Document recommendation methods, electronic devices and related apparatus
By recognizing user intent and contextual information, and combining file features to make accurate file recommendations, this technology solves the problem of inaccurate recommendation results in existing technologies, and improves file management efficiency and user experience.
Patent Information
- Application Number
- CN202510638340.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-16
- Publication Date
- 2026-11-17
AI Technical Summary
Existing file recommendation methods mainly rely on the time dimension information of files, resulting in inaccurate recommendation results that cannot meet the diverse needs of users.
By recognizing user contextual information and intent, and combining user characteristics and file characteristics, the system proactively displays files and/or folders that match the user's intent, and uses natural language processing and model prediction technologies for accurate recommendations.
It improves the accuracy and relevance of file recommendations to user intent, reduces the time users spend manually searching, and enhances file management efficiency and user experience.
Smart Images

Figure CN123115069A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of terminal technology, and in particular to a document recommendation method, electronic device, and related apparatus. Background Technology
[0002] In the daily use of electronic devices, users frequently come into contact with and manage a large number of files. Although the storage and processing capabilities of electronic devices are constantly improving, file management and file recommendation still face many challenges.
[0003] To improve file processing efficiency, file recommendation systems have emerged. Current file recommendation mechanisms are primarily based on the following approach: sorting files according to time-related information such as last access time, modification time, or creation time to recommend files to users.
[0004] However, this method of recommending files mostly relies on time-related information about the files, which can lead to inaccurate recommendation results. Summary of the Invention
[0005] This application provides a document recommendation method, an electronic device, and related apparatus.
[0006] In a first aspect, embodiments of this application provide a document recommendation method applied to electronic devices, the method comprising:
[0007] The first interface is displayed. Then, upon recognizing the first intent, at least one first recommendation item is displayed, which is a file and / or folder that matches the first intent.
[0008] The first intent is obtained based on contextual information, which includes at least one of the following: content displayed on the first interface, user actions performed on the first interface, the location of the electronic device, and the current time. Furthermore, the first intent in this application can also be used to determine the display location of at least one first recommended item.
[0009] In this way, when a primary intent is recognized, at least one primary recommendation can be proactively displayed on the screen of the electronic device. This primary recommendation is a file and / or folder that matches the primary intent. This allows for the recommendation of files corresponding to the intent, improving the accuracy and relevance of the recommended files to the user's intent, avoiding the need for users to spend excessive time filtering files, and improving user efficiency.
[0010] In one possible implementation, at least one first recommendation item is displayed, including:
[0011] At least one first recommendation item is displayed at a first position corresponding to the first element. The first element is an identifier or a first control for a first application. The first application is an application for file management. The identifier described here can be, for example, an icon or a text identifier, and this application does not limit this. The first control is a control for adding files.
[0012] In this way, at least one recommended item can be proactively recommended in the first position corresponding to the first element. This proactive recommendation method can promptly push related files to the user, avoiding the need for the user to actively discover the recommended files, thus improving the user experience. Furthermore, when the first element is the first control, displaying at least one recommended item in the position corresponding to the first control facilitates quick user interaction with the recommended items. And, when the first element is the identifier of the first application, displaying at least one recommended item in the position corresponding to the identifier of the first application, since the first application is a file management application, helps the user understand the context of the currently displayed recommended items and ensures the rationality of the interface display.
[0013] In one possible implementation, the method further includes:
[0014] A first prompt message is displayed at a second position corresponding to the first element. The first prompt message is used to indicate the existence of recommended files and / or folders.
[0015] Display at least one first recommended item at the first position corresponding to the first element, including:
[0016] In response to a first operation applied to the first element, at least one first recommended item is displayed at a first position corresponding to the first element. The first operation can be placing the cursor on the first element.
[0017] In this implementation, a primary prompt can be displayed first to inform the user of the currently existing recommended files and / or folders. Then, based on user interaction, the primary recommended item can be displayed, thus serving both as a prompt and providing the user with a degree of choice; that is, the user can choose whether or not to view at least one primary recommended item.
[0018] In one possible implementation, at least one first recommendation item is displayed, including:
[0019] In response to a second operation applied to the first element, the application window of the first application is displayed, and at least one first recommendation item is displayed in the application window. The first element is the identifier of the first application or a first control, the first control is a control for adding files, and the first application is an application for file management.
[0020] In this implementation, for example, a first prompt message could be displayed in a second position corresponding to the first element. This prompt message indicates the existence of recommended files and / or folders, thus informing the user that recommended files and / or folders currently exist. Subsequently, the user can perform a second operation on the first element to trigger the display of the application window for the first application.
[0021] In this implementation, the first recommended item can be displayed only after user interaction. This provides users with a degree of choice, allowing them to decide whether or not to view at least one of the first recommended items.
[0022] In one implementation, the first intent in this application can be used to determine a first element. For example, if the first intent is any of the following: file sharing, adding attachments, and file collaboration, the first element can be determined as a first control. Then, at least one first recommended item is displayed at the first position corresponding to the first control. Alternatively, for other cases where the first intent is different, the first element can be determined as the identifier of a first application, and then at least one first recommended item is displayed at the first position corresponding to the identifier of the first application. Thus, based on the determination of the display position of at least one first recommended item, the correspondence between the first intent and the first element can be determined according to actual needs, and this embodiment does not impose any limitations on this.
[0023] In one possible implementation, the method further includes:
[0024] Based on the reference information, at least one first recommended item is determined from multiple alternative documents. The reference information in this application may include contextual information. Furthermore, the reference information may also include, for example, user characteristic information.
[0025] In one possible implementation, based on reference information, at least one first recommended item is determined from a plurality of alternative files, including:
[0026] Based on some or all of the reference information, obtain the characteristics of the target file;
[0027] Among multiple candidate files, files whose file characteristics match those of the target file are searched to determine at least one first recommendation.
[0028] In this implementation, the technical solution of this application can integrate user characteristics, file characteristics and contextual information to determine the recommended files, thereby intelligently pushing documents to improve the accuracy of the recommended files and their matching with the actual scenario.
[0029] In one possible implementation, the context information includes the content of the first interface, and the target file features include the first target file features obtained based on the analysis of the content of the first interface.
[0030] In this implementation, electronic devices can accurately identify a user's intent in chat or email scenarios, for example, using natural language processing techniques. Compared to existing technologies, this implementation can more accurately capture user needs, reducing misjudgments and manual operations. Furthermore, the electronic device not only identifies user intent but also extracts keywords from the window content and combines these keywords for file matching. This improvement makes file recommendations more accurate, reducing the time users spend manually searching.
[0031] In one possible implementation, the context information includes the location of the electronic device and / or the current time;
[0032] The target file characteristics include second target file characteristics determined based on user characteristic information. These second target file characteristics are those corresponding to files whose access behavior within a first time period and / or a first geographical area satisfies a first condition. The access behavior may include at least one of the following: number of openings, editing duration, file foreground display duration, etc. The access behavior satisfying the first condition may include at least one of the following: the number of openings is greater than a first threshold, the editing duration is greater than a first duration threshold, or the file foreground display duration is greater than a second duration threshold, etc. The specific access behaviors and the implementation of the first condition can be extended according to actual needs.
[0033] Next, among multiple candidate files, files whose file feature information matches the target file features are searched to determine at least one first recommendation, including:
[0034] If the current time falls within a first time period, and / or if the location of the electronic device falls within a first geographical area, search among multiple candidate files for files whose file feature information matches the features of the second target file to determine at least one first recommendation.
[0035] In this implementation, personalized file recommendations can be provided by combining the time patterns and geographical locations reflected in user characteristic information.
[0036] In one possible implementation, the context information includes user actions taken during a second time period prior to the current time;
[0037] The target file characteristics include a third target file characteristic determined based on user actions during the second time period.
[0038] In this implementation, file matching can be performed by combining the features of the third target file of interest reflected in the user's operations in the interface, so as to improve the accuracy of the determined first recommendation.
[0039] In one possible implementation, the user actions during the second time period include multiple screenshot operations, and at least one first recommendation item includes multiple screenshots obtained from the multiple screenshot operations; and / or, at least one first recommendation item includes a file obtained by stitching together multiple screenshots.
[0040] This allows for the identification of consecutive screenshots of similar content (such as meeting screenshots, tutorial step screenshots, etc.) taken by the user, and the packaging (stacking) of these screenshots for presentation. Compared to implementations that present screenshots as individual files, which lack the identification and packaging processing of the correlation between consecutive screenshots, the technical solution of this application significantly improves the efficiency of screenshot management through intelligent packaging and stacking.
[0041] In one possible implementation, based on reference information, at least one first recommended item is determined from a plurality of alternative files, including:
[0042] The reference information and the first intent are input into the first model so that the first model outputs at least one recommendation file that matches the first intent;
[0043] Based on at least one recommendation document, determine at least one first recommendation item.
[0044] Based on this implementation method, the recommendation files can be determined simply and effectively, and by using model prediction, the technical solution of this application can be applied to various possible scenarios.
[0045] In one possible implementation, at least one first recommended item includes at least one of the following implementations: a first folder in the form of a folder, a first file in the form of a file, and a first overlay item in the form of an overlay of folders and files.
[0046] In one possible implementation, there are multiple recommended files for the first intent, and these multiple recommended files belong to a first folder. At least one of the first recommended items can include the first folder in the form of a folder, or it can also include a first superimposed item in the form of a combination of folders and files. These two cases are described below:
[0047] For example, the difference in target scores among multiple recommended documents could be a first threshold, in which case at least one first recommended item could include a first folder in the form of a folder.
[0048] In this way, compared to the method of needing multiple recommendation slots to display multiple recommended files in sequence, directly displaying the first folder can effectively save recommendation slots while effectively recommending relevant files to users.
[0049] For example, among multiple recommended files, there exists a target file with a target score greater than or equal to a third threshold, and the difference in target scores among the remaining recommended files (excluding the target file) is less than a first threshold. In this case, at least one first recommended item may include a first overlay item in the form of an overlay of folders and files. The first overlay item includes a first folder and the target file within that first folder.
[0050] In this way, compared to requiring multiple recommendation slots to display various recommended files sequentially, displaying the first overlay item effectively saves recommendation slots while still effectively recommending relevant files to users. Furthermore, the top-rated target file can be directly shown in the first recommendation item, allowing users to quickly understand that the target file is the first recommendation.
[0051] or,
[0052] If there are multiple recommended files for a first intention, and if the multiple recommended files do not belong to the same folder, then at least one first recommended item includes the first file, and the first file is one of the multiple recommended files;
[0053] or,
[0054] In the case where there is a single recommendation file for the first intent, at least one recommendation includes the first file, and the single recommendation file is the first file.
[0055] In this implementation, multiple recommended documents or a single recommended document can be displayed as an independent recommendation item, thereby improving the efficiency of users in finding the relevant recommended documents.
[0056] In one possible implementation, the recommended item corresponds to a target score;
[0057] The target score for a recommendation item is determined based on the importance of the primary intent and / or the degree of matching between the features of the recommended document and the target document.
[0058] As described above, recommendations can be in file format, in which case a recommendation item contains only one recommended file. Alternatively, recommendations can also be in folder format or stacked item format, in which case a recommendation item will contain multiple recommended files. These recommended files are all files contained within the folder of that recommendation item, but it should be understood that not all files within the folder are recommended files; the specific recommended files are still determined according to the implementation described above.
[0059] When a recommendation item includes multiple recommendation files, one approach is to first determine the target scores of these recommendation files, and then use these target scores to determine the target score of the recommendation item. For example, the target score of the recommendation item can be determined by averaging, medianing, or finding the maximum / minimum values of the target scores across the multiple recommendation files. Alternatively, the matching degree between the recommendation item and the target file can be determined based on the matching degree between each recommendation file and its features, and then the target score of the recommendation item can be determined based on the matching degree and / or the importance of the primary intent.
[0060] The following example demonstrates how to determine the target score for a recommendation file, and then, based on the target scores of the recommendation files contained within the recommended item, determine the target score for that item. In this implementation, the target score for the recommendation file must first be determined. For example, the target score for the recommendation file can be determined as follows:
[0061] For any given recommendation file, obtain an initial score for the recommendation file, wherein the initial score is determined based on the degree of matching between the recommendation file and the features of the target file, and the target file features are any one of the following: a first target file feature, a second target file feature, and a third target file feature; or, the initial score is determined based on a first model, wherein the first model is also used to output the initial score corresponding to at least one recommendation file.
[0062] The initial score of the recommended document is adjusted based on the first parameter of the first intent to determine the target score of the recommended document. The first parameter is used to indicate the importance of the first intent.
[0063] In one possible implementation, when there are multiple intents, each intent corresponds to at least one recommendation item; wherein, when displaying multiple recommendation items corresponding to multiple intents, the target ratings of the multiple recommendation items are sorted from high to low.
[0064] This allows for the display of recommendations corresponding to multiple intents within the first area, facilitating quick searching for relevant recommendations. Furthermore, sorting by target rating from highest to lowest improves the speed at which users find recommendations for their most pressing intent.
[0065] In one possible implementation, the method also includes:
[0066] Using a preset time period as the cycle, reduce the target score of at least one recommended item by the first magnitude.
[0067] Based on this processing method, the scores of the recommended items corresponding to each intent can be dynamically adjusted so that the recommended items corresponding to newly identified intents can be displayed in the first area for recommendation.
[0068] In one possible implementation, the method further includes:
[0069] Obtain user feedback information, which is used to indicate whether the first intent matches the first recommended item;
[0070] Adjust the model parameters of the first model based on the feedback information.
[0071] In one possible implementation, the user characteristic information includes at least one of the following: file access information, file editing frequency, file editing duration, contact priority, and contact interaction frequency;
[0072] File characteristic information includes at least one of the following: file name, file type, file creation time, file modification time, file size, file association information, file usage frequency, file keywords, file content, and file topic.
[0073] In one possible implementation, the first intent is any of the following: file editing, file searching, file sharing, adding attachments, information collection, and file collaboration.
[0074] In one possible implementation, where the first intent is any of the following: file sharing, adding attachments, file collaboration, the first element is the first control.
[0075] When the system detects intentions such as file sharing, adding attachments, or file collaboration, at least one recommended item can be displayed in the first position corresponding to the first control used to add files. This makes it easier for users to quickly add files, reduces the amount of user interaction, and improves convenience.
[0076] Furthermore, the method also includes:
[0077] In response to a third operation applied to a target recommendation item in at least one of the first recommendation items, the function corresponding to the first control is executed on the target recommendation item.
[0078] In other words, the technical solution of this application allows users to directly operate on at least one first recommended item, thereby improving the convenience of adding files. For example, in response to a third operation, the first control can directly add a recommended item as an attachment in an email, or directly send a recommended item in a chat window, reducing the user's operational steps.
[0079] In a second aspect, embodiments of this application provide an electronic device, including a processor and a memory, wherein the memory is used to store code instructions and the processor is used to execute the code instructions to perform the methods described in the first aspect or any possible implementation thereof.
[0080] Thirdly, embodiments of this application provide a computer-readable storage medium storing a computer program or instructions that, when executed on a computer, cause the computer to perform the methods described in the first aspect or any possible implementation thereof.
[0081] Fourthly, embodiments of this application provide a computer program product including a computer program, which, when run on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation of the first aspect.
[0082] Fifthly, this application provides a chip or chip system including at least one processor and a communication interface. The communication interface and the at least one processor are interconnected via a circuit. The at least one processor is used to run computer programs or instructions to perform the methods described in the first aspect or any possible implementation thereof. The communication interface in the chip can be an input / output interface, pins, or circuits, etc.
[0083] In one possible implementation, the chip or chip system described above in this application further includes at least one memory storing instructions. The memory can be an internal storage unit of the chip, such as a register or cache, or it can be a storage unit of the chip itself (e.g., read-only memory, random access memory, etc.).
[0084] It should be understood that the second to fifth aspects of this application correspond to the technical solutions of the first aspect of this application, and the beneficial effects achieved by each aspect and the corresponding feasible implementation are similar, and will not be repeated here. Attached Figure Description
[0085] Figure 1 The following is a schematic diagram illustrating the recommended implementation of the documents provided in this application embodiment;
[0086] Figure 2 A flowchart illustrating the document recommendation method provided in this application embodiment;
[0087] Figure 3 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 1 ;
[0088] Figure 4 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 2;
[0089] Figure 5 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 3 ;
[0090] Figure 6 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 4 ;
[0091] Figure 7 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 5 ;
[0092] Figure 8 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 6 ;
[0093] Figure 9 Implementation illustration of determining the recommended items provided in the embodiments of this application Figure 1 ;
[0094] Figure 10 Implementation illustration of determining the recommended items provided in the embodiments of this application Figure 2 ;
[0095] Figure 11 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 7 ;
[0096] Figure 12 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 8 ;
[0097] Figure 13 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 9 ;
[0098] Figure 14 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 10 ;
[0099] Figure 15 Implementation illustration of determining the recommended items provided in the embodiments of this application Figure 3 ;
[0100] Figure 16 A schematic diagram of an interface displaying multiple recommended items provided for an embodiment of this application;
[0101] Figure 17 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 10 one;
[0102] Figure 18 This is a schematic diagram illustrating the implementation of the form of the recommended item provided in the embodiments of this application;
[0103] Figure 19 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 10 two;
[0104] Figure 20 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation
[0105] To facilitate understanding, the relevant terms and concepts involved in the embodiments of this application will be introduced below:
[0106] 1. Electronic equipment
[0107] The electronic devices in this application embodiment can be referred to as user equipment (UE), terminal, etc. For example, electronic devices can be mobile phones, tablets, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, computing devices, in-vehicle devices or wearable devices, virtual reality (VR) terminal devices, augmented reality (AR) terminal devices, wireless terminals in industrial control, wireless terminals in smart homes, etc. The form of the electronic devices is not specifically limited in this application embodiment.
[0108] 2. Other terms
[0109] In the embodiments of this application, terms such as "first" and "second" are used to distinguish identical or similar items with substantially the same function and purpose. For example, "first chip" and "second chip" are used only to distinguish different chips and do not limit their order of execution. Those skilled in the art will understand that terms such as "first" and "second" do not limit the quantity or execution order, and that "first" and "second" do not necessarily imply that they are different.
[0110] It should be noted that, in the embodiments of this application, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.
[0111] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, a--c, bc, or abc, where a, b, and c can be single or multiple.
[0112] To better understand the technical solution of this application, the relevant technologies involved in this application are introduced below.
[0113] With the rapid development of information technology, electronic devices have become central to users' daily work and life. During daily use of electronic devices, users frequently access and manage large amounts of files, such as documents, emails, pictures, and videos. However, despite the continuous improvement in the storage and processing capabilities of electronic devices, file management and file recommendation still face many challenges.
[0114] Currently, most traditional file management systems rely on manual categorization and searching by users. With the surge in the number of files, problems such as repetitive work, wasted time, and information omissions may occur during file searching and sending. Especially in multitasking and multi-window work environments, users often need to frequently switch applications and search for files, resulting in low efficiency and low user satisfaction.
[0115] To address the problems described above, file recommendation systems have emerged. Currently, many operating systems and applications offer simple file recommendation functionality. The following section will combine... Figure 1 This section describes one possible implementation of file recommendation. Figure 1 The diagram illustrates the recommended implementation of the documents provided in this application.
[0116] In one implementation, while a user is using an electronic device, the device's operating system can, for example, recommend a series of recently opened files based on their opening times.
[0117] For example, you can refer to Figure 1 To understand, Figure 1This demonstrates the desktop interface of an electronic device. The desktop interface may include a taskbar, which corresponds to the area at the bottom of the screen in the image, specifically the area below the horizontal line. As shown, the taskbar can include icons for multiple applications, such as... Figure 1 The image exemplifies the icons of applications A through J, with the rounded rectangles representing the application icons.
[0118] Furthermore, the taskbar may also include a start control 101 as shown in the figure, where the start control 101 is a control used to trigger the display of the start menu. For example, in response to a user action on the start control 101, the following can be displayed on the desktop interface: Figure 1 The window 102 shown is the window corresponding to the Start Menu.
[0119] Referring to the illustration, it can be determined that window 102 may include a "Recommended Items" menu, under which at least one recommended file can be displayed. For example, these recommended files are files recently opened by the user, and they can be sorted sequentially according to the opening time of the files.
[0120] For example, Figure 1 File 1 was opened 9 minutes ago, file 2 was opened 30 minutes ago, file 3 was opened 31 minutes ago, file 4 was opened 1 hour ago, and file 5 was opened 2 hours ago. Users can also click the "More" button shown in the image to browse more recommended files. These more recommended files are also determined according to their opening time.
[0121] Figure 1 The electronic devices shown can be understood as personal computers. Those skilled in the art will understand that the technical solution described above is also applicable to other computing platforms and application scenarios such as smartphones and cloud storage services.
[0122] The method described above allows users to quickly access recently used files. It should also be noted that... Figure 1 The illustrated example uses the most recent opening time of a file as an example. In reality, file recommendation mechanisms are not limited to this. For example, file recommendations can also be based on multiple time dimensions such as file modification time and file creation time.
[0123] In other words, the current file recommendation mechanism in related technologies is mainly based on the following approach: sorting files according to time-related information such as last access time, modification time, or creation time in order to recommend files to users.
[0124] However, this method of recommending files mostly relies on the time-related information of the files, while ignoring the user's intent perception and contextual information, and lacks an understanding of user behavior and usage scenarios, resulting in inaccurate recommendation results that cannot meet the diverse needs of users.
[0125] To address the aforementioned issues, this application proposes a file recommendation method. This method integrates user feature information, file semantic features, and real-time contextual information to proactively perceive the user's work intent and accurately push the required files, thereby improving the intelligence and accuracy of file recommendation, enhancing file management efficiency, and optimizing user experience.
[0126] The document recommendation method provided in this application is described below with reference to specific embodiments. These embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.
[0127] Figure 2 This is a flowchart illustrating the document recommendation method provided in an embodiment of this application. (Refer to...) Figure 2 The document recommendation method provided in this application embodiment may include:
[0128] 1. Identify the primary intent.
[0129] In one implementation, intent recognition can be performed based on contextual information, which may include at least one of the following: content displayed on the first interface, user actions performed on the first interface, the location of the electronic device, and the current time.
[0130] The first interface can be understood as the interface currently displayed on the electronic device, such as a desktop interface. Alternatively, during the user's use of the electronic device, for example, if a chat application is opened, the first interface can also be the chat application's interface. This application does not limit the specific implementation of the first interface, which depends on the actual usage scenario of the electronic device. For example, the content in the first interface can include at least one of the following forms: text, images, controls, and files.
[0131] Taking a chat application's interface as an example, the first screen might contain contact names, chat text, chat images, and so on. Semantic analysis based on this content can then be performed to identify, for instance, the initial intent.
[0132] For example, a user can perform user operations on the first interface, and these operations can reflect the user's intent. Therefore, for instance, the initial intent can be identified based on the user operations performed on the first interface. The user operations can be performed in any area or location within the first interface; this embodiment does not impose any limitations on this.
[0133] For example, if a user frequently processes certain files within a specific time period or area, the primary intent can be identified based on the location of the electronic device and / or the current time.
[0134] In this embodiment, the first intent is an intent related to file processing. For example, the first intent includes at least one of the following: file editing, file searching, file sharing, information collection, and file collaboration. Alternatively, the specific implementation of the first intent can be extended according to actual needs. The first intent is not a specific intent, but rather an intent related to file processing identified based on the above description; the specific implementation of the first intent depends on the actual usage scenario of the electronic device.
[0135] It is also important to understand that in the implementation of identifying the primary intent based on the content described above, the specific intent recognition method can be selected according to actual needs. For example, a model for identifying intent can be pre-trained, and then the content described above can be input into the model to determine whether the primary intent has been recognized. Alternatively, intent recognition rules can be pre-established, and then based on the pre-established rules and the content described above, it can be determined whether the primary intent has been recognized.
[0136] 2. Display at least one primary recommendation item, which is a file and / or folder that matches the primary intent.
[0137] In this embodiment, when a first intent is recognized, at least one first recommendation item can be proactively displayed on the screen of the electronic device. The first recommendation item is a file and / or folder that matches the first intent. This improves both the accuracy and relevance of the recommended files to the user, reducing the time required for file filtering and increasing user efficiency. Furthermore, this proactive recommendation method allows for timely push of relevant files to the user, eliminating the need for active discovery and enhancing the user experience.
[0138] The following sections, with reference to the accompanying diagrams, describe several implementations of displaying at least one first recommendation item. Figure 3 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 1 , Figure 4 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 2 , Figure 5 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 3 , Figure 6 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 4 , Figure 7 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 5 , Figure 8 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 6 .
[0139] The first objective is to simultaneously display a first prompt message and at least one first recommended item, whereby the first prompt message indicates that a recommended item currently exists.
[0140] In this implementation, upon recognizing the first intent, the first prompt message and the first recommendation can be displayed simultaneously to indicate to the user that there are currently recommended files and / or folders, while directly showing the currently recommended files and / or folders to the user.
[0141] For example, at least one first recommended item can be displayed at a first position corresponding to the first element. In one implementation, for example, a first window can be displayed at the first position, and at least one first recommended item can be displayed within the first window. This makes it easier to integrate the recommended items into a container (i.e., the first window) for presentation.
[0142] The first element is either the identifier of the first application or the first control. For example, the first application could be a file management application, allowing users to easily perform subsequent file operations based on that application. The first control could be a control for adding files, facilitating quick and easy file addition by the user. For instance, chat or email applications typically include a file addition feature.
[0143] And when displaying the first prompt message, for example, the first prompt message can be displayed at a second position corresponding to the first element.
[0144] For example, you can refer to Figures 3 to 5 Understand the implementation of displaying the first prompt message and the first recommended item.
[0145] like Figure 3 As shown, application A can be, for example, the first application described above, and its icon (i.e., an icon containing the text "A" within a rounded rectangle) is displayed in the figure. Then, for example, a first prompt message (i.e., the black line shown in the figure) as shown in 301 can be displayed below the icon of application A to indicate the presence of recommended files and / or folders. That is, the position below the icon of application A is the second position described above.
[0146] For example, a window as shown in Figure 302 can be displayed above the icon of application A. This window contains multiple first recommended items; in this example, the first recommended items include File 1, File 2, and File 3. In this application, the window used to display the first recommended items is referred to as the first window, and the following description will focus on the first window. This allows users to quickly perform corresponding operations on the recommended files. Users can click the file icon in the first window 302 to open the corresponding file. Alternatively, users can drag the file icon in the first window 302 to send the corresponding file. In this example, the position above the icon of application A is the first position described above.
[0147] Refer to Figure 4 As shown, for example, the interface of a chat application includes, for example, the following Figure 4 The first control 401 shown is for adding files. For example, a first prompt message (the black line shown in the figure) as indicated by 402 can be displayed below the first control 401 to indicate the presence of recommended files and / or folders. That is, the position below the first control 401 is the second position described above.
[0148] Furthermore, for example, a first window, as shown in Figure 403, can be displayed above the first control 401. This first window contains multiple first recommended items; in this example, the multiple first recommended items include two files named "Event A Poster V3" and "Party B Poster." This allows users to quickly perform corresponding operations on the recommended files. For example, a user can click the file icon in the first window 403 to quickly send the corresponding file in the current chat window. In this example, the position above the first control 401 is the first position described above.
[0149] Refer to Figure 5 As shown, for example, the interface of an email application includes... Figure 5 The first control 501 shown is for adding files. For example, a first prompt message (the black line shown in the figure) as indicated by 502 can be displayed below the first control 501 to indicate the presence of recommended files and / or folders. That is, the position below the first control 501 is the second position described above.
[0150] Furthermore, for example, a first window, as shown in Figure 503, can be displayed above the first control 501. This first window contains a first recommended item; in this example, the first recommended item includes two files named "Project A Acceptance Report" and "Project A Mid-term Reference." This allows users to quickly perform corresponding operations on the recommended files. For instance, a user can click the file icon in the first window 503 to quickly add the corresponding file as an attachment to the current email window. In this example, the position above the first control 501 is the first position described above.
[0151] In the above combination Figures 3-5 In the examples described, the second position is located below the application icon or the first control, and the first position is located above the application icon or the first control. It's important to understand that in actual implementation, the specific implementation of the first and second positions can be set according to actual needs, as long as the first position (and the second position) corresponds to the application icon or the first control, allowing users to understand that a corresponding recommended file exists under the logic of certain applications or controls.
[0152] In the implementation described above, a thick horizontal line represents the first prompt message. In actual implementation, the specific form and visual effect of the first prompt message can be set according to actual needs. For example, the application icon or the first control can be set to display certain animation effects (such as bouncing up and down, highlighting, etc.) as the first prompt message described here. This embodiment does not limit the specific implementation of the first prompt message, as long as the first prompt message can achieve the effect of indicating the existence of recommended files and / or folders.
[0153] Furthermore, the specific presentation of at least one first recommended item can also be set according to actual needs. The above-described implementation displays at least one first recommended item in the first window. Alternatively, at least one first recommended item can be displayed directly without using a window.
[0154] In the implementation described above, on the one hand, displaying the initial prompt message allows for timely notification to the user upon recognizing the initial intent, ensuring the user is aware of the currently recommended files and / or folders. On the other hand, directly displaying at least one initial recommendation facilitates quick operations on the corresponding recommended files and / or folders, enhancing the user experience.
[0155] Implementation 2: First, display the first prompt message, then respond to the user's action and display at least one first recommended item.
[0156] The implementation of displaying the first prompt and the first recommended item is similar to that described above and will not be repeated here. The difference in the current implementation is that, upon recognizing the initial intent, the first prompt is displayed first, without displaying the first recommended item. The first recommended item is then displayed only after user interaction. This provides users with a degree of choice, allowing them to choose whether or not to view the first recommended item.
[0157] The current implementation will be described below with reference to the accompanying diagrams.
[0158] like Figure 6 As shown in (a), application A can be, for example, the first application described above, and its icon (i.e., an icon containing the text "A" within a rounded rectangle) is displayed in the figure. Then, for example, a first prompt message (i.e., the black line shown in the figure) as shown in 601 can be displayed below the icon of application A to indicate the presence of recommended files and / or folders. Referring to the illustration, it can be understood that initially only the first prompt message 601 is displayed in the interface, and the first recommended item is not displayed.
[0159] Subsequently, for example, in response to the first operation, at least one first recommended item can be displayed. Similarly, as described above, at least one first recommended item can be displayed at a first position corresponding to the first element. The specific implementation can be understood by referring to the above description.
[0160] For example, the first operation could be, for instance, placing the cursor on the first element. (See reference...) Figure 6 To understand (a) in this context, for example, if the user moves the cursor and places the table over the icon of application A, then for example, it can display something like... Figure 6 As shown in (b), the first window 602 is displayed in the first position, and the first window 602 contains the first recommended items, such as file 1, file 2 and file 3.
[0161] As can be seen from the illustration, the first window 602 may also include a smaller window 603 as shown. Typically, when the cursor is placed on the application's icon, a thumbnail of the application's window is displayed. For example, if application A is a file management application, then the smaller window 603 in the illustration could be a thumbnail of the file management application. In another implementation, the first window 602 and the thumbnail window of application A can also be two independent windows; this application does not limit the specific implementation.
[0162] And in response to the above Figure 4The implementation of displaying the first prompt and the first recommended item at the position corresponding to the first control in a chat application is similar. For example, upon recognizing the first intent, the first prompt can be displayed first, followed by a response to the first action, and then a first window can be displayed at the first position to show at least one first recommended item. The first action can be, for example, placing the cursor on the first control, which is understood similarly to the above and will not be elaborated upon here.
[0163] Similarly, regarding the above Figure 5 The implementation of displaying the first prompt message and the first recommended item at the position corresponding to the first control in the email application is similar. For example, upon recognizing the first intent, the first prompt message can be displayed first, followed by a response to the first action, and then a first window can be displayed at the first position to display at least one first recommended item. The first action can be, for example, placing the cursor on the first control, which is understood similarly to the above and will not be elaborated here.
[0164] It should be understood that the first operation can be, in addition to placing the cursor on the first element, it can also be set according to actual needs. For example, the first operation can also be a click operation on the first element, or it can be an operation to draw a preset path on the first element. This application does not limit the specific implementation of the first operation.
[0165] Understandably, in the implementation described above, which displays a first window in the first position and then at least one first recommended item within that first window, the first window is a dedicated window for displaying recommended items. The implementation of displaying the first window is in the form of a pop-up notification window, which improves user convenience and makes the recommended items more prominent. However, if the user closes this first window, they may be unable to access the files recommended in that window later. In other words, after closing the first window, the user cannot find the entry point for the recommended items.
[0166] Therefore, in another implementation, a specific area can be set within the application's window, dedicated to displaying recommended files and / or folders. In this approach, users can access the recommended files at any time within this designated area, offering the advantage of high stability.
[0167] The following is combined Figure 7 and Figure 8 This implementation will be introduced.
[0168] like Figure 7As shown in (a), application A can be, for example, the first application described above, and its icon (i.e., an icon containing the text "A" within a rounded rectangle) is displayed in the figure. Then, for example, a first prompt message (i.e., the black line shown in the figure) as shown in 701 can be displayed below the icon of application A to indicate the presence of recommended files and / or folders. Referring to the illustration, it can be understood that initially only the first prompt message 701 is displayed in the interface, and the first recommended item is not shown.
[0169] Subsequently, in response to a second operation applied to the first element, the application window of the first application can be displayed in the first interface, showing the first recommended item. For example, the second operation could be a click operation on the first element. And in the implementation of displaying the first recommended item in the application window, for example, a fixed first area can be set in the application window for displaying the recommended item, and then the first recommended item can be displayed within that first area.
[0170] Reference Figure 7 To understand (a) in this context, for example, if a user clicks the icon of application A, then the following might be displayed: Figure 7 The situation shown in (b) is the display of the application window 702 of the first application as shown in the figure. That is, the application window 702 is the window corresponding to application A. For example, if the first application is an application used to implement file management functions, then the application window 702 contains file management-related content.
[0171] Referring to the illustration, it can be determined that the content currently displayed in application window 702 may be, for example, the contents of drive D on the local disk, where folders K, L, M, N, O, P, Q, file 5, file 6, and file 7 are files stored on drive D. Referring also to the illustration, it can be determined that application window 702 may contain, for example, a first area 703, which is used to display recommended files and / or recommended folders. Therefore, upon recognizing the first intent, the first recommended item described above may be displayed, for example, within the first area 703 of the application window. For example, at least one first recommended item may include file 1, file 2, and file 3 as shown in the figure.
[0172] In one implementation, in a first area of the application window, in addition to displaying the first recommended item corresponding to the currently identified first intent, at least one recommended item corresponding to other previously identified intents may also be displayed, such as intents identified before the current time.
[0173] For example, the file 4 displayed in the first area 703 of the figure can be a recommended item corresponding to other intentions besides the first intention. This allows the user to perform corresponding operations on file 4 after the first window is closed, even if the user does not process file 4 when the first window recommends it. In this way, the user can choose to open the application window of the first application to quickly find file 4 in the first area and perform corresponding operations.
[0174] The above combination Figure 7 This section describes the implementation where the first element is the icon of the first application. In other implementations, after the user interacts with the first control (the control used to add files), for example, the application window of the first application can be launched. The subsequent processing in this case is similar and will not be elaborated here.
[0175] This allows users to easily browse current and historical recommended files in the first area according to their actual needs, thereby improving the stability and accessibility of the recommended files provided to users.
[0176] The first application can be any application besides file management, as long as the user can browse files within it. Additionally, users can browse recommended files through the Start Menu. The following section combines... Figure 8 To understand.
[0177] like Figure 8 As shown in (a) above, the Start Menu icon 800 is displayed in the figure. Similar to the above description, for example, a first prompt message (in a similar format to that described above, not shown in the figure) can be displayed below the Start Menu icon 800 to indicate the presence of recommended files and / or folders. Thus, initially, only the first prompt message is displayed in the interface, without showing the first recommended item.
[0178] Subsequently, for example, in response to a second operation acting on the Start Menu icon 800, the Start Menu window can be displayed again in the first interface, with at least one first recommended item displayed in a specific area of the Start Menu window. In one implementation, the Start Menu can also be understood as an application, and thus the Start Menu can be understood as the first application described above. Correspondingly, the Start Menu window can be understood as the application window described above, and the specific area of the Start Menu window can be understood as the first area described above. Based on this, the following description will be provided. For example, the second operation can be a click operation on the icon of the first application.
[0179] Reference Figure 8 To understand (a) in this context, for example, if a user clicks the Start menu icon 800, then the following might be displayed: Figure 8 As shown in (b) above, the application window 802 of the Start Menu is displayed as shown in the figure. Referring to the figure, it can be determined that the application window 802 may include a "Recent Apps" menu and a "Guess You Are Looking For" menu, where the area 803 corresponding to the "Guess You Are Looking For" menu is the first area described in this embodiment.
[0180] As shown in the figure, at least one of the first recommended items described above can be displayed in area 803. For example, at least one first recommended item may include File 1, File 2, and File 3 shown in the figure. In one implementation, in the first area of the application window, in addition to displaying the first recommended item corresponding to the currently identified first intent, recommended items corresponding to other previously identified intents can also be displayed, such as intents identified before the current time. For example, File 4 displayed in the first area 703 of the figure could be a recommended item corresponding to other intents besides the first intent, and its implementation is similar to that described above, and will not be repeated here.
[0181] Referring to the above introduction Figure 7 and Figure 8 In actual implementation, the specific application window in which the first area is set to display the recommended items can be arbitrarily set according to actual needs; this application is the first application described in this article. The specific setting of the first area can also be expanded according to actual needs, and this embodiment does not limit this.
[0182] Implementation 3: Directly display at least one first recommended item.
[0183] In this implementation, there's no need to display the first prompt message; instead, at least one first recommended item is displayed directly. Therefore, based on Implementation 1 and Implementation 2 above, by removing the implementation related to displaying the first prompt message and directly executing the implementation that displays at least one first recommended item in the first position, we obtain Implementation 3, which is described here.
[0184] Based on the above description, displaying at least one first recommended item in the first position can include the following two implementations:
[0185] In one implementation, at least one first recommended item may be displayed at a first position corresponding to the first element. The specific implementation can be referred to the description of the above embodiments, and will not be repeated here.
[0186] In another implementation, in response to a second operation applied to the first element, the application window of the first application can be displayed, and then at least one first recommendation item can be displayed in the application window of the first application.
[0187] For details on the specific implementation of these two implementations, please refer to the descriptions in Implementation 1 and Implementation 2 above, which will not be repeated here.
[0188] This approach of directly displaying the first recommended item allows users to quickly access and manage the recommended files and / or folders, thus improving the user experience.
[0189] Based on the above embodiments, the implementation of determining the first recommended item for the first intent in this application will be described below with reference to specific embodiments. Figure 9 Implementation illustration of determining the recommended items provided in the embodiments of this application Figure 1 , Figure 10 Implementation illustration of determining the recommended items provided in the embodiments of this application Figure 2 , Figure 11 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 7 , Figure 12 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 8 , Figure 13 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 9 , Figure 14 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 10 .
[0190] For example, you can refer to Figure 9 To understand this, the recommended document method provided in this application includes the following steps:
[0191] S901. Obtain reference information, which may include context information.
[0192] In this embodiment, for example, the first recommended item can be determined based on reference information. Figure 9 As shown, the reference information may include context information. The implementation of the context information can be referred to the description in the above embodiments. It can also be understood that the context information is used to reflect user needs in the current scenario, and any information used to reflect the usage scenario can be used as context information in this application.
[0193] In addition, the reference information may also include user characteristic information. This user characteristic information refers to relevant information reflecting the characteristics of a user using the electronic device. For example, user characteristic information may include at least one of the following: file access information, file editing frequency, file editing duration, contact priority, and contact interaction frequency.
[0194] The file access information, file editing frequency, and file editing duration can be recorded separately for multiple files accessed by the user, reflecting the user's file operation habits. For example, file access information may include the time, location, and frequency of file access, etc.; any information related to file access can be used as file access information in this application. The contact person can refer to contacts in a third-party application or contacts recorded in an electronic device; this embodiment does not impose any limitations on this, and it can reflect the user's social relationships.
[0195] It should be understood that the specific content of user characteristic information can be expanded according to actual needs. For example, user characteristic information can also include the user's application combination habits. For instance, if a user often opens application B immediately after opening application A, this combination habit may reflect certain user needs, and therefore, the user's application combination habits can also be recorded. In actual implementation, any feature related to user operations can be used as user characteristic information in this application.
[0196] In one implementation, for example, historical user behavior data can be collected and then analyzed using machine learning algorithms to create a user profile, which includes the user characteristic information described above. Alternatively, other strategies besides machine learning algorithms can be used for data analysis to determine user characteristic information; this application does not impose any limitations on this approach.
[0197] Regarding the implementation of the various reference information described above, it can be understood that user characteristic information is continuously collected and stored during the use of the electronic device, and context information is also continuously monitored during the current use of the electronic device. Therefore, for example, reference information can be obtained based on pre-stored user characteristic information and dynamically monitored and determined context information.
[0198] S902. Based on the reference information, determine at least one first recommended item from among the multiple alternative documents.
[0199] The multiple alternative files can include all files stored in the electronic device, or the multiple alternative files can also include files contained in a specified file path, where the specified file path can be preset or manually set by the user.
[0200] For any candidate file, the file characteristic information may include at least one of the following: filename, file type, file creation time, file modification time, file size, file association information, file usage frequency, file keywords, file content, file tags, and file topic. It should be understood that the file characteristic information can also be expanded according to actual needs; any information related to the file itself can be used as the file characteristic information in this application.
[0201] In one implementation, for example, file metadata can be analyzed to obtain brief information about the file, such as the filename and file type. Alternatively, techniques such as natural language processing and semantic understanding can be used to analyze the file and determine its semantic information. This semantic information may include, for example, keywords, content, topics, tags, and so on.
[0202] After obtaining the reference information, refer to Figure 9 Based on the reference information, the recommended file and / or folder can be determined from multiple alternative files, thereby identifying the first recommended item that matches the first intent.
[0203] In one implementation, in determining at least one first recommendation based on reference information, for example, target file features can be first obtained based on the reference information, such as based on part or all of the reference information. Then, among the plurality of candidate files, files whose file feature information matches the target file features are searched to determine at least one first recommendation.
[0204] In this implementation, the technical solution of this application can integrate user characteristics, file characteristics and contextual information to determine the recommended files, thereby intelligently pushing documents to improve the accuracy of the recommended files and their matching with the actual scenario.
[0205] Based on the above introduction, the following sections will explain several implementation methods for determining at least one first recommended item from multiple candidate files based on reference information, using several scenarios as examples.
[0206] In one implementation, the context information may include content from the first interface. Then, when determining at least one first recommended item from multiple candidate files based on reference information, this can be referred to... Figure 10 The implementation of (a) in the figure:
[0207] First, the content in the first interface is analyzed to obtain the characteristics of the first target file. Then, files whose characteristics match the first target file characteristics are searched among multiple candidate files to determine the first recommended option. The target file characteristics described above can include the first target file characteristics described here.
[0208] The content in the first interface may include text, images, files, and controls, etc., and this embodiment does not limit this; it depends on the specific implementation of the first interface. For example, text can be extracted from the content of the first interface, and the extracted text can be analyzed to obtain the features of the first target file. For example, the extracted text can be analyzed using natural language processing techniques to obtain information related to file features, thereby obtaining the features of the first target file.
[0209] The extracted first target file features may include at least one of the following: filename, file type, file creation time, file modification time, file size, file association information, file usage frequency, file keywords, file content, file theme, etc. The specific implementation of the first target file features depends on the content of the first interface. This application does not limit the specific implementation of the first target file features; it can be determined according to actual needs.
[0210] The following is combined Figure 11 and Figure 12 This implementation will be introduced using two specific scenarios.
[0211] First, combine Figure 11 Let's take a scenario as an example, for instance, on the first screen... Figure 11 Taking the chat interface shown in (a) as an example, it can be understood by referring to the illustration that the first interface is the interface corresponding to the chat application, and the first interface may contain text and controls.
[0212] As illustrated in the chat scenario, a message has been sent by a contact named "Xiao Hong" stating, "I've arrived at the client's site. The client says the third version of the poster for event A is better, and they'd also like to refer to the poster they made for party B last year," and "Hurry, hurry, hurry!" The user of the device has replied with two messages: "Okay, coming right away" and "One minute."
[0213] Based on the content contained in the first interface, such as identifying the primary intent of "file sharing," it is necessary to determine the first matching recommendation item for this primary intent. For example, analysis can be performed on the content contained in the first interface to extract file-related information, thereby obtaining the first target file features. If the message content contains the text information "The third version of the poster for event A is better" and "Reference...the poster image made for party B," then, for example, the first target file features of "the third version of the event A poster" and "the poster image for party B" can be extracted. These two first target file features can be understood as features at the filename dimension.
[0214] Then, based on the characteristics of the first target file and the file characteristic information of multiple candidate files, a match can be made to find two files named "Event A Poster V3" and "Party B Poster", thus obtaining two first recommendation items that match the first intent. For example, it can then be displayed as follows: Figure 11 The interface shown in (b) displays the first prompt message and at least one first recommended item. (Regarding...) Figure 11 The details contained in (b) above can be found in the above. Figure 4 This will be understood in more detail here.
[0215] Based on this implementation of displaying at least one first recommended item, the user can then, for example, interact with the recommended files in the first interface 403 to quickly send files within the current chat window. (See reference...) Figure 11 As shown in (b), for example, if a user clicks on a file named "Event A Poster V3" in the first interface 403, it can display something like this. Figure 11 The case shown in (c) is that the file is sent in the chat window.
[0216] and reference Figure 11 As shown in (c), the user can then click on the file named "Party B Poster" in the first interface 403 to display the following... Figure 11 As shown in (d), this refers to sending the file in the chat window. Further, it can be understood that after the user has clicked on all the files in the first interface 403, the display of the first prompt message and at least one first recommended item on the first interface can be stopped to avoid the prolonged presence of the first prompt message and at least one first recommended item causing inconvenience to the user.
[0217] In addition to the methods described above for stopping the display of the first prompt message and at least one first recommended item, the display of the first prompt message and at least one first recommended item can also be stopped when any of the following stopping conditions are met. Stopping conditions may include: the display duration of the first prompt message and at least one first recommended item reaches a preset duration (e.g., stopping display after 30 seconds); the user clicks any position other than the first and second positions (this may indicate that the user does not want to process these recommended files now, so the user clicks a blank space outside these recommended positions to close the first prompt message and at least one first recommended item); the user clicks the close control in the first interface (for example, a close control can be set in the first interface so that the user can close the first window used to display recommended files and close the first prompt message as needed). In actual implementation, the conditions for stopping the display of the first prompt message and at least one first recommended item can also be set according to actual needs, and this embodiment does not limit this.
[0218] And it is also necessary to understand the above combination Figure 11 The described scenario uses the position corresponding to the first control 401 as an example. In this implementation, user operations on the file can be clicks, allowing for convenient and quick file processing. Alternatively, users can drag and drop files, moving them from the first interface to the desired location as needed.
[0219] In another implementation, where the first position corresponds to the icon of the first application, for example, the user can drag and drop a file into the chat box of the chat application window to send the file. This can also achieve the purpose of file sharing in the current usage scenario. No further illustrations will be provided here.
[0220] Secondly, combine Figure 12 To illustrate this further, for example, the first interface is... Figure 12 The email interface shown in (a) is, as can be understood from the illustration, the first interface is the interface corresponding to the email application, and the first interface may contain text and controls.
[0221] As shown in the email editing scenario illustrated in the figure, the subject of the current email is "Regarding the achievement rate of the 'Project A Acceptance Report'", and the email content is "Dear colleagues in the project team: The achievement rate of this project is 98%. Please check and review the attached 'Project A Acceptance Report'. We look forward to continuing our cooperation next time!"
[0222] Based on the content contained in the first interface, such as identifying the primary intent of "add attachment," it is necessary to determine at least one matching primary recommendation item for this primary intent. For example, analysis can be performed based on the content contained in the first interface to extract file-related information, thereby obtaining the primary target file feature. If the email subject and content contain the text information "Project A Acceptance Report," then, for example, the primary target file feature of "Project A Acceptance Report" can be extracted. This primary target file feature can be understood as a feature at the filename level.
[0223] Then, based on the characteristics of the first target file and the file characteristic information of multiple candidate files, a match can be made to find the file named "Project A Acceptance Report", and for example, the associated file "Project A Mid-term Reference", thus obtaining two first recommendation items that match the first intent. Then, for example, the following can be displayed: Figure 11 The first prompt message and the first target file feature, as shown in (a) of the document, are the first recommended items. (Related information...) Figure 11The details contained in (a) can be found above. Figure 5 This will be understood in more detail here.
[0224] Based on this implementation of displaying the first target file feature and the first recommended item, the user can then, for example, operate on the recommended files in the first interface 503 to quickly add attachments to the current email window. (See reference...) Figure 11 As shown in (a) above, for example, if a user clicks on a file named "Project A Acceptance Report" in the first interface 503, the following will be displayed: Figure 11 The case shown in (b) is that the file is added as an attachment in the email.
[0225] Furthermore, after adding an attachment, if another recommended file is one the user doesn't want to add, the user can, for example, click on it in the interface. Figure 11 The position shown in (b) (that is, the position indicated by the mouse cursor) is then referred to. Figure 11 As shown in (c), for example, the display of the first prompt message and the first recommended item of the first target file feature can be stopped in the first interface to avoid the long-term existence of the first prompt message and the first recommended item of the first target file feature causing inconvenience to the user.
[0226] Similar to the methods described above, besides stopping the display of the first prompt message and the first recommended item of the first target file feature, it is also possible to stop displaying the first prompt message and the first recommended item of the first target file feature when any of the stopping conditions are met. The implementation of the stopping conditions is similar to that described above and will not be repeated here.
[0227] And it is also necessary to understand the above combination Figure 12 The described scenario uses the position corresponding to the first control 501 as an example. In this implementation, user operations on the file can be clicks, allowing for convenient and quick file processing. Alternatively, users can drag and drop files, moving them from the first interface to the desired location as needed.
[0228] In another implementation, where the first position corresponds to the icon of the first application, for example, users can drag and drop files into the email application window to add attachments. This also achieves the purpose of file sharing in the current usage scenario. No further illustrations are provided here.
[0229] It is understandable that the above combination Figure 11 and Figure 12In the two examples presented, the first target file features obtained are both filename-level features. However, in actual implementation, features from various possible dimensions may be obtained. For example, continuing with the chat scenario described above, if a contact sends a message saying, "Send me all the PDF files related to project A," then based on this message content, we can extract the file topic dimension "related to project A" and the file type dimension "PDF." Other possible scenarios will not be illustrated here, but it is understood that the extracted first target file features depend on the actual scenario, and this embodiment does not impose any limitations on this.
[0230] In this implementation, electronic devices can accurately identify a user's intent in chat or email scenarios, for example, using natural language processing techniques. Compared to existing technologies, this implementation can more accurately capture user needs, reducing misjudgments and manual operations. Furthermore, the electronic device not only identifies user intent but also extracts keywords from the window content and combines these keywords for file matching. This improvement makes file recommendations more accurate, reducing the time users spend manually searching.
[0231] In another implementation, the context information may include the location of the electronic device and / or the current time. Then, when determining at least one first recommended item from multiple alternative files based on reference information, it can be referred to... Figure 10 The implementation of (b) shown in the figure:
[0232] Based on the above introduction, it can be determined that user characteristic information can include file access information. Based on this access information, information such as the time, location, and frequency of file access can be determined. Therefore, for example, a second target file characteristic can be determined based on the user characteristic information. This second target file characteristic is the file characteristic corresponding to files whose access behavior within a first time period and / or a first geographical area meets the first condition. The target file characteristics described above can include the second target file characteristic described here. The access behavior meeting the first condition can include, for example, at least one of the following: the number of times the file is opened is greater than a first threshold, the editing time is greater than a first duration threshold, and the foreground display time of the file is greater than a second duration threshold. The specific access behavior and the first condition can be set according to actual needs. The following description uses the case where the number of times the file is opened is greater than the first threshold as an example.
[0233] For example, if a user opens file A in the first position for N consecutive days, then the file characteristics of file A can be used as the second target file characteristics described here. As another example, if a user opens .txt files at 7:30 PM for N consecutive days (for instance, opening files B, C, and D, all of which are .txt), then the file type ".txt" can be used as the second target file characteristics described here. The second target file characteristics are similar to those described above and can include at least one of the following: filename, file type, file creation time, file modification time, file size, file association information, file usage frequency, file keywords, file content, file theme, etc.
[0234] For example, different feature selection strategies can be set for different situations. In the first situation described above, the filename is determined as the second target file feature, and in the second situation described above, the file type is determined as the second target file feature. The correspondence between the situation and the feature selection strategy can be determined according to actual needs. Alternatively, user feature information can be analyzed based on machine learning or neural networks to extract the second target file feature. This application does not limit the implementation of determining the second target file feature.
[0235] Furthermore, the location of the electronic device and / or the current time can be determined based on contextual information. If the current time belongs to a first time period, and / or if the location of the electronic device belongs to a first geographical area, files whose file feature information matches the features of the second target file can be searched among multiple candidate files to determine at least one first recommendation.
[0236] The following is combined Figure 13 This implementation will be introduced using a specific scenario.
[0237] For example, the desktop interface may have an application window 702 of a first application, and the application window 702 may include a first area 703, which is used to display at least one recommendation.
[0238] For example, assuming a user opens file 1 around 19:30 for three consecutive days, the file characteristics of file 1 can be identified as the second target file characteristics, and the time period from 19:20 to 19:40 can be identified as the first time period. Then, based on the current time contained in the context information, if the current time belongs to the first time period described above, the first intent of "file access" can be identified, for example. Then, for this first intent, files corresponding to the second target file characteristics are matched among multiple candidate files, thereby determining file 1 as the recommended file. Then, file 1 can be displayed in the first area 703, where file 1 is also the first recommended item corresponding to the first intent.
[0239] And it is also necessary to understand the above combination Figure 13 The scenario described above is based on displaying the first recommended item in the application window. In reality, the first recommended item can also be displayed in the first position, and the implementation is similar, so it will not be elaborated here.
[0240] In this implementation, personalized file recommendations can be provided by combining the time patterns and geographical locations reflected in user characteristic information.
[0241] In another implementation, the context information can include user actions within a second time period prior to the current time. For example, an electronic device can record and analyze user actions at preset intervals, where the second time period could be, for instance, 5ms. The electronic device could record user actions at 5ms intervals and analyze the user actions within those 5ms.
[0242] Subsequently, when determining at least one primary recommendation from multiple alternative documents based on the reference information, one can refer to... Figure 10 The implementation of (b) shown in the figure:
[0243] First, based on user actions during the second time period, the characteristics of the third target file are determined. It can be understood that user actions can, to some extent, reflect the user's intentions regarding file operations; therefore, the characteristics of the third target file can be determined based on user actions during the second time period. For example, the characteristics of the third target file can be determined based on a pre-defined relationship table between user actions and file characteristics. The pre-defined relationship table can be understood with reference to the following table:
[0244]
[0245] As can be understood from the table above, user actions can reflect the user's underlying intentions. Therefore, for example, the characteristics of the third target file corresponding to the user actions can be determined based on the table. Furthermore, the table above only shows some possible implementation methods; this application does not limit the specific implementation of the preset relationship table, which can be determined according to actual needs.
[0246] Alternatively, in another implementation, user operations during the second time period can be analyzed based on machine learning or neural networks to extract features of the third target file. This application does not limit the implementation of determining the features of the third target file.
[0247] Furthermore, files whose file feature information matches the features of the third target file can be searched among multiple candidate files to determine at least one first recommendation.
[0248] The following is combined Figure 14 This implementation will be introduced using a specific scenario.
[0249] For example, the desktop interface may have an application window 702 of a first application, and the application window 702 may include a first area 703, which is used to display recommended files.
[0250] For example, assuming the user's actions during the second time period include taking consecutive screenshots, then based on the above description, a primary intent to share files can be identified. Subsequently, at least one primary recommendation needs to be determined based on this primary intent. For example, multiple screenshots obtained from multiple screenshot operations can be used as at least one primary recommendation. (Refer to...) Figure 14 As shown, multiple screenshots can be stacked for display, resulting in a screenshot collection as illustrated. This method of stacking multiple images saves recommendation space. For example, the number of screenshots currently stacked can also be displayed on the stacked screenshots to help users quickly understand the screenshot content.
[0251] Furthermore, to facilitate user processing of multiple screenshots, this application can also stitch multiple screenshots together to obtain a stitched file. For example, the stitched file can be in PDF format. Therefore, at least one first recommended item can, for example, include the stitched file, that is, it can display as shown... Figure 14 The screenshot shown is a .pdf file. In actual implementation, the format of the stitched file can be set according to actual needs, and it can be any image format, for example. Furthermore, since the stitched file is automatically generated by the system, it can be automatically deleted if no user action is detected on it within a certain period after it is recommended, to avoid unnecessary encroachment on storage space.
[0252] exist Figure 14 In the example, it can be understood that both the screenshot collection and the screenshot.pdf file are actually the first recommended items.
[0253] And it is also necessary to understand the above combination Figure 14 The scenario described is based on the implementation of displaying at least one first recommended item in the application window of the first application. In fact, the first recommended item can also be displayed in the first position, and the implementation is similar, so it will not be described in detail here.
[0254] This allows for the identification of consecutive screenshots of similar content (such as meeting screenshots, tutorial step screenshots, etc.) taken by the user, and the packaging (stacking) of these screenshots for presentation. Compared to implementations that present screenshots as individual files, which lack the identification and packaging processing of the correlation between consecutive screenshots, the technical solution of this application significantly improves the efficiency of screenshot management through intelligent packaging and stacking.
[0255] Referring to the embodiments described above, it can be determined that the various implementation methods described above all determine the target file characteristics based on reference information, and then match the target file characteristics with the file characteristic information of multiple candidate files to determine at least one first recommendation. For example, matching can be performed based on file characteristics, and then files with a matching degree greater than a preset threshold can be used as recommended files, thereby determining at least one first recommendation.
[0256] In addition, the following will be combined Figure 15 Another implementation for determining the first recommendation is introduced. Figure 15 Implementation illustration of determining the recommended items provided in the embodiments of this application Figure 3 .
[0257] like Figure 15 As shown, in another implementation, the technical solution of this application can also be implemented in the following way:
[0258] Reference information and a first intent are input into a first model, causing the first model to output at least one recommended file that matches the first intent. The reference information can be understood by referring to the description in the above embodiments, and will not be repeated here.
[0259] For example, in this embodiment, the first intent can also be input into the first model so that the first model can output recommended files that match the first intent. Alternatively, in some implementations, the reference information may not need to include the first intent, and the first model may analyze the user's intent based on context information to make file recommendations.
[0260] Furthermore, the first model can be a pre-trained model. For example, the first model can analyze the reference information and the first intent to output at least one recommended file that matches the first intent. The output of the first model can be, for example, a file identifier (such as a filename) of the recommended file. The specific implementation details of the first model can be set according to actual needs, as long as the purpose described here is achieved.
[0261] Subsequently, at least one first recommendation can be determined based on at least one recommendation file output by the first model. For example, all at least one recommendation file output by the first model can be determined as first recommendation items. Alternatively, at least one recommendation file output by the first model can be further integrated or reduced according to certain rules to obtain at least one first recommendation item, and this application does not impose any restrictions on this.
[0262] and reference Figure 15 It is also understood that this application may collect user feedback on the recommended documents, whereby the feedback information indicates the relationship between the recommended documents and the first intent (e.g., whether they are related or the degree of relatedness). Subsequently, for example, new training data may be constructed based on the feedback information, and the first model may be adjusted based on the new training data to improve the accuracy of the recommended documents output by the first model during device use.
[0263] The specific implementation of collecting feedback information can be set according to actual needs. For example, corresponding controls can be set in the user interface to collect user feedback information on recommended files. This embodiment does not limit the specific setting form of the controls.
[0264] Based on this implementation method, the recommendation files can be determined simply and effectively, and by using model prediction, the technical solution of this application can be applied to various possible scenarios.
[0265] Understandably, because the first model's prediction of recommended items relies heavily on reference data, for a new electronic device, in the early stages when there isn't much user feature data, directly predicting recommended items based on the first model might not guarantee accuracy. Therefore, in one implementation, during the stage of collecting user feature data (e.g., within a specific time period after the electronic device is activated), the feature matching method described above can be used to determine the recommended items. After collecting sufficient user feature data (e.g., after a specific time period after the electronic device is activated), the model processing method described above can be used to determine the recommended items.
[0266] Alternatively, either the feature matching method or the model processing method described above can be consistently used to determine the recommendation items; this embodiment does not impose any restrictions on this. In summary, both methods described above can achieve the determination of recommendation items.
[0267] The above embodiments describe the relevant implementation for determining a matching first recommendation item based on a first intent. Based on the above description, it can be determined that, regardless of whether the method is based on target file feature matching or inference using a first model, multiple recommendation files can be determined, and then at least one first recommendation item can be determined based on these recommendation files. In one implementation, a corresponding target score can be determined for each of these recommendation files, and then the display format of the first recommendation item can be determined based on these target scores.
[0268] The target score for the recommended document can be determined, for example, based on the importance of the primary intent and the degree of matching between the features of the recommended document and the target document.
[0269] In one implementation, determining the target score for the recommendation file can be understood by referring to the following steps:
[0270] 1. For any given recommended document, obtain its initial rating.
[0271] The recommended file can be determined, for example, by matching the characteristics of the target file with the file feature information of multiple candidate files. In one implementation, for example, the initial score of the recommended file can be determined based on the degree of matching between the file feature information of the recommended file and the characteristics of the target file. The initial score and the degree of matching can be proportional, for example, and the specific correspondence between these two parameters can be set according to actual needs; this embodiment does not limit this.
[0272] Alternatively, the recommended files could be the output of the first model. In one implementation, the initial ratings for the recommended files could be determined based on the first model. For example, in addition to outputting at least one recommended file (file identifier), the first model could also output at least one initial rating for each recommended file.
[0273] For example, each recommended file can have an initial score greater than a preset threshold. In other words, only files with an initial score greater than a certain threshold will be identified as recommended files. This can improve the relevance between the first recommended item and the first intention displayed later.
[0274] 2. Adjust the initial score of the recommended document based on the first parameter of the first intent to determine the target score of the recommended document. The first parameter is used to indicate the importance of the first intent.
[0275] After determining the initial score of the recommended document, the initial score can be further adjusted based on the first parameter corresponding to the first intent. The first parameter indicates the importance or urgency of the first intent. For example, the higher the importance of the first intent, the larger the first parameter can be, in order to subsequently improve the target score of the corresponding recommended document.
[0276] The importance of the primary intent can be determined based on contextual information, such as the preceding text. Figure 11 In the described scenario, when identifying the primary intent, it can be determined that the application window contains the message "Hurry, hurry, hurry, urgent!". Based on this message content, it can be determined that the primary intent is of high importance, and therefore the primary parameter corresponding to the primary intent can be used to increase the initial score. For example, the primary parameter can be multiplied by the initial score to adjust the score. Alternatively, it can be added to adjust the score; this embodiment does not limit the specific adjustment strategy.
[0277] It's understandable that in real-world electronic device usage scenarios, multiple user intentions might exist simultaneously. The above describes the implementation of displaying recommended items corresponding to the first intention. Therefore, when multiple intentions exist, for example, recommended items corresponding to each intention can be displayed separately. For instance, the target rating of the displayed recommended items can be set to be sorted from high to low, thus displaying recommended items for more important intentions first, making it easier for users to browse and operate.
[0278] For example, this can be combined with Figure 16 To understand, Figure 16 This is a schematic diagram of an interface displaying multiple recommended items, provided for an embodiment of this application.
[0279] like Figure 16 As shown in (a), for example, the application window of the first application can be displayed in the first interface, and the application window includes a first area 703 as shown in the figure (refer to the above for understanding, and will not be repeated here). The first area 703 is used to display at least one recommendation item corresponding to multiple intentions, as shown in the figure. For example, the first area 703 may include file 1, file 2, file 3 and file 4.
[0280] For example, files 1 and 2 could be recommendations matching intent A, file 3 could be a recommendation matching intent B, and file 4 could be a recommendation matching intent C. And in the example in the diagram, for instance, file 1 corresponds to a target score of 95, file 2 to a target score of 90, file 3 to a target score of 85, and file 4 to a target score of 81. Then, referring to... Figure 4It can be confirmed that the order of these four recommended documents in the first region 703 is: document 1 → document 2 → document 3 → document 4, which means that the target scores of these four recommended documents are sorted from high to low.
[0281] This allows for the display of recommendations corresponding to multiple intents within the first area, facilitating quick searching for relevant recommendations. Furthermore, sorting by target rating from highest to lowest improves the speed at which users find recommendations for their most pressing intent.
[0282] Furthermore, it can be understood that the importance of an intent actually decreases over time. For example, if a user currently wants to find a file corresponding to intent A, the recommended file for intent A might have the highest target score at this point. However, after 10 minutes, the user's intent B might become a more important intent, and intent A would no longer be as important. However, if the recommended file for intent A still maintains a high target score, it will consistently occupy a prominent position in the first area. Therefore, this application also proposes the following strategy:
[0283] Using a preset duration as the cycle, the target score corresponding to each recommendation item is reduced by the first magnitude.
[0284] For example, the preset duration can be 5 minutes, and the initial increment can be 3. This means that every 5 minutes, the target rating for each recommended item on the first screen will be reduced by 3. In this way, the recommended items displayed in the first area will always be those that the user needs most.
[0285] For example, you can refer to Figure 16 To understand, for example in Figure 16 Based on the content shown in (a), after a preset time period, the recommended items displayed in the first area 703 become... Figure 16 The situation is shown in (b) above. Referring to the above description, after a preset time period, the target score of each recommendation item in the first area needs to be reduced by a first increment, assuming the first increment is 3. Then, referring to... Figure 16 As shown in (b), the target score for document 1 is adjusted to 92 points, the target score for document 2 is adjusted to 87 points, and the target score for document 3 is adjusted to 82 points.
[0286] And assuming that within this preset time period, the electronic device recognizes a new intent D, and the first recommended item corresponding to intent D is document 5, and the target score of document 5 is, for example, 95 points, then document 5 can be sorted at the top, so that users can quickly find the recommended document corresponding to the new intent D.
[0287] Therefore, based on this processing method, the scores of the recommended items corresponding to each intent can be dynamically adjusted so that the recommended items corresponding to newly identified intents can be displayed in the first area for recommendation.
[0288] Based on the description of the above embodiments, it can be understood that the technical solution of this application can display the first recommended item in a first location or application window. The location of the recommended item in the first location or application window can be set according to actual needs. However, the location of the recommended item in the first location or application window is always limited. Therefore, the recommended files can also be integrated in the following ways in this application.
[0289] The following description will still take the first recommended item corresponding to the first intent as an example. In one implementation, the first recommended item may include at least one of the following implementations: a first folder in the form of a folder, a first file in the form of a file, and a first superimposed item in the form of a combination of folders and files.
[0290] For example, you can refer to Figure 17 and Figure 18 To understand, Figure 17 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 10 one, Figure 18 This is a schematic diagram illustrating the implementation of the form of the recommended item provided in the embodiments of this application.
[0291] For example, the first recommended item can be a file as shown in 1701, meaning the first recommended item is File 1 as shown in the figure (also known as the first file mentioned above). The first recommended item can also be a folder as shown in 1702, meaning the first recommended item is Folder K as shown in the figure (also known as the first folder mentioned above). Furthermore, the first recommended item can also be a combination of a folder and a file as shown in 1703 (meaning the icons of the file and folder are superimposed), meaning the first recommended item is File 2 + Folder L as shown in the figure (also known as the first combined item mentioned above).
[0292] The following explanation addresses each of these three scenarios:
[0293] Scenario 1:
[0294] If there are multiple recommended files for a given first intent, and the difference in the target scores of the multiple recommended files is less than a first threshold, and the multiple recommended files belong to a first folder, then at least one first recommended item includes the first folder.
[0295] The difference between the target scores of multiple recommendation files is less than the first threshold, which can be understood as the difference between the target scores of any two recommendation files being less than the first threshold.
[0296] For example, you can refer to Figure 18 Let's interpret (a) as follows: For example, there are four files as shown in the figure, namely file 1, file 2, file 3, and file 4, for the first intent. Among them, the target score of file 1 is 81 points, the target score of file 2 is 82 points, the target score of file 3 is 80 points, and the score of file 4 is 80 points. Assuming that the first threshold is set to 5, it can be understood that the difference in target scores between these four recommended files is less than the first threshold (5).
[0297] Meanwhile, for example, if all four recommended files belong to folder A, then referring to the diagram, folder A can be directly identified as the first recommended item, where folder A is also the first folder introduced in this embodiment.
[0298] In this way, compared to the method of needing 4 recommendation slots to display these 4 recommended files in sequence, directly displaying the first folder can effectively save recommendation slots while still effectively recommending relevant files to users.
[0299] In this implementation, when the recommended items are in the form of folders, for example, the target ratings of multiple recommended files can be averaged to determine the target rating corresponding to the recommended item, which makes it easier to determine the display position of the recommended item later.
[0300] Scenario 2:
[0301] If there are multiple recommended files for the first intent, and if there is a target file among the multiple recommended files with a target score greater than or equal to the third threshold, and the difference in target scores among the remaining recommended files excluding the target file is less than the first threshold, and the multiple recommended files belong to the first folder, then at least one first recommended item includes a first superimposed item in superimposed form, and the first superimposed item includes the target file and the first folder.
[0302] The understanding of the difference in target scores among the remaining recommendation files besides the target file is similar to the understanding of the difference in target scores among multiple recommendation files described above, and will not be repeated here.
[0303] For example, you can refer to Figure 18To understand (b) in the diagram, for example, there are four files as shown in the figure for the first intent: file 1, file 2, file 3, and file 4. File 1 has a target score of 99, file 2 has a target score of 82, file 3 has a target score of 80, and file 4 has a target score of 80. Assuming the third threshold is set to 95 and the first threshold remains at 5, it can be understood that the target score of file 1 is greater than the third threshold (95), and the difference between the target scores of the other three recommended files (excluding file 1) is less than the third threshold (5).
[0304] This means that among the multiple recommended files, there exists one file with a very high rating (i.e., the target file mentioned above), and several other recommended files with similar ratings. Furthermore, for example, if all four recommended files belong to folder A, then referring to the diagram, the superimposed item of file 1 + folder A can be directly determined as the first recommended item, where the superimposed item of file 1 + folder A is also the first superimposed item described in this embodiment.
[0305] In this way, compared to requiring four recommendation slots to display the four recommended files sequentially, displaying the first overlay item effectively saves recommendation slots while still effectively recommending relevant files to the user. Furthermore, the top-rated target file can be directly shown in the first recommendation item, allowing users to quickly understand that the target file is the first recommendation.
[0306] In this implementation, when the recommended item is a combination of the target file and the first folder, for example, the target scores of multiple recommended files can be averaged to determine the target score corresponding to the recommended item. This makes it easier to determine the display position of the recommended item later.
[0307] Furthermore, it can respond to user actions on different locations within the first overlay item, thereby handling different objects. For example, if the user action is on a folder within the overlay item, then operations on the folder, such as opening the folder, can be implemented. Similarly, if the user action is on a file within the overlay item, then operations on the file, such as opening, moving, or sending the file, can be implemented.
[0308] Scenario 3:
[0309] If there are multiple recommended files for a given first intent, and if the multiple recommended files do not belong to the same folder, then at least one first recommended item includes a first file, which is a file contained in the multiple recommended files; or, if there is a single recommended file for a given first intent, at least one first recommended item includes a first file, and the single recommended file is also the first file.
[0310] For example, you can refer to Figure 18 To understand (c) in this context, consider the following scenario: For example, given two files as shown in the diagram, File 1 and File 2, the target score for File 1 is 81 points, and the target score for File 2 is 82 points. Also, if these two recommended files belong to different folders, such as File 1 in folder A and File 2 in folder B, then integrated recommendation is not supported. Therefore, referring to the diagram, File 1 and File 2 can be used as separate first recommendation items.
[0311] Alternatively, if there is only a single recommended document for the first intent, then there is no need to integrate them. We can simply determine that single recommended document as the first recommended item. The single recommended document mentioned here is the first document mentioned above.
[0312] Furthermore, when multiple recommended files do not belong to the same folder, it's possible that some files within multiple recommended files may belong to the same folder. For example, consider three recommended files: file a1, file a2, and file b. Files a1 and a2 belong to folder A, while file b belongs to folder B. In this case, we can further calculate the proportion of recommended files belonging to folder A relative to the total number of files in folder A. If this proportion exceeds a certain threshold, folder A can be designated as the first recommended item. Otherwise, the multiple recommended files belonging to folder A can be treated as separate first recommended items.
[0313] For example, if the threshold for the weighting is set to 50%, in the example described above, we can determine that the number of recommended files belonging to folder A is 2 (file a1 and file a2), and assuming the total number of files in folder A is 3, then the weighting is determined to be 2 / 3. This weighting is greater than the threshold of 50%. Therefore, folder A can be considered as the first recommended item. As another example, assuming the total number of files in folder A is 10, then the weighting is determined to be 2 / 10. This weighting is less than the threshold of 50%. Therefore, file a1 and file a2 can each be considered as the first recommended items.
[0314] Furthermore, in the example described above, since file b does not belong to the same folder as the other files, file b can be considered as a first recommended item on its own.
[0315] Further reference Figure 17It is also understood that, when displaying the first recommended item, the technical solution of this application may, for example, display the recommendation reason corresponding to the first recommended item in the surrounding position, so as to help users understand why these files are recommended. The recommendation reason may be, for example, output by the first model, or it may be determined based on the target file features during file matching; this application does not limit this.
[0316] and reference Figure 17 It can also be understood that, for example, this application can also set in the application window (such as the first area) of the first application, such as... Figure 17 The control 1704 shown is used to adjust the file range corresponding to multiple candidate files. For example, when "Global Recommendation" is selected as shown, all files in the electronic device can be selected as candidate files as described above. As another example, when "Recommendation Only in Current Path" is selected as shown, files in the file path corresponding to the current application window can be selected as candidate files as described above. In this way, users can set the specific range of files to recommend according to their actual needs, thereby improving the configurability of file recommendation.
[0317] Furthermore, control 1704 can also include a "Disable Recommendations" option as shown in the figure. For example, in response to the operation applied to this option, the function of recommending files can be turned off, meaning the electronic device will no longer execute the technical solution described above. In this way, users can choose whether or not to perform file recommendations based on their actual needs.
[0318] In actual implementation, the control form of the related adjustment function corresponding to control 1704 can be expanded according to actual needs, and is not limited to the form described above. This embodiment does not impose any restrictions on this.
[0319] The above embodiments describe the implementation of displaying a first recommended item matching the first intent when the first intent is recognized. Referring to the above embodiments, this application can set a first area within the application window of a first application. This first area is specifically used to display recommended files and / or recommended folders. In other words, this first area can exist as a fixed entry point for recommended files, allowing users to easily access recommended files within this area.
[0320] In another scenario, if no intent is detected, recommended files and / or recommended folders can be displayed in the first area in a certain way to avoid confusion for the user due to the absence of content in the first area.
[0321] For example, if no intent is identified, a second recommended item can be determined according to fixed rules, and then displayed in the first area. See [reference needed]. Figure 19 Understanding fixed rules Figure 19 A schematic diagram of the recommended interface for the documents provided in this application embodiment. Figure 10 two.
[0322] like Figure 19 As shown, for example, if no intent is recognized, a second recommended item as shown in the figure can be displayed in the first area 703 (for details, please refer to the above embodiment, which will not be repeated here). The second recommended item includes four files, namely file 7, file 8, file 9 and file 10.
[0323] For example, one fixed rule could be the number of times a file is viewed. For instance, file 7 has been viewed 12 times, making it the most viewed file, and thus file 7 could be designated as the second recommended file. Another fixed rule could be the longest editing time. For instance, file 8 has been edited for 21 hours, making it the longest edited file, and thus file 8 could be designated as the second recommended file. Yet another fixed rule could be the number of times a file is shared. For instance, file 9 has been shared 4 times, making it the most shared file, and thus file 9 could be designated as the second recommended file. Yet another fixed rule could be the most recent viewing time. For instance, file 10 was viewed one minute ago, making it the most recently viewed file, and thus file 10 could be designated as the second recommended file.
[0324] In other words, multiple fixed rules can be set, and then recommended files are selected based on these fixed rules to determine the second recommendation. Alternatively, the fixed rules can also be: select the top N files in terms of view count / edit time / share count / view time as the second recommendation, and then sort them in the first area in descending order of view count / edit time / share count / view time to display multiple second recommendations. Alternatively, the fixed rules can be set according to other dimensions, which is not limited in this embodiment, and can be set arbitrarily according to actual needs.
[0325] In summary, the technical solution of this application can integrate user characteristics, file features and contextual information to achieve intelligent document push, thereby improving file processing efficiency and user experience.
[0326] The technical solution of this application can analyze users' historical behavior, work habits, and personalized needs, capture users' intentions in real time, and push relevant documents accordingly. The system can identify the work patterns of different users and customize recommendation strategies based on their preferences and needs, providing personalized recommendation results.
[0327] Furthermore, unlike traditional recommendation methods based on file metadata, this invention deeply analyzes the content and characteristics of files (such as file type, file tags, and file content semantics) to ensure that recommended files are highly relevant to the user's current tasks, providing more accurate push results. Simultaneously, the technical solution of this application proactively perceives user needs by combining contextual information such as the user's current work environment, ongoing tasks, and applications being used, as well as time patterns and habits, and automatically adjusts the recommendation strategy according to different application scenarios, ensuring a high degree of alignment between recommended files and user needs.
[0328] Furthermore, the strategy mentioned in this application's technical solution also possesses self-learning capabilities, enabling dynamic optimization based on users' constantly changing needs, work habits, and scenarios. As usage time increases, the system will increasingly accurately predict users' file needs, reducing false and duplicate recommendations, and improving the accuracy of the recommendation system and user satisfaction.
[0329] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.
[0330] In one implementation, this application provides an electronic device. Figure 20 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application.
[0331] like Figure 20 As shown, the electronic device 200 includes: a processor 2001 and a memory 2002; the memory 2002 stores computer execution instructions; the processor 2001 executes the computer execution instructions stored in the memory 2002, causing the electronic device 200 to perform the above-described method.
[0332] When the memory 2002 is set up independently, the electronic device also includes a bus 2003 for connecting the memory 2002 and the processor 2001.
[0333] This application provides a chip. The chip includes a processor, which is used to call a computer program in memory to execute the technical solutions in the above embodiments. Its implementation principle and technical effects are similar to those in the related embodiments described above, and will not be repeated here.
[0334] This application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, it implements the methods described above. The methods described in the above embodiments can be implemented wholly or partially by software, hardware, firmware, or any combination thereof. If implemented in software, the functionality can be stored as one or more instructions or code on or transmitted over the computer-readable medium. The computer-readable medium can include computer storage media and communication media, and can also include any medium that can transfer a computer program from one place to another. The storage medium can be any target medium accessible by a computer.
[0335] In one possible implementation, a computer-readable medium may include random access memory (RAM), read-only memory (ROM), compact discread-only memory (CD-ROM) or other optical disc storage, magnetic disk storage or other magnetic storage devices, or any other medium targeted to carry or to store required program code in the form of instructions or data structures, and accessible by a computer. Furthermore, any connection is appropriately referred to as a computer-readable medium. For example, if software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. As used herein, disks and optical discs include optical discs, laser discs, optical discs, digital versatile discs (DVDs), floppy disks, and Blu-ray discs, where disks typically reproduce data magnetically, while optical discs optically reproduce data using lasers. Combinations of the above should also be included within the scope of computer-readable media.
[0336] This application provides a computer program product, which includes a computer program that, when run, causes the computer to perform the above-described method.
[0337] It should be noted that the modules or components described in the above embodiments can be one or more integrated circuits configured to implement the above methods, such as one or more application-specific integrated circuits (ASICs), one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs), etc. Furthermore, when a module is implemented through processing element scheduler code, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processors capable of calling program code, such as a controller. Additionally, these modules can be integrated together to implement a system-on-a-chip (SOC).
[0338] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state disk (SSD)).
[0339] The term "multiple" in this document refers to two or more. The term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Furthermore, the character " / " in this document generally indicates an "or" relationship between the preceding and following related objects; in formulas, " / " indicates a "division" relationship. Additionally, it should be understood that in the description of this application, words such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.
[0340] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application.
[0341] It is understood that, in the embodiments of this application, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
Claims
1. A document recommendation method, characterized in that, Applied to electronic devices, the method includes: Display the first interface; Upon recognizing a first intent, at least one first recommendation item is displayed, wherein the first recommendation item is a file and / or folder that matches the first intent; wherein the first intent is recognized based on contextual information, the contextual information including at least one of the following: content displayed on the first interface, user actions performed on the first interface, the location of the electronic device, and the current time.
2. The method according to claim 1, characterized in that, The display of at least one first recommended item includes: At least one first recommendation item is displayed at a first position corresponding to the first element, wherein the first element is an identifier of a first application or a first control, the first application is an application for file management, and the first control is a control for adding files.
3. The method according to claim 2, characterized in that, The method further includes: A first prompt message is displayed at a second position corresponding to the first element, the first prompt message being used to indicate the existence of recommended files and / or folders; Displaying the at least one first recommended item at a first position corresponding to the first element includes: In response to a first operation applied to the first element, at least one first recommended item is displayed at a first position corresponding to the first element.
4. The method according to claim 3, characterized in that, The first operation is to place the cursor on the first element.
5. The method according to claim 1, characterized in that, The display of at least one first recommended item includes: In response to a second operation applied to the first element, an application window of the first application is displayed, in which at least one first recommendation item is displayed, the first element being an identifier or a first control of the first application, the first control being a control for adding files, and the first application being an application for file management.
6. The method according to any one of claims 1-5, characterized in that, The method further includes: Based on reference information, at least one first recommended item is determined from multiple alternative files, the reference information including the context information.
7. The method according to claim 6, characterized in that, The reference information also includes user characteristic information.
8. The method according to claim 6 or 7, characterized in that, The step of determining at least one first recommended item from multiple candidate files based on the reference information includes: Based on some or all of the reference information, obtain the characteristics of the target file; Among the multiple candidate files, files whose file feature information matches the target file feature are searched to determine at least one first recommendation.
9. The method according to claim 8, characterized in that, The context information includes the content of the first interface, and the target file features include the first target file features obtained based on the analysis of the content of the first interface.
10. The method according to claim 8, characterized in that, The reference information includes user characteristic information, and the context information includes the location of the electronic device and / or the current time; The target file features include a second target file feature determined based on the user feature information. The second target file feature is the file feature corresponding to a file whose access behavior within a first time period and / or within a first geographical range meets the first condition. The step of searching among the multiple candidate files for files whose file feature information matches the target file features to determine the at least one first recommendation includes: If the current time falls within the first time period, and / or if the location of the electronic device falls within the first geographical area, a file whose file feature information matches the second target file feature is searched among the plurality of candidate files to determine the at least one first recommendation.
11. The method according to claim 8, characterized in that, The context information includes user actions performed during a second time period prior to the current time; The target file features include a third target file feature determined based on user actions during the second time period.
12. The method according to claim 11, characterized in that, The user operations during the second time period include multiple screenshot operations, and the at least one first recommendation item includes multiple screenshots obtained from the multiple screenshot operations; and / or, the at least one first recommendation item includes a file obtained by stitching together the multiple screenshots.
13. The method according to any one of claims 6-12, characterized in that, The step of determining at least one first recommended item from the plurality of candidate files based on the reference information includes: The reference information and the first intent are input into the first model so that the first model outputs at least one recommended file that matches the first intent; The at least one first recommendation is determined based on the at least one recommendation document.
14. The method according to claim 13, characterized in that, The at least one first recommendation includes at least one of the following implementations: a first folder in the form of a folder, a first file in the form of a file, and a first overlay item in the form of an overlay of folders and files.
15. The method according to claim 14, characterized in that, There are multiple recommended files for the first intent, the multiple recommended files belong to a first folder, and the at least one first recommended item includes the first folder in the form of the folder, or includes a first superimposed item in the form of a superposition of the folder and the file; or, For the first intent, there are multiple recommended files, which do not belong to the same folder, and the at least one first recommended item includes a first file in the file format, wherein the first file is one of the multiple recommended files; or, For the first intent, there exists a single recommended file, and the at least one recommended item includes a first file in the form of the file, wherein the single recommended file is the first file.
16. The method according to claim 15, characterized in that, Each recommendation item corresponds to a target score; The target score of the recommendation item is determined based on the importance of the first intent and / or the degree of matching between the recommended documents included in the recommendation item and the features of the target document.
17. The method according to claim 16, characterized in that, In the presence of multiple intents, each intent corresponds to at least one recommendation item; wherein, when displaying multiple recommendation items corresponding to the multiple intents, the target ratings of the multiple recommendation items are sorted from high to low.
18. The method according to any one of claims 16-17, characterized in that, The method further includes: The target score corresponding to each of the at least one recommended item is reduced by a first magnitude, with a preset time period as the cycle.
19. The method according to any one of claims 7-18, characterized in that, The user characteristic information includes at least one of the following: file access information, file editing frequency, file editing duration, contact priority, and contact interaction frequency; The file feature information includes at least one of the following: file name, file type, file creation time, file modification time, file size, file association information, file usage frequency, file keywords, file content, and file theme.
20. The method according to any one of claims 1-19, characterized in that, The first intention is any of the following: file editing, file searching, file sharing, adding attachments, information collection, and file collaboration.
21. The method according to any one of claims 2-20, characterized in that, In the case where the first intent is any of the following: file sharing, adding attachments, file collaboration, the first element is the first control; The method further includes: In response to a third operation applied to a target recommendation item among the at least one first recommendation item, the function corresponding to the first control is executed on the target recommendation item.
22. An electronic device, characterized in that, The electronic device includes: one or more processors and memory; The memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the electronic device to perform the method as described in any one of claims 1-21.
23. A chip system, characterized in that, The chip system is applied to an electronic device, the chip system including one or more processors, the one or more processors being used to invoke computer instructions to cause the electronic device to perform the method as described in any one of claims 1-21.
24. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes computer instructions that, when executed on an electronic device, cause the electronic device to perform the method as described in any one of claims 1-21.
25. A computer program product, characterized in that, The computer program product includes computer program code that, when run on an electronic device, causes the electronic device to perform the method as described in any one of claims 1-21.
