Interaction method and electronic equipment

Abstract

The invention discloses an interaction method and electronic equipment, and the method comprises the steps: outputting an interaction interface, and displaying at least one feature node in the interaction interface; each feature node is determined based on at least one fact node associated with the feature node, the feature node comprises description of personalized information reflected by the at least one fact node associated with the feature node, and the fact node comprises description of at least one semantic unit in the source data; in response to a first trigger operation on a target feature node in the interactive interface, displaying a fact node associated with the target feature node in the interactive interface in a first mode; wherein the first mode is a mode capable of enhancing the visual relevance between the target feature node and each fact node associated with the target feature node.

Description

Technical Field

[0001] This application relates primarily to the field of artificial intelligence technology, and more specifically to an interaction method and electronic device. Background Technology

[0002] With the application and rapid development of artificial intelligence technology, the application scenarios of intelligent agents using artificial intelligence models to process tasks are becoming increasingly widespread. They can use models to perform feature analysis on the large amount of data generated by users using various smart devices, obtain basic information such as user attributes and background, as well as personalized information such as user interests and preferences in relevant businesses and personalized styles, in order to meet the data needs of downstream tasks.

[0003] Currently, the various personalized information mentioned above is usually edited by memorizing predefined static tags, which results in poor interactive flexibility, lack of self-awareness assistance value, and reduced user experience. Summary of the Invention

[0004] To address the above problems, this application provides the following technical solution:

[0005] A first aspect of this application provides an interaction method, the method comprising:

[0006] An interactive interface is output, in which at least one feature node is displayed; each feature node is determined based on at least one fact node associated with it, and the feature node contains a description of personalized information reflected by at least one fact node associated with it, and the fact node contains a description of at least one semantic unit in the source data.

[0007] In response to a first triggering operation on a target feature node in the interactive interface, the fact node associated with the target feature node is displayed in the interactive interface in a first manner;

[0008] The first method is a way to enhance the visual correlation between the target feature node and its associated fact nodes.

[0009] Optionally, at least one fact node may be displayed in the interactive interface while displaying the at least one feature node; the fact nodes and feature nodes of different categories in the interactive interface may be displayed in different visual styles or in different display areas.

[0010] Optionally, the fact node associated with the target feature node is displayed in the interactive interface in a first manner, including at least one of the following implementation methods:

[0011] The interactive interface displays, for the first time, the fact nodes associated with the target feature node in a first-view style;

[0012] Change the second visual style of the fact nodes associated with or not associated with the target feature node already displayed in the interactive interface;

[0013] Change the visual style of the node association relationship between the target feature node and its associated fact nodes in the interactive interface.

[0014] Change the spatial layout relationship between the target feature node and its associated fact nodes in the interactive interface.

[0015] A second aspect of this application provides an electronic device, the electronic device comprising:

[0016] At least one memory, and a computer program stored in the memory;

[0017] At least one processor capable of running intelligent agents;

[0018] The intelligent agent can execute the computer program through the processor to implement the steps of the interaction method proposed in the first aspect of this application. Attached Figure Description

[0019] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0020] Figure 1 This is a flowchart illustrating the interaction method proposed in Embodiment 1 of this application;

[0021] Figure 2 A schematic diagram of an interactive interface for displaying feature nodes in the interactive method proposed in the embodiments of this application;

[0022] Figure 3 This is a schematic diagram of the first display result after triggering a feature node in the interactive interface in the interactive method proposed in the embodiments of this application.

[0023] Figure 4 This is a schematic diagram of the second display result after triggering a feature node in the interactive interface in the interactive method proposed in the embodiments of this application.

[0024] Figure 5 This is a schematic diagram of a third display result after a feature node in the interactive interface is triggered in the interactive method proposed in this application embodiment;

[0025] Figure 6This is a schematic diagram illustrating the differentiated display of different categories of feature nodes on the interactive interface in the interactive method proposed in the embodiments of this application.

[0026] Figure 7 This is a flowchart illustrating the interaction method proposed in Embodiment 2 of this application;

[0027] Figure 8 This is a schematic diagram of the fourth display result after triggering a feature node in the interactive interface in the interactive method proposed in the embodiments of this application;

[0028] Figure 9 This is a flowchart illustrating the interaction method proposed in Embodiment 3 of this application;

[0029] Figure 10 This is a schematic diagram illustrating the classification and display of different categories of nodes in the interactive interface in the interactive method proposed in the embodiments of this application;

[0030] Figure 11 This is a schematic diagram of a display result after triggering a fact node in the interactive interface in the interactive method proposed in the embodiments of this application;

[0031] Figure 12 This is a flowchart illustrating the interaction method proposed in Embodiment 4 of this application;

[0032] Figure 13 This is a schematic diagram illustrating the display of source data fragments associated with fact nodes on the interactive interface in the interactive method proposed in the embodiments of this application.

[0033] Figure 14 This is a flowchart illustrating the interaction method proposed in Embodiment 5 of this application;

[0034] Figure 15 This is a schematic diagram of a process for generating fact nodes in the interaction method proposed in an embodiment of this application;

[0035] Figure 16 This is a schematic diagram of a process for generating feature nodes in the interaction method proposed in the embodiments of this application;

[0036] Figure 17 This is a schematic diagram of the structure of an interactive device proposed in an embodiment of this application;

[0037] Figure 18 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0038] The embodiments of this application are described below with reference to the accompanying drawings. The terminology used in the implementation section of this application is only for explaining specific embodiments and is not intended to limit the application. The embodiments of this application are described below with reference to the accompanying drawings. It will be understood by those skilled in the art that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0039] The terms "first," "second," etc., used in the context of this application and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of units is not necessarily limited to those units, but may include other units not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0040] It is understood that before using the technical solutions disclosed in the embodiments of this application, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this application in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained. For example, in response to receiving a user's active request, a pop-up window may be used, and a textual prompt message may be presented in the pop-up window to clearly inform the user that the requested operation will require the acquisition and use of the user's personal information. The user can choose whether to provide personal information to the electronic device, application, server, or storage medium or other software or hardware that performs the operation of the technical solution of this application based on the prompt message. This application does not limit the prompt message and the method of user authorization implementation. The data involved in this technical solution (including but not limited to the data itself, the acquisition or use of data) shall comply with the requirements of relevant laws and regulations and related provisions.

[0041] Regarding the technical problems described in the background section, this application proposes an interactive method. This method uses a visual interactive interface to display feature nodes containing descriptions of personalized information, allowing users to intuitively see "the system's view of themselves," thus achieving self-awareness. Since each feature node is associated with at least one fact node, and the fact node contains a description of at least one semantic unit in the source data, when a user needs to specifically view the extraction of a target feature node and the reasons behind it, the fact nodes associated with the target feature node can be displayed on the interactive interface in a way that enhances visual relevance. This intuitively realizes the interpretability of feature nodes, helping users better understand the formation mechanisms of their behavioral habits and preferences, and has value for self-awareness and reflection, reducing the user's cognitive burden. The interactive method proposed in this application will be described in detail below with reference to the accompanying drawings.

[0042] It should be noted that the models involved in this application (such as the first model, the second model, etc. described below) can be general-purpose AI (Artificial Intelligence) models deployed on the edge. They can adopt, but are not limited to, Transformer or its architectural variants (such as using only encoder-only / decoder-only, encoder-decoder, or MoE (Mixture of Experts, a neural network architecture) or other basic architectures. They learn the features and rules of natural language by training on a large amount of diverse data, thereby being able to understand and generate natural language. They typically have hundreds of millions to hundreds of billions of model parameters (model parameters are variables that control the behavior of the target model) and are able to capture complex relationships and patterns in natural language.

[0043] The AI ​​models may include, but are not limited to, generative models and generative language models (GLMs). For example, one or more of the following: large language model (LLM), GPT (Generative Pre-trained Transformer) series models, T5 (Text to Text Transfer Transformer) models, large visual models, and multimodal large models.

[0044] Depending on actual needs, the models involved in this application's embodiments can also be expert large models fine-tuned from general AI models based on task requirements in real-world scenarios. For example, they can be proprietary models (such as domain models) trained on sample datasets from general AI models (pre-trained models) within specific scenarios (such as professional scenarios in vertical fields like medicine, finance, law, biology, or remote sensing, or general scenarios). The models involved in this application can also be lightweight models compressed from general AI models or expert large models through lightweight methods such as quantization, knowledge distillation, or pruning to better meet the deployment needs of mobile devices with limited computing resources. This application does not limit the type of model involved in the context and can flexibly determine it according to actual task requirements.

[0045] Reference Figure 1 This is a flowchart illustrating the interaction method proposed in Embodiment 1 of this application. The method proposed in this embodiment can be applied to electronic devices, which are terminal devices with data processing capabilities and display functions, such as smartphones, laptops, augmented reality (AR) / virtual reality (VR) devices, robots, smart home devices, in-vehicle terminals, or other business terminals, etc., without limitation.

[0046] In this application, an intelligent program can run in the electronic device to execute the interactive method of this application. This intelligent program can be built into the operating system or application program of the computer device as part of the application program or operating system, or it can be used as a standalone application without being built into the operating system; this application does not impose any restrictions on this. The intelligent program can be an intelligent agent (artificial intelligence agent) or other program capable of providing intelligent functions in a conversational manner, capable of autonomously perceiving the environment, analyzing data based on powerful model capabilities, and performing tasks to provide users with a natural and intelligent conversational interactive experience. It can also be an artificial intelligence assistant, which can be woken up and started through voice or specific operations to implement the method provided in this application; this application does not impose any restrictions on this, and this application only uses an intelligent agent as an example for illustration.

[0047] In one possible implementation, the agent has the ability to invoke various models (such as the first and second models mentioned in the context of this application), such as through interface calls or other interactive methods. Different models can be invoked through corresponding different interfaces. Optionally, the agent may also have one or more models, meaning that the model is part of the agent. During the processor's operation of the agent, the agent can launch one or more models to handle corresponding tasks according to actual needs. This application does not impose any restrictions on this.

[0048] Based on the above analysis, such as Figure 1 As shown, the interaction method proposed in this application embodiment may include:

[0049] Step S11: Output an interactive interface and display at least one feature node in the interactive interface; each feature node is determined based on at least one fact node associated with it, and the feature node contains a description of personalized information reflected by at least one fact node associated with it, and the fact node contains a description of at least one semantic unit in the source data.

[0050] The source data includes at least one of text data, audio data, video data, and image data, which typically reflects user behavior, scene events, or interests and preferences. It can be acquired through one or more methods, such as user-initiated uploads, user-authorized collection, or user-associated cloud storage. There are no restrictions on the transmission protocol or collection interface of each source data, such as Bluetooth, Wi-Fi, or API (Application Programming Interface) calls. The source scenarios and acquisition methods of different source data can be flexibly determined based on the application scenarios and needs of the actual interactive system (such as local memory assistants, cloud-based personalized content platforms, etc.). This application does not restrict the source channels or carrier forms of the source data.

[0051] In some embodiments, source data can be obtained through active input sources, i.e., users directly upload at least one type of locally stored source data via the input interface of an electronic device, such as memo text, chat logs, recorded voice files, captured videos / photos, etc.; or through real-time acquisition sources, i.e., data collected in real time through built-in or external data acquisition modules of the electronic device (such as microphones, cameras, text input boxes, etc.), such as natural language commands input by the user in the intelligent agent, real-time recorded audio or video data, real-time captured on-site videos / images, and other user interaction data. It can also be obtained through the cloud, i.e., user-associated cloud storage data, such as online data synchronized from the user's bound cloud account (such as cloud storage, social platforms, or content platforms, etc.) or external devices via the network, such as at least one of the following cloud-synchronized emails, social updates, podcast articles, audio listening / video viewing records, album content, etc., without limitation.

[0052] Subsequently, this application can perform multimodal parsing on the source data to extract factual memories from the source data and generate multiple fact nodes, denoted as Activity nodes. Each fact node can include a description of at least one semantic unit in the source data. The semantic unit can be a user behavior or scene event, or any information block with independent semantics, such as a state or entity relationship. The definition and content of the semantic units can differ for different modalities of source data.

[0053] In some embodiments, the semantic units of text data can be text fragments with independent meanings, such as one or more of a single sentence, paragraph, keyword group, or complete event description statement, which can be the direct basis for extracting fact nodes. The semantic units of audio data can be text content converted from speech recognition, or unconverted audio fragments with specific characteristics (such as specific voice commands or iconic environmental sound effects), which can correspond to the descriptions of "voice behavior" or "audio events" in the fact nodes. The semantic units in video data can be visual content corresponding to video frames (such as character actions or scene images), audio content corresponding to accompanying sound, or text content corresponding to screen subtitles, which can be extracted individually or in combination as descriptions of fact nodes. The semantic units of image data can be visual content obtained through image recognition (such as object names, scene types, or character postures), which can correspond to the descriptions of "image content" or "visual events" in the fact nodes. However, they are not limited to the semantic unit content described in this paragraph.

[0054] Therefore, regardless of the type of source data, the description of a semantic unit can be one or more of the following: an event description of the corresponding episodic memory, a conceptual description of the corresponding semantic memory, and an operational description of the corresponding procedural memory, to correspond to a user behavior or scene event. It usually includes one or more dimensions of information such as time, location, action, and screen operation. This application does not limit the description content of each semantic unit.

[0055] Based on the above description of fact nodes, this application can perform cluster analysis on the generated fact nodes to generate at least one feature node, such as a personalized feature node. Each feature node can contain a description of one or more personalized information of the user (which can be a general description, such as a love of cooking or family meals). This feature node is associated with at least one fact node that can reflect this personalized information, usually fact nodes belonging to the same cluster, such as the feature node "loves cooking" being associated with fact nodes that reflect the user's love of cooking, and the feature node "family meals" being associated with fact nodes that reflect the user's experience of having meals with family. This application does not limit the implementation method of generating at least one feature node based on multiple fact nodes.

[0056] Therefore, this application adopts a two-layer structure of feature nodes and fact nodes, which ensures both the abstraction level of the data and the preservation of the original information of the data. This enables the system to simultaneously support high-level feature retrieval and low-level fact tracing, significantly improving the efficiency of data organization and management. For example, if the source data includes 152 video clips of kitchen activities, application usage logs (such as browsing history of a food app), and text content (such as recipe collections), after analysis, event descriptions such as "chopping vegetables," "stir-frying," and "baking" are extracted; behavioral pattern recognition is performed on the application logs to extract behavioral tags such as "browsing the vegetable market on weekends" and "collecting food videos"; semantic analysis is performed on the text content to extract interest tags such as "likes Sichuan cuisine" and "skilled in making noodles" to generate multiple corresponding fact nodes. The fact nodes such as "chopping vegetables," "stir-frying," and "baking" are aggregated into the "love cooking" feature node, and the fact nodes such as "browsing the vegetable market on weekends" and "collecting food videos" are aggregated into the "food exploration" feature node. Thus, the "love cooking" feature node can include, but is not limited to, a description of the user's love for cooking, such as "impressions of activities in the kitchen generated based on 152 videos."

[0057] Subsequently, this application can visually output an interactive interface, intuitively displaying at least one feature node. This visualization makes personalized information such as user behavior patterns, interests, preferences, styles, inherent attributes, and background explicit, helping users better understand their habits and preferences, reducing their cognitive burden, and providing conditions for diverse interactive experiences. It should be noted that this application does not limit the method of displaying feature nodes on the interactive interface; it can adopt, but is not limited to, graph-based, tree-structured, list-based, region-based, or other network structure forms, and can be flexibly configured or dynamically adjusted.

[0058] In some embodiments, the currently output interactive interface may only display feature nodes, while fact nodes may not be displayed; that is, in scenarios where the interactive interface initially has no fact nodes, such as... Figure 2 As shown in Figure (a), each large sphere (which is only used as an example and can also be represented by other shapes or patterns) represents a feature node. In order to help users understand the description of these feature nodes, a summary text can be generated based on the descriptions contained in each of the displayed feature nodes and presented as a memory summary on the interactive interface. This application does not limit the content of the memory summary, and it can be dynamically updated as the number, category and / or description of the feature nodes displayed on the interactive interface change. The implementation process is not detailed in this application.

[0059] Optionally, around each feature node displayed on the interactive interface, summary information generated based on the description contained in that feature node can be displayed through pop-ups or prompts. This could include category tags or key attributes of the personalized information described by the feature node. In this way, users can see the various feature nodes displayed on the interactive interface without interacting with them, allowing them to understand the content of the feature nodes, distinguish between different feature nodes, and subsequently select the target feature node they need to learn more about. This application does not limit the display method or content of feature nodes on the interactive interface. Figure 2 Figure (a) is only one example of how feature nodes are displayed, and is not limited to this display method.

[0060] Furthermore, if multiple feature nodes have been generated according to the method described above, all of these feature nodes can be displayed in the interactive interface, or some feature nodes can be displayed in the interactive interface according to pre-defined display rules, such as at least one feature node of the same category. This application does not impose any restrictions on this. Figure 2 The way the feature nodes are displayed in Figure (a) does not constitute a limitation on step S11 of this application.

[0061] In other embodiments, at least one fact node can be displayed simultaneously with at least one feature node in the interactive interface, i.e., in scenarios where the interactive interface initially has a fact node, such as... Figure 2 As shown in Figure (b), each large sphere represents a feature node, and each small sphere represents a fact node. However, this representation is not limited to this method; it is used here only as an example. It should be understood that all generated (or extracted from source data) fact nodes can be displayed; a subset of fact nodes can be displayed, such as all or a subset of fact nodes associated with each simultaneously displayed feature node; or a randomly selected subset of fact nodes can be displayed, in which case it is not required that the displayed fact nodes be associated with feature nodes. This display method is used to intuitively inform the user of the association between feature nodes and fact nodes. This application does not limit the display method of simultaneously displaying at least one feature node and at least one fact node in the interactive interface, including but not limited to... Figure 2 The method shown in Figure (b) is merely an example of a display method and does not limit the number of feature nodes and fact nodes that can be displayed simultaneously on the interactive interface.

[0062] Optionally, in scenarios where the interactive interface described above initially has or does not have fact nodes, other visualization methods can also be used to display the corresponding nodes, such as... Figure 2In the illustrated display method, the display distance between different nodes can represent the strength of the association between two corresponding nodes that are related. Of course, it can also mean nothing. Other visual styles / display methods can also be used to display information such as the association relationships and association attributes between different nodes; these will not be detailed here. Furthermore, for each node displayed in the interactive interface, its visual style can be dynamically adjusted by controlling the node's visual attributes, thereby controlling the interactive interface to display the corresponding nodes according to the appropriate visual style. The implementation process will not be detailed here.

[0063] Step S12, in response to the first trigger operation on the target feature node in the interactive interface, the fact node associated with the target feature node is displayed in the interactive interface in a first manner; the first manner is a way that can enhance the visual association between the target feature node and its associated fact nodes.

[0064] The target feature node can be any feature node displayed on the current interactive interface, and the first triggering operation can be a selection and viewing operation of the currently displayed target feature node, such as the user using the mouse to hover over the node icon of the target feature node (e.g., ...). Figure 2 The large sphere in the image can be clicked or double-clicked, or the node icon can be touched with a stylus. Alternatively, the target feature node and its associated fact nodes can be selected and viewed by speaking a voice signal, such as "view the fact nodes associated with 'love cooking'", to generate corresponding instructions. This application does not restrict the implementation method of the first trigger operation.

[0065] In one possible implementation, the application can be implemented by an intelligent agent directly responding to the first trigger operation, or by an input component supporting the first trigger operation sensing the first trigger operation on the target feature node and sending a corresponding first instruction to the intelligent agent. The intelligent agent then responds to / executes the first instruction to determine the fact nodes associated with the target feature node and controls the interactive interface to display at least one fact node associated with the target feature node in a first manner. Since the first manner can enhance the visual correlation between the target feature node and its associated fact nodes, the user viewing the interactive interface can focus on the target feature node and its associated fact nodes, intuitively demonstrating the interpretability of the target feature node and reducing the user's cognitive burden.

[0066] Using the example scenario of the source data mentioned above as an example, the explanation will be based on, but is not limited to, the following: Figure 2 As shown, the interactive interface displays feature nodes such as "Passionate about Cooking" and "Food Exploration." When a user clicks on the "Passionate about Cooking" feature node, a differentiated visual style (e.g., ...) appears around the "Passionate about Cooking" feature node in the interactive interface. Figure 2The large and small spheres shown represent feature nodes and fact nodes, respectively. They display fact nodes such as "chopping vegetables", "stir-frying", and "baking", and can enhance the visual association between these fact nodes and the feature node in a first way.

[0067] Based on the above analysis, in the process of displaying the fact node associated with the target feature node in the interactive interface in the first manner, in the first possible implementation, if the interactive interface in step S11 does not display the fact node, such as Figure 3 As shown, the fact nodes associated with the target feature node can be initially displayed in the interactive interface in a first-person visual style. At this point, the first-person visual style differs from the visual style of the target feature node, allowing users to intuitively distinguish between feature nodes and fact nodes. By displaying fact nodes from scratch, the salient visual presentation of the fact node can be achieved, drawing the user's attention to it.

[0068] To distinguish between target feature nodes and non-target feature nodes, the visual style for differentiating them can be adjusted by controlling at least one visual attribute of the node, such as color, hidden state (transparent, collapsed, etc.), or icon, so that target feature nodes and non-target feature nodes in the interactive interface are displayed with differentiated visual styles. Optionally, this application can also differentiate target feature nodes and non-target feature nodes in different display areas to ensure that users can quickly focus on the triggered target feature node and its associated fact nodes. This application does not limit the implementation method of differentiated visual styles or the implementation method of distinguishing target feature nodes and non-target feature nodes, and can be determined as needed, or can be customized / flexibly configured by the user.

[0069] In the second possible implementation, if the interactive interface in step S11 has already displayed fact nodes, the second visual style of the fact nodes associated with or not associated with the target feature node displayed in the interactive interface can be changed (e.g., changing the node's color, size, pattern, show / hide state, etc.). This achieves a differentiated display between the fact nodes associated with and not associated with the target feature node, thereby strengthening the visual association between the target feature node and its associated fact nodes, and weakening the visual association between the target feature node and its unassociated fact nodes. This differentiated display method can be implemented using differentiated visual styles or different display areas, and the implementation process can refer to the relevant description above of the differentiated display method between target feature nodes and non-target feature nodes. This embodiment will not be detailed here.

[0070] In the third possible implementation method, if the fact node has been displayed in the interactive interface in step S11, the visual style of the node association relationship between the target feature node and its associated fact nodes in the interactive interface can be changed. For example, the visual style can be changed by means of connecting lines, arrows, dynamic flow effects and color coding (based on at least one dimension such as the association strength and association type of the node association relationship, and different colors represent different dimensions, etc.), so as to strengthen the visual association between the target feature node and its associated fact nodes.

[0071] Furthermore, such as Figure 4 As shown, the node relationships between the target feature node and its associated fact nodes are displayed using connecting lines. Of course, other visual styles can also be used to represent these node relationships.

[0072] This application can also weaken the visual association between target feature nodes and non-target feature nodes, as well as fact nodes that are not associated with target feature nodes, such as... Figure 4 As shown, the transparency of these non-target feature nodes and unrelated fact nodes is increased to make these nodes transparent (a hidden state, making them appear to fade out of the interface), thereby highlighting the target feature nodes and their associated fact nodes. Optionally, in addition to adjusting the visual attribute of node transparency, these nodes can also be controlled to switch from a displayed state to a collapsed state (i.e., a hidden state, which is another form of hiding). However, this is not a limitation; the visual attributes of the corresponding nodes can be flexibly adjusted according to the actual situation to change the current visual style of the display.

[0073] In a fourth possible implementation, this application can also change the spatial layout relationship between the target feature node and its associated fact nodes in the interactive interface, such as arranging the target feature node and its associated fact nodes in a ring (e.g., Figure 5 As shown, with the target feature node as the center, its associated fact nodes are arranged around the center (the specific arrangement method is not limited), aggregated in the same visual container / display area, or the corresponding connecting line style is adjusted, etc., to achieve the effect of enhancing visual relevance.

[0074] Therefore, when a user performs a first-trigger operation on a specific feature node (target feature node) displayed on the interactive interface, the electronic device can enhance visual relevance by employing differentiated visual styles and dynamically adjusting spatial layout relationships. This simultaneously weakens the visual relationships between corresponding other nodes, making the information hierarchy immediately clear and attracting the user's attention to the factual nodes associated with the target feature node, thus reducing the user's cognitive burden. Furthermore, this application, through this dynamic interaction method, enables the interactive interface to adjust the displayed content in real time according to the user's operational intentions, enhancing the responsiveness and interactivity of the interactive interface and improving the user experience.

[0075] In some embodiments, the multiple feature nodes in the above embodiments may include feature nodes of different categories. These different categories of feature nodes may include descriptions of different personalized information from user profile information (such as basic user attributes and background information, which can be denoted as Profile), user preference information (such as user preferences, interests, etc., reflecting the user's inclination or choice towards specific content, functions, or services, which can be denoted as Preference), and user characteristic information (which may be comprehensive information reflecting the user's personality, style, or behavioral patterns, which can be denoted as Personality). Feature nodes containing Profile descriptions can be denoted as first-category feature nodes, feature nodes containing Preference descriptions as second-category feature nodes, and feature nodes containing Personality descriptions as third-category feature nodes. Thus, when different categories of feature nodes are displayed simultaneously on the interactive interface, the different categories of feature nodes are differentiated visually (e.g., ...). Figure 6 The node patterns shown may differ, but are not limited to these. Differentiation can also be achieved by changing colors, sizes, etc., or they can be displayed separately in different display areas to show nodes of the same category in the same display area.

[0076] Similarly, when displaying different categories of feature nodes and fact nodes simultaneously on the interactive interface (all fact nodes are configured as nodes of the same category), the fact nodes and feature nodes of different categories in the interactive interface are distinguished by differentiated visual styles or displayed in different display areas. The implementation process can be referred to the description of the corresponding part of the above embodiment, and will not be repeated here. It should be noted that the differentiated visual styles between different categories of nodes can be represented by one or more dimensions such as color, size, and pattern, including but not limited to the differentiated visual styles shown in the attached figures.

[0077] In some embodiments, different categories of feature nodes may have association relationships, which can be obtained based on the fact nodes associated with the corresponding feature nodes. The analysis process is not detailed in this application. Based on this, this application can also display at least one non-target feature node associated with the target feature node in the interactive interface to enhance self-awareness. Combined with the above description of different categories of feature nodes, this method can help users intuitively understand "why I have this preference" and "how this behavior is formed," thereby enhancing self-awareness, rather than just knowing "I have this preference." For example, the "love cooking" feature node can be understood in this way as follows: you (the user) like baking because it is driven by "pursuit of precision" (personality trait) and "enjoyment of a sense of accomplishment" (psychological need).

[0078] Based on this, refer to Figure 7 The flowchart shown in Embodiment 2 of this application illustrates the interaction method. The method proposed in this embodiment is still applicable to electronic devices, such as... Figure 7 As shown, the interaction method proposed in this embodiment may include:

[0079] Step S71: Output the interactive interface, and display the feature nodes of different categories in the interactive interface with differentiated visual styles.

[0080] The methods for acquiring and distinguishing different categories of feature nodes (which may include differentiated visual styles or different display areas, etc.) can be referred to the descriptions in the corresponding sections of the above embodiments, and will not be detailed in this embodiment. The different categories of feature nodes may include two or three of the following: user profile type (first category) feature nodes, user preference type (second category) feature nodes, and user characteristic type (third category) feature nodes. This achieves a structured organization of multi-dimensional user feature information. Since different categories of feature nodes carry different categories of personalized information, the system can understand and describe users from multiple perspectives, facilitating unified management of multiple categories of feature nodes.

[0081] In this embodiment, the relationships between feature nodes of different categories are analyzed based on multiple fact nodes. These include semantic relationships established based on shared fact nodes, temporal relationships established based on time series, spatial relationships established based on spatial location, and behavioral relationships established based on behavioral patterns. This cross-category feature node relationship analysis can reveal the deep connections between user profiles, preferences, and features. For example, it can reveal how a user's personality traits (user feature category) influence their content preferences (user preference category), or how a user's behavioral patterns (user profile category) reflect their aesthetic preferences (user feature category). This cross-category relationship analysis provides a more comprehensive and in-depth understanding of users.

[0082] In some embodiments, when displaying feature nodes of different categories in the interactive interface, feature nodes of different categories and their relationships can also be visualized in the form of graphs or other network forms based on the above-mentioned relationships. Visual styles such as node visual attributes and connection line styles can also be used to display feature node categories and relationship strengths, making complex cross-category relationship networks intuitive and easy to understand. Users can see the relationship between feature nodes of different categories at a glance, reducing cognitive burden and improving information acquisition efficiency.

[0083] It should be understood that, based on the above-mentioned content, at least one fact node can also be displayed. Preferably, fact nodes associated with different categories of feature nodes can be displayed. Furthermore, information such as the node association relationship and association strength between feature nodes and fact nodes can be displayed through differentiated visual styles. This application does not limit the implementation method.

[0084] Step S72, in response to the second triggering operation on the target feature node in the interactive interface, at least one non-target feature node that is associated with the target feature node is displayed in the interactive interface in a second manner; the non-target feature node and the target feature node are of the same category or different categories.

[0085] In the dynamic association display process, the second trigger operation on the target feature node is different from the first trigger operation described above. This allows for the identification of the trigger operation type when executing a trigger operation on the same feature node, determining what to display in the interactive interface. The second trigger operation can also be completed using an input component, and its implementation method can be similar to the first trigger operation; this embodiment will not elaborate further.

[0086] After obtaining the trigger operation on the target feature node within the display, we can first identify if the trigger operation is the first trigger operation, as described in the corresponding section above; if the trigger operation is identified as the second trigger operation, such as... Figure 8 As shown, at least one non-target feature node associated with the target feature node can be displayed in the interactive interface in a second manner. This includes other feature nodes belonging to the same category as the target feature node, as well as other feature nodes deployed in the same category, allowing the user to focus on the non-target feature node. The second method can be the same as or different from the first method, but its implementation is similar.

[0087] Based on the above description of the first method, the visual correlation between target feature nodes and non-target feature nodes can be enhanced by adjusting visual styles and spatial layout relationships. If necessary, the fact nodes associated with the target feature nodes can also be displayed simultaneously in the interactive interface using the first method, such as... Figure 8As shown, in the same interactive interface, other nodes associated with the target feature node can be displayed by enhancing visual relevance. These nodes include both non-target feature nodes and target feature nodes. This application does not limit the implementation method of enhancing the visual relevance of each node.

[0088] Reference Figure 9 This is a flowchart illustrating the interaction method proposed in Embodiment 3 of this application. The method proposed in this embodiment is still applicable to electronic devices, such as... Figure 9 As shown, the interaction method proposed in this embodiment may include:

[0089] Step S91: Output the interactive interface, which displays different categories of feature nodes and fact nodes, as well as corresponding node category identifiers, in a differentiated visual style.

[0090] Each node category identifier is associated with each node of the corresponding category, and each fact node is a node of the same category. This application does not restrict the differentiated visual styles of different node category identifiers, and can use the visual styles of the corresponding nodes on the interactive interface, but is not limited to this.

[0091] Step S92, in response to the fourth trigger operation on the target node category identifier in the interactive interface, display each node of the target category associated with the target node category identifier in the interactive interface in a fourth manner;

[0092] The fourth method is a way to achieve visual differences between nodes of the target category and nodes of non-target categories. It can be the same as or different from the first method mentioned above, and there are no restrictions on this.

[0093] Based on this, such as Figure 10 As shown, when a user needs to statistically view nodes of the same category, a fourth trigger operation, such as clicking or touching, can be performed on the corresponding node category identifier in the interactive interface. The interactive interface then displays nodes of the corresponding category in a way that enhances visual relevance, such as feature nodes or fact nodes of the same category. This application does not limit the implementation method of this display. For example, Figure 10 As shown, for other nodes associated with non-target node category identifiers, their display can be weakened by changing their style, or they can be hidden, etc., without any restrictions.

[0094] Step S93: In response to a third trigger operation on the target fact node in the interactive interface, at least one feature node associated with the target fact node is displayed in the interactive interface in a third manner.

[0095] If the fact node is already displayed in the current interactive interface, you can also choose to view the feature nodes associated with the target fact node, such as... Figure 11As shown, the third method is a way to enhance the visual correlation between the target fact node and its associated feature nodes. Specifically, the method can be the same as or similar to the display implementation method of the first method, which will not be described in detail here.

[0096] In summary, this application enhances visual relationships to achieve dynamic and interconnected display, enabling the interface to adjust its content in real time based on the user's operational intent. This enhances the interface's responsiveness and interactivity, thereby improving the user experience. Furthermore, the method in this application supports flexible classification and association analysis of feature nodes. Different categories can be defined according to actual needs, or the dimensions of association analysis can be adjusted. Thus, through feature node classification, cross-category associations, visual display, and dynamic interconnected display, unified management and intelligent interaction of multiple categories of feature nodes are achieved, further improving the user experience.

[0097] In some embodiments, for the interaction methods described in the above embodiments, the nodes displayed on the interactive interface may include the aforementioned feature nodes and / or fact nodes, and the descriptions contained in the nodes may be displayed in a progressive manner. Therefore, in response to a first type of interactive operation on a target node in the interactive interface, summary information of the target node is displayed in the interactive interface; in response to a second type of interactive operation on the target node, detailed information of the target node is displayed in the interactive interface, etc., but it is not limited to these two levels of progressiveness.

[0098] The summary information and detailed information can be generated based on the description of the semantic units contained in the corresponding fact node, or the description of the personalized information contained in the corresponding feature node. They can be generated by calling a model or flexibly configured manually. This application does not restrict the content or generation method of these two types of progressive information. The semantic content of the detailed information is more extensive than that of the summary information. The first type of interactive operation can be that the cursor hovers over the node display position, and the summary information of the node is displayed through text or voice broadcast, which can be topic tags, key attributes, etc., as shown in the floating prompt box in the attached figure above. The second type of interactive operation can be that the cursor selects the currently displayed fact node and associates it with the source data fragment corresponding to the semantic units it contains. The source data fragment comes from the node of the source data, and can display a summary descriptive text of the node, etc. This application implements the source and control display method for the summary information and detailed information of the same node.

[0099] Furthermore, in the interaction methods described in the above embodiments, the fact node is associated with the source data fragment corresponding to the semantic unit it contains. The source data fragment comes from the source data, and this application does not limit the acquisition method. Based on this, in addition to the above embodiments, in different interaction processes, in response to the triggering operation of the target fact node displayed on the interaction interface, the source data fragment associated with the target fact node can also be displayed in the interaction interface, enhancing the interpretability of the content extraction of the feature node and realizing the tracing of the feature node.

[0100] In one possible implementation, such as Figure 12 The flowchart shown in this application embodiment illustrates the interaction method, which may include:

[0101] Step S121: Output the interactive interface, and display the fact nodes and feature nodes in a differentiated visual style in the interactive interface.

[0102] Step S122: In response to the triggering operation of the target fact node displayed on the interactive interface, the source data fragment associated with the target fact node includes audio or video fragments. The playback control is started in the interactive interface, and the audio or video fragments are played through the playback control.

[0103] In this embodiment, triggering an operation on a target fact node can generate an audio / video playback view. The video playback view can be controlled to position and play the audio / video segment associated with the target fact node. Activating the playback control in the interactive interface can include at least one of the following: a play / pause button, a progress bar, a volume control, a full-screen switching control, etc. Thus, the activation method of the playback control can include at least one of the following: popping up a floating playback window near the target fact node; displaying a player panel at a fixed position in the interactive interface; switching the interactive interface to full-screen playback mode, etc. For example, Figure 11 and Figure 13 As shown, an audio / video playback window is displayed in the interactive interface, and audio / video clips are played within this window.

[0104] During the playback process described above, the playback status of audio or video segments is controlled in response to user actions on the playback controls. These actions may include at least one of the following: play, pause, fast forward, rewind, volume adjustment, full-screen switching, etc. Upon completion of playback, the playback controls are hidden or removed in response to a close operation, restoring the original display state of the interactive interface. In some embodiments, the window where the playback controls play audio / video segments may also display detailed information about the target fact point, such as summary text, to allow users to quickly and intuitively understand the fact point.

[0105] Therefore, this application achieves a complete tracing path from abstract features to specific facts and then to the original data by triggering the playback of source data segments through fact nodes. Users can directly view the original evidence supporting feature nodes and fact nodes, enhancing the interpretability and credibility of the system. For example, when the system displays the "loves cooking" feature node, users can trace back to a specific kitchen video segment through the fact node to verify the accuracy of the system's judgment.

[0106] Furthermore, the source data fragment display method proposed in this application provides users with an immersive data exploration experience. Users can directly view the raw data within the interactive interface without switching to other applications or interfaces, maintaining the continuity and smoothness of the interaction and achieving seamless interaction. Users can complete data tracing and viewing without complex operations. This smooth interaction method improves operational efficiency, allowing users to focus more on data exploration itself. This embodiment supports the playback of audio and video fragments, adapting to different types of data sources. Whether it's video recordings of kitchen activities, audio fragments from voice memos, or screen recordings, all can be played through a unified playback control, enhancing the system's practicality and applicability.

[0107] Step S123: In response to the source data fragment associated with the target fact node containing text content, the text content is displayed in a differentiated visual style in the interactive interface.

[0108] In this embodiment of the application, in response to a triggering operation on a target fact node displayed on the interactive interface, it is determined that the source data fragment associated with the target fact node contains text content, such as screen content in a video, text documents, instant messaging records, note content, web page content, text information in application interaction records, etc., at least one of the following: The differentiated visual style for displaying the text content may include, but is not limited to, at least one of the following:

[0109] A floating text box pops up near the target fact node to display the text content; a text panel is displayed in a fixed position in the interactive interface to display the text content; the interactive interface is switched to text reading mode to display the text content in full screen; key information in the text content is highlighted using at least one of the following methods: highlighting, zooming, bolding, color changing, etc.; the text content is displayed in a structured manner using at least one of the following methods: paragraphing, indentation, list, quotation format, etc.

[0110] In some embodiments, during the display of text content on the interactive interface, corresponding functions can be executed in response to operations on the text content. Different operations may execute different functions. For example: in response to scrolling the text content, the display position of the text content can be adjusted; in response to copying the text content, the selected text content can be copied to the clipboard; in response to searching the text content, a specified keyword can be searched within the text content; in response to annotation of the text content, the text content can be highlighted, underlined, or annotated; in response to closing the text content, the text display area can be hidden or removed, restoring the original display state of the interactive interface, etc., thereby reducing the cost of information acquisition and improving reading efficiency.

[0111] Therefore, this application enhances the interpretability and credibility of the system by displaying text content through differentiated visual styles, allowing users to directly view the original textual evidence supporting the factual nodes. For example, when the system displays the feature node "likes science and technology articles," users can trace back to the specific article content through the factual node to verify the accuracy of the system's judgment. Furthermore, it provides flexible text display methods to adapt to different scenario needs, dynamically selecting the appropriate display format based on the length of the text content and the user's browsing habits, thus improving the system's usability and user experience.

[0112] The display method of the text content is consistent with the overall design of the interactive interface, maintaining visual unity. Users can view the text content without leaving the current interface, reducing cognitive load and operational complexity, and making the interaction process more natural and fluid. This application can process multiple types of data, such as text, audio, and video, through a unified interaction method, achieving unified management of multimodal data. Whether it is text content or audio / video clips, they can be traced and viewed through similar interaction methods, improving the integrity and consistency of the system.

[0113] In some embodiments, during the determination of each feature node based on at least one fact node associated with it, fact nodes can be generated based on a first model, and feature nodes can be generated based on a second model. Based on this, such as... Figure 14 The flowchart of the interaction method proposed in this application embodiment shown below illustrates that one possible implementation method for determining feature nodes may include:

[0114] Step S141: Process at least one source data based on the first model to generate multiple fact nodes; the source data includes at least one of text data, audio data, video data, and image data, and the video data includes screen content in the video;

[0115] The source of the source data and the method of obtaining it can be referred to the description in the corresponding part of the above embodiment, and will not be repeated in this embodiment.

[0116] In one possible implementation, video data can be decomposed into two streams: a video stream (used to capture scenes, actions, and events) and image frames containing screen content (keyframes). This allows the system to simultaneously acquire natural scene information and screen content information, ensuring the integrity and reliability of factual memory. In this case, the first model can be a general AI model or a composite model consisting of a large video model and a text analysis model, such as... Figure 15 As shown, event descriptions are extracted from the video stream using a large video model, and the correlations between different events are identified for subsequent aggregation analysis. A text analysis model is used to analyze screen content, identifying device interfaces, text information, and interactive elements to supplement the semantic information of the video scene, but this is not limited to the data processing methods described in this paragraph.

[0117] For data in other modalities, the corresponding modal model can be used to process the data, and the processing results of various modal source data can be combined to generate multiple fact nodes. Optionally, this application can also use a multimodal large model as the first model to perform multimodal parsing on multiple source data to generate structured fact nodes. This application does not limit the processing methods for each modal data. After completing the extraction of various types of information, this information can be integrated into fact nodes based on a data integration model consistent with the display form of multiple fact nodes (such as knowledge graphs), unifying the semantic representation of different data types, and establishing semantic relationships between different fact nodes to form a basic factual knowledge graph (Fact KG). The data integration process for other display forms of fact nodes is similar and will not be detailed. The relationships between different fact nodes can reflect the logical order of events, causal relationships, or contextual relationships, providing a structured foundation for subsequent feature node extraction.

[0118] After that, as Figure 16 As shown, multiple fact nodes can be processed based on the second model to aggregate and generate at least one feature node; each feature node is associated with the fact node that generated it. The second model can be a general generative model, or a pre-trained feature node extraction model / personalized extraction model, etc., without restriction.

[0119] Step S142: Analyze the description of the semantic units contained in each fact node based on the second model, aggregate the fact nodes with related relationships, and generate at least one feature node and a description of the personalized information contained in each feature node.

[0120] Step S143: Establish the node association relationship between each feature node and the fact node that aggregates to generate the feature node, so as to display the fact node associated with the target feature node in the interactive interface in a first manner based on the node association relationship.

[0121] In this embodiment, the generated fact nodes can be used as input to the second model to generate multiple feature nodes, and a network structure consistent with the display form of the multiple feature nodes (such as a knowledge graph) can be established, such as a personalized feature graph (Personalized KG). In conjunction with the interaction method described in the above embodiment, in order to enhance the interpretability and credibility of feature node extraction, the association between feature nodes and fact nodes can be established in this network structure so that each feature node can be traced back to the source data fragment associated with the fact node, such as specific user behavior and video fragments. The implementation method is not detailed in this application. The implementation process of displaying the fact nodes associated with the target feature node in the first way in the interactive interface can be referred to the description of the corresponding part of the above embodiment, which will not be repeated here.

[0122] In the process of aggregating fact nodes, related fact nodes can be aggregated based on dimensions such as semantic similarity, temporal proximity, and spatial relevance to discover deep connections between data, generate more accurate and meaningful user feature descriptions, and improve the accuracy and interpretability of feature extraction.

[0123] Furthermore, this application is not dependent on specific data source types or application scenarios. Through multimodal parsing and a two-layer node structure, it can adapt to different data sources and application requirements. It also supports the processing of large-scale data, possesses good scalability, and can meet the needs of applications of different scales.

[0124] Reference Figure 17 This is a schematic diagram of the structure of an interactive device proposed in an embodiment of this application. The interactive device may include:

[0125] An interactive interface output module 171 is used to output an interactive interface, in which at least one feature node is displayed; each feature node is determined based on at least one fact node associated with it, and the feature node contains a description of personalized information reflected by at least one fact node associated with it, and the fact node contains a description of at least one semantic unit in the source data.

[0126] The first trigger display module 172 is used to respond to a first trigger operation on a target feature node in the interactive interface and display the fact nodes associated with the target feature node in the interactive interface in a first manner; wherein, the first manner is a manner that can enhance the visual relevance between the target feature node and its associated fact nodes.

[0127] Optionally, the first trigger display module 172 may include at least one of the following processing units:

[0128] The first processing unit is used to display, for the first time in the interactive interface, each fact node associated with the target feature node in a first visual style;

[0129] The second processing unit is used to change the second visual style of the fact nodes associated with or not associated with the target feature node displayed in the interactive interface.

[0130] The third processing unit is used to change the visual style of the node association relationship between the target feature node and its associated fact nodes in the interactive interface.

[0131] The fourth processing unit is used to change the spatial layout relationship between the target feature node and its associated fact nodes in the interactive interface.

[0132] Optionally, at least one fact node may be displayed simultaneously with the at least one feature node in the interactive interface; the fact nodes and feature nodes of different categories in the interactive interface may be displayed in different visual styles or in different display areas.

[0133] Among them, different categories of feature nodes contain descriptions of different personalized information in user profile information, user preference information, and user characteristic information; different categories of feature nodes have association relationships obtained based on the fact nodes associated with the corresponding feature nodes. Based on this, the above-mentioned interactive device may further include:

[0134] The second trigger display module is used to respond to a second trigger operation on a target feature node in the interactive interface, and to display at least one non-target feature node that is associated with the target feature node in the interactive interface in a second manner; the non-target feature node and the target feature node are of the same category or different categories; the second manner is a way to enhance the visual association between the target feature node and the at least one non-target feature node.

[0135] Optionally, the fact node is associated with the source data segment corresponding to the semantic unit it contains, and the source data segment comes from the source data; the description of the semantic unit is an event description of the context memory in the corresponding source video segment, a concept description of the corresponding semantic memory, or an operation description of the corresponding process memory; based on this, the above interactive device may further include:

[0136] The source tracing and display module is used to respond to a trigger operation on the target fact node displayed on the interactive interface and to display the source data fragment associated with the target fact node in the interactive interface.

[0137] Optionally, the traceability display module may include at least one of the following display units:

[0138] The first display unit is configured to, in response to the source data segment associated with the target fact node containing an audio or video segment, activate a playback control in the interactive interface and play the audio or video segment through the playback control.

[0139] The second display unit is used to display the text content in a differentiated visual style in the interactive interface in response to the source data fragment associated with the target fact node containing text content.

[0140] In some embodiments, the interactive device may further include:

[0141] The third trigger display module is used to respond to a third trigger operation on the target fact node in the interactive interface and display at least one of the feature nodes associated with the target fact node in the interactive interface in a third manner; the third manner is a way that can enhance the visual relevance between the target fact node and its associated feature nodes.

[0142] In some embodiments, the interactive device may further include:

[0143] A node category identifier display module is used to display at least one node category identifier in the interactive interface; each node category identifier is associated with each node of the corresponding category, and each of the fact nodes is a node of the same category;

[0144] The fourth trigger display module is used to respond to a fourth trigger operation on the target node category identifier in the interactive interface, and to display each node of the target category associated with the target node category identifier in the interactive interface in a fourth manner; the fourth manner is a way to achieve visual difference between the target category node and the non-target category node.

[0145] Optionally, the above-mentioned interactive device may further include:

[0146] The first interactive display module is used to respond to a first type of interactive operation on a target node in the interactive interface and display summary information of the target node in the interactive interface.

[0147] The second interactive display module is used to respond to the second type of interactive operation on the target node and display detailed information of the target node in the interactive interface;

[0148] The summary information and the detailed information are generated based on the description of the semantic units contained in the corresponding fact nodes, or the description of the personalized information contained in the corresponding feature nodes; the semantic content of the detailed information is greater than that of the summary information.

[0149] In some embodiments, the interactive device may further include:

[0150] A feature node determination module is used to determine a feature node based on at least one fact node associated with the feature node. Optionally, the feature node determination module may include:

[0151] The fact node generation unit is used to process at least one source data based on the first model to generate multiple fact nodes; the source data includes at least one of text data, audio data, video data, and image data, and the video data includes screen content in a video;

[0152] The feature node generation unit is used to process the multiple fact nodes based on the second model and aggregate them to generate at least one feature node; each feature node is associated with the fact node that aggregated to generate the feature node.

[0153] Optionally, the feature node generation unit may include:

[0154] The description generation unit is used to analyze the description of the semantic units contained in each of the fact nodes based on the second model, aggregate the fact nodes with related relationships, and generate at least one feature node and a description of the personalized information contained in each feature node.

[0155] A relationship establishment unit is used to establish a node association relationship between each feature node and the fact node that aggregates to generate the feature node, so as to display the fact node associated with the target feature node in the interactive interface in a first manner based on the node association relationship.

[0156] This application also provides a computer program product including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements any of the interaction methods provided in this application. The computer program product can be stored in a readable storage medium, such as a computer floppy disk, USB flash drive, portable hard drive, ROM (Read-Only Memory), RAM (Random Access Memory), magnetic disk, or optical disk, and includes several instructions to cause an electronic device to execute the interaction methods described in the various embodiments of this application.

[0157] This application also provides a computer-readable storage medium carrying one or more computer programs. When these programs are executed by an electronic device, the electronic device can implement any of the interaction methods provided in this application. This application does not limit the product form of the computer-readable storage medium.

[0158] Reference Figure 18This is a schematic diagram of the hardware structure of an electronic device proposed in an embodiment of this application. The electronic device may include at least one memory 181, a computer program stored in the memory 181, and at least one processor 183 capable of running an intelligent agent 182.

[0159] The intelligent agent 182 can execute computer programs through the processor 183 to implement the interaction methods described in the corresponding embodiments of this application above. The implementation process will not be described in detail in this application.

[0160] Furthermore, the electronic device may also include at least one input component 184, such as an audio collector, stylus, mouse and keyboard, or joystick, for the user to operate the input component 184 to trigger operations on the interactive interface. This application does not describe in detail the type of input component 184 or its input principle. The electronic device may also include at least one output component 185, such as an audio player or display screen, for outputting the interactive interface, playing audio data, etc., without limitation.

[0161] It should be understood that, Figure 18 The structure of the electronic device shown does not constitute a limitation on the computer device in the embodiments of this application. In practical applications, the computer device may include more than Figure 18 The application does not provide detailed examples of the more or fewer components shown, or combinations of certain components, such as gyroscopes, accelerometers, and gravity sensors used to obtain sensing parameters, power management modules, antennas, or other communication elements.

[0162] Finally, it should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the accompanying drawings of the device embodiments provided in this application, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines.

[0163] In the above embodiments, the invention can be implemented entirely or partially by software, hardware, firmware, or any combination thereof. Through the description of the above embodiments, those skilled in the art can clearly understand that this application can be implemented using software plus necessary general-purpose hardware, or it can be implemented using dedicated hardware including dedicated integrated circuits, dedicated CPUs, dedicated memory, dedicated components, etc. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The various embodiments in this specification are described in a progressive or parallel manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to mutually. For the apparatus and computer equipment disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

Claims

1. An interaction method, the method comprising: Output an interactive interface, displaying at least one feature node; each The feature node is determined based on at least one fact node associated with it, and the feature node contains a description of personalized information reflected by at least one fact node associated with it, and the fact node contains a description of at least one semantic unit in the source data. In response to a first triggering operation on a target feature node in the interactive interface, the fact node associated with the target feature node is displayed in the interactive interface in a first manner; The first method is a way to enhance the visual correlation between the target feature node and its associated fact nodes.

2. The method of claim 1, wherein, While displaying the at least one feature node in the interactive interface, at least one fact node is also displayed; The fact nodes and feature nodes of different categories in the interactive interface are displayed in different visual styles or in different display areas.

3. The method of claim 1, wherein, The interactive interface displays the fact nodes associated with the target feature node in a first manner, including at least one of the following implementation methods: The interactive interface displays, for the first time, the fact nodes associated with the target feature node in a first-view style; Change the second visual style of the fact nodes associated with or not associated with the target feature node already displayed in the interactive interface; Change the visual style of the node association relationship between the target feature node and its associated fact nodes in the interactive interface. Change the spatial layout relationship between the target feature node and its associated fact nodes in the interactive interface.

4. The method of claim 2, wherein, Different categories of feature nodes contain descriptions of different personalized information in user profile information, user preference information, and user characteristic information; Different categories of feature nodes have association relationships obtained from the analysis of the fact nodes associated with the corresponding feature nodes; The method further includes: In response to a second triggering operation on a target feature node in the interactive interface, at least one non-target feature node that is associated with the target feature node is displayed in the interactive interface in a second manner; the non-target feature node and the target feature node are of the same category or different categories. The second approach is a method that can enhance the visual correlation between the target feature node and the at least one non-target feature node.

5. The method according to any one of claims 1 to 4, wherein, The fact node is associated with the source data fragment corresponding to the semantic unit it contains, and the source data fragment comes from the source data; The description of the semantic unit is an event description of the context memory in the corresponding source video segment, a conceptual description of the corresponding semantic memory, or an operational description of the corresponding procedural memory. The method further includes: In response to a trigger operation on the target fact node displayed in the interactive interface, the source data fragment associated with the target fact node is displayed in the interactive interface.

6. The method according to claim 5, wherein displaying the source data fragment associated with the target fact node in the interactive interface includes at least one of the following methods: In response to the source data segment associated with the target fact node containing an audio or video segment, a playback control is initiated in the interactive interface, and the audio or video segment is played through the playback control; In response to the source data fragment associated with the target fact node containing text content, the text content is displayed in the interactive interface in a differentiated visual style.

7. The method according to any one of claims 1-4, further comprising: In response to a third triggering operation on a target fact node in the interactive interface, at least one of the feature nodes associated with the target fact node is displayed in the interactive interface in a third manner. The third method is a way to enhance the visual correlation between the target fact node and its associated feature nodes.

8. The method according to any one of claims 1-4, further comprising: At least one node category identifier is displayed in the interactive interface; Each node category identifier is associated with each node of the corresponding category, and all the fact nodes are nodes of the same category; In response to a fourth trigger operation on the target node category identifier in the interactive interface, each node of the target category associated with the target node category identifier is displayed in the interactive interface in a fourth manner; The fourth method is a way to achieve visual differences between nodes of the target category and nodes of non-target categories.

9. The method according to any one of claims 1-4, wherein, The process of determining each feature node based on at least one fact node associated with it includes: Based on the first model, at least one source data is processed to generate multiple fact nodes; the source data includes at least one of text data, audio data, video data, and image data, and the video data includes screen content in the video; The multiple fact nodes are processed based on the second model, and at least one feature node is generated by aggregation; each feature node is associated with the fact node that generated the feature node.

10. An electronic device, the electronic device comprising: At least one memory, and a computer program stored in the memory; At least one processor capable of running intelligent agents; The intelligent agent can execute the computer program through the processor to implement the steps of the interaction method as described in any one of claims 1 to 9.

Images