Webpage browsing assisting method and device, equipment and storage medium

Abstract

The invention relates to a webpage browsing assisting method and device, equipment and a storage medium, the method is suitable for a browser plug-in, and the method comprises the steps that webpage content of a webpage currently browsed by a user and intention information of the user are acquired; retrieving target content related to the intention information in the webpage content; inputting the target content into a pre-trained large language model, and generating an information extraction result corresponding to the intention information by the large language model; and displaying the information extraction result in the webpage. According to the method, the target content related to the intention information of the user is retrieved in the webpage content, and the information extraction result of the webpage is generated according to the target content and displayed, so that required information can be quickly and accurately provided for the user when the user browses the webpage, the information acquisition efficiency is improved, the webpage browsing experience of the user is improved, and the user experience is improved. Meanwhile, the user does not need to input accurate keywords, retrieval can be directly carried out through a natural language, and the requirement for professional knowledge possessed by the user is lowered.

Description

Technical Field

[0001] The present disclosure relates to the field of information generation technology, and in particular to a web browsing auxiliary method, device, equipment and storage medium. Background Art

[0002] With the rapid development of the Internet, people are increasingly relying on the Internet to obtain information. However, the lengthy and complex articles and occasional advertisements on web pages require users to spend a lot of time browsing to obtain the information they want, which reduces the efficiency of information acquisition and seriously affects the user's web browsing experience. Although users can use keyword search to quickly locate, this method is highly dependent on the accuracy of the input keywords. Once the input keywords are deviated, the content of interest may not be found. Most non-professional users find it difficult to accurately determine the keywords, and they cannot quickly obtain the required information through keyword search. Therefore, how to help users quickly obtain the required information when browsing the web is a technical problem that needs to be solved urgently. Summary of the invention

[0003] In order to solve the above technical problems, the present disclosure provides a web browsing assistance method, device, equipment and storage medium.

[0004] A first aspect of an embodiment of the present disclosure provides a webpage browsing assistance method, which is applicable to a browser plug-in, and includes:

[0005] Obtain the webpage content of the webpage currently browsed by the user and the user's intention information;

[0006] Retrieving target content related to the intention information from the webpage content;

[0007] Inputting the target content into a pre-trained large language model, and generating an information extraction result corresponding to the intent information by the large language model;

[0008] The information extraction result is displayed on the webpage.

[0009] A second aspect of an embodiment of the present disclosure provides a web browsing auxiliary device, which is applicable to a browser plug-in, and includes:

[0010] An acquisition module is used to acquire the webpage content of the webpage currently browsed by the user and the user's intention information;

[0011] A retrieval module, used to retrieve target content related to the intention information in the webpage content;

[0012] A generation module for inputting the target content into a pre-trained large language model, and generating an information extraction result corresponding to the intent information by the large language model;

[0013] A display module for displaying the information extraction result in the web page.

[0014] A third aspect of the embodiments of the present disclosure provides a computer device, including a memory, a processor, and a computer program, wherein the computer program is stored in the memory, and when the computer program is executed by the processor, the web page browsing assistance method as described in the first aspect above is implemented.

[0015] A fourth aspect of the embodiments of the present disclosure provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by the processor, the web page browsing assistance method as described in the first aspect above is implemented.

[0016] The technical solutions provided by the embodiments of the present disclosure have the following advantages compared with the prior art:

[0017] In the web page browsing assistance method, device, device and storage medium provided by the embodiments of the present disclosure, by obtaining the web page content of the web page currently browsed by the user and the intent information of the user, retrieving the target content related to the intent information in the web page content, inputting the target content into a pre-trained large language model, generating an information extraction result corresponding to the intent information by the large language model, and displaying the information extraction result in the web page, it is possible to retrieve the target content related to the intent information of the user in the web page content, generate and display the information extraction result of the web page according to the target content, so as to quickly and accurately provide the required information for the user when the user browses the web page, improve the efficiency of information acquisition, improve the user's web page browsing experience, and at the same time do not require the user to input accurate keywords, and can directly retrieve with natural language, reducing the requirements for the professional knowledge of the user. Description of the Drawings

[0018] The drawings here are incorporated into the specification and form a part of this specification, showing the embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure.

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

[0020] Figure 1 It is a flowchart of a web page browsing assistance method provided by the embodiments of the present disclosure;

[0021] Figure 2 It is a flowchart of a method for obtaining web page content provided by an embodiment of the present disclosure;

[0022] Figure 3 It is a flowchart of a method for obtaining intent information provided by an embodiment of the present disclosure;

[0023] Figure 4 It is a flowchart of a method for retrieving target content provided by an embodiment of the present disclosure;

[0024] Figure 5 It is a flowchart of a method for generating an information extraction result provided by an embodiment of the present disclosure;

[0025] Figure 6 It is a schematic structural diagram of a web page browsing assistance device provided by an embodiment of the present disclosure;

[0026] Figure 7 It is a schematic structural diagram of a computer device provided by an embodiment of the present disclosure. Detailed implementation manners

[0027] In order to be able to more clearly understand the above-mentioned objects, features and advantages of the present disclosure, the solutions of the present disclosure will be further described below. It should be noted that, without conflict, the embodiments of the present disclosure and the features in the embodiments may be combined with each other.

[0028] Many specific details are set forth in the following description in order to fully understand the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only a part of the embodiments of the present disclosure, rather than all the embodiments.

[0029] It should be understood that the various steps recorded in the method embodiments of the present disclosure may be executed in different orders and / or executed in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this regard.

[0030] Figure 1 It is a flowchart of a web page browsing assistance method provided by an embodiment of the present disclosure. This method can be executed by a web page browsing assistance device, and the web page browsing assistance device can be set in a browser plugin. As Figure 1 shown, the web page browsing assistance method provided by this embodiment includes the following steps:

[0031] S101. Obtain the web page content of the web page currently browsed by the user and the intent information of the user.

[0032] The web page content in the embodiments of the present disclosure may be all the content included in the web page, such as text, pictures, videos, etc., or may only include text content, which is not limited herein.

[0033] The intention information in the embodiments of the present disclosure can be understood as information used to describe the content that the user hopes to obtain from the web page.

[0034] In the embodiments of the present disclosure, the web page browsing assistance device can obtain the web page content of the web page currently browsed by the user and the user's intention information when the user browses the web page.

[0035] In an exemplary implementation manner of the embodiments of the present disclosure, the web page browsing assistance device can obtain the web page source code through pre-written program code, parse the Document Object Model Tree (DOM Tree) from the web page source code, and extract the web page content according to the document object model tree.

[0036] In another exemplary implementation manner of the embodiments of the present disclosure, the web page browsing assistance device can obtain the user input statement, perform intention recognition on the user input statement based on Natural Language Processing (NLP). Specifically, it can perform preprocessing such as word segmentation and stop word removal on the user input statement, then use machine learning or deep learning models to extract text features, and finally identify the user's intention information through an intention recognition algorithm.

[0037] S102. Retrieve the target content related to the intention information in the web page content.

[0038] In the embodiments of the present disclosure, after obtaining the web page content and the intention information, the web page browsing assistance device can retrieve the content related to the intention information in the web page content and determine it as the target content.

[0039] In an exemplary implementation manner of the embodiments of the present disclosure, the web page browsing assistance device can input the web page content and the intention information into a pre-trained semantic analysis model, deeply analyze the web page content through the model, combine lexical matching, syntactic analysis, and semantic reasoning to understand the context and meaning of the web page content, and then identify the target text paragraphs or sentences that match or are highly relevant to the intention information and determine them as the target content.

[0040] S103. Input the target content into a pre-trained large language model, and the large language model generates an information extraction result corresponding to the intention information.

[0041] The large language model in the embodiments of the present disclosure can be understood as an artificial intelligence model with large-scale parameters and training data that can understand and generate natural language.

[0042] In the embodiments of the present disclosure, after determining the target content related to the intent information in the web page content, the web page browsing assistance device may input the target content into a pre-trained large language model. The large language model will integrate and refine the target content, generate an information extraction result corresponding to the intent information, and return the information extraction result to the web page browsing assistance device.

[0043] S104. Display the information extraction result in the web page.

[0044] In the embodiments of the present disclosure, after obtaining the information extraction result, the web page browsing assistance device may display the information extraction result in the web page currently browsed by the user.

[0045] In an exemplary implementation manner of the embodiments of the present disclosure, the web page browsing assistance device may directly display the information extraction result in the web page currently browsed by the user, or may also display the information extraction result in the display interface of the browser plug-in, which is not limited herein.

[0046] The embodiments of the present disclosure obtain the web page content of the web page currently browsed by the user and the intent information of the user, retrieve the target content related to the intent information in the web page content, input the target content into a pre-trained large language model, generate an information extraction result corresponding to the intent information by the large language model, and display the information extraction result in the web page. It can retrieve the target content related to the intent information of the user in the web page content, generate and display the information extraction result of the web page according to the target content, so as to quickly and accurately provide the required information for the user when the user browses the web page, improve the efficiency of information acquisition, improve the user's web page browsing experience, and at the same time do not require the user to input accurate keywords, and can directly perform retrieval in natural language, reducing the requirements for the professional knowledge possessed by the user.

[0047] Figure 2 is a flowchart of a method for obtaining web page content provided by the embodiments of the present disclosure. As Figure 2 shown, on the basis of the above embodiments, the web page content can be obtained by the following method.

[0048] S201. In response to the user's trigger operation, obtain the uniform resource locator of the web page.

[0049] The trigger operation in the embodiments of the present disclosure can be understood as the user's click, long press and other trigger operations on the shortcut key or preset control.

[0050] In the embodiments of the present disclosure, the web browsing assistance device can monitor the triggering operations of the user on the shortcut keys or preset controls corresponding to the browser plug-in, and when receiving the triggering operation, call the Application Programming Interface (API) of the browser to obtain the Uniform Resource Locator (URL) of the web page.

[0051] S202. Obtain the source code of the web page based on the uniform resource locator.

[0052] In the embodiments of the present disclosure, after obtaining the uniform resource locator of the web page, the web browsing assistance device can send a network request including the uniform resource locator through a network protocol, and download the page source code of the web page based on the network request.

[0053] S203. Extract the web page content from the source code.

[0054] In the embodiments of the present disclosure, after obtaining the source code of the web page, the web browsing assistance device can remove the tags in the source code, extract the text content within the tags, and determine it as the web page content.

[0055] In the embodiments of the present disclosure, by responding to the triggering operation of the user, obtaining the uniform resource locator of the web page, obtaining the source code of the web page based on the uniform resource locator, and extracting the web page content from the source code, it is possible to obtain the web page content without the user manually inputting the web page address or performing more complex operations, with simple operations, and further improving the user's web browsing experience.

[0056] In some embodiments, the web page content may include web page text information and multimedia text information. S203 may specifically include: inputting the source code and the first guiding information into a large language model, and the large language model extracts the web page text information and multimedia resource links from the source code based on the first guiding information; obtaining the multimedia resource files based on the multimedia resource links, and performing text conversion processing on the multimedia resource files to obtain the multimedia text information corresponding to the multimedia resource files.

[0057] Among them, the first guiding information can be understood as a preset prompt for guiding the large language model to extract the web page text information and multimedia resource links from the source code. By way of example, the first guiding information can be: extract the text content within the tags and multimedia resource links in the above code. The web page text information can be understood as the text information directly included in the source code tags, and the multimedia resource links can be the links of audio and video multimedia files with formats such as mp3, mp4, wav, etc.

[0058] Specifically, after obtaining the source code of a web page, the web page browsing assistance device can splice and aggregate the source code and the first guiding information, input the aggregation result into a large language model. The large language model extracts web page text information and multimedia resource links from the input source code according to the first guiding information, accesses the link after extracting the multimedia resource link to obtain a multimedia resource file, and performs text conversion processing on the multimedia resource file to convert the audio in the multimedia file into text, obtaining the multimedia text information corresponding to the multimedia resource file. Optionally, after extracting the web page text information and the multimedia text information, the web page browsing assistance device can save the web page text information and the multimedia text information to the text field and the href field in a json file respectively, waiting for subsequent retrieval of target content related to the intent information.

[0059] Figure 3 is a flowchart of a method for obtaining intent information provided by an embodiment of the present disclosure. As Figure 3 shown, based on the above embodiment, the intent information can be obtained through the following method.

[0060] S301. Obtain a user input statement.

[0061] In the embodiment of the present disclosure, the web page browsing assistance device can obtain a user input statement after the user initiates a retrieval input.

[0062] In an exemplary implementation manner of the embodiment of the present disclosure, the web page browsing assistance device can obtain a user input statement input by the user in text form, or can obtain the audio input by the user in voice form and perform text conversion processing on the audio to obtain a user input statement.

[0063] S302. Perform anaphora resolution and semantic expansion processing on the user input statement to obtain intent information.

[0064] The anaphora resolution in the embodiment of the present disclosure can be understood as the process of identifying and processing anaphoric words (such as "he", "it", "this", etc.) in text in natural language processing. In a conversation, anaphoric words usually refer to a noun or noun phrase mentioned previously. The purpose of anaphora resolution is to determine the specific object that these anaphoric words refer to, so as to more accurately understand the meaning of a sentence or conversation. For example, in the sentence "Xiaoming bought a dog and it is very cute", "it" refers to "a dog".

[0065] The semantic expansion in the embodiments of the present disclosure can be understood as enriching the semantics of the user input statement by adding additional information or context during the process of understanding the user's intention, making it more explicit and specific. It can be achieved in various ways, such as adding synonyms, hyponyms, related concepts, etc. Semantic expansion helps to more comprehensively understand the user's intention and provide more accurate answers. For example, if the user asks "How about a certain mobile phone?", the question expansion may include "the performance of a certain mobile phone", "the user experience of a certain mobile phone", "the price of a certain mobile phone", etc., so as to more comprehensively reply to the user's input statement.

[0066] In the embodiments of the present disclosure, the web browsing assistance device can, after obtaining the user input statement, perform anaphora resolution and semantic expansion processing on the user input statement. Specifically, it can first perform anaphora resolution processing on the user input statement, and then perform semantic expansion on the processing result to obtain the user's intention information.

[0067] In an exemplary implementation manner of the embodiments of the present disclosure, the web browsing assistance device can input the user input statement and the second guidance information into a large language model, and the large language model performs anaphora resolution and semantic expansion processing on the user input statement based on the second guidance information to obtain the intention information.

[0068] Among them, the second guidance information can be understood as a pre-set prompt for guiding the large language model to perform anaphora resolution and semantic expansion processing on the user input statement. Exemplarily, the second guidance information can be: perform anaphora resolution and semantic expansion on the above content.

[0069] Specifically, the web browsing assistance device can, after obtaining the user input statement, splice and aggregate the user input statement and the second guidance information, input the aggregation result into the large language model, the large language model performs anaphora resolution and semantic expansion processing on the user input statement according to the second guidance information, and returns the processing result to the web browsing assistance device. The web browsing assistance device determines the received processing result as the user's intention information.

[0070] By obtaining the user input statement, performing anaphora resolution and semantic expansion processing on the user input statement, and obtaining the intention information, the embodiments of the present disclosure can improve the system's understanding ability when retrieving web content based on the intention information subsequently, thereby improving the accuracy of the retrieval result, providing content that better meets the user's needs, and further enhancing the user experience.

[0071] Figure 4 It is a flowchart of a method for retrieving target content provided by the embodiments of the present disclosure. As Figure 4 shown, based on the above embodiments, the target content can be retrieved through the following method.

[0072] S401. Convert the web page content into a content vector and store it in a vector library, and convert the intent information into an intent vector.

[0073] In the embodiments of the present disclosure, when the web page browsing assistance device retrieves the target content related to the intent information in the web page content, the web page content and the intent information can be respectively converted into vector forms to obtain the content vector corresponding to the web page content and the intent vector corresponding to the intent information, and the content vector is stored in the pre-established vector library.

[0074] In an exemplary implementation manner of the embodiments of the present disclosure, the web page browsing assistance device can implement the vector conversion of the web page content and the intent information through a pre-trained embedding model.

[0075] S402. Retrieve the target content vector in the vector library whose vector similarity with the intent vector is greater than or equal to a preset threshold, and return the target content corresponding to the target content vector.

[0076] The preset threshold in the embodiments of the present disclosure can be a preset threshold of vector similarity, or the vector similarity at a preset position after sorting the vector similarities of the content vector and the intent vector from high to low, which is not limited here.

[0077] In the embodiments of the present disclosure, the web page browsing assistance device can match the content vectors stored in the vector library with the intent vector respectively, determine the vector similarity between each content vector and the intent vector, and determine the content vector whose vector similarity is greater than or equal to the preset threshold as the target content vector, and determine the web page content corresponding to the target content vector as the target content to return.

[0078] By converting the web page content into a content vector and storing it in the vector library, converting the intent information into an intent vector, retrieving the target content vector in the vector library whose vector similarity with the intent vector is greater than or equal to the preset threshold, and returning the target content corresponding to the target content vector, the embodiments of the present disclosure can retrieve the target content related to the intent information through the vector similarity, ensure the matching degree of the subsequent generated information extraction result with the user's needs, and improve the user experience.

[0079] Figure 5 It is a flowchart of a method for generating an information extraction result provided by the embodiments of the present disclosure. As Figure 5 shown, on the basis of the above embodiments, the information extraction result can be generated through the following method.

[0080] S501. Reorder the target content vector to obtain a first sorting result.

[0081] In the disclosed embodiment, the web browsing assistance device can input the target content and the intent information into the Rerank model after determining the target content related to the intent information, and use the Rerank model to calculate the similarity between the target content and the intent information, determine the similarity ranking of the target content and the intent information based on the similarity calculation result, and determine the first ranking result after the target content vector is re-ranked in combination with the correspondence between the target content and the target content vector.

[0082] S502 : Based on a reciprocal sorting fusion algorithm, the vector similarity between the target content vector and the intent vector and the first sorting result are fused to obtain a second sorting result.

[0083] In an embodiment of the present disclosure, after obtaining the reordered first sorting result, the web browsing auxiliary device can use a reciprocal rank fusion (RRF) algorithm to fuse the vector similarity sorting result between the target content vector and the intent vector with the first sorting result, evaluate the comprehensive score of the two sorting results, and merge them to generate a second sorting result of the target content vector.

[0084] S503: Input the target content, the second sorting result, the intent information and the third guiding information into the large language model, and the large language model integrates and refines the target content based on the third guiding information to generate an information extraction result corresponding to the intent information.

[0085] The third guidance information in the embodiment of the present disclosure may be understood as a pre-set prompt for guiding the large language model to integrate and refine the target content. For example, the third guidance information may be: integrating and refining the above content to generate a description text.

[0086] In the disclosed embodiment, after obtaining the second sorting result of the target content vector, the web browsing auxiliary device may concatenate and aggregate the target content, the second sorting result, the intent information and the third guidance information, and input the aggregated result into the large language model. The large language model integrates and refines the target content according to the third guidance information, generates a generation result corresponding to the intent information, and returns the generation result to the web browsing auxiliary device. The web browsing auxiliary device determines the generation result as the information extraction result.

[0087] In an exemplary implementation of the disclosed embodiment, the web browsing auxiliary device can input a preset connecting word into the large language model together with the target content, the second sorting result, the intention information and the third guiding information into the large language model, and the large language model comprehensively considers the current context information to generate an information extraction result corresponding to the intention information.

[0088] In another exemplary implementation manner of the embodiments of the present disclosure, the web browsing assistance device may, after determining the information extraction result, perform post-processing on the information extraction result, including text deduplication, grammar error correction, etc., and display the information extraction result after post-processing.

[0089] In the embodiments of the present disclosure, the target content vectors are re-sorted to obtain a first sorting result. Based on the reverse sorting fusion algorithm, the vector similarity between the target content vectors and the intent vectors and the first sorting result are fused to obtain a second sorting result. The target content, the second sorting result, the intent information, and the third guiding information are input into a large language model. The large language model integrates and refines the target content based on the third guiding information to generate an information extraction result corresponding to the intent information. When generating the information extraction result, the relevance ranking between the target content and the intent information can be considered, further improving the matching degree between the content presented to the user and the user's needs, and enhancing the user experience.

[0090] Figure 6 FIG. is a schematic structural diagram of a web browsing assistance device provided by an embodiment of the present disclosure. The device is applicable to a browser plugin. As Figure 6 shown, the web browsing assistance device 600 includes: an acquisition module 610, a retrieval module 620, a generation module 630, and a display module 640. Among them, the acquisition module 610 is configured to acquire the web page content of the web page currently browsed by the user and the intent information of the user; the retrieval module 620 is configured to retrieve target content related to the intent information in the web page content; the generation module 630 is configured to input the target content into a pre-trained large language model, and the large language model generates an information extraction result corresponding to the intent information; the display module 640 is configured to display the information extraction result in the web page.

[0091] Optionally, the acquisition module 610 includes: a first acquisition unit configured to acquire the uniform resource locator of the web page in response to a trigger operation of the user; a second acquisition unit configured to acquire the source code of the web page based on the uniform resource locator; and an extraction unit configured to extract the web page content from the source code.

[0092] Optionally, the web page content includes web page text information and multimedia text information. The extraction unit includes: an extraction subunit configured to input the source code and the first guiding information into the large language model, and the large language model extracts web page text information and multimedia resource links from the source code based on the first guiding information; and a conversion subunit configured to acquire a multimedia resource file based on the multimedia resource link and perform text conversion processing on the multimedia resource file to obtain multimedia text information corresponding to the multimedia resource file.

[0093] Optionally, the obtaining module 610 includes: a third obtaining unit, configured to obtain a user input statement; and a processing unit, configured to perform anaphora resolution and semantic expansion processing on the user input statement to obtain the intention information.

[0094] Optionally, the processing unit is specifically configured to input the user input statement and the second guiding information into the large language model, and the large language model performs anaphora resolution and semantic expansion processing on the user input statement based on the second guiding information to obtain the intention information.

[0095] Optionally, the retrieval module 620 includes: a conversion unit, configured to convert the web page content into a content vector and store it in a vector library, and convert the intention information into an intention vector; and a retrieval unit, configured to retrieve, in the vector library, a target content vector whose vector similarity with the intention vector is greater than or equal to a preset threshold, and return the target content corresponding to the target content vector.

[0096] Optionally, the web page browsing assistance device 600 further includes: a sorting module, configured to re-sort the target content vectors to obtain a first sorting result; a fusion module, configured to fuse the vector similarity between the target content vector and the intention vector and the first sorting result based on a reverse sorting fusion algorithm to obtain a second sorting result; the generating module 630 is specifically configured to input the target content, the second sorting result, the intention information, and the third guiding information into the large language model, and the large language model integrates and refines the target content based on the third guiding information to generate the information extraction result corresponding to the intention information.

[0097] The web page browsing assistance device provided in this embodiment can execute the method described in any of the above embodiments, and its execution manner and beneficial effects are similar, which will not be elaborated here.

[0098] Figure 7 It is a schematic structural diagram of a computer device provided by an embodiment of the present disclosure.

[0099] As Figure 7 shown, the computer device may include a processor 710 and a memory 720 storing computer program instructions.

[0100] Specifically, the above-mentioned processor 710 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0101] Memory 720 may include a mass storage for information or instructions. By way of example and not limitation, memory 720 may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disc, a magneto-optical disc, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 720 may include removable or non-removable (or fixed) media. Where appropriate, memory 720 may be internal or external to the integrated gateway device. In a particular embodiment, memory 720 is a non-volatile solid-state memory. In a particular embodiment, memory 720 includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0102] Processor 710 reads and executes computer program instructions stored in memory 720 to perform the steps of the web browsing assistance method provided by the embodiments of the present disclosure.

[0103] In one example, the computer device may further include a transceiver 730 and a bus 740. Among them, as Figure 7 shown, the processor 710, the memory 720, and the transceiver 730 are connected through the bus 740 and complete communication with each other.

[0104] Bus 740 includes hardware, software, or both. By way of example and not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side BUS (FSB), a Hyper Transport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable bus or a combination of two or more of these. Where appropriate, bus 740 may include one or more buses. Although embodiments of the present application describe and illustrate specific buses, the present application contemplates any suitable bus or interconnect.

[0105] Embodiments of the present disclosure also provide a computer-readable storage medium that may store a computer program, which when executed by a processor causes the processor to implement the web browsing assistance method provided by the embodiments of the present disclosure.

[0106] The above storage medium may include, for example, a memory 720 storing computer program instructions executable by a processor 710 of the web browsing assistance device to implement the web browsing assistance method provided by the embodiments of the present disclosure. Optionally, the storage medium may be a non-transitory computer-readable storage medium. For example, the non-transitory computer-readable storage medium may be a ROM, a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc. The above computer program may be written in any combination of one or more programming languages to write program code for performing the operations of the embodiments of the present disclosure. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, executed as a stand-alone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0107] It should be noted that in this document, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0108] The above are only specific embodiments of the present disclosure, enabling those skilled in the art to understand or implement the present disclosure. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A web browsing assistance method, characterized in that: The method is applicable to a browser plug-in, and the method comprises: Obtain the webpage content of the webpage currently browsed by the user and the user's intention information; Retrieving target content related to the intention information from the webpage content; Inputting the target content into a pre-trained large language model, and generating an information extraction result corresponding to the intent information by the large language model; The information extraction result is displayed on the webpage.

2. The method according to claim 1, characterized in that The step of obtaining the webpage content of the webpage currently browsed by the user includes: In response to a trigger operation by a user, obtaining a uniform resource locator of the webpage; Obtaining source code of the webpage based on the uniform resource locator; The webpage content is extracted from the source code.

3. The method according to claim 2, characterized in that The webpage content includes webpage text information and multimedia text information, and the step of extracting the webpage content from the source code includes: Inputting the source code and the first guide information into the large language model, and extracting web page text information and multimedia resource links from the source code by the large language model based on the first guide information; A multimedia resource file is acquired based on the multimedia resource link, and text conversion processing is performed on the multimedia resource file to obtain multimedia text information corresponding to the multimedia resource file.

4. The method according to claim 1, characterized in that: Get user intent information, including: Get user input sentence; The user input sentence is subjected to reference elimination and semantic expansion processing to obtain the intention information.

5. The method according to claim 4, characterized in that The performing reference elimination and semantic expansion processing on the user input sentence to obtain the intention information includes: The user input sentence and the second guide information are input into the large language model, and the large language model performs reference elimination and semantic expansion processing on the user input sentence based on the second guide information to obtain the intention information.

6. The method according to claim 1, characterized in that The step of retrieving target content related to the intention information from the webpage content includes: Convert the webpage content into a content vector and store it in a vector library, and convert the intent information into an intent vector; A target content vector whose vector similarity with the intention vector is greater than or equal to a preset threshold is retrieved from the vector library, and a target content corresponding to the target content vector is returned.

7. The method according to claim 6, characterized in that Before inputting the target content into a pre-trained large language model and generating an information extraction result corresponding to the intent information by the large language model, the method further includes: Reordering the target content vectors to obtain a first ordering result; Based on a reciprocal sorting fusion algorithm, the vector similarity between the target content vector and the intent vector and the first sorting result are fused to obtain a second sorting result; The step of inputting the target content into a pre-trained large language model, and generating an information extraction result corresponding to the intent information by the large language model, includes: The target content, the second ranking result, the intention information and the third guidance information are input into the large language model, and the large language model integrates and refines the target content based on the third guidance information to generate the information extraction result corresponding to the intention information.

8. A web browsing auxiliary device, characterized in that: The device is applicable to a browser plug-in, and the device comprises: An acquisition module is used to acquire the webpage content of the webpage currently browsed by the user and the user's intention information; A retrieval module, used to retrieve target content related to the intention information in the webpage content; A generation module, used for inputting the target content into a pre-trained large language model, and generating an information extraction result corresponding to the intent information by the large language model; A display module is used to display the information extraction result in the web page.

9. A computer device, characterized in that: include: Memory; processor; and a computer program; wherein the computer program is stored in the memory and is configured to be executed by the processor to implement the web browsing assistance method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the web browsing assistance method according to any one of claims 1 to 7 is implemented.

Images