CN116992183B - Domestic browser adaptation compatible method based on deep learning technology - Google Patents
Domestic browser adaptation compatible method based on deep learning technology Download PDFInfo
- Publication number
- CN116992183B CN116992183B CN202311245727.1A CN202311245727A CN116992183B CN 116992183 B CN116992183 B CN 116992183B CN 202311245727 A CN202311245727 A CN 202311245727A CN 116992183 B CN116992183 B CN 116992183B
- Authority
- CN
- China
- Prior art keywords
- tree
- dom
- node
- interaction
- event
- 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
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000005516 engineering process Methods 0.000 title claims abstract description 20
- 238000013135 deep learning Methods 0.000 title claims abstract description 14
- 230000006978 adaptation Effects 0.000 title claims abstract description 11
- 238000009877 rendering Methods 0.000 claims abstract description 31
- 230000003993 interaction Effects 0.000 claims description 97
- 230000002452 interceptive effect Effects 0.000 claims description 33
- 238000013528 artificial neural network Methods 0.000 claims description 11
- 230000004913 activation Effects 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 abstract description 3
- 238000006243 chemical reaction Methods 0.000 abstract description 2
- 230000010365 information processing Effects 0.000 abstract description 2
- VYZAMTAEIAYCRO-UHFFFAOYSA-N Chromium Chemical group [Cr] VYZAMTAEIAYCRO-UHFFFAOYSA-N 0.000 description 4
- 230000004044 response Effects 0.000 description 3
- 229910052804 chromium Inorganic materials 0.000 description 2
- 239000011651 chromium Substances 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/957—Browsing optimisation, e.g. caching or content distillation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention relates to the technical field of network information processing, and discloses a home-made browser adaptation compatible method based on a deep learning technology, which comprises the following steps: the client starts a domestic browser, inputs a website and sends a request to the cloud server; the cloud server receives the request and downloads the webpage content, and then analyzes the webpage content to render and generate an image cache; the cloud server sends the generated image cache to the client, and the client receives the image cache and then directly displays the image cache in the browser; according to the invention, a cloud service conversion technology is adopted, a webpage accessed by a user in a domestic browser is transmitted to a cloud service for analysis and rendering by utilizing a cloud computing technology, and the processed webpage is returned to the user, so that the method has strong universality and adaptability, and the problem of incompatibility of the webpage is avoided.
Description
Technical Field
The invention relates to the technical field of network information processing, in particular to a home-made browser adaptation compatible method based on a deep learning technology.
Background
Because the kernels used by different browsers and the webpage language standards such as HTML supported by the kernels are different, the same webpage has different display effects in different browsers. At present, in the aspect of solving the compatibility problem of domestic browsers, a dual-core browser technical means is generally adopted to realize the switching between an IE core and a Chrome core, so that the compatibility of the browser is improved. However, the dual-core technology is only applicable to two cores, and cannot cope with the compatibility problem of multiple browser cores, and the dual-core browser consumes more memory and system resources, so that the user experience and the running speed are affected.
Disclosure of Invention
The invention provides a home browser adaptation compatibility method based on a deep learning technology, which solves the technical problem of low compatibility of home browser cores in the related technology.
The invention provides a home-made browser adaptation compatible method based on a deep learning technology, which comprises the following steps: step 101, a client starts a domestic browser and inputs a website, and sends a request to a cloud server; the cloud server receives the request and downloads the webpage content, and then analyzes the webpage content to render and generate an image cache; step 102, the cloud server sends the generated image cache to a client, and the client receives the image cache and then directly displays the image cache in a browser; step 103, identifying the interaction event type of the interaction event of the client, wherein the interaction event type of the interaction event comprises link interaction and non-link interaction; if the identification result is a link interaction, go to step 104, and if the identification result is a non-link interaction, go to step 107; 104, extracting DOM tree, CSSOM tree and rendering tree of the current webpage, generating a global tree based on the DOM tree, CSSOM tree and rendering tree, wherein the global tree comprises all nodes of the DOM tree, CSSOM tree and rendering tree, and the connection relation of the nodes in the DOM tree, CSSOM tree and rendering tree is reserved; encoding and generating a first node vector for all nodes in the global tree; step 105, inputting the global tree and the first node vector into a pre-rendering neural network, wherein a first hiding layer of the pre-rendering neural network updates the first node vector of a node in the global tree to generate a second node vector, a first classifier, a second classifier, a third classifier and a fourth classifier of the pre-rendering neural network input the second node vector of the node of the DOM element, the first classifier outputs two classification labels which represent the hiding and the non-hiding of the DOM element, and the second classifier outputs a classification label which represents the DOM element size scaling; the third classifier outputs a classification label representing the distance of the DOM element moving along the abscissa; the fourth classifier outputs a classification label representing the distance of the DOM element moving along the ordinate; step 106, updating a rendering tree based on the display hiding state, the size scaling, the abscissa moving distance and the ordinate moving distance of DOM elements output by the prerendering neural network, rendering the updated rendering tree to generate an image cache, and sending the image cache to the client; and 107, the cloud server analyzes the address link corresponding to the link interaction access to download the webpage content, generates a new image cache and sends the new image cache to the client.
Further, the interaction events include mouse events and keyboard events.
Further, discretizing the value range of the size scale to obtain scalar quantities of the size scale, wherein each scalar quantity corresponds to a classification label representing the size scale of the DOM element.
Further, the calculation formula of the first hidden layer is as follows:
。
wherein the method comprises the steps of，/>，/>，/>A second node vector representing the ith node, a second node vector representing the ith node>And->First node vector representing the i and j-th nodes, respectively,>weight parameter representing the first hidden layer, +.>Weight vector representing the first hidden layer, T representing the transpose,/->Representing a set of nodes directly connected to the ith node,/->Representing an activation function->Represents the LeakyReLU activation function.
Further, when generating nodes for the interaction event, the elements contained in the interaction event need to be analyzed, a corresponding interaction event node is generated for each element, wherein the interaction event node corresponding to the interaction event type is connected with the nodes in the global tree corresponding to the DOM element, and the rest of interaction event nodes are connected with the interaction event nodes corresponding to the interaction event type.
Further, the nodes of the interaction event comprise interaction event nodes corresponding to the interaction event types.
Further, step 107 includes: analyzing the interactive event of the current link interactive type, extracting the interactive events of N link interactive types occurring before the current interactive event, respectively generating local trees for the interactive time of the N+1 link interactive types, wherein the local trees are generated based on DOM trees of web pages when the interactive event occurs, comprise all nodes and connection relations of the DOM trees, and simultaneously comprise nodes for representing the interactive event; the nodes of the interaction event are connected with the nodes of the corresponding DOM element.
Generating a third node vector for all node codes in the local tree, inputting the local tree and the third node vector into a prediction neural network model, wherein the prediction neural network model comprises a third hidden layer, a fourth hidden layer and a fifth classifier, the third hidden layer updates the third node vector of the node to generate a fourth node vector, the t-th time step of the fourth hidden layer inputs the fourth node vector of the node corresponding to the interaction event type of the t-th interaction event, and the type label representing the attribute of the HTML DOM Anchor object is output in the (n+1) -th time step.
And searching the HTML DOM Anchor objects with the corresponding attributes from the DOM tree of the current webpage according to the attributes of the outputted HTML DOM Anchor objects, and accessing and downloading the addresses corresponding to the HTML DOM Anchor objects to generate webpage content, and caching the webpage content in an intermediate space.
If the HTML DOM Anchor object pointed by the interaction event of the next link interaction type belongs to the HTML DOM Anchor object obtained according to the interaction event of the last link interaction type, directly extracting a corresponding image cache from the intermediate space and sending the image cache to the client, otherwise, accessing and downloading webpage content according to an address corresponding to the HTML DOM Anchor object pointed by the interaction event of the current link interaction type to generate the image cache and sending the image cache to the client.
Further, the t-th interaction event refers to the t-th interaction event after the interaction events are ordered according to the sequence of the occurrence time.
Further, a type tag corresponds to an attribute of an HTML DOM Anchor object.
Further, the fourth hidden layer includes an RNN unit, and a t-th time step of the RNN unit inputs a fourth node vector of nodes corresponding to the interactivity event type of the t-th interactivity event.
The invention has the beneficial effects that: according to the invention, a cloud service conversion technology is adopted, a webpage accessed by a user in a domestic browser is transmitted to a cloud service for analysis and rendering by utilizing a cloud computing technology, and the processed webpage is returned to the user, so that the method has strong universality and adaptability, and the problem of incompatibility of the webpage is avoided.
In addition, the method for assisting the quick response reduces delay caused by the increase of information transmission links caused by taking cloud service as an intermediary.
Drawings
Fig. 1 is a flowchart of a home browser adaptation compatible method based on the deep learning technology.
Fig. 2 is a flow chart II of a home browser adaptation compatible method based on the deep learning technology.
Detailed Description
The subject matter described herein will now be discussed with reference to example embodiments. It is to be understood that these embodiments are merely discussed so that those skilled in the art may better understand and implement the subject matter described herein and that changes may be made in the function and arrangement of the elements discussed without departing from the scope of the disclosure herein. Various examples may omit, replace, or add various procedures or components as desired. In addition, features described with respect to some examples may be combined in other examples as well.
As shown in fig. 1 and 2, a home browser adaptation compatible method based on a deep learning technology includes the following steps:
step 101, establishing a communication link between a client and a cloud server; the client starts a domestic browser, inputs a website and sends a request to the cloud server; the cloud server receives the request and downloads the webpage content, and then analyzes the webpage content to render and generate an image cache; the communication link between the client and the cloud server is WebSocket connection.
The cloud server downloads the web page content using the HTTP protocol.
And installing and configuring a Chromium browser engine module on the cloud server, analyzing the downloaded webpage content by using the Chromium browser engine module, separating the HTML document from the CSS style sheet, and analyzing the HTML document into a DOM tree and a CSSOM tree. After generating the DOM tree and the CSSOM tree, typesetting and rendering are performed by using a chrome browser engine, and the rendered result is stored as an Image Cache (Image Cache) in a picture format.
According to different webpage contents, the cloud server can select different browser engines for analysis.
And 102, the cloud server sends the generated image cache to the client, and the client receives the image cache and then directly displays the image cache in the browser.
Step 103, identifying the interactive event type of the interactive event of the client, wherein the interactive event type of the interactive event comprises link interaction and non-link interaction, the link interaction represents that a mouse or a keyboard triggers address links such as URL links, and the other interactive events belong to the non-link interaction; if the recognition result is a link interaction, go to step 104, and if the recognition result is a non-link interaction, go to step 107.
The user can interact with the page by means of a mouse, a keyboard and the like, and the interaction events comprise mouse events and keyboard events.
For example, when a user performs a mouse operation in a browser, the Web application captures a mouse event through JavaScript code and transmits the mouse event to the cloud server using WebSocket.
104, extracting DOM tree, CSSOM tree and rendering tree of the current webpage, generating a global tree based on the DOM tree, CSSOM tree and rendering tree, wherein the global tree comprises all nodes of the DOM tree, CSSOM tree and rendering tree, and the connection relation of the nodes in the DOM tree, CSSOM tree and rendering tree is reserved; a first node vector is generated for all node encodings in the global tree.
Step 105, inputting the global tree and the first node vector into a prerendering neural network, wherein the prerendering neural network comprises a first hiding layer, a first classifier, a second classifier, a third classifier and a fourth classifier, the first hiding layer updates the first node vector of the node in the global tree to generate a second node vector, the first classifier inputs the second node vector of the node of the DOM element, outputs two classification labels representing hidden and non-hidden DOM element, the second classifier inputs the second node vector of the node of the DOM element, and outputs a classification label representing the DOM element size scaling (specifically, a value domain of the size scaling can be discretized to obtain a scalar of the size scaling, and each scalar corresponds to one classification label representing the DOM element size scaling); the third classifier inputs a second node vector of the node of the DOM element and outputs a classification label representing the distance of the DOM element moving along the abscissa; the fourth classifier inputs a second node vector of the node of the DOM element, and outputs a classification label indicating a distance that the DOM element moves along the ordinate.
Wherein, the abscissa and the ordinate are coordinate systems referred by webpage rendering generation, a classifying label of a distance moving along the abscissa corresponds to a point value of a distance, and a classifying label of a distance moving along the ordinate corresponds to a point value of a distance.
In one embodiment of the present invention, the calculation formula of the first hidden layer is as follows:。
wherein the method comprises the steps of，/>，/>，/>A second node vector representing the ith node, a second node vector representing the ith node>And->First node vector representing the i and j-th nodes, respectively,>weight parameter representing the first hidden layer, +.>Weight vector representing the first hidden layer, T representing the transpose,/->Representing a set of nodes directly connected to the ith node,/->Representing an activation function->Represents the LeakyReLU activation function.
And step 106, updating a rendering tree based on the display hiding state, the size scaling, the abscissa moving distance and the ordinate moving distance of DOM elements output by the prerendering neural network, rendering the updated rendering tree to generate an image buffer, and sending the image buffer to the client.
And 107, the cloud server analyzes the address link corresponding to the link interaction access to download the webpage content, generates a new image cache and sends the new image cache to the client.
In one embodiment of the invention, encoding nodes in the global tree may be a method employing a one-hot encoding or other natural semantic encoding process.
In one embodiment of the invention, when the interaction event is click of a mouse event, the interaction event node is connected with a node in the global tree corresponding to the DOM element clicked by the mouse.
In another embodiment, when the interactivity event is mouseout of the mouse event, five interactivity event nodes are generated, and the interactivity event nodes are respectively connected with nodes in the global tree corresponding to the DOM element of the mouse movement, wherein the five interactivity event nodes correspond to the interactivity event type, the abscissa of the starting point of the mouse movement, the ordinate of the starting point of the mouse movement, the abscissa of the ending point of the mouse movement, and the ordinate of the ending point of the mouse movement.
That is, when the node is generated by the interaction event, the elements contained in the interaction event need to be analyzed, a corresponding interaction event node is generated for each element, wherein the interaction event node corresponding to the interaction event type is connected with the nodes in the global tree corresponding to the DOM element, and the rest of the interaction event nodes are connected with the interaction event nodes corresponding to the interaction event type, and of course, only the interaction event nodes corresponding to the interaction event type can be generated.
By the method, quick response can be performed after the user interaction information is received, the original address of the webpage is not required to be accessed, and the received webpage content is re-analyzed to re-render.
In order to further increase the response speed of the browser, the following method is provided, which comprises the following steps:
analyzing the interactive event of the current link interactive type, extracting the interactive events of N link interactive types occurring before the current interactive event, respectively generating local trees for the interactive time of the N+1 link interactive types, wherein the local trees are generated based on DOM trees of web pages when the interactive event occurs, comprise all nodes and connection relations of the DOM trees, and simultaneously comprise nodes for representing the interactive event; the nodes of the interaction event are connected with the nodes of the corresponding DOM element.
Generating a third node vector for all node codes in the local tree, inputting the local tree and the third node vector into a prediction neural network model, wherein the prediction neural network model comprises a third hidden layer, a fourth hidden layer and a fifth classifier, the third hidden layer updates the third node vector of the node to generate a fourth node vector, the t-th time step of the fourth hidden layer inputs the fourth node vector of the node corresponding to the interaction event type of the t-th interaction event, and the type label representing the attribute of the HTML DOM Anchor object is output in the (n+1) -th time step; the t interaction event refers to the t interaction event after the interaction events are ordered according to the order of occurrence time.
The calculation formula of the third hidden layer is the same as that of the first hidden layer, and the fourth hidden layer comprises an RNN unit, and a fourth node vector of a node corresponding to the interaction event type of the t interaction event is input in the t time step of the RNN unit.
And searching the HTML DOM Anchor objects with the corresponding attributes from the DOM tree of the current webpage according to the attributes of the outputted HTML DOM Anchor objects, and accessing and downloading the addresses corresponding to the HTML DOM Anchor objects to generate webpage content, and caching the webpage content in an intermediate space.
If the HTML DOM Anchor object pointed by the interaction event of the next link interaction type belongs to the HTML DOM Anchor object obtained according to the interaction event of the last link interaction type, the corresponding image cache is directly extracted from the intermediate space and sent to the client.
In the embodiment described above, the document object model (Document Object Model, abbreviated as DOM) is a standard programming interface recommended by the W3C organization to process extensible markup language (HTML or XML), and the type tags of the attributes of the HTML DOM Anchor objects are organized according to the W3C standard, where one type tag corresponds to an attribute of an HTML DOM Anchor object.
The embodiment has been described above with reference to the embodiment, but the embodiment is not limited to the above-described specific implementation, which is only illustrative and not restrictive, and many forms can be made by those of ordinary skill in the art, given the benefit of this disclosure, are within the scope of this embodiment.
Claims (9)
1. A home-made browser adaptation compatible method based on a deep learning technology is characterized by comprising the following steps: step 101, a client starts a domestic browser and inputs a website, and sends a request to a cloud server; the cloud server receives the request and downloads the webpage content, and then analyzes the webpage content to render and generate an image cache; step 102, the cloud server sends the generated image cache to a client, and the client receives the image cache and then directly displays the image cache in a browser; step 103, identifying the interaction event type of the interaction event of the client, wherein the interaction event type of the interaction event comprises link interaction and non-link interaction; if the identification result is a link interaction, go to step 104, and if the identification result is a non-link interaction, go to step 107; 104, extracting DOM tree, CSSOM tree and rendering tree of the current webpage, generating a global tree based on the DOM tree, CSSOM tree and rendering tree, wherein the global tree comprises all nodes of the DOM tree, CSSOM tree and rendering tree, and the connection relation of the nodes in the DOM tree, CSSOM tree and rendering tree is reserved; encoding and generating a first node vector for all nodes in the global tree; step 105, inputting the global tree and the first node vector into a pre-rendering neural network, wherein a first hiding layer of the pre-rendering neural network updates the first node vector of a node in the global tree to generate a second node vector, a first classifier, a second classifier, a third classifier and a fourth classifier of the pre-rendering neural network input the second node vector of the node of the DOM element, the first classifier outputs two classification labels which represent the hiding and the non-hiding of the DOM element, and the second classifier outputs a classification label which represents the DOM element size scaling; the third classifier outputs a classification label representing the distance of the DOM element moving along the abscissa; the fourth classifier outputs a classification label representing the distance of the DOM element moving along the ordinate; the calculation formula of the first hidden layer is as follows:
；
wherein the method comprises the steps of，/>，/>，/>A second node vector representing the ith node, a second node vector representing the ith node>And->First node vector representing the i and j-th nodes, respectively,>weight parameter representing the first hidden layer, +.>Weight vector representing the first hidden layer, T representing the transpose,/->Representing a set of nodes directly connected to the ith node,/->Representing an activation function->Represents a LeakyReLU activation function;
step 106, updating a rendering tree based on the display hiding state, the size scaling, the abscissa moving distance and the ordinate moving distance of DOM elements output by the prerendering neural network, rendering the updated rendering tree to generate an image cache, and sending the image cache to the client; and 107, the cloud server analyzes the address link corresponding to the link interaction access to download the webpage content, generates a new image cache and sends the new image cache to the client.
2. The method for adapting and compatible home browser based on deep learning technology as claimed in claim 1, wherein the interactive event includes a mouse event and a keyboard event.
3. The method of claim 1, wherein discretizing the value range of the size scale to obtain scalar quantities of the size scale, each scalar quantity corresponding to a class label representing the size scale of the DOM element.
4. The method for adapting and compatible home-made browser based on deep learning technology according to claim 1, wherein elements included in an interaction event need to be analyzed when generating nodes for the interaction event, a corresponding interaction event node is generated for each element, wherein the interaction event node corresponding to the interaction event type is connected with nodes in a global tree corresponding to the DOM element, and the rest of the interaction event nodes are connected with the interaction event nodes corresponding to the interaction event type.
5. The method of claim 4, wherein the node of an interactivity event comprises an interactivity event node corresponding to an interactivity event type.
6. The method for home-made browser adaptation compatibility based on deep learning technology of claim 1, wherein step 107 comprises:
analyzing the interactive event of the current link interactive type, extracting the interactive events of N link interactive types occurring before the current interactive event, respectively generating local trees for the interactive time of the N+1 link interactive types, wherein the local trees are generated based on DOM trees of web pages when the interactive event occurs, comprise all nodes and connection relations of the DOM trees, and simultaneously comprise nodes for representing the interactive event; the nodes of the interaction event are connected with the nodes of the corresponding DOM elements;
generating a third node vector for all node codes in the local tree, inputting the local tree and the third node vector into a prediction neural network model, wherein the prediction neural network model comprises a third hidden layer, a fourth hidden layer and a fifth classifier, the third hidden layer updates the third node vector of the node to generate a fourth node vector, the t-th time step of the fourth hidden layer inputs the fourth node vector of the node corresponding to the interaction event type of the t-th interaction event, and the type label representing the attribute of the HTML DOM Anchor object is output in the (n+1) -th time step;
searching HTML DOM Anchor objects with corresponding attributes from a DOM tree of a current webpage according to the attributes of the outputted HTML DOM Anchor objects, accessing and downloading webpage content to generate images and caching the images in an intermediate space;
if the HTML DOM Anchor object pointed by the interaction event of the next link interaction type belongs to the HTML DOM Anchor object obtained according to the interaction event of the last link interaction type, directly extracting a corresponding image cache from the intermediate space and sending the image cache to the client, otherwise, accessing and downloading webpage content according to an address corresponding to the HTML DOM Anchor object pointed by the interaction event of the current link interaction type to generate the image cache and sending the image cache to the client.
7. The method for adapting and compatible home-made browser based on deep learning technology as claimed in claim 6, wherein the t-th interaction event is the t-th interaction event after the interaction events are ordered according to the order of occurrence time.
8. The method of claim 6, wherein a type tag corresponds to an attribute of an HTML DOM Anchor object.
9. The method of claim 6, wherein the fourth hidden layer includes an RNN unit, and the t-th time step of the RNN unit inputs a fourth node vector of nodes corresponding to the interaction event type of the t-th interaction event.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311245727.1A CN116992183B (en) | 2023-09-26 | 2023-09-26 | Domestic browser adaptation compatible method based on deep learning technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311245727.1A CN116992183B (en) | 2023-09-26 | 2023-09-26 | Domestic browser adaptation compatible method based on deep learning technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116992183A CN116992183A (en) | 2023-11-03 |
| CN116992183B true CN116992183B (en) | 2024-01-09 |
Family
ID=88523511
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311245727.1A Active CN116992183B (en) | 2023-09-26 | 2023-09-26 | Domestic browser adaptation compatible method based on deep learning technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116992183B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103077239A (en) * | 2013-01-10 | 2013-05-01 | 上海电力学院 | IFrame embedded Web3D (Web three-dimensional) system based on cloud rendering |
| CN104063407A (en) * | 2013-03-22 | 2014-09-24 | 谈天霆 | Browser architecture based on cloud computing and analysis method |
| US9262396B1 (en) * | 2010-03-26 | 2016-02-16 | Amazon Technologies, Inc. | Browser compatibility checker tool |
| CN112099844A (en) * | 2020-09-16 | 2020-12-18 | 国网甘肃省电力公司信息通信公司 | Multi-kernel compatible intelligent browsing system for state network service system |
| CN115470426A (en) * | 2022-09-02 | 2022-12-13 | 北京海泰方圆科技股份有限公司 | Browser kernel determining method and device, computer equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102346770B (en) * | 2011-09-21 | 2014-05-14 | 晨星软件研发(深圳)有限公司 | WebKit browser webpage content loading method and device |
-
2023
- 2023-09-26 CN CN202311245727.1A patent/CN116992183B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9262396B1 (en) * | 2010-03-26 | 2016-02-16 | Amazon Technologies, Inc. | Browser compatibility checker tool |
| CN103077239A (en) * | 2013-01-10 | 2013-05-01 | 上海电力学院 | IFrame embedded Web3D (Web three-dimensional) system based on cloud rendering |
| CN104063407A (en) * | 2013-03-22 | 2014-09-24 | 谈天霆 | Browser architecture based on cloud computing and analysis method |
| CN112099844A (en) * | 2020-09-16 | 2020-12-18 | 国网甘肃省电力公司信息通信公司 | Multi-kernel compatible intelligent browsing system for state network service system |
| CN115470426A (en) * | 2022-09-02 | 2022-12-13 | 北京海泰方圆科技股份有限公司 | Browser kernel determining method and device, computer equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 国产安全浏览器技术研究;张俊贤;汪丽;安晓江;;信息安全研究(01);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116992183A (en) | 2023-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2018133452A1 (en) | Webpage rendering method and related device | |
| CN109033358B (en) | Method for associating news aggregation with intelligent entity | |
| US20160283499A1 (en) | Webpage advertisement interception method, device and browser | |
| US6865593B1 (en) | Dynamic integration of web sites | |
| US8560956B2 (en) | Processing model of an application wiki | |
| US7865544B2 (en) | Method and system for providing XML-based asynchronous and interactive feeds for web applications | |
| US8977653B1 (en) | Modifying web pages to reduce retrieval latency | |
| CN108399150B (en) | Text processing method and device, computer equipment and storage medium | |
| US20070061700A1 (en) | Initial server-side content rendering for client-script web pages | |
| US20090112824A1 (en) | Method and apparatus for generating presentation configuration file of document content | |
| JP2001344169A (en) | Network system, server, web server, web page, data processing method, storage medium, and program transmitting device | |
| EP2041674A1 (en) | Method for defining a wiki page layout using a wiki page | |
| US20090006471A1 (en) | Exposing Specific Metadata in Digital Images | |
| CN107153716B (en) | Webpage content extraction method and device | |
| KR100456022B1 (en) | An XML-based method of supplying Web-pages and its system for non-PC information terminals | |
| CN103049439A (en) | Processing method for markup language documents, browser and network operating system | |
| JP4771915B2 (en) | Apparatus, method, and program for converting HTML text | |
| CN112417338B (en) | Page adaptation method, system and equipment | |
| US8140575B2 (en) | Apparatus, method, and program product for information processing | |
| US20030158894A1 (en) | Multiterminal publishing system and corresponding method for using same | |
| JP5151696B2 (en) | Program to rewrite uniform resource locator information | |
| CN101576885B (en) | Technical scheme for extracting dynamic generation web page contents | |
| CN116992183B (en) | Domestic browser adaptation compatible method based on deep learning technology | |
| EP2711838A1 (en) | Documentation parser | |
| CN113806667B (en) | Method and system for supporting webpage classification |
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 |