internet of things Download PDFInfo
- Publication number
- CN103177094B CN103177094B CN201310081635.4A CN201310081635A CN103177094B CN 103177094 B CN103177094 B CN 103177094B CN 201310081635 A CN201310081635 A CN 201310081635A CN 103177094 B CN103177094 B CN 103177094B
- Authority
- CN
- China
- Prior art keywords
- data
- node
- internet
- things
- udb
- 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
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a cleaning method of data of the internet of things, and relates to the field of data process of the internet of things. The method includes the following steps: (1) operators obtains original data of the internet of things through web service, (2) the operators reconstruct the original data and store the original data in a memory, (3) the operators reassemble the reconstructed original data into a UDB tree, and (4) a preset rule is read and data cleaning is carried out. The data cleaning is carried out according to the preset rule so as to reduce input/output (I/O) operations to the largest extent, and therefore traditional performance bottleneck problems can be solved fundamentally.
Description
Technical field
The present invention relates to Internet of Things data process field, more particularly, to a kind of Internet of Things data cleaning method.
Background technology
With sensor, RF identification（RFID）, global positioning system, infrared inductor, laser scanner, gas sensing
The development of the various device such as device and technology, all things in reality may by Real-time Collection any need monitoring, connect,
Interactive object or during, gather the information of the various needs such as its sound, light, heat, electricity, mechanics, chemistry, biology, position, and mutual
The huge network that networking combines to form, a kind of such giant grid is referred to as Internet of Things.
Due to the above characteristic of Internet of Things, during gathered data, producing in a large number, various forms are different, meaning is different
Mass data, and how from these data according to user need data is carried out, to reach data effectively utilizes
Purpose become the emphasis of Internet of Things information processing.
Traditional data cleansing mode, the overwhelming majority is based on data in magnetic disk file, along with right in a large number in cleaning process
The I/O operation of disk, although query optimization can improve efficiency to a certain extent, when in the face of mass data, frequently
I/O operation is still directly becoming the bottleneck of impact performance.The present invention is directed to above problem, has invented a kind of new data cleansing
Method, Internet of Things data is loaded onto in server memory, then Uniform data format data structure（Uniform-
Delicate B-Tree）, the cleaning algorithm that is directed to using data structure, on multiple servers, according to rule set in advance
To data cleansing, farthest reduce I/O operation, thus fundamentally solving the problems, such as traditional performance bottleneck.
Content of the invention
The present invention is directed to above problem, has invented a kind of new Data Cleaning Method, Internet of Things data is loaded onto clothes
In business device internal memory, then Uniform data format data structure（Uniform-Delicate B-Tree）, using data structure pin
To cleaning algorithm, on multiple servers, according to rule set in advance to data cleansing, farthest reduce I/O behaviour
Make, thus fundamentally solving the problems, such as traditional performance bottleneck.
For achieving the above object, the present invention takes technical scheme below：A kind of Internet of Things data cleaning method, including following
Step：The first step：Enforcement personnel obtain Internet of Things initial data by Web Service；Second step：Enforcement personnel are by original number
After reconstruct, store it in internal memory；3rd step：Initial data after reconstructing is reassembled as UDB tree by enforcement personnel；4th
Step：Read rule set in advance, carry out data cleansing.
Preferred version：The detailed process of the first step comprises the steps：Using the mode of Web Service, externally provide
Database service interface, by data uniform transmission to central database, central database adopts traditional relevant database, for every
The equipment of type creates respective tables of data, and device numbering will store as unique identifier, the initial data of all collections
In respective tables of data.
Preferred version：The detailed process of second step comprises the steps：After obtaining all data from the central database, implement
These initial data are assembled into as data block personnel；After completing data reconstruction, then these data blocks are loaded onto in internal memory.
Preferred version：The detailed process of the 3rd step comprises the steps：Using search algorithm UDB, by the number in second step
Start to index according to keyword according to block, after finding corresponding position, will be corresponding for data block insertion according still further to UDB insertion algorithm
Node.
Preferred version：The detailed process of the 4th step comprises the steps：In data cleansing, operating personnel can pass through soft
Part configures, or utilizes configuration file, and cleaning rule is defined, during whole data cleansing, will be in cleaning rule
When being carried out under conditions of agreement, and cleaning, one or more step can be divided into, each step can formulate difference
Cleaning rule.
Preferred version：The interior internal memory saving as computer cluster composition described in second step.
Preferred version：When low memory is to accommodate all of data, initial data will be retained in disk, only will every time
Partial data assembling is data block and is loaded onto internal memory, after the data block in internal memory completes to process, then loads surplus from disk
Remaining data.
Preferred version：Described cleaning rule can be created using data base query language or script and be compiled
Volume.
In sum, due to employing technique scheme, the concrete beneficial effect of the present invention is：Make full use of internal memory
Capacity storage basic data, thus decreasing I/O expense, improves access efficiency；Using unified data form, be conducive to counting
According to fast resolving although more time can be consumed during Uniform data format, but when parsing data, it will significantly carry
At high speed, thus improving bulk velocity；Using the data structure for Internet of Things, using searching algorithm with strong points, improve
Retrieval rate；Using the advantage of computer cluster, carry out data cleansing process parallel, thus fundamentally solving traditional performance
Bottleneck problem.
Brief description
Fig. 1 is intermediate node schematic diagram；
Fig. 2 is back end schematic diagram；
Fig. 3 is data block schematic diagram；
Fig. 4 is UDB tree schematic diagram：
Fig. 5 data cleansing flow chart；
Fig. 6 is data cleansing flow chart.
Specific embodiment
All features disclosed in this specification, or disclosed all methods or during step, except mutually exclusive
Feature and/or step beyond, all can combine by any way.
This specification（Including any accessory claim, summary and accompanying drawing）Disclosed in any feature, except non-specifically is chatted
State, all can be replaced by other alternative features equivalent or that there is similar purpose.I.e., unless specifically stated otherwise, each feature
It is a series of equivalent or one of similar characteristics example.
Due to needing to be directed to the present invention newly-established data structure using some in specific implementation step and its being related to
Algorithm, therefore first illustrate these data structures and its algorithm being related to.
Concept 1.UDB (Uniform-Delicate B-Tree) is set,
Definition：This is a species B- tree, meets the general definition of B- tree construction, because B- tree is to reach common understanding in this area
A kind of data structure, therefore do not elaborate, the present invention is described in detail just for the special construction of UDB：
UDB tree includes 2 kinds of nodes, and they are defined as follows：
The first node is intermediate node, as shown in figure 1, this node only includes classification information, builds intermediate node
Purpose be retrieval data for convenience, itself does not comprise data.The structure of intermediate node contains N, and pointer and Key are crucial
Word：
N：Node number, is the node number including in this node.
*：It is used for the data structure connecting between pointer, as node.
Key：Keyword, as indicates the unique identifier of node.It is type number.
The identification number of intermediate node, needs to accomplish globally unique in type, for example, cell phone apparatus and electronic tag, then and
It is necessary to be distinguished with uniquely numbering, numbering can not repeat the cell phone apparatus equipment different with electronic tag 2 class, such as SJ, BQ, and
The concrete equipment under concrete cell phone apparatus and electronic tag under cell phone apparatus, then can be identically numbered, below SJ
Below 001 and BQ 001.
Second node is back end, is also called terminal node, leaf node, as shown in Figure 2.Terminal node contains one
Individual HASH hash table, HASH hash table utilizes the hashed value quick-searching data block of keyword.
The structure of data block is as shown in figure 3, wherein：
Key (keyword) is the mark of data block, and data block Key under same HASH hashed value cannot repeat, and presses
Connected in order using pointer according to lexicographic order data block, generally type.
Data represents the True Data of data block storage.
* represent pointer, as between data block, be used for the data structure connecting.
Both the above node constitutes the structure of UDB tree, as shown in Figure 4.Specifically it is summarized as follows：UDB tree includes middle node
Point data node, described intermediate node includes root node and non-leaf nodes, and described back end is leaf node, described in
Intermediate node includes first structure and the number of described first structure, and described first structure specifically includes for presentation class relation
Pointer and numbering；Described back end includes data block chain and the HASH hash table for retrieving described data block chain, described number
Specifically include the data block that several have linking relationship according to block chain, described data block includes data block data, is used for representing data
The pointer of relation at the same level and numbering.
Due to during setting up UDB tree, needing to use corresponding search algorithm and insertion algorithm, now this 2 kinds are calculated
Method is described as follows：
Algorithm 1.UDB tree query algorithm.
1) according to the KEY value of node, in whole tree, from the beginning all nodes in node have started, middle in these
Enter line retrieval in node, according to the Query Result in node, enter in corresponding region and carry out interval query.
2) after obtaining the Query Result of subregion in node, need the child node pointer deposited according to subregion head, plus
On leave relative displacement in index node data block in, this side-play amount is the node serial number of first node relatively.Circulation is straight
To finding indexed node.
3) using the HASH function setting, after calculating corresponding HASH value, visit again data block chain.
It is eventually found data block to be looked for.
Algorithm 2.UDB tree insertion algorithm.
During to the new node of UDB tree insertion.
The leaf node that inquiry KEY should insert.After leaf node is positioned, need to judge whether this node has enough
Space accommodating new index entry.If insufficient space, need to divide leaf node.
When leaf node needs division, need to consider there are enough spaces when father node.If father node P has enough
Space, f is the child node group pointer on some subregion in P.G is the child node group of f indication.Certain node in g divides
When splitting, need all nodes in bigger child node group g' of distribution ratio g to be all copied in g, include being split off producing simultaneously
New node.Then by child node group g' new for sensing, the space of g will be released f.The corresponding index entry of new node is inserted into
To in father node P.
If father node does not have redundant space, then its ability also will be split off.Hypothesis P is father node, and f is a certain in P
Child node group pointer on individual subregion, g is the child node group of f indication.During division, distribute new child node group g, and multiple from g
Make all of child node and new split vertexes.P itself also will be split off, and generate new node P ', and replicate the rope of half from P
Draw item.Meanwhile, the node group that P is located is also required to redistribute space and replica node data is accommodating P'.If necessary, also will
Continue to divide father's node of P.Meanwhile, the division of every minor node also needs to adjust index in node.
Web Service：It is a kind of General Model building application program, can be in any operation supporting network service
Implement in system to run;It is a kind of new web application branch, is self-contained, self-described, modular application, permissible
Issue, position, called by web.Web Service be one application assembly, its logicality for other application programs provide
Data and service.Each application program passes through procotol and some standard data formats of regulation（Http, XML, Soap) come to visit
Ask Web Service, results needed is obtained by execution inside Web Service.Web Service can execute from simple
Ask any function of complicated business processing.Once after deployment, other Web Service application programs can be found that and adjust
The service disposed with it.
After fully understanding above-mentioned special concept and algorithm, further introduce being embodied as of being related in the present invention
Journey, overall flow figure such as Fig. 5：
The first step：Enforcement personnel obtain Internet of Things initial data by Web Service.Because the data volume of Internet of Things is huge
Greatly, Internet of Things data will be stored in different network nodes, and the data type difference of these data is huge, such as relationship type number
According to storehouse, non-relational database, non-structural data（As XML, Excel etc.）, these data will be disposed on the network node, utilization
The mode of Web Service, externally provides database service interface, by data uniform transmission to central database, central database
Using traditional relevant database, it is that each type of equipment creates respective tables of data, device numbering is as unique identification
Number, the initial data of all collections will be stored in respective tables of data.While in order to ensure real-time property, do not give network
Bandwidth brings performance bottleneck, and the frequency of transmission data is once a day.
Second step：After initial data is reconstructed by enforcement personnel, store it in internal memory.Obtain all from the central database
After data, these initial data are assembled into the data block for referring in concept 1 by enforcement personnel, now, pointer in data block
For sky, data is data, can become having of Key keyword：
1. device type, the as type of gathered data equipment.
2. the numbering of node serial number, as network node.
3. data form, the data of as same data form is as a class.
After completing data reconstruction, then by these data blocks be loaded onto one, multiple stage, even computer cluster composition in
In depositing.But the mass data in view of Internet of Things may make low memory, the buffering using " disk-internal memory " is carried out
Process, that is,：When low memory is to accommodate all of data, initial data will be retained in disk, every time only will be a certain amount of
Data assembling is data block and is loaded onto internal memory, after the data block in internal memory completes to process, then loads remaining number from disk
According to.
3rd step：Initial data after reconstructing is reassembled as UDB tree by enforcement personnel, and this UDB tree is the number described in concept 1
According to structure.Using search algorithm UDB of algorithm 1 description, the data block in second step is started to index according to keyword, is finding
Behind corresponding position, insert corresponding node according still further to the UDB insertion algorithm of algorithm 2 description, by data block, the data needing
After block is inserted into corresponding position, these data blocks are just constructed as UDB tree.Wherein, root node represents total type, intermediate node table
Show classifying type, the data within each intermediate node represents the species number of same level, terminal node represents the true of this type
Data.
4th step：Read rule set in advance, carry out data cleansing.In data cleansing, operating personnel can pass through
Software merit rating, or utilize configuration file, cleaning rule is defined, this rule can use data base query language
(SQL), script (as JavaScript) is created and is edited.During whole data cleansing, will be in cleaning rule
When being carried out under conditions of agreement, and cleaning, multiple steps can be divided into, each step can formulate cleaning rule, such as
Fig. 6.
The basic principles, principal features and advantages of the present invention have been shown and described above.Above embodiment is only in order to retouch
State technical scheme rather than technical method limited, the present invention may extend away in application for other modifications,
Change and application, and think that all such modifications, change and application both fall within scope of the claimed invention.
Claims (5)
1. a kind of Internet of Things data cleaning method it is characterised in that：Comprise the following steps：
The first step：Enforcement personnel obtain Internet of Things initial data by Web Service；
The detailed process of the described first step comprises the steps：Using the mode of Web Service, data, services are externally provided to connect
Mouthful, by data uniform transmission to central database, central database adopts traditional relevant database, is each type of setting
The respective tables of data of standby establishment, device numbering will be stored in respective number as unique identifier, the initial data of all collections
According in table；
Second step：After initial data is reconstructed by enforcement personnel, store it in internal memory；
The detailed process of described second step comprises the steps：After obtaining all data from the central database, implement personnel by this
A little initial data are assembled into as data block；After completing data reconstruction, then these data blocks are loaded onto in internal memory；
3rd step：Initial data after reconstructing is reassembled as UDB tree by enforcement personnel；
The detailed process of described 3rd step comprises the steps：Using search algorithm UDB, by the data block in second step according to pass
Key word starts to index, and after finding corresponding position, according still further to UDB insertion algorithm, data block is inserted corresponding node；
Described search algorithm UDB comprises the steps：
1) according to the key value of node, in whole tree, from the beginning all nodes in node have started, middle node in these
Enter line retrieval in point, according to the Query Result in node, enter in corresponding region and carry out interval query；
2) after obtaining the Query Result of subregion in node, need the child node pointer deposited according to subregion head, add and deposit
It is placed on the relative displacement in index node data block, this side-play amount is the node serial number of first node relatively；Circulation step 1),
2) until finding indexed node；
3) using the HASH function setting, after calculating indexed node corresponding HASH value, visit again data block chain；
Described UDB insertion algorithm comprises the steps：
During to the new node of UDB tree insertion, the leaf node that key word of the inquiry should insert；After leaf node is positioned, need
Judge whether this node has enough spaces to accommodate new index entry；If insufficient space, need to divide leaf node；
When leaf node needs division, need to consider whether father node has enough spaces；If father node P has enough skies
Between, f is the child node group pointer on some subregion in P；G is the child node group of f indication；During certain node split in g,
Need all nodes in bigger child node group g' of distribution ratio g to be all copied in g, include the new section being split off producing simultaneously
Point；Then by child node group g' new for sensing, the space of g will be released f；The corresponding index entry of new node is inserted into father's section
In point P；
If father node does not have redundant space, then itself also will be split off；Hypothesis P is father node, and f is some point in P
Child node group pointer in area, g is the child node group of f indication；During division, distribute new child node group g', and replicate institute from g
Some child nodes and new split vertexes；P itself also will be split off, and generate new node P ', and replicate the index of half from P
?；Meanwhile, the node group that P is located is also required to redistribute space and replica node data is accommodating P'；If necessary, also will
Continue father's node of division P；Meanwhile, the division of every minor node also needs to adjust index in node；
4th step：Read rule set in advance, carry out data cleansing.
2. a kind of Internet of Things data cleaning method according to claim 1 it is characterised in that：The detailed process bag of the 4th step
Include following steps：In data cleansing, operating personnel can pass through software merit rating, or utilizes configuration file, to cleaning rule
It is defined, during whole data cleansing, will be carried out under conditions of cleaning rule agreement, and when cleaning, permissible
It is divided into multiple steps, each step can formulate different cleaning rules.
3. a kind of Internet of Things data cleaning method according to claim 1 it is characterised in that：Interior described in second step save as
The internal memory of computer cluster composition.
4. a kind of Internet of Things data cleaning method according to claim 3 it is characterised in that：When low memory is to accommodate
During some data, initial data will be retained in disk, only for data block and the data assembling of part is loaded onto internal memory every time,
After data block in internal memory completes to process, then load remaining data from disk.
5. a kind of Internet of Things data cleaning method according to claim 2 it is characterised in that：Described cleaning rule uses
Data base query language or script are created and are edited.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310081635.4A CN103177094B (en) | 2013-03-14 | 2013-03-14 | Cleaning method of data of internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310081635.4A CN103177094B (en) | 2013-03-14 | 2013-03-14 | Cleaning method of data of internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103177094A CN103177094A (en) | 2013-06-26 |
| CN103177094B true CN103177094B (en) | 2017-02-22 |
Family
ID=48636955
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310081635.4A Active CN103177094B (en) | 2013-03-14 | 2013-03-14 | Cleaning method of data of internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103177094B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103561019B (en) * | 2013-10-30 | 2018-02-06 | 上海斐讯数据通信技术有限公司 | One kind is directed to TR069 data access methods |
| CN105550315A (en) * | 2015-12-11 | 2016-05-04 | 广州精点计算机科技有限公司 | Data repetition removal method and device, and data cleaning method and device |
| CN107229662B (en) * | 2016-03-25 | 2022-02-25 | 阿里巴巴集团控股有限公司 | Data cleaning method and device |
| CN105913316B (en) * | 2016-04-15 | 2019-12-10 | 中国银行股份有限公司 | Rule configuration method and device |
| CN106294745A (en) * | 2016-08-10 | 2017-01-04 | 东方网力科技股份有限公司 | Big data cleaning method and device |
| CN108153747A (en) * | 2016-12-02 | 2018-06-12 | 航天星图科技（北京）有限公司 | A kind of parallel data cleaning system |
| CN106790491B (en) * | 2016-12-14 | 2019-10-15 | 日照职业技术学院 | The implementation method of the Internet of Things movable termination intelligent platform of data-oriented |
| CN108319609A (en) * | 2017-01-16 | 2018-07-24 | 医渡云（北京）技术有限公司 | ETL data processing methods and system, data cleaning method and device |
| CN106933990A (en) * | 2017-02-21 | 2017-07-07 | 南京朴厚生态科技有限公司 | A kind of sensing data cleaning method |
| CN109241045A (en) * | 2018-08-29 | 2019-01-18 | 宜人恒业科技发展（北京）有限公司 | A kind of method and apparatus of preprocessed data |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102142027A (en) * | 2011-02-23 | 2011-08-03 | 南京智尚丰软件有限公司 | Adaptive method for data integration |
| CN102722582A (en) * | 2012-06-07 | 2012-10-10 | 陈浩 | System and method for integrating data on basis of reverse clearing |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7634464B2 (en) * | 2006-06-14 | 2009-12-15 | Microsoft Corporation | Designing record matching queries utilizing examples |
-
2013
- 2013-03-14 CN CN201310081635.4A patent/CN103177094B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102142027A (en) * | 2011-02-23 | 2011-08-03 | 南京智尚丰软件有限公司 | Adaptive method for data integration |
| CN102722582A (en) * | 2012-06-07 | 2012-10-10 | 陈浩 | System and method for integrating data on basis of reverse clearing |
Non-Patent Citations (1)
| Title |
|---|
| 在数据清洗过程中基于MMDB的数据匹配技术研究;史习一;《中国优秀硕士学位论文全文数据库》;20070331;正文第23页第5段、第24页第5-7段、第47页第1-3段、第48页第1-6段、第50页第1-3段、 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103177094A (en) | 2013-06-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103177094B (en) | Cleaning method of data of internet of things | |
| EP3365808B1 (en) | Proxy databases | |
| Maccioni et al. | Scalable pattern matching over compressed graphs via dedensification | |
| CN102270232B (en) | Semantic data query system with optimized storage | |
| Brunkhorst et al. | Distributed queries and query optimization in schema-based P2P-systems | |
| CN103003813B (en) | Columnar storage representations of records | |
| CN102332030A (en) | Data storing, managing and inquiring method and system for distributed key-value storage system | |
| CN103116625A (en) | Volume radio direction finde (RDF) data distribution type query processing method based on Hadoop | |
| CN106452450A (en) | Data compression | |
| WO2009031915A1 (en) | Method and a system for storing, retrieving and extracting information on the basis of low-organised and decentralised datasets | |
| CN103646032A (en) | Database query method based on body and restricted natural language processing | |
| CN104239572A (en) | System and method for achieving metadata analysis based on distributed cache | |
| CN107451225A (en) | Scalable analysis platform for semi-structured data | |
| CN104205039A (en) | Interest-driven business intelligence systems and methods of data analysis using interest-driven data pipelines | |
| WO2017070590A1 (en) | Proxy databases | |
| CN103678550A (en) | Mass data real-time query method based on dynamic index structure | |
| CN102999563A (en) | Network resource semantic retrieval method and system based on resource description framework | |
| CN107247799A (en) | Data processing method, system and its modeling method of compatible a variety of big data storages | |
| CN105608228B (en) | A kind of efficient distributed RDF data storage method | |
| CN103646051A (en) | Big-data parallel processing system and method based on column storage | |
| CN101710336A (en) | Method for accelerating data processing by using relational middleware | |
| Ma et al. | G-SQL: Fast query processing via graph exploration | |
| Wang et al. | TSMH Graph Cube: A novel framework for large scale multi-dimensional network analysis | |
| CN112231351A (en) | Real-time query method and device for PB-level mass data | |
| CN105550351B (en) | The extemporaneous inquiry system of passenger's run-length data and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent for invention or patent application | ||
| CB02 | Change of applicant information |
Address after: 610054 information industry building, 159 East Ring Road, Chengdu, Chenghua District, Sichuan Applicant after: Chengdu Comsys Information Technology Co., Ltd. Address before: 610054 information industry building, 159 East Ring Road, Chengdu, Chenghua District, Sichuan Applicant before: Uestc Comsys Information Co., Ltd. |
|
| CB03 | Change of inventor or designer information |
Inventor after: Tang Xuefei Inventor after: Chen Ke Inventor after: Shi Li Inventor after: Han Chunmei Inventor before: Tang Xuefei Inventor before: Chen Ke Inventor before: Shi Li |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: APPLICANT; FROM: CHENGDU KANGSAI INFORMATION TECHNOLOGY CO., LTD. OF UESTC TO: CHENGDU COMSYS INFORMATION TECHNOLOGY CO., LTD. Free format text: CORRECT: INVENTOR; FROM: TANG XUEFEI CHEN KE SHI LI TO: TANG XUEFEI CHEN KE SHI LI HAN CHUNMEI |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |