CN109918602A - A kind of web data preloading method and system - Google Patents
A kind of web data preloading method and system Download PDFInfo
- Publication number
- CN109918602A CN109918602A CN201910141291.9A CN201910141291A CN109918602A CN 109918602 A CN109918602 A CN 109918602A CN 201910141291 A CN201910141291 A CN 201910141291A CN 109918602 A CN109918602 A CN 109918602A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- page
- zone
- hot
- 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
Landscapes
- Information Transfer Between Computers (AREA)
Abstract
The invention belongs to data load domains, disclose a kind of web data preloading method and system, and experience of the user in browsing pages data is improved in the hot-zone that prediction user may may click next time；Front and back is kept to communicate by establishing long connection, it is responsible for that user behavior data is acquired and is tracked in foreground, and the prediction result on backstage is executed, the user behavior data of foreground acquisition is periodically reported to backstage, behavior is clicked as separator using user in backstage, carrys out the behavioral data that process cycle reports.The present invention uses study of the machine learning to user's history behavioral data, and according to the local weighted algorithm such as time, user, page type, predicts the hot-zone that user may click next time, notice front end executes；Web data pre-add support method provided by the invention promotes data loading, and more fitting user habit.
Description
Technical field
The invention belongs to data load domains more particularly to a kind of web data to preload method and system
Background technique
With the development of the emerging technologies such as mobile Internet, Internet of Things, user data is exponentially increased, with data volume
Increase, the loading velocity of the page is also slower and slower, user need to expend more times come waiting system response.
Have benefited from the application of caching technology, data loading has a certain upgrade, but due to being cached with certain hit
Probability and timeliness cause data load sometimes fast and sometimes slow.It is difficult to provide the user with a stable experience.
Machine learning is the learning behavior specialized in computer and how to simulate or realize the mankind, with obtain new knowledge or
Technical ability reorganizes the existing structure of knowledge and is allowed to constantly improve the performance of itself.Utilize this characteristic of machine learning, the present invention
Web data can be preloaded to behavior and give machine learning to predict, traditional Web is solved with this and loads speed in big data
The problem of degree is slow, poor user experience.
In conclusion problem of the existing technology is:
(1) the problem of tradition Web loading velocity in big data is slow, poor user experience.
(2) tradition preload not can be reduced TCP shake hands upper waste time and to server generate pressure, cause data
Stabilization transmission performance it is poor.
(3) traditional caching technology hit rate is low, and caching breakdown occurs often, and the concurrent pressure of database server is big.
(4) traditional preloading technology point of use (individual element) is as object is preloaded, usually with keyword for triggering side
Formula, preloading accuracy rate is low, is separated with user behavior habit.
(5) legacy user's tracking technique is immature, can not track user operation habits comprehensively.
Solve the difficulty and meaning of above-mentioned technical problem:
Human information level improves year by year, and big data uses and be treated as certainty, since hardware device limits, greatly
Data processing all brings to database server, application server and consumes for a long time, so as to cause the decline of integrity service performance.
The present invention promotes big data under the premise of not increasing hardware device investment, by having information technology and means
Loading velocity reduces the waiting time that user is wasted when using big data, promotes user while promoting economic benefit
Experience.Frequent TCP is solved by long interconnection technique to shake hands, and guarantees the real-time tracking of user behavior.It is right by machine learning algorithm
Existing subscriber's behavioral data carries out machine learning, with behavior prediction behavior, improves cache hit rate, it is concurrent to reduce database service
Number.In the case where existing service architecture is constant, this technology can be transformed traditional services without intrusive mood, optimize user's body
It tests.
Summary of the invention
In view of the problems of the existing technology, the present invention provides a kind of web datas to preload method and system, utilizes machine
Device learns this characteristic, and web data can be preloaded behavior and give machine learning to predict by the present invention, solves tradition with this
The problem of Web loading velocity in big data is slow, poor user experience.
The invention is realized in this way a kind of web data pre-add support method includes:
Step 1: being that active user uses one long connection of creation between machine and server in frame page；
Step 2: the interface that front end script currently browses user in a grid formation, since the page upper left corner, according to from
It is left-to-right, mode from top to bottom be cut into same size, etc. pixels square hot-zone；
Step 3: tracking user's mouse motion track, mouse-over hot-zone, mousebutton operation, source page, the page stop
Stay the operation behaviors such as time；It is logical to the user's operation behavior, page feature ID, User ID (the case where if there is logging in) that are collected into
Too long interface channel is sent to backstage；
Step 4: separate training set data using each click of user, and user's operation behavioral data packet is carried out
It extracts；
Step 5: when including that user clicks in the user's operation behavior extracted, i.e., user completes clicks from first time
Second is arrived later and clicks the behavioral data terminated, as a complete training set, inserts data into machine learning training
It concentrates；Behavioral data before operation behavior is second of click, then judge whether data are consistent with the last time, if consistent
It does nothing, otherwise needs machine learning according to the training such as users personal data, personal data time, page relevance
Collection carries out local weighted calculating and makes the prediction that user clicks hot-zone next time；
Step 6: pushing to user front end for the hot-zone of prediction, and front end script checks the data resource within the scope of hot-zone, with
The mode of Ajax to backstage, complete that data can be added to buffer area after resource preloads in advance by request resource, backstage, and user clicks
When predicting hot-zone, data will be returned directly from buffer area.
Further, in step 2, the page cutting, which divides, includes:
The page is cut into the separation on logical meaning, does not show and has an impact to front end page；Separate thermal zone is according to reality
Border ratio zooms in and out, only related in the page, unrelated with browser window size and screen resolution.
Further, in step 3, the page feature ID includes:
Page feature ID is can to identify that page generic and the page are uniquely worth；Such as user is taken from the background just clear
The page feature ID look at, can learn that user browsing is user's blogroll page.
Further, in step 4, the training set data includes: when user starts after clicking in hot-zone first time
And second of all operation behavior between click in hot-zone terminates is referred to as a complete training set, it is complete at one
Training set does not occur to be called forecast set when event is clicked in second of hot-zone；
Divide according to data consumer, there are user and other people in the source of training set.
Further, in step 4, the data packet extraction includes:
(1) it is partitioned into the user behavior data packet of report cycle, is input to prediction input parameter.
(2) it is separated out complete user's operation behavioral data packet each time, training set is input to, is provided as machine learning.
Further, in step 5, the hot-zone prediction technique includes:
(1) behavioral data that the continuous receiving front-end in backstage is collected, and record as training set data；
(2) after obtaining certain training set data, using training set data as the sample of machine learning, mode knowledge is carried out
Not, wherein user behavior is as input object, and the hot-zone that user clicks is as desired output；
(3) deduction ability is generated by analyzing and training data, forecast set data is inferred, obtain prediction hot-zone.
Further, in step 5, the local weighted calculating includes:
Because there are user and other people in the source of training set；Self data has biggish behavior meaning, other numbers
According to biggish data sense；Behavior meaning refers to operating habit and Behavior preference of user etc., and data sense refers to
Data attention rate and data sensibility etc.；Therefore there need to be individual prediction algorithm to each user, engineering is guaranteed with this
The prediction algorithm that acquistion is arrived can introduce local weighted calculation method closer to individual consumer；
Local weighted calculation method principle: self data's weight is greater than other people data weightings, in person Recent data weight
Greater than my data weighting at a specified future date, related pages weight is greater than irrelevant page weight.
Another object of the present invention is to provide a kind of web data preloadings for implementing the web data pre-add support method
Control system.
Another object of the present invention is to provide a kind of web datas to preload computer program, and the web data preloads
Computer program realizes the web data pre-add support method.
Another object of the present invention is to provide a kind of terminal, the terminal, which is at least carried, realizes the web data pre-add
The server of support method.
Another object of the present invention is to provide a kind of computer readable storage mediums, including instruction, when it is in computer
When upper operation, so that computer executes the web data pre-add support method.
In conclusion advantages of the present invention and good effect are as follows:
The long connection of present invention creation keeps front and back end to communicate simultaneously real-time report user's operation behavior and executes preload operation,
Can reduce TCP shake hands upper waste time and to server generate pressure, guarantee data stablize transmission.
The hot-zone that the page is divided into block of pixels is predicted and is tracked by front end script, and compared with traditional technology, this technology is not
It relies on page layout, depend on page control, reduce improvement cost.
The separation as user behavior data is clicked using user, machine learning algorithm is introduced and user behavior is learnt
And analysis, and field is segmented according to time, user, page type etc. and carries out local weighted prediction, machine learning is constantly being predicted
In constantly learn, improve machine learning prediction accuracy.
By promoting data loading to web data pre-add support method, caching breakdown possibility is reduced, user is optimized
Experience, fitting user's habit.
Present invention introduces machine learning algorithms, guarantee big data loading velocity by prediction mode, with space for time
Mode compare, this technology cost is lower, and applicability is wider.
The present invention does not need to be adjusted prior art framework, and invasive is low, can Quick thread and offline when needed.
Detailed description of the invention
Fig. 1 is web data pre-add support method schematic diagram of logic principle provided in an embodiment of the present invention；
Fig. 2 is that front end page hot-zone provided in an embodiment of the present invention divides schematic diagram；
Fig. 3 is that web data provided in an embodiment of the present invention preloads method flow diagram.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to
Limit the present invention.
The problem of traditional Web loading velocity in big data is slow, poor user experience.In the prior art, it not can be reduced
TCP shake hands upper waste time and to server generate pressure, cause the stabilization transmission performance of data poor；Machine learning prediction
Accuracy it is poor.
In order to solve the above technical problems, being described in detail below with reference to concrete scheme to application principle of the invention.
As shown in Figure 1 to Figure 3, web data pre-add support method provided in an embodiment of the present invention includes:
S101: being that active user uses one long connection of creation between machine and server in frame page.
S102: the interface that front end script currently browses user in a grid formation, since the page upper left corner, according to from a left side
To the right side, mode from top to bottom be cut into same size, etc. pixels square hot-zone.
S103: tracking user's mouse motion track, mouse-over hot-zone, mousebutton operation, source page, the page stop
The operation behaviors such as time；The user's operation behavior that is collected into, page feature ID, User ID (the case where if there is logging in) are passed through
Long interface channel is sent to backstage.
S104: using each click of user to separate training set data, and user's operation behavioral data packet is mentioned
It takes.
S105: when including that user clicks in the user's operation behavior extracted, i.e., user completes from clicking it for the first time
The behavioral data that second is clicked end is arrived afterwards inserts data into machine learning training set as a complete training set
In；Behavioral data before operation behavior is second of click, then judge whether data are consistent with the last time, if consistent not
Any operation is carried out, otherwise needs machine learning according to training sets such as users personal data, personal data time, page relevances
It carries out local weighted calculating and makes the prediction that user clicks hot-zone next time.
S106: pushing to user front end for the hot-zone of prediction, and front end script checks the data resource within the scope of hot-zone, with
The mode of Ajax to backstage, complete that data can be added to buffer area after resource preloads in advance by request resource, backstage, and user clicks
When predicting hot-zone, data will be returned directly from buffer area.
In step S102, page cutting provided in an embodiment of the present invention, which divides, includes:
The page is cut into the separation on logical meaning, does not show and has an impact to front end page；Separate thermal zone is according to reality
Border ratio zooms in and out, only related in the page, unrelated with browser window size and screen resolution.
In step S103, page feature ID provided in an embodiment of the present invention includes:
Page feature ID is can to identify that page generic and the page are uniquely worth；Such as user is taken from the background just clear
The page feature ID look at, can learn that user browsing is user's blogroll page.
In step S104, training set data provided in an embodiment of the present invention includes:
When user first time start after being clicked in hot-zone and second clicked in hot-zone terminate between it is all
Operation behavior is referred to as a complete training set, and quilt when event is clicked in second of hot-zone does not occur in a complete training set
Referred to as forecast set.
Divide according to data consumer, there are user and other people in the source of training set.
In step S104, data packet extraction provided in an embodiment of the present invention includes:
(1) it is partitioned into the user behavior data packet of report cycle, is input to prediction input parameter.
(2) it is separated out complete user's operation behavioral data packet each time, training set is input to, is provided as machine learning.
In step S105, hot-zone prediction technique provided in an embodiment of the present invention includes:
(1) behavioral data that the continuous receiving front-end in backstage is collected, and record as training set data.
(2) after obtaining certain training set data, using training set data as the sample of machine learning, mode knowledge is carried out
Not, wherein user behavior is as input object, and the hot-zone that user clicks is as desired output.
(3) deduction ability is generated by analyzing and training data, forecast set data is inferred, obtain prediction hot-zone.
In step S105, local weighted calculating provided in an embodiment of the present invention includes:
Because there are user and other people in the source of training set；Self data has biggish behavior meaning, other numbers
According to biggish data sense；Behavior meaning refers to operating habit and Behavior preference of user etc., and data sense refers to
Data attention rate and data sensibility etc.；Therefore there need to be individual prediction algorithm to each user, engineering is guaranteed with this
The prediction algorithm that acquistion is arrived can introduce local weighted calculation method closer to individual consumer.
Local weighted calculation method principle: self data's weight is greater than other people data weightings, in person Recent data weight
Greater than my data weighting at a specified future date, related pages weight is greater than irrelevant page weight.
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or any combination thereof real
It is existing.When using entirely or partly realizing in the form of a computer program product, the computer program product include one or
Multiple computer instructions.When loading on computers or executing the computer program instructions, entirely or partly generate according to
Process described in the embodiment of the present invention or function.The computer can be general purpose computer, special purpose computer, computer network
Network or other programmable devices.The computer instruction may be stored in a computer readable storage medium, or from one
Computer readable storage medium is transmitted to another computer readable storage medium, for example, the computer instruction can be from one
A web-site, computer, server or data center pass through wired (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)
Or wireless (such as infrared, wireless, microwave etc.) mode is carried out to another web-site, computer, server or data center
Transmission).The computer-readable storage medium can be any usable medium or include one that computer can access
The data storage devices such as a or multiple usable mediums integrated server, data center.The usable medium can be magnetic Jie
Matter, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disk Solid
State Disk (SSD)) etc..
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (9)
1. a kind of web data pre-add support method, which is characterized in that the web data pre-add support method includes:
Step 1: being that active user uses the long connection of creation between machine and server in frame page；
Step 2: the interface that front end script currently browses user in a grid formation, since the page upper left corner, according to from a left side to
The right side, mode from top to bottom be cut into same size, etc. pixels square hot-zone；
Step 3: when tracking user's mouse motion track, mouse-over hot-zone, mousebutton operation, source page, the page stop
Between operation behavior；To be collected into user's operation behavior, page feature ID, User ID by long interface channel to backstage send；
Step 4: using each click of user to separate training set data, and user's operation behavioral data packet is extracted；
Step 5: when including that user clicks in the user's operation behavior extracted, user is completed after clicking from first time to the
The secondary behavioral data terminated of clicking inserts data into machine learning training set as a complete training set；Work as behaviour
Make the behavioral data before behavior is second of click, then judge whether data are consistent with the last time, without appointing if consistent
What is operated, and machine learning is otherwise needed to carry out part according to users personal data, personal data time, page relevance training set
Weighted calculation makes the prediction that user clicks hot-zone next time；
Step 6: the hot-zone of prediction is pushed into user front end, front end script checks the data resource within the scope of hot-zone, with Ajax
Mode to backstage, request resource, backstage complete that data buffer area can be added after resource preloads in advance, user clicks prediction
When hot-zone, data will be returned directly from buffer area.
2. web data pre-add support method as described in claim 1, which is characterized in that in step 2, the page cutting is divided
Include:
The page is cut into the separation on logical meaning, does not show and has an impact to front end page；Separate thermal zone is according to practical ratio
Example zooms in and out, only related in the page, unrelated with browser window size and screen resolution；
In step 3, the page feature ID includes:
Page feature ID is can to identify that page generic and the page are uniquely worth.
3. web data pre-add support method as described in claim 1, which is characterized in that in step 4, the training set data packet
It includes:
When user starts after clicking in hot-zone in first time and is clicking all operations between terminating in hot-zone for the second time
Behavior is referred to as a complete training set, does not occur to be referred to as when event is clicked in second of hot-zone in a complete training set
For forecast set；
Divide according to data consumer, there are user and other people in the source of training set；
In step 4, the data packet extraction includes:
(1) it is partitioned into the user behavior data packet of report cycle, is input to prediction input parameter.
(2) it is separated out complete user's operation behavioral data packet each time, training set is input to, is provided as machine learning.
4. web data pre-add support method as described in claim 1, which is characterized in that in step 5, the hot-zone prediction technique
Include:
(1) behavioral data that the continuous receiving front-end in backstage is collected, and record as training set data；
(2) after obtaining certain training set data, using training set data as the sample of machine learning, pattern-recognition is carried out,
Middle user behavior is as input object, and the hot-zone that user clicks is as desired output；
(3) deduction ability is generated by analyzing and training data, forecast set data is inferred, obtain prediction hot-zone.
5. web data pre-add support method as described in claim 1, which is characterized in that in step 5, the local weighted calculating
Include:
Local weighted calculation method principle includes: that self data's weight is greater than other people data weightings, in person Recent data weight
Greater than my data weighting at a specified future date, related pages weight is greater than irrelevant page weight.
6. a kind of web data pre-add borne control system for implementing web data pre-add support method described in claim 1.
7. a kind of web data preloads computer program, which is characterized in that the web data preloads computer program and realizes
Web data pre-add support method described in Claims 1 to 5 any one.
8. a kind of terminal, which is characterized in that the terminal, which is at least carried, realizes web data described in Claims 1 to 5 any one
The server of pre-add support method.
9. a kind of computer readable storage medium, including instruction, when run on a computer, so that computer is executed as weighed
Benefit requires web data pre-add support method described in 1-5 any one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910141291.9A CN109918602B (en) | 2019-02-26 | 2019-02-26 | Web data preloading method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910141291.9A CN109918602B (en) | 2019-02-26 | 2019-02-26 | Web data preloading method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109918602A true CN109918602A (en) | 2019-06-21 |
| CN109918602B CN109918602B (en) | 2021-04-30 |
Family
ID=66962265
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910141291.9A Active CN109918602B (en) | 2019-02-26 | 2019-02-26 | Web data preloading method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109918602B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889061A (en) * | 2019-11-12 | 2020-03-17 | 望海康信(北京)科技股份公司 | Webpage loading method and device |
| CN111881391A (en) * | 2020-06-12 | 2020-11-03 | 马上消费金融股份有限公司 | Static network resource preloading method, data model training method and device |
| CN112733044A (en) * | 2021-03-30 | 2021-04-30 | 腾讯科技（深圳）有限公司 | Recommended image processing method, apparatus, device and computer-readable storage medium |
| CN117009690A (en) * | 2023-07-03 | 2023-11-07 | 唯科终端技术(深圳)有限公司 | Method and system for preloading content |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101604393A (en) * | 2009-07-10 | 2009-12-16 | 华南理工大学 | A kind of Chinese-character stroke feature extracting method that is used for on-line handwritten Chinese character identification |
| CN102186103A (en) * | 2011-04-20 | 2011-09-14 | 深圳市同洲软件有限公司 | Program on demand request transmission method and system |
| US20150193395A1 (en) * | 2012-07-30 | 2015-07-09 | Google Inc. | Predictive link pre-loading |
| CN106202368A (en) * | 2016-07-07 | 2016-12-07 | 贵州白山云科技有限公司 | Prestrain method and apparatus |
| CN106779817A (en) * | 2016-11-29 | 2017-05-31 | 竹间智能科技（上海）有限公司 | Intension recognizing method and system based on various dimensions information |
| CN107632985A (en) * | 2016-07-18 | 2018-01-26 | 腾讯科技（北京）有限公司 | Webpage preloads method and device |
| CN107995259A (en) * | 2017-11-14 | 2018-05-04 | 北京思特奇信息技术股份有限公司 | A kind of method and device handled cross-domain request |
| CN108153565A (en) * | 2016-12-02 | 2018-06-12 | 阿里巴巴集团控股有限公司 | The method and device of page info is provided |
| CN109271015A (en) * | 2018-10-10 | 2019-01-25 | 杭州电子科技大学 | A method of reducing large-scale distributed machine learning system energy consumption |
-
2019
- 2019-02-26 CN CN201910141291.9A patent/CN109918602B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101604393A (en) * | 2009-07-10 | 2009-12-16 | 华南理工大学 | A kind of Chinese-character stroke feature extracting method that is used for on-line handwritten Chinese character identification |
| CN102186103A (en) * | 2011-04-20 | 2011-09-14 | 深圳市同洲软件有限公司 | Program on demand request transmission method and system |
| US20150193395A1 (en) * | 2012-07-30 | 2015-07-09 | Google Inc. | Predictive link pre-loading |
| CN106202368A (en) * | 2016-07-07 | 2016-12-07 | 贵州白山云科技有限公司 | Prestrain method and apparatus |
| CN107632985A (en) * | 2016-07-18 | 2018-01-26 | 腾讯科技（北京）有限公司 | Webpage preloads method and device |
| CN106779817A (en) * | 2016-11-29 | 2017-05-31 | 竹间智能科技（上海）有限公司 | Intension recognizing method and system based on various dimensions information |
| CN108153565A (en) * | 2016-12-02 | 2018-06-12 | 阿里巴巴集团控股有限公司 | The method and device of page info is provided |
| CN107995259A (en) * | 2017-11-14 | 2018-05-04 | 北京思特奇信息技术股份有限公司 | A kind of method and device handled cross-domain request |
| CN109271015A (en) * | 2018-10-10 | 2019-01-25 | 杭州电子科技大学 | A method of reducing large-scale distributed machine learning system energy consumption |
Non-Patent Citations (1)
| Title |
|---|
| 王懿: "Web页面加载性能监测平台的设计与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑（月刊）》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889061A (en) * | 2019-11-12 | 2020-03-17 | 望海康信(北京)科技股份公司 | Webpage loading method and device |
| CN111881391A (en) * | 2020-06-12 | 2020-11-03 | 马上消费金融股份有限公司 | Static network resource preloading method, data model training method and device |
| CN112733044A (en) * | 2021-03-30 | 2021-04-30 | 腾讯科技（深圳）有限公司 | Recommended image processing method, apparatus, device and computer-readable storage medium |
| CN112733044B (en) * | 2021-03-30 | 2021-07-16 | 腾讯科技（深圳）有限公司 | Recommended image processing method, apparatus, device and computer-readable storage medium |
| CN117009690A (en) * | 2023-07-03 | 2023-11-07 | 唯科终端技术(深圳)有限公司 | Method and system for preloading content |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109918602B (en) | 2021-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109918602A (en) | A kind of web data preloading method and system | |
| US9367601B2 (en) | Cost-based optimization of configuration parameters and cluster sizing for hadoop | |
| EP2705438B1 (en) | Predicting user navigation events | |
| US9070046B2 (en) | Learning-based image webpage index selection | |
| US9590880B2 (en) | Dynamic collection analysis and reporting of telemetry data | |
| Ali et al. | Intelligent Naïve Bayes-based approaches for Web proxy caching | |
| JP7267305B2 (en) | Advanced web page content management | |
| TWI676913B (en) | Data processing method, data transmission method, risk identification method and device | |
| US20070271519A1 (en) | System and Method for Collecting User Interest Data | |
| CN105069087B (en) | The web information flow method excavated based on web log data | |
| US20190311114A1 (en) | Man-machine identification method and device for captcha | |
| EP2972804A1 (en) | Personalized community model for surfacing commands within productivity application user interfaces | |
| TW200925970A (en) | Customized today module | |
| CN104798072A (en) | Information management and display in web browsers | |
| CN103176985A (en) | Timely and high-efficiency crawling method for internet information | |
| EP2904509A1 (en) | Improving access to network content | |
| CN102254004A (en) | Method and system for modeling Web in weblog excavation | |
| Apaolaza et al. | Understanding users in the wild | |
| Kirsh et al. | Splitting the web analytics atom: from page metrics and KPIs to sub-page metrics and KPIs | |
| CN110020273B (en) | Method, device and system for generating thermodynamic diagram | |
| US10467255B2 (en) | Methods and systems for analyzing reading logs and documents thereof | |
| Wang et al. | Information classification algorithm based on decision tree optimization | |
| US11108802B2 (en) | Method of and system for identifying abnormal site visits | |
| Diaz et al. | Search result prefetching using cursor movement | |
| US20120166926A1 (en) | Hyperlink display method based on visit history accumulation |
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 |