A content delivery method, apparatus, device, and storage medium

Abstract

This disclosure provides a content delivery method, apparatus, device, and storage medium, relating to the field of computer technology, particularly to search engines, search and information client technologies, and large-scale models, and can be used in application scenarios such as generative search, intelligent document editing, intelligent assistants, virtual assistants, and intelligent e-commerce. The method includes: crawling multiple content data from multiple data sources; clustering the multiple content data to obtain multiple events and a content list for each event; the content list of each event contains multiple content data for that event; extracting the multiple events with the highest popularity values ​​from the multiple events, and summarizing the content data corresponding to each extracted event to obtain a summary content for each event. This disclosure can meet users' needs for quickly consuming trending events.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to the fields of search engines, search and information client technology, and large models. Background Technology

[0002] When using search and news clients, users have a need to grasp and understand the day's major trending events during their fragmented time. How to meet this need and provide users with timely event information is a technical problem that needs to be solved. Summary of the Invention

[0003] This disclosure provides a content delivery method, apparatus, device, and storage medium.

[0004] According to one aspect of this disclosure, a content delivery method is provided, comprising:

[0005] Fetch multiple content data from multiple data sources;

[0006] Multiple data items are clustered to obtain multiple events and a list of content for each event; the content list of an event contains multiple data items for the corresponding event.

[0007] Extract the events with the highest popularity from a pool of events, and summarize the content data corresponding to each extracted event to obtain the summary content for each event.

[0008] According to another aspect of this disclosure, a content providing apparatus is provided, comprising:

[0009] The crawling module is used to crawl multiple content data from multiple data sources;

[0010] The summary module is used to cluster multiple content data to obtain multiple events and a content list for each event; the content list of an event contains multiple content data for the corresponding event; the module extracts the multiple events with the highest popularity values ​​from the multiple events, and summarizes the content data corresponding to each extracted event to obtain the summary content for each event.

[0011] According to another aspect of this disclosure, an electronic device is provided, comprising:

[0012] At least one processor; and

[0013] The memory is communicatively connected to the at least one processor; wherein,

[0014] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform any of the methods described in the present disclosure.

[0015] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform any of the methods according to embodiments of this disclosure.

[0016] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the methods according to embodiments of this disclosure.

[0017] The content provision method disclosed herein can obtain multiple events and summary content for each event by crawling content data from multiple data sources and clustering and summarizing the crawled data content, thereby meeting users' needs for quickly consuming trending events.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0019] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0020] Figure 1 This is a schematic flowchart of a content provision method according to an embodiment of the present disclosure;

[0021] Figure 2 This is a diagram illustrating the process of providing content by triggering a gesture;

[0022] Figure 3 This is a schematic diagram illustrating the usage flow of a content provision method according to an embodiment of this disclosure;

[0023] Figure 4A This is a display effect diagram of the relevant APP during the startup phase in a content provision method according to an embodiment of this disclosure;

[0024] Figure 4B This is a content provisioning method according to an embodiment of the present disclosure, showing the display effect of the relevant APP during the content provisioning implementation stage. Figure 1 ;

[0025] Figure 4C This is a content provisioning method according to an embodiment of the present disclosure, showing the display effect of the relevant APP during the content provisioning implementation stage. Figure 2 ;

[0026] Figure 4D This is an embodiment of the content provisioning method disclosed herein, showing the display effect of the relevant APP during the content display phase. Figure 1 ;

[0027] Figure 4E This is an embodiment of the content provisioning method disclosed herein, showing the display effect of the relevant APP during the content display phase. Figure 2 ;

[0028] Figure 5 This is a schematic diagram illustrating the implementation of hot topic mining according to one embodiment of the present disclosure;

[0029] Figure 6 This is a schematic diagram of the structure of a content providing device 600 according to an embodiment of the present disclosure;

[0030] Figure 7 This is a schematic diagram of the structure of a content providing apparatus 700 according to an embodiment of the present disclosure;

[0031] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. Detailed Implementation

[0032] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0033] The term "and / or" in this disclosure indicates that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. The term "at least one" in this document means any combination of at least two of a plurality of options, such as including at least one of A, B, and C, which can mean including any one or more elements selected from the set of A, B, and C. The terms "first" and "second" in this document refer to and distinguish multiple similar technical terms, and do not imply a specific order or a limitation to only two. For example, "first feature" and "second feature" refer to two types / two features; the first feature can be one or more, and the second feature can also be one or more.

[0034] Users have a need to grasp and understand the day's major trending events when using search and news clients, such as search and news applications (APPs). Currently, this need can be met through the trending topics channels of search and news apps. However, this method has at least the following problems:

[0035] First, the access point is unstable; the trending topics channel is configured individually, and some users do not see the trending topics channel on their home screen. Second, the content of the trending topics channel is mainly selected and operated manually, which has a certain lag and subjectivity, and may not be what users are interested in or concerned about. Third, when users want to learn more about the content related to trending keywords, they need to click on the trending keywords to view the details, which results in low consumption efficiency.

[0036] Figure 1 This is a flowchart illustrating a content provision method according to an embodiment of the present disclosure. The method can be applied to a data processing apparatus, which can be deployed in an electronic device. The electronic device may be a single-machine or multi-machine terminal, server, or other processing device. The terminal may be a mobile device, a personal digital assistant (PDA), a handheld device, a computing device, an in-vehicle device, a wearable device, or other user equipment (UE). In some possible implementations, the electronic device may be applied to a data center; for example, the electronic device may be a computing engine device in a data center. In some possible implementations, the method may also be implemented by a processor calling computer-readable instructions stored in memory. Figure 1 As shown, the method may include:

[0037] S110. Retrieve multiple content data from multiple data sources;

[0038] S120. Cluster the multiple content data to obtain multiple events and a content list for each event; the content list of an event contains multiple content data corresponding to that event;

[0039] S130. Extract the events with the highest popularity values ​​from the multiple events, and summarize the content data corresponding to each extracted event to obtain the summary content corresponding to each event.

[0040] In some examples, multiple data sources include information from various internal channels, push notifications sent by the app, overall app rankings, overall app push notifications, and information manually submitted by product lines. Overall app rankings mainly include comprehensive rankings, vertical rankings (such as finance rankings, entertainment rankings, technology rankings, etc.), and regional rankings.

[0041] After obtaining the summary content for each event, the summary content for each event can be displayed on the terminal device for users to view.

[0042] In some examples, a pre-defined trigger gesture can be used to initiate the process of providing the above content. Figure 2 This is a diagram illustrating the process of providing content by triggering a gesture, such as... Figure 2As shown, in Figure 1 Before the process shown, the following also includes:

[0043] S210, Recognize touch events;

[0044] S220: If the touch event satisfies the preset trigger gesture, extract the page display data;

[0045] S230. Based on the data displayed on this page, determine the corresponding content delivery plan.

[0046] In some examples, the trigger gesture includes a pinch gesture. A pinch gesture can be a multi-finger pinch-swipe gesture initiated from the edge of the display screen, where "multi-finger" refers to two or more fingers, and "pinch" means that two or more fingers converge in the swiping direction of the touchscreen, that is, the swiping positions of the two or more fingers move closer together as the swiping action occurs. A certain area on the screen can be set to belong to the screen edge.

[0047] In some examples, the page display data may include the location of the page content within the current app. For instance, if the page content is located on the app's homepage, step S230 can determine, based on the page display data, that the corresponding content provision method is the content provision method proposed in this embodiment, and then the content data crawling and clustering process continues. If the page content is not located on the app's homepage, other corresponding methods can be executed; for example, if a pinch gesture is received on the message reading page, a summary of the message can be generated and displayed.

[0048] Figure 3 This is a schematic diagram illustrating the usage flow of a content provision method according to an embodiment of this disclosure. Figure 3 As shown, the process includes the following steps: (1) The user opens the homepage of the search and information APP; (2) The user performs a pinch gesture on the homepage; (3) Based on the pinch gesture, the content provision method proposed in this embodiment is used to mine important events (i.e. events with high popularity); (4) A list of hot events and summary content are generated.

[0049] Figures 4A-4E This is a display effect diagram of the relevant APP at various stages in a content provision method according to an embodiment of this disclosure. Figure 4AThis is an illustration of the relevant app's homepage. When the app's homepage is displayed on the screen, the user can make a pinch gesture to trigger the page pinching action. Releasing the pinch after a certain distance triggers a page-specific event. During background operations within the app, an icon or animation indicating that content summarization is in progress can be displayed on the screen. During major events (such as major holidays or sporting events), the app can summarize relevant content about the event. In this case, during background operations, an icon or animation indicating that the major event is being summarized will be displayed on the screen, such as... Figure 4B As shown, the screen displays an icon or animation indicating that a "Highlights of the Match" summary is in progress. During normal times, the app can summarize relevant content from current trending events. In this case, while the app's background system is performing operations, the screen displays an icon or animation indicating that a summary of current / dayly trending events is in progress, such as... Figure 4C As shown, the screen displays an icon or animation indicating that the "Daily Morning Briefing" summary is in progress. The screen displays... Figure 4B The / 4C phase can be called the transition phase. After the transition phase, the summarized content can be displayed on the screen. For example, during a major event, relevant content about the major event summarized using the method proposed in the embodiments of this disclosure can be displayed, such as... Figure 4D As shown, the system can display a tournament leaderboard and multiple tournament highlights. For each highlight, a summary of its characteristics can be displayed, providing users with concise tournament information and helping them easily understand the highlights. Users can click on the highlight's summary to view its detailed content. For example, during normal times, after a transition phase, the screen can display relevant content summarizing current / dayly major events using the method proposed in this embodiment, such as... Figure 4E As shown, it can include multiple daily hot topics; for each daily hot topic, it can display news headlines and news summaries related to that hot topic, thereby providing users with brief content information and helping users easily understand the content of each hot topic event; when users want to know the details of the event, they can click on the summary of the hot topic event to display the detailed content of the hot topic.

[0050] The content delivery method proposed in this disclosure offers several advantages. First, users can access summaries of current trending events through pinch gestures, providing a stable entry point. Second, the pinch gesture is a global gesture, ensuring a consistent triggering method for summaries. Furthermore, the pinch gesture is generally understood to have a compression connotation, and this disclosure's use of it as the trigger for content extraction and summarization aligns with common understanding of pinch gestures, resulting in low learning costs for users and quickly meeting their needs for rapid consumption of trending events. This disclosure automatically mines lists of important trending events from internal and external data sources such as ranking lists, promptly delivering recent major events to users. In terms of important content filtering, only the most popular trending events are displayed to users, while the corresponding content data for each event is summarized into concise summaries. This allows users to quickly understand the event overview, and if they wish to learn more related content, they can click on specific articles for further consumption. In some examples, a large model can be used to summarize the content data corresponding to the event, thereby obtaining the event's summary content.

[0051] In some implementations, step S110 includes the following steps:

[0052] Determine the type of content data to be crawled based on at least one of the following: current major events, user's geographic location, and user's personalized preferences;

[0053] Based on the type, content data is crawled from the multiple data sources.

[0054] By employing the above methods, personalized content services can be provided to users based on factors such as their geographical location, personal preferences, and interactions with major events (such as major holidays and sporting events). Combining internal and external rankings and trending searches, the app can uncover key events that users care about and generate summaries of trending events using AI models, intelligently connecting users to these events to help them quickly understand current hot topics. Furthermore, the app's engaging nature can be enhanced through a new touchscreen gesture.

[0055] In some implementations, before clustering multiple pieces of content data, basic filtering and validation can be performed on the crawled content data to remove erroneous data; deduplication can also be performed on multiple pieces of content data, for example:

[0056] The crawled content data is deduplicated by comparing it with existing data in the dataset. If the content data is duplicated (meaning it has been crawled before), the crawling information for the existing data is updated. If the content data is unique (meaning it has not been crawled before), the content data and its crawling information are saved to the dataset. Thus, the dataset contains content data from multiple crawling sessions, along with the crawling information for each session. This deduplication process avoids using duplicate data from multiple crawling sessions when clustering data, thereby preventing errors in the clustering process.

[0057] In some implementations, the content data crawling information includes at least one of crawling time, source, and popularity value. The popularity value of the content data can be related to at least one of the following: the source of the content data, the popularity of the source, the number of related content items, the readership of the related content, and the number of user searches. It is easy to understand that the more important the source of the content data or the larger the user base, the higher the popularity of the source, the larger the number of related content items, the higher the readership of the related content, and the more user searches, the higher the popularity value of the content data. Using the aforementioned crawling information makes it easier to determine the popularity of content data, thereby providing users with the most currently popular data and meeting their need to quickly understand trending information.

[0058] In some implementations, clustering the multiple content data may include:

[0059] For each piece of content data, extract the keyword vector and / or corresponding entity for that content data;

[0060] Based on the keyword vectors and / or corresponding entities of each content data, the multiple content data are initially clustered to obtain multiple clusters, each cluster including multiple content data, and each cluster corresponding to one event;

[0061] The similarity of content data within each cluster is determined, and dissimilar content data is removed from the cluster. For example, a large language model can be used to determine the similarity of content data within a cluster and remove dissimilar content data.

[0062] The aforementioned clustering process consists of two stages: a preliminary clustering stage and a secondary review stage. The preliminary screening stage performs initial clustering of content data based on keyword vectors and / or corresponding entities. Since the acquisition and comparison of keyword vectors and / or corresponding entities are relatively simple, preliminary clustering can be performed quickly. The secondary review stage can correct some errors in the preliminary clustering and improve the accuracy of content data clustering.

[0063] In some implementations, before extracting the events with the highest popularity values ​​from multiple events, the popularity value of each event can be determined separately for the popularity values ​​of multiple content data corresponding to each event; wherein,

[0064] The popularity value of content data is related to at least one of the following: the source of the content data, the popularity of the source of the content data, the number of related content, the number of views of the related content, and the number of user searches.

[0065] For example, when determining the popularity value of content data, a source importance score is determined based on the importance of the source of the content data or the number of users; a source popularity score is determined based on the popularity of the source of the content data; a number of associations score is determined based on the number of related content; a number of related reads score is determined based on the number of reads of the related content; and a number of search scores is determined based on the number of user searches for the content data. The aforementioned scores are then weighted and summed to obtain the popularity value of the content data.

[0066] When determining the popularity value of an event, the average popularity value of multiple content events corresponding to that event can be used as the popularity value of the event.

[0067] Based on the popularity value of each event, users can be shown the most popular events and a summary of each event.

[0068] Figure 5 This is a schematic diagram illustrating the implementation of hot topic mining according to one embodiment of the present disclosure.

[0069] The process of uncovering trending topics generally includes two stages: the intent identification stage and the important content mining stage. Among them:

[0070] The first stage is the intent recognition stage, which determines local major news, major industry news, and holiday activities based on the user's geographical location, personalized preferences, and interactions during major holidays, thereby determining which types of content data to collect.

[0071] The second stage is the important content mining stage, which combines internal and external rankings, author posts, and trending user searches to determine the list of important content. For example... Figure 5 As shown, the process includes the following:

[0072] 1. Data Source Mining: Connect with intervention information from various channels / push messages sent by the app, and monitor the overall network rankings (including comprehensive rankings / vertical rankings / regional rankings, etc.) and push messages across the entire network, as well as information manually reported by the product line, to extract multiple content data (also known as signals, information, resources, etc.) from multiple data sources.

[0073] 2. Signal Capture: Perform basic filtering and verification on the acquired signals (i.e., content data) to remove erroneous signals; deduplicate the processed signals with existing signals to generate new signal IDs, and update the capture information (such as time / source / popularity value and other auxiliary signals); save the captured signals and their capture information into the data set, i.e., perform signal storage.

[0074] 3. Resource clustering and hierarchical classification:

[0075] This phase includes at least the following processes:

[0076] (1) Periodically merge newly captured signals and existing events (or signals corresponding to events): Perform unified formatting on data from multiple sources to facilitate subsequent processing.

[0077] (2) Event clustering: The same event may occur repeatedly in different sources or different events from the same source, so the most similar events need to be clustered. Entities in the extracted content data can be extracted through vector and content understanding, and the entities for the same event can be initially clustered based on the extracted entities; then, the similarity of resources within the cluster can be determined by a large language model, and dissimilar content can be removed.

[0078] (3) Generate new events: For new events that cannot be clustered, call the large language model to generate the title of the new event, which will be used for subsequent event clustering and manual reading list.

[0079] The large language model used in the process of event clustering and generating new events can be provided by external services.

[0080] (4) Event-related resources: Similar resources already existing in the database can be associated through implicit vector recall or explicit entity recall. The same event can be associated with multiple resources. Among them, the operations such as vector recall or entity recall can be provided by external services. For example, the keyword vector and / or corresponding entity of the content data (i.e., resources) can be extracted by external functional entities. The keyword vector and / or corresponding entity are used for resource clustering.

[0081] (5) Event Domain Determination: Multiple methods can be used to determine the domain of an event. One method involves understanding the content of the resources corresponding to the event to determine their classification / region, thereby identifying the event's classification / region. Another method involves determining the domain of an event by classifying and voting on the classifications of its resources, such as by statistically analyzing the classifications / regions of each resource corresponding to an event and identifying the classification / region with the highest number of statistical votes. This can be achieved through external services providing an understanding of the event's resource content; for example, external functional entities can determine the resource's classification / region.

[0082] (6) Event Classification: Events are classified based on multiple signal sources, the popularity of those sources, the number of associated resources, the readership of related resources, and the number of clicks on related user queries. In this step, external functional entities can provide relevant statistical services, including statistics on resource readership and the number of search queries.

[0083] Through the above process, the final event name, event level, event domain, resource list, and other information are obtained and saved into a database, i.e., the event is written to the database. The updated database contains both the updated existing events and the newly added events.

[0084] After the event list is extracted, a summary can be generated by calling an AI model and returned to the app. The app displays a list of trending events, summaries of each event, and other content related to the current trending events, meeting users' needs for quickly accessing trending events. For example, for major sporting events, the app can display the medal table and other event highlights.

[0085] This disclosure also proposes a content providing device. Figure 6 This is a schematic diagram of the structure of a content providing apparatus 600 according to an embodiment of the present disclosure, including:

[0086] The crawling module 610 is used to crawl multiple content data from multiple data sources;

[0087] The summary module 620 is used to cluster multiple content data to obtain multiple events and a content list for each event; the content list of an event contains multiple content data for the corresponding event; the multiple events with the highest popularity values ​​are extracted from the multiple events, and the content data corresponding to each extracted event is summarized to obtain the summary content for each event.

[0088] In some implementations, the grasping module 610 is used for:

[0089] Determine the type of content data to be crawled based on at least one of the following: current major events, user's geographic location, and user's personalized preferences;

[0090] Based on this type, content data is scraped from multiple data sources.

[0091] In some implementations, the summarizing module 620 is also used for:

[0092] The crawled content data is deduplicated from the existing data in the dataset. If the content data is duplicated, the crawling information of the existing data is updated. If the content data is not duplicated, the content data and its crawling information are saved into the dataset.

[0093] In some implementations, the content data crawling information includes at least one of crawling time, source, and popularity value.

[0094] In some implementations, the summary module 620 is used for:

[0095] For each piece of content data, extract the keyword vector and / or corresponding entity for that content data;

[0096] Based on the keyword vectors and / or corresponding entities of each content data, multiple content data are initially clustered to obtain multiple clusters. Each cluster includes multiple content data, and each cluster corresponds to an event.

[0097] Determine the similarity of content data in each cluster and delete dissimilar content data from the cluster.

[0098] In some implementations, the summarizing module 620 is also used for:

[0099] For each event, a specific popularity value is determined based on the popularity values ​​of multiple content data points corresponding to that event; among them,

[0100] The popularity value of content data is related to at least one of the following: the source of the content data, the popularity of the source of the content data, the number of related content, the number of views of the related content, and the number of user searches.

[0101] Figure 7 This is a schematic diagram of the structure of a content providing apparatus 700 according to an embodiment of the present disclosure, as shown below. Figure 7 As shown, in some embodiments, it also includes:

[0102] The gesture recognition module 730 is used to recognize touch events. When the touch event meets the preset trigger gesture, the grasping module 610 is activated.

[0103] The crawling module 610 extracts the page display data and determines the corresponding content delivery solution based on the page display data.

[0104] In some implementations, the triggering gesture includes a pinch gesture.

[0105] In some implementations, page display data includes the position of the page content within the current application;

[0106] The crawling module 610 is used to crawl multiple content data from multiple data sources when the content displayed on the page is in the homepage position of the current application.

[0107] The specific functions and examples of each module and submodule of the apparatus in this disclosure can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.

[0108] The acquisition, storage, and application of personal information by users involved in this technical solution comply with relevant laws and regulations and do not violate public order and good morals.

[0109] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0110] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0111] like Figure 8 As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0112] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0113] The computing unit 801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as detection methods. For example, in some embodiments, the detection method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the detection method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform the detection method by any other suitable means (e.g., by means of firmware).

[0114] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0115] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0116] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0117] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0118] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0119] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0120] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0121] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A content delivery method, comprising: Recognize touch events; Extract page display data when the touch event meets the preset trigger gesture; The page display data includes the position of the page content in the current application; When the content displayed on the page is located on the homepage of the current application, the system determines local news, industry news, and holiday activities based on at least one of the user's geographical location, personalized preferences, and major holiday interactions. Based on the determination results, the system retrieves multiple content data from multiple data sources. The data sources include push messages sent by the application, online rankings, and product line reports. The multiple content data are clustered to obtain multiple events and a content list for each event; the content list of each event contains multiple content data corresponding to that event. The multiple events are associated with resources in the resource library through vector recall or entity recall; By performing content understanding on the resources, the category and / or region of the events corresponding to the resources are determined, and the domain of the multiple events is determined based on the category and / or region; wherein, when the categories and / or regions of multiple resources for an event are different, the categories and / or regions of the resources corresponding to the event are counted, and the category and / or region with the most counts is taken as the domain of the event; the level of the multiple events is determined by at least one of the multiple signal sources, source popularity, number of associated resources, resource views, and number of query clicks; the multiple events, as well as the corresponding resources, domains, and levels, are saved into a dataset; Extract the events with the highest popularity values ​​from the multiple events in the dataset, and use a large model to summarize the content data corresponding to each extracted event to obtain the summary content corresponding to each event. If the application is currently on a message reading page, a message summary is generated and displayed.

2. The method according to claim 1, wherein, Before clustering the multiple content data, the method further includes: The captured content data is deduplicated from the existing data in the dataset. If the content data is duplicated, the capture information of the existing data is updated. If the content data is not duplicated, the content data and its capture information are saved into the dataset.

3. The method according to claim 2, wherein, The content data crawling information includes at least one of the following: crawling time, source, and popularity value.

4. The method according to any one of claims 1-3, wherein, The process of clustering the multiple content data includes: For each of the aforementioned content data, extract the keyword vector and / or corresponding entity of that content data; Based on the keyword vectors and / or corresponding entities of each of the content data, the multiple content data are initially clustered to obtain multiple clusters, each cluster including multiple content data, and each cluster corresponds to one of the events; Determine the similarity of content data in each cluster, and delete dissimilar content data from the cluster.

5. The method according to claim 1, wherein, Before extracting the events with the highest popularity values ​​from the plurality of events in the dataset, the method further includes: For each of the aforementioned events, a popularity value is determined based on the popularity values ​​of multiple content data corresponding to each event; wherein, The popularity value of the content data is related to at least one of the following: the source of the content data, the popularity of the source of the content data, the number of related contents, the number of views of the related contents, and the number of user searches.

6. The method according to claim 1, wherein, The triggering gestures include pinching gestures.

7. A content providing apparatus, comprising: Gesture recognition module, used to recognize touch events; Extract page display data when the touch event meets the preset trigger gesture; The page display data includes the position of the page content in the current application; The crawling module is used to determine local news, industry news, and holiday activities based on at least one of the following: the user's geographical location, personalized preferences, and interactions during major holidays, when the content displayed on the page is located on the homepage of the current application. Based on the determination results, it crawls multiple content data from multiple data sources. The data sources include push messages sent by the application, ranking lists across the entire network, and product line reporting information. The module is also used to generate a message summary and display the summary when the content displayed on the page is located on a message reading page of the current application. The summary module is used to cluster the multiple content data to obtain multiple events and a content list for each event; the content list of each event contains multiple content data corresponding to that event; The multiple events are associated with resources in the resource library through vector recall or entity recall; By performing content understanding on the resources, the classification and / or region of the events corresponding to the resources are determined, and the domain of the multiple events is determined based on the classification and / or region. Where multiple resources for an event have different classifications and / or regions, the classifications and / or regions corresponding to the event are statistically analyzed, and the classification and / or region with the most statistically analyzed values ​​is taken as the domain of the event. The level of the multiple events is determined by at least one of the following: multiple signal sources, source popularity, number of associated resources, resource readership, and query click count. The multiple events, along with their corresponding resources, domains, and levels, are saved into a dataset. The multiple events with the highest popularity values ​​are extracted from the multiple events in the dataset, and the content data corresponding to each extracted event is summarized using a large model to obtain the summary content corresponding to each event.

8. The apparatus according to claim 7, wherein, The summary module is also used for: The captured content data is deduplicated from the existing data in the dataset. If the content data is duplicated, the capture information of the existing data is updated. If the content data is not duplicated, the content data and its capture information are saved into the dataset.

9. The apparatus according to claim 8, wherein, The content data crawling information includes at least one of the following: crawling time, source, and popularity value.

10. The apparatus according to any one of claims 7-9, wherein, The summary module is used for: For each of the aforementioned content data, extract the keyword vector and / or corresponding entity of that content data; Based on the keyword vectors and / or corresponding entities of each of the content data, the multiple content data are initially clustered to obtain multiple clusters, each cluster including multiple content data, and each cluster corresponds to one of the events; Determine the similarity of content data in each cluster, and delete dissimilar content data from the cluster.

11. The apparatus according to claim 7, wherein, The summary module is also used for: For each of the aforementioned events, a popularity value is determined based on the popularity values ​​of multiple content data corresponding to each event; wherein, The popularity value of the content data is related to at least one of the following: the source of the content data, the popularity of the source of the content data, the number of related contents, the number of views of the related contents, and the number of user searches.

12. The apparatus according to claim 7, wherein, The triggering gestures include pinching gestures.

13. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

14. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-6.

15. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-6.

Images