Predictive caching and highlighting of web pages

Predictive caching and highlighting of web pages

Technical field

The present invention is related to web browser and server technology, and specifically to web browsing technology that provides browsing capabilities that prioritize users' personal interests.

Background technique

The Internet and the World Wide Web have become an indispensable part of business operations, personal life, and educational processes.

The center of Internet technology is web browser technology and Internet server technology.

Internet servers store "content" such as documents, image or graphic files, tables, audio clips, etc., all of which are available to systems and browsers with Internet connections.

A web browser or "client" computer can request a document from a web address, and the appropriate web server responds by sending one or more web documents, image or graphic files, forms, audio clips, etc. The most common protocol for sending web documents and content from a server to a browser is the Hypertext Transfer Protocol ("HTTP").

Figure 1 shows the basic client-server configuration for Internet and intranet communications. The client browser computer (1) is equipped with Internet access (2) to the World Wide Web (3) via common devices such as dial-up telephone lines and modems, cable modems or local area networks ("LAN"). The web browser computer (1) is also equipped with appropriate web browsing software such as Netscape's Navigator or Microsoft's Explorer. The web server computer (5) is also equipped with Internet access (4) that uses similar devices or high-bandwidth devices such as T1 and T3 data lines and a web server software suite to reach the World Wide Web (3). Alternatively, the client and server may be interconnected via an intranet (6) such as a corporate LAN. These configurations are well known in this technical field.

The most common type of Internet content or document is the Hypertext Markup Language ("HTML") document, but other formats are also well known in this technical field, such as the Adobe Portable Document Format ("PDF"). HTML, PDF, and other web documents provide "hyperlinks" within the documents that allow users to select another document or website for viewing. Hyperlinks are specially marked words or areas in a document, and when selected by the user, they instruct the browser software to retrieve or obtain the pointed document.

Generally, when the user selects a common hyperlink, the current page displayed in the graphical user interface ("GUI") window of the web browser disappears, and the newly received page is displayed. If the master page is an index, such as the IBM website www.patents.ibm.com, and the user wants to access each subsequent link (for example, read a document with tips on how to use the site), then the master page or index page Disappear, and this new page (such as a help page) is displayed.

As the computing power of web browser computers increases and the communication bandwidth of web browser computers increases significantly, a problem for organizations that provide Internet websites and content is to consider these large processing and throughput speeds to transmit and filter this Kind of content.

This is especially true in the field of web-based applications and in developing better and more effective ways to deliver user-appropriate information to the desktop or client.

However, some current web browsers are usually non-smart software packages. As currently existing web browsers, they generally require users to manually search for any articles or documents they are interested in. This is usually very cumbersome, because they often need to download many documents to find a closely related document.

Search engines provide a certain degree of "intelligence" for browsing. The user can point his non-smart browser to a search engine address, enter some search keywords, and select hyperlinks in the search results or manually The web browser points to the provided web address, reviewing each returned document one at a time. However, search engines do not actually search the entire Internet, they only search their own Internet content indexes that are usually established by search engine operators through a process of reviewing manual submissions by other website operators. Therefore, users often need to use several search engines to find information on specific topics, because each search engine will return different results according to their own index content.

In order to partially solve this problem, two other technologies that are well known in this technical field have been developed. The first technology is called the "metasearch engine", which is one of multiple search engines. The ultimate search engine does not hold its own index, but submits queries to multiple search engines at the same time, and returns the highest returned by each of these search engines to the user. Although this is more useful than manually visiting each search engine queried one by one, the results are usually not as satisfactory as expected. Generally, the few returned items at the top of the list that match the search keywords are not the most interesting, so users often visit the sites listed in the middle or the end of the returned list. Although the ultimate search engine can return the top 5 items from the 4 search engines, it may filter out information that is likely to be of interest.

The second way to solve this problem is called the web "crawler" (crawler) engine. These servers periodically contact other servers to "re-index" previously indexed website content, which keeps them up-to-date, and incorporates any information recently available from a website into their index. However, since thousands of new websites are posted online every day, it is practically impossible for a crawler browser to access these new sites. Therefore, even web crawler browsers may not provide complete coverage of Internet content.

Several other attempts have been proposed in various US patents, including the creation of a "community of some intelligent agents", which uses server-based interactive classification and filtering to trigger a client-side "intelligent assistant" when a special logo is encountered in a web document. ", and the automatic "bookmark" function. In a nutshell, these proposed technologies and methods all require a certain amount of cooperation between the server side and the client side, which makes it difficult to promote and apply these technologies on a large scale.

A few years ago, a client-side technology was introduced to download all web pages within a hyperlink of the web page currently loaded by the browser. Since all documents directly linked from the currently visited webpage are collected, no matter which document the user selects, it is immediately available from the cache in the local memory, so there is no need to wait for the server to send the newly selected webpage to the browser. . When the user finishes reading the next page (now it is the current page) and selects a follow-up document, the follow-up document has been cached, so it can also be displayed without transmission delay. However, this approach has some disadvantages when visiting a "link-rich" web page. For example, a webpage of a popular news site may have more than 60 documents linked directly from the homepage of the new business. Therefore, the communication network serving the web browser computer may present a bottleneck or time-limiting factor for loading all 60 directly linked documents when the user reads the homepage and before the user selects a hyperlink on the homepage. Therefore, only a few of these directly linked web pages can be successfully downloaded within the time it takes for the reader to find the home page and decide to view the next document. Unfortunately, these webpages that were successfully downloaded during the review of the homepage may not be of interest to the user, because the download function does not classify the webpages or determine which webpage may or may not be of interest.

Summary of the invention

Therefore, in the first aspect, the present invention provides a method for predictively browsing web documents that may be of interest to users of a web browser system having a user display, a user input device, and a persistent A storage device, where the web document contains some words and can be accessed from the browser system through a link address, and the method includes the following steps: receiving a part of a web document from a link address; determining the part of a web document Whether it contains one or more predetermined words that are of interest to the user; and in response to determining that the one or more words of interest exist in the document part, the entire web document is received and stored.

The method of the first aspect of the present invention preferably further includes repeating the receiving of a part of a web document for a plurality of web documents accessible within a predetermined number of link addresses from a first web document, and determining whether the part contains The words of interest and the overall steps of receiving and storing a web document.

The method of the first aspect of the invention preferably further includes the step of providing a dedicated user highlighting of a link to a web document on the web browser display.

The method of the first aspect of the invention preferably further includes creating a quick link to the stored web document on the browser display.

In a second aspect, the present invention provides a computer program, which includes program code that, when loaded into a computer system for execution, causes the computer system to perform all steps of the method of the first aspect of the present invention.

In the third aspect, the present invention provides an enhanced web browser system that can predictably browse web documents that may be of interest to a user of a web browser system. The web document contains some words and can be accessed through a link address. Accessed from the browser system, the system includes: a processor that executes program code; a user display that displays information for the user; a user input device that receives user input; a persistent storage device that stores data and information, It includes a list of items of interest of the user stored therein, the list of items of interest containing some words of interest to the user; and a predictive interest-based browser program that can be executed by the processor, and the browser program uses To receive a part of a web document from a link address, determine whether the part of a web document contains one or more item words of interest, and make a decision on whether one or more item words of interest are found in the document part Responsive to receiving and storing the entire web document.

Preferably, the predictive interest-based browser program of the system of the third aspect of the present invention further includes a standard web browser program with a browser plug-in for receiving a web from a link address A part of a document, determining whether the part of a web document contains one or more item words of interest, and responding to determining that one or more item words of interest are found in the document part, receiving and storing the web document Overall.

Preferably, the web document includes an HTML document.

Preferably, the system of the third aspect of the present invention further includes a link highlighting display for highlighting a link to a stored web document on the web browser display.

Preferably, the system of the third aspect of the present invention further includes a quick link creator, and the quick link points to the stored web document on the web browser display.

Therefore, the present invention suitably and desirably enables the web browser to predictably and automatically search for web documents containing items of interest to the web browser user that are linked to the currently displayed web page. The linked document containing the item of interest is appropriately automatically retrieved and stored while the user is viewing the current document, so that if the user selects a link to the stored document, the document can be displayed without waiting for it download. In order to further help users find documents that contain items of interest to users, the links to the documents of interest in the current page can be highlighted, and special quick links to these web pages can be created and displayed for better reminders and user convenience use.

Therefore, the preferred embodiments of the present invention beneficially provide a web browsing method and system, which can predictively retrieve information from computer network servers such as the World Wide Web and distributed databases based on user-interested items or keyword lists. In addition, it is beneficial that this new system and method are compatible with widely used browser technologies such as personal computers, web-enabled telephones, Internet devices, personal digital assistants, and pocket PCs, with very little or no Support or cooperation technology on the server side is required. In addition, it is beneficial that this new system and method highlight predictive cached information or links to such information on the user's display, so that the user can conveniently and quickly view the predictive cached information.

Some preferred embodiments may also preferably provide a system and method for configuring a browser system to include a list of items of interest to the user. This method provides a list of keywords that users search most frequently. This list can be used for other software programs on the same client's web browser computer.

Description of the drawings

A preferred embodiment of the present invention will be exemplified below with reference to the drawings. In these drawings: Figure 1 shows a well-known configuration between an Internet client or web browser system, a web server system and a communication network; 2 illustrates the well-known architecture of a web browser and a web server system; FIG. 3 shows a typical tree structure of some hyperlink documents on a website; and FIG. 4 discloses the configuration of a preferred embodiment of the present invention.

detailed description

For the purposes of this description, it is assumed that all tasks associated with searching and loading web pages are performed by a web browser application such as Netscape's Navigator or Microsoft's Explorer. In fact, the embodiments of the present invention described here can be implemented by software associated with a web browser, which may or may not be a part of the browser itself, such as a collaborative independent software application or a browser plug-in module such as. Therefore, those familiar with the technical field can realize that the compilation of the list of items of interest can be implemented by any software as described herein, and the results can be used for other browser-related functions and software.

Figure 2 shows the general hardware and software architecture of a typical web server and web browser computer system. The web browser computer (20) communicates with the web server computer (22) via the Internet or an intranet (21). The web browser system includes standard user interface devices (23) such as a computer display or monitor, keyboard and mouse. The hardware platform of the web browser computer (20) includes a central processing unit ("CPU") (24), a disk drive (25), a user interface device I/O (26), and a network interface card ("NIC") (27) . The NIC can be one of several well-known varieties in the technical field, including a dial-up Internet modem, a local area network ("LAN") card, or a cable modem interface. The software executed by the web browser computer (20) may include some device drivers and a basic input/output system ("BIOS") (28), as well as the operating system (203), application programs (202) and applet interpreters ( 29) and small applications (201). A web browser program, such as Netscape's Navigator, is an application program that can be executed by the CPU (24). This architecture and configuration of a computer with a web server is well known in the technical field.

In this preferred embodiment, the standard web browser application software program is modified to include some logic and functional enhancements. These functional enhancements take advantage of some existing capabilities of existing web browsers, such as: (1) interpreting received web documents; (2) enabling all or part of a web document to be displayed in the current web browsers display window ; (3) Display user option icons, drop-down lists or other mode control indicators in the web browser display window; (4) Receive user feedback on the user option icons, drop-down lists and other modes displayed in the web browser receiving window The selection of the control indicator; and (5) the establishment, storage and access of data items such as documents, records and cookies in the system memory, especially in the persistent memory such as hard disk drives and non-volatile RAM or ROM .

Since the general configuration and architecture of the above-mentioned web browser system are well known in this technical field, the rest of the description of the preferred embodiment of the invention will desirably be used as a system under Microsofts Windows[TM] operating system on an IBM compatible computer. The steps and functions of the browser plug-in of the running Netscape Navigator are given. However, those familiar with the relevant technical fields can recognize that, without departing from the scope of the present invention, other operating systems such as UNIX, Linux, and Sun Microsystems Solaris, such as IBMs RS6000, Apples iMac(TM) Other computer hardware such as personal digital assistants and web-enabled telephones, and other software implementations such as Java scripts or compilers can also be used. In some other embodiments, the servlet or program of the web server can maintain a list of items of interest, so that this list can be used by client-side programs and plug-ins upon request.

In summary, the preferred embodiment of the present invention improves the original concepts and functions of the web browser. It is desirable for such a web browser to determine which keywords may be of interest to the web browser user. These items of interest are desirably stored in the persistent storage of the system and can be accessed by the present invention as a flat text document. Other embodiments of the list of items of interest may also be used, such as records in a database, all of which are easily accessible by other programs including the browser plug-in of the preferred embodiment of the present invention.

Other methods or systems for establishing a list of items of interest can be used in conjunction with the preferred embodiment of the present invention, but the systems and methods described above provide some useful methods for generating a list of items of interest.

Table 1 shows an example after the embodiment of the list of items of interest is generated. The list of user-interested items in this example is given in a comma-separated variable ("CSV") format, where a colon ":" is specified to indicate a general category with some subcategories. If there is no colon after a category or item, it is assumed that all subcategories and items available under this category are of interest. Table 1: Examples of document list of items of interest to users. Politics <CR> Sports: Baseball, Professional Basketball, Motorcycle Sports <CR><EOF>

The users list of items of interest is preferably directly editable by the user. Therefore, if a user wishes to delete an item of interest that may have been added before, he can do so easily with a common text document editor or database program. Similarly, if a user wants to add an item of interest later, he can recall the menu or directly edit a document.

The preferred embodiment of the present invention provides two user-selectable procedures for predictively retrieving and caching information from the web server based on the user's list of items of interest. In the first process, only the dedicated hyperlink information of the "item of interest" will be cached first, thereby improving the well-known process of caching all "1 hop" web pages by web browsers. The second process disclosed here highlights any hyperlinks leading to information containing items of interest to the user, such as by highlighting text or images on the web browser display, in a separate web browser window, or in the original A glance within a dedicated frame of the web browser window makes the user pay attention to these links.

In order to be clearer and more specific in the following detailed description, the following terms are used: "interest term" is a self-evident word or phrase that the end user is interested in; "N hop scan" Indicates the link space in which the web browser will try to predictively load and check web pages and associated text; "interest links" are those within "N jump scans" containing items of interest An accessible hyperlink; a "fast link" is a highly visible link extracted from the cluttered background of an ordinary web page that contains a direct link to a web page found containing an item of interest; a "deep link" (deep-linking) is a commonly accepted term that refers to pulling web content from the depths of an organizations website or retrieving data through a series of URLs without loading or accessing the intervening web pages; "contemplation time (contemplation time) ) is defined as the time the user spends on a given webpage, which is the time available for the web browser system to determine and highlight any interesting link branches of the currently loaded webpage; and "TB is the time for browsing The length of the text downloaded when the scanner scans a web page to search for items of interest, for example, in bytes.

N jump scans, as discussed above, are predictive scans or retrievals of documents accessible within "N" hyperlinks from the starting point. Figure 3 shows a typical tree structure or representation of the content of a website. Each webpage has some hyperlinked webpages from it, and these hyperlinks are shown as arrow lines from one webpage to another webpage or some other webpages. The variable "N" represents the depth or space of the search information relative to the starting point.

For example, a one-time jump scan (for example, N=1) (51) retrieves all hyperlink documents accessible by a single "click" or hyperlink from the current web page (50), in this example page 2 , 3 and 4, scan the web content of these documents to see if there are any items of interest to users.

Similarly, 2 jump scans (N=2) (52) retrieve all documents accessible from the hyperlink of the current webpage through the second "click", for example, in this example, add pages to all pages scanned by 1 jump 2a, 3a, 3b, 4a and 4b.

It can be seen from the tree expansion of this diagram that the amount of data to be considered may increase exponentially with respect to the value of N. Higher-order scanning will be more practical, but it requires larger computer network communication bandwidth and higher processor speed of the web browser computer. .

Look at Figure 4, which shows the implementation structure of the preferred embodiment. The predictive scanner plug-in for items of interest (43) runs in the web browser program (40) environment on the web browser computer (20), and the user I/O (23) of the web browser computer is used to display the highlighted Links, quick links, and the resulting display frame are as described in the following description. A simple text document or a list of items of interest to the user (42) in some database records is accessed from its storage medium (41), such as stored on a hard drive or in the persistent storage of the web browser system (20) Inside. Alternatively, the list of items of interest to the user (42) can also be accessed from a web or web server accessible by the web browser system (20).

The predictive scanner plug-in for items of interest (43) also utilizes the communication capabilities of the web browser computer (20) (such as its network interface card and communication protocol (TCP/IP)) and the web browser program (40) Communication and display capabilities (such as HTTP) selectively retrieve parts of web documents from the Internet (3) or other computer networks.

The preferred embodiment of the present invention operates during the gaze time of a current webpage, and predictively retrieves the hyperlinked documents in the N jump scan space according to the items of interest of the user. It is assumed that the keywords of interest can be stored in the web browser system and/or associated software. Next, the "read-ahead" predictive download utilizes knowledge of such items of interest.

Once a web browser loads any web page after a user's selection of the web page or other actions (such as selecting a bookmark, navigation buttons, etc.) to load the web page, the predictive caching process begins immediately. The currently loaded and viewed web page is set as the starting point of N jump scans, or "current page".

The preferred embodiment of the present invention then analyzes the source of the current page, such as the HTML of the current page, and starts to download all web pages that are directly linked to the current page called the 1-jump page. The download of each web page is terminated after successfully receiving a predetermined amount of data, such as the number of bytes or kilobytes given by TB.

Next, scan the download portion of each web page to determine whether they contain any items of interest to the user. If after downloading a predetermined number of bytes, no user-interested items are found in the plaintext or meta-words of this webpage, stop the download. Since the download of the entire webpage is stopped, the browser saves network bandwidth and time, and the saved resources can then be used to scan the next possible webpage of interest. If an item of interest is found, the download is resumed and completed, and the entire linked web page is stored in the cache.

When the user continues to look at the currently loaded webpage, the browser checks the next one-jump webpage, and then the next one, until all the one-jump webpages are scanned and cached as needed.

If all 1 jump web pages are scanned before the user finishes reviewing the current page, increase the hop level, by downloading each 2 jumps and then 3 jumps to scan the successive depths of the web pages. Layer site content, search for keywords, and cache the entire web page if an item of interest is found, as explained above.

This predictive scanning process can be described by the pseudo code in Table 2.

Table 2: Pseudo-code of the predictive scanning process <pre listing-type="program-listing">UNTIL (user selects the link in current_page): FOR hop=1 to N: Scan_page=current_page catalog all referenced_links from current_page randomly order from first to last all referenced_links FROM first TO last referenced_link: download document portion at referenced_link scan portion far occurrences of interest terms IF occurrences found, THEN: complete download of document store document in cache highlight referenced_link create "fast link" to cached document (optional) ELSE discard portion of document NEXT referenced link/*Search for the next part of the document linked in this jump*/ NEXT hop/* Search for the next set of documents that jump again from the current page*/</pre>

When it is found that some one-time jump web pages contain items of interest to the user, any one of several methods is used to make the user pay attention to these web pages. First, the hyperlinks or links leading to web pages containing items of interest can be highlighted in the display of the current page, such as by changing the color, font, or size of these hyperlinks. In an enhanced embodiment of the present invention, a "quick link" can be established in an independent frame along the side, top, or bottom of the current page in the current window or in an independent web browser window.

This provides users with an improved web browser display, which can highlight links to documents that are likely to be of interest to users according to the list of items of interest to users, so that users can browse the current website more efficiently.

It should be pointed out that if quick link display is used, the link of interest that jumps multiple times can be one jump, that is, the next step in the path leading to the link of interest is shown. The next step in this path, etc., or the link of interest that jumps multiple times can also be "deep-linked." In the latter case, the first link shown in the frame, window, etc. is a deep link to the item of interest, even if it is only accessible through multiple jumps, and the display of the first link at the top level can be highlight. In a further improved embodiment of the present invention, a highlighting method can be used to highlight the one-time jump link leading to the document of interest, and another highlighting method can be used to highlight the document of interest. The multiple jump links. For example, a link of interest for one jump can be set to flashing red text, and multiple jump links to a document of interest can be shown or highlighted in stable red text. The HTML codes for setting the color, font, and blinking attributes are well known, so the browser plug-in of this preferred embodiment only needs to change these attributes displayed by the web browser of this part of the current web page.

It should also be noted that the preferred embodiment of the present invention performs "breadth first search" instead of deep drilling through "N jumps" from a given starting point. Alternatively, a "depth-first search" can also be performed, although this search is not very practical and efficient from the view of the inventor of the present invention, because some links that are not included in the originally downloaded document part may be missed or skipped . Any search technique can be used, and the principles disclosed here are universally applicable.

It should also be understood that, if necessary, a common list of "quick links" to web pages of interest can be maintained in a separate window or frame, even if the user goes down a specific path. For example, consider the situation where a user on the page "A" lists a list of links "B" and "C" of interest. It can be easily seen that the user will go to the link "B" of interest, while still maintaining a quick link to the web page "C" in a separate frame or window. After reading "B", if some interesting links arrive while reading, keep "C" in the quick link window so that the user can immediately jump back to this other path that was not taken before.

Incorporating the preferred embodiment of the present invention into a web browsing system or product, a more intelligent device can be obtained for browsing the World Wide Web and a large amount of content of its website according to the user's interest.

Although some specific examples and descriptions related to a preferred embodiment are given above, those skilled in the art can recognize that various substitutions and engineering choices can be made without departing from the scope of the present invention, including but Not limited to implementing this method as an application, portable language script, server-side program or script, or browser enhancement, use a different web browser such as a web-enabled phone, Internet device, or personal digital assistant Computer, and use other operating systems such as Windows[TM]CE.