Method for managing memory and electronic apparatus

Abstract

An operation method of an electronic apparatus according to an embodiment of the present disclosure comprises: a step for acquiring pairs of interaction data related to respective interaction events and associated data related to the interaction data in response to the occurrence of the interaction events between an agent application installed in an electronic apparatus and a user; a step for storing, in a memory, the pairs of the interaction data related to the respective interaction events and the associated data; a step for determining, using an AI model, priority scores for the pairs of the interaction data related to the respective interaction events and the associated data; a step for storing the priority scores in the memory; and a step for managing the pairs of the interaction data and the associated data stored in the memory on the basis of the priority scores if the capacity of the memory exceeds the critical ratio.

Description

Methods for managing memory and electronic devices

[0001] The present disclosure relates to a method and electronic device for managing memory, and more specifically, to a method for managing pairs of interaction data and associated data related to the interaction data stored in memory.

[0002] Recent advancements in Artificial Intelligence (AI) are progressing toward AI agents driven by improvements in the performance of Large Language Models (LLMs). AI agent services recognize user requests, environments, and situations, and perform functions such as decision-making, planning, and task execution.

[0003] As AI agents handle various functions, the frequency of interactions with users is increasing rapidly, and the volume of interaction data is skyrocketing. Furthermore, while providing standardized responses was common in the past, personalization by AI agents is now emerging as a critical requirement. In other words, generating appropriate responses tailored to interactions with different users has become a key challenge.

[0004] However, due to the limitations of limited memory size in electronic devices, which makes it difficult to store all specific data about individual users and accurately identify their preferences, AI agents generally store partial information based on predefined rules and provide differentiated responses accordingly.

[0005] A method of operation of an electronic device according to one embodiment of the present disclosure includes the step of acquiring pairs of interaction data and associated data related to each interaction event in response to the occurrence of an interaction event between an agent application mounted on the electronic device and a user. The method of operation of the electronic device includes the step of storing the pairs of interaction data and associated data related to each interaction event in memory. The method of operation of the electronic device includes the step of determining a priority score for the pairs of interaction data and associated data related to each interaction event using an artificial intelligence model. The method of operation of the electronic device includes the step of managing the pairs of interaction data and associated data stored in memory based on the priority score when the capacity of the memory exceeds a threshold ratio.

[0006] An electronic device according to one embodiment of the present disclosure comprises: a communication interface; a memory for storing at least one instruction; and at least one processor including a circuit device. By executing at least one instruction individually or collectively by at least one processor, the electronic device acquires pairs of interaction data and associated data associated with each interaction event in response to the occurrence of an interaction event between an agent application mounted on the electronic device and a user. By executing at least one instruction individually or collectively by at least one processor, the electronic device stores the pairs of interaction data and associated data associated with each interaction event and priority scores in memory. By executing at least one instruction individually or collectively by at least one processor, the electronic device determines priority scores for the pairs of interaction data and associated data associated with each interaction event using an artificial intelligence model. By executing at least one instruction individually or collectively by said at least one processor, the electronic device manages the pairs of interaction data and associated data stored in memory based on priority scores when the capacity of memory exceeds a threshold ratio.

[0007] According to one embodiment of the present disclosure, a computer-readable recording medium is provided on which a program for performing any one of the methods described above or below is recorded.

[0008] The present invention can be easily understood from the combination of the following detailed description and the accompanying drawings, where reference numerals denote structural elements.

[0009] FIG. 1 is a schematic diagram of a method of operation of an electronic device according to one embodiment of the present disclosure.

[0010] FIG. 2 is a flowchart for explaining the operation of an electronic device according to one embodiment of the present disclosure.

[0011] FIG. 3a is a reference diagram for explaining interaction data and association data according to one embodiment of the present disclosure.

[0012] FIG. 3b is a reference diagram for explaining the operation of connecting pairs of interaction data and associated data according to one embodiment of the present disclosure.

[0013] FIG. 4 is a reference diagram for explaining a method of storing interaction data according to one embodiment of the present disclosure.

[0014] FIG. 5a is a reference diagram for explaining a method for determining priority scores according to one embodiment of the present disclosure.

[0015] FIG. 5b is a reference diagram for explaining a method of determining priority scores through a weighted sum according to one embodiment of the present disclosure.

[0016] FIG. 5c is a reference diagram for explaining a method of updating a priority score (530) according to one embodiment of the present disclosure.

[0017] FIG. 6 is a reference diagram for explaining a method for determining priority scores according to one embodiment of the present disclosure.

[0018] FIG. 7 is a reference diagram for explaining a method of managing the memory of an electronic device according to one embodiment of the present disclosure.

[0019] FIG. 8 is a flowchart illustrating an operation to merge pairs of interaction data and association data according to one embodiment of the present disclosure.

[0020] FIG. 9 is a reference diagram for further explaining the operation of merging pairs of interaction data and association data according to one embodiment of the present disclosure.

[0021] FIG. 10 is a flowchart illustrating a method for an electronic device to output response information through an agent application according to one embodiment of the present disclosure.

[0022] FIG. 11 is a reference diagram illustrating an example in which an electronic device outputs response information through an agent application according to one embodiment of the present disclosure.

[0023] FIG. 12 is a block diagram illustrating the configuration of an electronic device according to one embodiment of the present disclosure.

[0024] The various embodiments of this document and the terms used therein are not intended to limit the technical features described in this document to specific embodiments, and should be understood to include various modifications, equivalents, or substitutions of said embodiments.

[0025] In relation to the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of the noun corresponding to an item may include one or more of said items unless the relevant context clearly indicates otherwise.

[0026] In this document, each of the phrases such as "A or B", "at least one of A and B", "at least one of A or B", "A, B or C", "at least one of A, B and C", and "at least one of A, B, or C" may include any one of the items listed together in the corresponding phrase, or all possible combinations thereof.

[0027] The term “and / or” includes a combination of multiple related described components or any of the multiple related described components.

[0028] Terms such as "first," "second," or "first" or "second" may be used simply to distinguish a component from another corresponding component and do not limit the components in other aspects (e.g., importance or order).

[0029] Where any (e.g., 1st) component is referred to as "coupled" or "connected" to another (e.g., 2nd) component, with or without the terms "functionally" or "communicationly," it means that said any component may be connected to said other component directly (e.g., via a wire), wirelessly, or through a third component.

[0030] Terms such as “include” or “have” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in this document, and do not preclude the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0031] When it is said that one component is “connected,” “combined,” “supported,” or “in contact” with another component, this includes not only cases where the components are directly connected, combined, supported, or in contact, but also cases where they are indirectly connected, combined, supported, or in contact through a third component.

[0032] When it is said that a component is located “on” another component, this includes not only cases where one component is in contact with the other, but also cases where another component exists between the two components.

[0033] It should be understood that the blocks in each flowchart and combinations of flowcharts can be executed by one or more computer programs containing computer-executable instructions. One or more computer programs may be stored all in a single memory or may be partitioned and stored in multiple different memories.

[0034] One embodiment of the present disclosure may be represented by functional block configurations and various processing steps. Some or all of these functional blocks may be implemented by various numbers of hardware and / or software configurations that execute specific functions. For example, the functional blocks of the present disclosure may be implemented by one or more microprocessors or by circuit configurations for a specific function. Additionally, for example, the functional blocks of the present disclosure may be implemented in various programming or scripting languages. The functional blocks may be implemented as algorithms executed on one or more processors. Furthermore, the present disclosure may employ prior art for electronic configuration, signal processing, and / or data processing, etc.

[0035] All functions or operations, including functions related to artificial intelligence according to the present disclosure, are operated through a processor and memory. The processor may be composed of one or more processors. A single processor or a combination of processors may include circuitry that performs processing, such as an AP (Application Processor), CP (Communication Processor), GPU (Graphical Processing Unit), NPU (Neural Processing Unit), MPU (Microprocessor Unit), SoC (System on Chip), IC (Integrated Chip), etc.

[0036] The present disclosure will be described in detail below with reference to the attached drawings.

[0037] FIG. 1 is a schematic diagram of a method of operation of an electronic device according to one embodiment of the present disclosure.

[0038] The electronic device (100) may be a device that identifies interaction events between an application mounted on the electronic device (100) and a user. The electronic device (100) may be implemented in various forms. The electronic device (100) may be any form of device that performs functions including a processor and memory. The electronic device (100) may be a stationary or portable device. For example, the electronic device (100) may include, but is not limited to, a digital TV, an AR (Augmented Reality) device, a VR (Virtual Reality) device, a vehicle control unit (e.g., Advanced Driver-Assistance System (ADAS)), a wearable device, a mobile terminal (e.g., a smartphone, a tablet PC), or a PC.

[0039] The electronic device (100) may be equipped with an agent application. In the present disclosure, the "agent application" or "agent" may represent a software program that is executed on the electronic device (100), processes user input, and provides a response corresponding to the user. The agent application may utilize artificial intelligence technologies such as speech recognition, natural language processing, and machine learning to understand user requests and provide appropriate information or perform specific functions. Additionally, the agent application may represent a program that not only performs simple commands but also learns the user's usage patterns to actively provide services and information. The agent application may be described as a virtual assistant or an artificial intelligence assistant.

[0040] According to one embodiment of the present disclosure, a user may converse with an agent application and receive a response from the agent application. For example, the user may operate the agent application and receive a service corresponding to the user's operation from the agent application. In the present disclosure, "user" may refer to a person who uses the agent application. The user may use the agent application. The user may interact with the agent application in various ways by utilizing the agent application. For example, the user may converse with the agent application.

[0041] According to one embodiment of the present disclosure, a user may request information or services from an agent application and receive the requested information or services from the agent application. For example, the user may execute a specific function of an electronic device (100) by inputting a command into the agent application. Additionally, the user may perform a desired task by directly operating the agent application, and accordingly, the agent application may provide a service or function that appropriately responds to the user's operation.

[0042] The electronic device (100) can identify that an interaction event (10) occurs between an agent application and a user. According to one embodiment, the 'interaction event' may represent a dialogue or action occurring between an agent application mounted on the electronic device (100) and a user. For example, the interaction event may include not only actions where the user manipulates the agent application, but also the response result of the agent application according to said actions.

[0043] According to one embodiment of the present disclosure, an interaction event may occur regardless of how the user uses the agent application. For example, the interaction event may represent two-way communication that takes place when the user uses or operates the agent application, or when the agent application provides information or services to the user. Here, 'operate' may include not only individual UI actions such as simple clicks or button presses, but also the entire process of actually utilizing various functions within the application.

[0044] According to one embodiment of the present disclosure, the interaction event may include various types of events. For example, command input through user voice recognition, text input using a keyboard and touch screen, button click or menu selection of an application UI, screen scrolling of an electronic device (100), menu selection, execution of application functions for reservation, search, payment, etc., and state change or notification display of an agent application according to the user environment (location, sensor information, etc.) may also be included in the interaction event.

[0045] In one embodiment of the present disclosure, in response to the occurrence of an interaction event between an agent application and a user, an electronic device (100) may acquire interaction data associated with each interaction event. For example, the electronic device (100) may acquire at least one of the content of the interaction event, the time of occurrence of the interaction event, or the location information of the electronic device (100) at the time of the interaction event as interaction data.

[0046] In one embodiment of the present disclosure, in response to the occurrence of an interaction event between an agent application and a user, an electronic device (100) may acquire associated data related to the interaction data. For example, the associated data may include factual information representing actual objective facts corresponding to information included in the interaction event, usage history information related to the user's agent application or interaction data, and user response information included in the interaction data.

[0047] The electronic device (100) can acquire pairs (20) of interaction data and associated data. The electronic device (100) can store pairs of interaction data and associated data associated with each interaction event in memory (or, database (30)). In one embodiment of the present disclosure, the pairs of interaction data and associated data may correspond to memory chunks. The pairs of interaction data and associated data may correspond to interaction data storage units in memory.

[0048] The electronic device (100) can determine (40) priority scores for pairs of interaction data and associated data related to each interaction event using an artificial intelligence model (50). The priority scores may indicate how important the user considers the interaction data to be, based on predetermined criteria, the user's interests over time, or the user's response intensity. The electronic device (100) can store the determined priority scores in memory.

[0049] When the capacity of the memory exceeds a threshold ratio, the electronic device (100) can manage (60) pairs of interactive data and related data stored in memory based on priority scores. That is, the electronic device (100) can organize data of low importance to efficiently utilize storage space. For example, memory can be optimized by deleting pairs with low priority scores to efficiently manage the storage space of memory. For example, data that is frequently viewed or repeatedly interacted with by a user can be continuously stored while maintaining a high priority score, while data that is rarely used can be automatically deleted after a certain period. In addition, rather than simply deleting data based on priority scores, methods such as merging highly related data or summarizing and storing only essential information can also be utilized.

[0050] The electronic device (100) can provide a customized experience by maintaining data optimized for individual users by reflecting the user's preference, response intensity, and frequency of use, thereby providing response information (70) tailored to the user's characteristics when the user subsequently inputs actions or conversations. Below, we will examine in detail the method of the electronic device (100) according to an embodiment of the present disclosure managing pairs of interaction data and associated data, providing response information based on the managed pairs, and the electronic device (100) performing the same.

[0051] FIG. 2 is a flowchart for explaining the operation of an electronic device according to one embodiment of the present disclosure.

[0052] Referring to FIG. 2, the method of an electronic device (100) managing memory may include operations 210 to 250. In one embodiment of the present disclosure, operations 210 to 250 may be executed by at least one processor included in the electronic device (100). The method of an electronic device (100) managing memory is not limited to that illustrated in FIG. 2, and in one or more embodiments, operations not illustrated in FIG. 2 may be further included, or some operations may be omitted.

[0053] In operation 210, the electronic device (100) can identify interaction events occurring between an agent application mounted on the electronic device (100) and a user.

[0054] In one embodiment of the present disclosure, the electronic device (100) can identify that a user is conversing with an agent application. For example, the electronic device (100) can identify that the user is performing at least one of question-and-answer with the agent application via voice commands, input via a text-based chat interface, or interaction on a graphical user interface.

[0055] In one embodiment of the present disclosure, an electronic device (100) can identify an action in which a user operates an agent application. For example, the electronic device (100) can identify (or detect) at least one of an action in which a user clicks the user interface of the agent application, presses a button, touches the screen, or selects a menu. For example, the electronic device (100) can identify (or detect) at least one of an action in which a user inputs a keyboard of the electronic device (100) equipped with the agent application, inputs a gesture using a touchscreen or camera, or drags or scrolls the touchscreen. For example, the electronic device (100) can identify an action in which a user performs a specific function such as making a reservation, searching, or making a payment in the agent application.

[0056] In one embodiment of the present disclosure, the electronic device (100) can acquire interaction data associated with each interaction event. In one embodiment of the present disclosure, the electronic device (100) can acquire interaction data associated with each interaction event in response to the occurrence of an interaction event between an agent application mounted on the electronic device and a user.

[0057] FIG. 3a is a reference diagram for explaining interaction data and association data according to one embodiment of the present disclosure.

[0058] Referring to the interaction data (310) of FIG. 3, the interaction data (310) may include at least one of the content of the interaction event (312), the time of occurrence of the interaction event (314), or the location information (316) of the electronic device (100) at the time of the interaction event. The content of the interaction event (312) may include at least one of the content of a conversation or command entered by the user into the agent application or the content of an action performed by the user to operate the agent application, but is not limited thereto. The time of occurrence of the interaction event (314) may include at least one of the date information or time information at the time the interaction event occurred, but is not limited thereto. The location information (316) of the electronic device (100) at the time of the interaction event may include at least one of the country information, city information, or address information where the electronic device (100) is located at the time the interaction event occurred, but is not limited thereto.

[0059] For example, if a user inputs a conversation (e.g., text input, voice input) to an agent application saying "What is the weather like today?" and the agent application outputs a response saying "It is very cold today," the electronic device (100) can obtain at least one of the content of the conversation or the response of the agent application as the content of the interaction event (312). Additionally, the electronic device (100) can obtain at least one of the time when the user inputs the conversation to the agent application or the time when the agent application responds as the time of occurrence of the interaction event (314). The electronic device (100) can obtain the location of the electronic device (100) at the time the user inputs the conversation to the agent application or the time when the agent application outputs the response (e.g., home, office, out, specific location value, etc.) as the location information (316) of the electronic device (100) at the time the interaction event occurs.

[0060] For example, if a user changes the wake-up alarm time to 7:00 AM in the "Alarm Settings" menu of the agent application, the electronic device (100) can obtain the notification time change event as the content (312) of the interaction event. For example, the electronic device (100) can obtain the time when the user changed the alarm time as the time of occurrence (314) of the interaction event. For example, the electronic device (100) can obtain the location of the electronic device (100) at the time when the user changed the alarm time as the location information (316) of the electronic device (100) at the time of occurrence of the interaction event.

[0061] According to one embodiment of the present disclosure, an electronic device (100) may collect all data related to an interaction event as interaction data. According to one embodiment of the present disclosure, the electronic device (100) may collect some or all of the data related to the interaction event as summarized data through an artificial intelligence model. For example, the electronic device (100) may input all data related to the interaction event into an artificial intelligence model and obtain summarized data output from the artificial intelligence model as interaction data.

[0062] In one embodiment of the present disclosure, an electronic device (100) may acquire associated data associated with each interaction event in response to the occurrence of an interaction event between an agent application mounted on the electronic device (100) and a user. Referring to the associated data (320) of FIG. 3, the associated data (320) may include at least one of fact information (322), usage history information (324), or reaction information (326), but is not limited thereto.

[0063] In one embodiment of the present disclosure, an electronic device (100) may obtain fact information (322) representing actual facts corresponding to information included in an interaction event as associated data. For example, the electronic device (100) may identify (or search for) actual facts (e.g., objective facts, universal facts, or general common sense) corresponding to information included in an interaction event from an external database or external server through a communication interface included in the electronic device (100), and obtain the identified actual facts (e.g., objective facts, universal facts, or general common sense) as fact information (322). If the information included in the interaction event contains information different from actual facts (e.g., objective facts, universal facts, or general common sense), the information included in the interaction event and the fact information (322) may not match each other. The fact information (322) may represent information stored by taking into account the difference between the content entered by the user and the actual objective facts. That is, if the user expresses something differently from reality, the electronic device (100) can obtain actual facts (e.g., objective facts, universal facts, or general common sense) as factual information (322) by utilizing a reliable database or external information source through a communication interface.

[0064] For example, when a user inputs "China is the largest country in the world" into an agent application and an interaction event occurs, the electronic device (100) can refer to an international geographic information database through a communication interface to confirm that the country with the largest area in the world is Russia. The electronic device (100) can obtain the actual fact that "Russia is the largest country in the world, and China is the fourth largest country in the world" as factual information (322).

[0065] In one embodiment of the present disclosure, the electronic device (100) may obtain user usage history information (324) regarding a subject or agent application related to an interaction event as associated data.

[0066] The usage history information (324) may represent information generated based on records of interactions between the user and the agent application. For example, the usage history information (324) may include at least one of a record of conversations the user had with the agent application regarding a specific topic or a pattern of repeated searches the user performed using the agent application regarding a specific topic. For example, the usage history information (324) may include at least one of a history of the user running the agent application (e.g., time and frequency) or a history of the user using a specific function of the agent application.

[0067] In one embodiment of the present disclosure, the electronic device (100) may obtain user usage history information (324) regarding a topic related to an interaction event as associated data. For example, the electronic device (100) may identify a topic included in a conversation and obtain usage history information (324) regarding the user's conversation with the agent application about the identified topic.

[0068] For example, when a user inputs a conversation "What is the weather like today?" into an agent application, the electronic device (100) identifies that the topic of the conversation is weather and can obtain or identify a history of the user having previously had a conversation related to weather with the agent application. For example, the electronic device (100) can obtain usage history information (324) that has asked questions such as "When is the rainy day?" or "What is the weather like this weekend?" over the past few days.

[0069] In one embodiment of the present disclosure, an electronic device (100) may obtain user usage history information (324) for an agent application. For example, the electronic device (100) may identify an agent application where an interaction event occurred and obtain user usage history information (324) for the identified agent application.

[0070] For example, if a user inputs a command to an agent application saying, "Play a song similar to the song I recently listened to," the electronic device (100) can obtain or identify usage history information (324) about the song recently played by the user from the agent application.

[0071] In one embodiment of the present disclosure, the electronic device (100) may obtain user response information (326) based on at least one of the user's response intensity, the user's positive degree, or the user's negative degree included in the interaction data.

[0072] In one embodiment of the present disclosure, an electronic device (100) may obtain user response information (326) by identifying at least one of the user's response intensity, the user's degree of positivity, or the user's degree of negativity included in the interaction data in various ways. For example, the electronic device (100) may obtain user response information (326) by identifying the user's response intensity based on at least one of the user's voice loudness (volume), speech speed, or intonation change included in the interaction data. The electronic device (100) may obtain user response information (326) that the response intensity is high if the volume of the user's voice included in the interaction data is loud or the speed is fast. The electronic device (100) may obtain user response information (326) that the response intensity is low if the volume of the user's voice included in the interaction data is quiet or the speed is slow.

[0073] For example, the electronic device (100) can obtain user response information (326) by identifying the user’s response intensity, the user’s positive degree, or the user’s negative degree based on at least one of the keywords, emojis, or punctuation marks included in the interaction data. For example, the electronic device (100) can obtain user response information (326) by identifying at least one of the keywords “Oh my,” “What is this,” or the emoji “ㅠㅠ” included in the interaction data “Oh my, what is this weather today? ㅠㅠ”, and identifying that the user’s response intensity is high and the user’s negative degree is high.

[0074] For example, the electronic device (100) can obtain user response information (326) by identifying the user's response intensity, the user's positive degree, or the user's negative degree based on at least one of the user's response speed or the number of times the user responds in interaction data that manipulates a user interface. For example, the electronic device (100) can obtain user response information (326) by identifying that the user's response intensity is high because the user clicked the button quickly multiple times and the user's positive degree is high because the user pressed the Like button, based on interaction data where the user clicked the Like button quickly multiple times.

[0075] In one embodiment of the present disclosure, the electronic device (100) may obtain pairs (330) of interaction data and associated data. In one embodiment of the present disclosure, the electronic device (100) may obtain the interaction data described above and pairs of associated data corresponding to the interaction data.

[0076] Returning to FIG. 2, in operation 220, the electronic device (100) may store pairs of interaction data and associated data associated with each interaction event in memory. In one embodiment of the present disclosure, the memory may be a memory included in the electronic device (100). According to one embodiment of the present disclosure, the method by which the electronic device (100) stores interaction data will be described in more detail later with reference to FIG. 4.

[0077] In operation 230, an electronic device (100) according to one embodiment of the present disclosure can determine priority scores for pairs of interaction data and associated data associated with each interaction event using an artificial intelligence model.

[0078] In one embodiment of the present disclosure, the artificial intelligence model may be a learning-based model that analyzes user interaction data and association data and evaluates the importance of pairs of interaction data and association data. For example, the artificial intelligence model may include deep learning, machine learning, rule-based algorithms, or large language models. The artificial intelligence model may be included in an agent application or may exist separately from the agent application.

[0079] In one embodiment of the present disclosure, a priority score may represent a value indicating the importance of pairs of interaction data and association data. In one embodiment of the present disclosure, an electronic device (100) may determine a priority score for pairs of interaction data and association data based on at least one of a predefined criterion, a user's interest that changes over time, and a user's individual characteristics.

[0080] In one embodiment of the present disclosure, the electronic device (100) may update a priority score stored in memory based on satisfying an update condition indicating a condition for updating the priority score. For example, the update condition may include at least one of the idle state of the electronic device (100), a specified time, or a specified location. According to one embodiment of the present disclosure, the method by which the electronic device (100) determines the priority score will be described in more detail with reference to FIGS. 5 and 6.

[0081] An electronic device (100) according to one embodiment of the present disclosure may store priority scores in memory. In one embodiment of the present disclosure, the memory may be a memory included in the electronic device (100). In one embodiment of the present disclosure, the memory may be a memory such as a memory in which pairs of interaction data and association data are stored, or a different memory. According to one embodiment of the present disclosure, the electronic device (100) may store priority scores in memory by mapping them to pairs of interaction data and association data, or may store priority scores in memory separately from pairs of interaction data and association data.

[0082] In operation 240, an electronic device (100) according to one embodiment of the present disclosure can identify whether the capacity of memory exceeds a threshold ratio. In one embodiment of the present disclosure, the threshold ratio may represent a value that serves as a memory management standard for the electronic device (100). The threshold ratio may be a value set in advance or a value set by a user. For example, if the threshold ratio is 80%, the electronic device (100) can monitor whether the capacity of memory exceeds 80%.

[0083] In one embodiment of the present disclosure, the electronic device (100) can identify whether the capacity of the memory exceeds a threshold ratio at regular intervals. For example, the electronic device (100) can identify whether the capacity of the memory exceeds a threshold ratio at 2 a.m. every day.

[0084] In one embodiment of the present disclosure, the electronic device (100) can identify whether the capacity of memory exceeds a threshold ratio whenever new interaction data is added. For example, when a new interaction event occurs, the electronic device (100) can identify whether the capacity of memory exceeds a threshold ratio before storing pairs of interaction data and associated data related to the new interaction event.

[0085] According to one embodiment of the present disclosure, the electronic device (100) can continue to collect and store pairs of interaction data and associated data related to interaction events in memory when the capacity of memory does not exceed a threshold ratio (No of operation 250).

[0086] In operation 250, an electronic device (100) according to one embodiment of the present disclosure may manage pairs of interaction data and associated data stored in memory based on a priority score when the capacity of memory exceeds a threshold ratio (Yes of operation 250). In one embodiment of the present disclosure, the electronic device (100) may determine whether to maintain or delete the storage state for each pair of interaction data and associated data stored in memory.

[0087] For example, the electronic device (100) may delete at least one pair of interaction data and associated data with a low priority score. For example, the electronic device (100) may delete at least one pair of interaction data and associated data with a low priority score until the capacity of the memory is below a threshold ratio. According to one embodiment of the present disclosure, the method by which the electronic device (100) manages memory will be described in more detail later with reference to FIG. 7.

[0088] Through operations according to one embodiment of the present disclosure, the electronic device (100) can identify the context of the interaction data by acquiring associated data related to the interaction data. The electronic device (100) can efficiently manage information stored in memory by reflecting user characteristics by managing pairs of interaction data and associated data based on priority scores. The electronic device (100) can establish an efficient memory management system that reflects the individual characteristics and preferences of the user, rather than simply accumulating and storing data upon the occurrence of an interaction event and deleting data according to a set rule.

[0089] FIG. 3b is a reference diagram for explaining the operation of connecting pairs of interaction data and associated data according to one embodiment of the present disclosure.

[0090] According to one embodiment of the present disclosure, an electronic device (100) can connect a pair (340) of interaction data and association data with at least one related pair (350).

[0091] In one embodiment of the present disclosure, when an electronic device (100) acquires a pair (340) of interaction data and association data, it may identify (or search, extract) at least one pair (350) associated with the acquired pair (340) in memory (120). For example, the electronic device (100) may identify at least one pair (350) associated with the interaction data (342) included in the acquired pair (450) (e.g., at least one of the content of the interaction data, the time at which the interaction data occurred, and the location at which the interaction data occurred) in memory (120). For example, the electronic device (100) may identify at least one pair (350) associated with the association data (342) included in the acquired pair (340) (e.g., at least one of fact information, usage history information, or reaction information) in memory (120).

[0092] In one embodiment of the present disclosure, when an electronic device (100) acquires a pair (340) of interaction data and association data, it can identify (or search, extract) the similarity between the acquired pair (340) and the pairs (350, 360, 370) stored in memory. The electronic device (100) can identify (or search, extract) the similarity between the pair (340) acquired by the electronic device (100) and the pairs (350, 360, 370) stored in memory based on at least one of the interaction data (342) or the association data (344).

[0093] In one embodiment of the present disclosure, an electronic device (100) may connect (380) a pair (340) of acquired interaction data and associated data with at least one associated pair (350). The electronic device (100) may connect (380) the acquired pair (340) and the stored pair (350) by utilizing a data structure. For example, the electronic device (100) may create links between the associated pairs to connect them. The electronic device (100) may connect the associated pairs by utilizing a node and edge structure. The electronic device (100) may connect the associated pairs by key-value mapping or hierarchically configuring them. The electronic device (100) may apply various data structures such as a graph database, a hash table, or a tree structure, but the present disclosure is not limited to the described embodiments.

[0094] For example, when an interaction event occurs in which a user performs multiple searches on the electronic device (100) on the topic of "overseas travel," the electronic device (100) can store interaction data and related data corresponding to the interaction event. Subsequently, when an interaction event occurs in which flight information for a specific region is searched on the electronic device (100), the electronic device (100) can connect a pair corresponding to the interaction event "overseas travel search" with a pair corresponding to the interaction event "searching flight information for a specific region." The electronic device (100) can determine that the two pairs are related and connect the pairs based on the content of the interaction data included in the pair corresponding to the interaction event "overseas travel search" and the content of the interaction data included in the pair corresponding to the interaction event "searching flight information for a specific region." Accordingly, the electronic device (100) can quickly search for information included in the pair corresponding to the "overseas travel search."

[0095] In this way, the electronic device (100) can improve data retrieval speed and optimize memory utilization by connecting and storing related pairs, and provide an environment that facilitates future exploration and utilization of related information.

[0096] FIG. 4 is a reference diagram for explaining a method of storing interaction data according to one embodiment of the present disclosure.

[0097] According to one embodiment of the present disclosure, the electronic device (100) can store interaction data in various formats.

[0098] In one embodiment of the present disclosure, the electronic device (100) may store interaction data associated with each interaction event. For example, the electronic device (100) may store original interaction data associated with each interaction event. For example, the electronic device (100) may store all interaction data associated with each interaction event without compression.

[0099] In one embodiment of the present disclosure, the electronic device (100) may store all interaction data associated with high-importance interaction events (e.g., without summarizing). For example, the electronic device (100) may store all interaction data associated with each interaction event (e.g., without summarizing) in at least one of the following situations: when the loss of interaction data is critical, when a high-importance conversation is taking place, or when the user is expected to stay on one screen for a long time.

[0100] Referring to the first storage method (410) of FIG. 4, the electronic device (100) can store the contents of the interaction event (400), such as user input (402) (e.g., "I have an important meeting in the morning but I can't sleep... I'm worried I won't be able to wake up, so please set an alarm for early morning") and the response (404) of the agent application (hereinafter referred to as the agent response) (e.g., "I have set an alarm for 6 AM."). The electronic device (100) can store the time of occurrence of the interaction event (400) (e.g., 1:00 AM) and the location of occurrence of the interaction event (e.g., home) in response to the occurrence of the interaction event (400). The electronic device (100) can store the processing content of the agent application corresponding to the interaction event (e.g., NLU analysis result → alarm request alarm database update: (time=06:00, title="Notification: User request")) in response to the occurrence of an interaction event (400).

[0101] In one embodiment of the present disclosure, an electronic device (100) may store pairs of collected interaction data and associated data corresponding to the collected interaction data in memory. For example, the electronic device (100) may store first associated data corresponding to the interaction data stored according to a first storage method (410) in memory. For example, the electronic device (100) may store at least one of the user's previous alarm setting history or the user's wake-up and sleep time history as first associated data in memory. For example, the electronic device (100) may store at least one of the user's reaction information, the user's positive intensity, or the user's negative intensity included in the content of the interaction data stored according to the first storage method (410) as first associated data in memory.

[0102] In one embodiment of the present disclosure, the electronic device (100) can summarize and store interaction data related to each interaction event through an artificial intelligence model.

[0103] In one embodiment of the present disclosure, the electronic device (100) may summarize and store interaction data related to interaction events of low importance. For example, the electronic device (100) may summarize and store interaction data related to each interaction event in at least one of the following situations: when the intermediate process of the interaction data is not important, or when a conversation or action of low importance is performed.

[0104] In one embodiment of the present disclosure, the electronic device (100) can store interaction data related to each interaction event in a specified format by using an artificial intelligence model. The electronic device (100) can extract some data as summary data from the entire data related to the interaction event, or obtain summary data by processing the entire data related to the interaction event. For example, when an interaction event occurs between an agent application and a user, the electronic device (100) can use an artificial intelligence model to extract summary data from the entire data related to the interaction event that indicates how the user's action or conversation started (Starting Point), went through what process (Actions), and finally transitioned to what state (End Point).

[0105] For example, when an interaction event occurs between an agent application and a user, the electronic device (100) can use an artificial intelligence model to summarize the entire data related to the interaction event in a 5W1H format.

[0106] In one embodiment of the present disclosure, the electronic device (100) can summarize and store interaction data related to each interaction event by storing at least one of the state of the agent application at the start of the interaction event, the main action or conversation that occurred in the interaction event, or the state of the agent application at the end of the interaction event.

[0107] For example, the electronic device (100) can extract data related to the entry into the settings menu of the agent application (Start Point), the change of options (Action), and the state of returning to the home screen of the agent application (End Point) in an interaction event where the user changes a specific option on the settings screen of the agent application and then returns to the home screen, and can summarize and store the interaction data by excluding unnecessary data such as intermediate screen movement information.

[0108] Referring to the second storage method (420) of FIG. 4, the electronic device (100) can store the content of an interaction event, such as a summary of user input ("Set an alarm for early morning") or an agent response ("Set an alarm for 6 AM"), in response to the occurrence of an interaction event (400). The electronic device (100) can store the time of occurrence of the interaction event (AM 01:00) and the location of occurrence of the interaction event (home) in response to the occurrence of the interaction event (400).

[0109] In one embodiment of the present disclosure, the electronic device (100) may store pairs of summarized interaction data and associated data corresponding to the summarized interaction data in memory. Since the method of storing associated data overlaps with the method of storing associated data of interaction data according to the first storage method (410), it is omitted here.

[0110] In one embodiment of the present disclosure, the electronic device (100) may first accumulate and store interaction data associated with each interaction event, and then update the stored interaction data by summarizing and storing the stored interaction data.

[0111] FIG. 5a is a reference diagram for explaining a method for determining priority scores according to one embodiment of the present disclosure.

[0112] According to one embodiment of the present disclosure, an electronic device (100) can determine a priority score (530) for a pair (520) of interaction data and associated data related to the interaction event in response to the occurrence of an interaction event (510).

[0113] In one embodiment of the present disclosure, an electronic device (100) may determine a priority score (530) based on at least one of a predefined standard, a user's interests that change over time, and a user's individual characteristics. Hereinafter, an embodiment of the present disclosure will be described by designating a priority score based on a predefined standard as a first score (532), a priority score based on a user's interests that change over time as a second score (534), and a priority score based on a user's individual characteristics as a third score (536). However, the electronic device (100) is not required to determine all of the first score (532), the second score (534), and the third score (536) to determine the priority score (530), and the electronic device (100) may determine the priority score (530) based on at least one of the first score (532), the second score (534), or the third score (536).

[0114] In one embodiment of the present disclosure, the electronic device (100) may determine a first score (532) based on a predefined criterion. The electronic device (100) may determine the first score (532) based on at least one of the purpose of the agent application or general common sense.

[0115] For example, the electronic device (100) can determine a first score (532) based on a set rule based on the purpose of the agent application. For example, the electronic device (100) can evaluate the importance of data and assign a score according to a pre-set rule, taking into account the function performed by the agent application and the purpose of use.

[0116] For example, the electronic device (100) can identify the main function of the agent application and determine a first score (532) based on the extent to which interaction events affect the main function. For example, in an agent application for increasing work productivity, pairs of interaction data and associated data related to work productivity, such as schedule registration or file sharing, may receive a higher first score than pairs of interaction data and associated data related to scrolling actions or schedule checking. For example, in a social media-based agent application, pairs of interaction data and associated data related to social activities, such as user posting or sharing activities, may receive a higher first score (532) than pairs of interaction data and associated data related to viewing posts or moving categories.

[0117] For example, the electronic device (100) can determine a first score (532) based on common sense. The electronic device (100) can determine the first score (532) in a Commonsense (CS)-based manner. For example, the electronic device (100) can determine the importance of pairs of interaction data and association data using an artificial intelligence model that has learned common loss.

[0118] For example, the electronic device (100) may assign a high first score (532) to a specific data type based on general common sense. For example, the electronic device (100) may determine a high importance for pairs of interaction data and associated data that include information of generally high importance, such as personal information, birth date, death date, or user schedule.

[0119] In one embodiment of the present disclosure, the electronic device (100) may determine a first score (532) by combining the purpose of the agent application and general common sense. For example, when a user enters "My birthday is June 19" into an agent application for schedule management, the electronic device (100) may determine it as a general schedule registration based on the purpose of the agent application and assign a first score (532) of a moderate level, but may determine that the date of birth information is important personal information based on general common sense and adjust the first score (532) to a high value.

[0120] In one embodiment of the present disclosure, the electronic device (100) may determine a second score (534) based on the user's interests that change over time. By determining the second score (534), the electronic device (100) may reflect the user's interests that change over time.

[0121] In one embodiment of the present disclosure, the electronic device (100) may determine a second score (534) to be high for a pair containing an interaction event that has a high similarity to the most recently acquired interaction data among the pairs stored in memory. The most recently acquired interaction data may represent interaction data acquired during a specified recent period (e.g., three months, one month, or one week). In one embodiment of the present disclosure, the electronic device (100) may determine the second score (534) based on usage history information included in the associated data. The electronic device (100) may reflect the user's interests that change over time in the priority score by continuously updating the second score (534) (e.g., at regular intervals).

[0122] In one embodiment of the present disclosure, an electronic device (100) can determine the similarity between pairs stored in memory and the most recently acquired pairs. For example, the electronic device (100) can convert pairs stored in memory into vectors and determine the similarity (e.g., cosine similarity) between the vectors corresponding to the pairs stored in memory and the vectors corresponding to the most recently acquired pairs.

[0123] In one embodiment of the present disclosure, the electronic device (100) may determine a third score (536) based on the individual characteristics of the user. For example, the electronic device (100) may determine the third score (536) based on factual information included in the associated data and the user's reaction information.

[0124] In one embodiment of the present disclosure, the electronic device (100) determines the similarity between the content of the interaction data and the factual information, and can determine the third score (536) as a higher value as the similarity is lower. Since the factual information has been described with reference to FIG. 2, a description thereof will be omitted here.

[0125] For example, if the electronic device (100) has a high degree of similarity between the content of the interaction data and the factual information, it may determine that the content of the interaction data corresponds to general common sense or common knowledge and determine the third score (536) to a low value. For example, if the electronic device (100) has a low degree of similarity between the content of the interaction data and the factual information, it may determine that the user's individual characteristics (e.g., tendencies or perspectives) are reflected and assign a high third score (536) to the pair of interaction data and related data corresponding thereto.

[0126] In one embodiment of the present disclosure, the electronic device (100) may determine a third score (536) to a higher value as at least one of the user's reaction intensity, the user's positive intensity, or the user's negative degree increases, based on the user's reaction information. Since the reaction information has been described with reference to FIG. 2, a description thereof will be omitted here. The electronic device (100) may determine that the interaction reflects individual characteristics more as the user's reaction (i.e., emotional expression or opinion) is strongly revealed in the interaction data, and thus determine the importance corresponding to the pair of interaction data and related data to a higher value.

[0127] In one embodiment of the present disclosure, the electronic device (100) may obtain a priority score (530) by summing (540) at least one of a first score (532), a second score (534), or a third score (536).

[0128] FIG. 5b is a reference diagram for explaining a method of determining priority scores through a weighted sum according to one embodiment of the present disclosure.

[0129] In one embodiment of the present disclosure, the electronic device (100) may determine a priority score (530) by applying different weights to each of the first score (532), the second score (534), and the third score (536). The electronic device (100) may determine a priority score (530) by applying different weights to each of the predefined criteria, the user's interests, and the user's individual characteristics.

[0130] The electronic device (100) can identify a first weight (552) corresponding to a predefined criterion (or, a first score (532)), a second weight (554) corresponding to a user's interest (or, a second score (534)), and a third weight (556) corresponding to a user's individual characteristics (or, a third score (536)). The electronic device (100) can determine at least one of the first weight (552), the second weight (554), or the third weight (556) by receiving input from the user regarding the importance of each of the predefined criterion, the user's interest, or the user's individual characteristics through a user interface. The electronic device (100) can determine at least one of the first weight (552), the second weight (554), or the third weight (556) through an artificial intelligence model based on the type of agent application or the interaction data between the agent application and the user.

[0131] For example, when determining priority scores based on general knowledge, based on user preferences or characteristics of the agent application, the electronic device (100) may determine the first weight (552) to be higher than the second weight (554) or the third weight (556). For example, when determining priority scores based on the characteristics of the individual user, the electronic device (100) may determine the second weight (554) and / or the third weight (556) to be higher than the first weight (552).

[0132] The electronic device (100) can determine a priority score by combining a predefined standard, the user's interests, and the user's individual characteristics with a first weight (552), a second weight (554), and a third weight (556). The electronic device (100) can determine the priority score (530) by multiplying the first score (532) by the first weight (552), the second score (534) by the second weight (554), and the third score (536) by the third weight (556).

[0133] FIG. 5c is a reference diagram for explaining a method of updating a priority score (530) according to one embodiment of the present disclosure.

[0134] In one embodiment of the present disclosure, the electronic device (100) can obtain an updated priority score (580) by updating the priority score (570) when it satisfies an update condition (560) indicating a condition for updating the priority score. In one embodiment of the present disclosure, the electronic device (100) can maintain or change the priority score by updating the priority score.

[0135] In one embodiment of the present disclosure, the electronic device (100) can identify whether the update condition (560) is satisfied.

[0136] For example, the update condition (560) may include an idle state of the electronic device (100). In one embodiment of the present disclosure, the electronic device (100) requires a large amount of memory resources to calculate the priority score. The electronic device (100) can perform optimization work in the background without impairing the user experience by updating the priority score (570) while the electronic device (100) is not in use. For example, the electronic device (100) can update the priority score (570) when the user does not operate the electronic device for a certain period of time, or when certain conditions (e.g., charging, screen off, etc.) are met.

[0137] For example, the update condition (560) may include at least one of a specified time or a specified location. The electronic device (100) may update the priority score (570) at a specified location at a specified time interval. At least one of the specified time or the specified location may be specified by the user or set by the electronic device (100). The electronic device (100) may analyze the user's electronic device (100) usage pattern to identify at least one of the times or locations where the user rarely uses the electronic device (100) and designate this as the update condition. For example, the electronic device (100) may update the priority score (570) at the user's home between 3:00 AM and 5:00 AM every Tuesday.

[0138] In one embodiment of the present disclosure, the electronic device (100) may determine a priority score (580) corresponding to a pair of interaction data and association data corresponding to the interaction data based on at least one of a predefined criterion, a user's interest, or a user's individual characteristic, based on interaction data and association data obtained between the time when the existing priority score (530) is determined and the time when the priority score is updated. The electronic device (100) may optimize the priority score (580) by analyzing interaction data accumulated over a certain period and reflecting a pre-set criterion or a user's individual behavioral pattern.

[0139] In one embodiment of the present disclosure, the electronic device (100) may update at least one of the first score (532), the second score (534), or the third score (536) described above with reference to FIGS. 5a and 5b. The electronic device (100) may not update the first score (532), which is determined according to a predetermined standard, in order to optimize memory usage. The electronic device (100) may not update the third score (536), which is determined based on user response information, in order to optimize memory usage. The electronic device (100) may update the second score (534), which is determined based on the user's interests, in order to reflect changes in the user's interests over time.

[0140] In one embodiment of the present disclosure, when the electronic device (100) satisfies the update condition, the priority score can be re-determined by performing the operation of determining the priority score described above with reference to FIG. 5a and FIG. 5b. Since the operation of the electronic device (100) determining the priority score has been described above, a redundant description is omitted.

[0141] In one embodiment of the present disclosure, the electronic device (100) may update at least one of the first weight (552), the second weight (554), or the third weight (556) described above with reference to FIGS. 5a and 5b. The electronic device (100) may update at least one of the first weight (552), the second weight (554), or the third weight (556) in response to a change in the priority score determination criteria of the user. For example, if the electronic device (100) previously determined the priority score (530) based on the user's personal characteristics, but then changed to determining the priority score (530) based on general knowledge, the first weight (552) may be updated to a higher value than before, and the second weight (554) and the third weight (556) may be updated to lower values ​​than before. In one embodiment of the present disclosure, the electronic device (100) can update the priority score (530) by updating at least one of the first weight (552), the second weight (554), or the third weight (556).

[0142] FIG. 6 is a reference diagram for explaining a method of managing the memory of an electronic device (100) according to one embodiment of the present disclosure.

[0143] The electronic device (100) can identify pairs (610) of interaction data and association data stored in memory and priority scores (612) corresponding to the pairs. The electronic device (100) can acquire new pairs (620) of interaction data and association data. Upon acquiring new pairs (620) of interaction data and association data, the electronic device (100) can identify that the capacity of memory has exceeded a threshold ratio (e.g., 80%) (630).

[0144] For example, the total memory capacity is 10 slots, and data was previously filled in 8 slots, which is 80% of the capacity. Upon acquiring a new pair of interaction data and associated data (620), the electronic device (100) can identify that data is filled in 9 slots, which is more than 80% of the memory capacity.

[0145] The electronic device (100) can update the priority score (640) when the update condition is met. For example, the electronic device (100) can determine the priority score for a newly entered pair. For example, the electronic device (100) can re-determine the priority score for pairs previously stored in memory.

[0146] The electronic device (100) can delete (650) at least one of the pairs based on a priority score. For example, the electronic device (100) can delete at least one of the pairs with a lower priority score until the capacity of the memory is below a threshold ratio. For example, the electronic device (100) can delete the pair with the lowest priority score.

[0147] FIG. 7 is a flowchart illustrating an operation to merge pairs of interaction data and association data according to one embodiment of the present disclosure.

[0148] Referring to operation 710, an electronic device (100) according to one embodiment of the present disclosure can identify whether pairs of interaction data and association data are merged, and whether specified conditions are satisfied.

[0149] For example, a specified condition for merging pairs may include an idle state of the electronic device (100). In one embodiment of the present disclosure, the electronic device (100) requires a large amount of memory resources to merge pairs. The electronic device (100) can perform optimization work in the background without impairing the user experience by merging pairs while the electronic device (100) is not in use. For example, if the user does not operate the electronic device for a certain period of time, or if certain conditions (e.g., charging, screen off, etc.) are met, it may be identified (or determined) that a specified condition for merging pairs is met.

[0150] For example, the conditions for merging pairs may include at least one of a specified time or a specified location. The electronic device (100) may merge pairs at a specified location at a specified time interval. At least one of the specified time or a specified location may be specified by the user or set by the electronic device (100). The electronic device (100) may analyze the user's electronic device (100) usage pattern to identify at least one of the times or locations where the user rarely uses the electronic device (100) and designate (or set) this as the specified condition for merging pairs. For example, the electronic device (100) may merge pairs at the user's home between 3:00 AM and 5:00 AM every Tuesday.

[0151] In one embodiment of the present disclosure, the specified condition for merging pairs may be the same as or different from the update condition (560) described with reference to FIG. 5. That is, the specified time or specified location included in the update condition and the specified condition for merging described with reference to FIG. 7 may be the same condition or a different condition.

[0152] Referring to operation 720, an electronic device (100) according to one embodiment of the present disclosure may merge pairs of interaction data and associated data where the similarity is greater than or equal to a threshold value, based on satisfying a specified condition for merging pairs of interaction data and associated data. In one embodiment of the present disclosure, the electronic device (100) may effectively manage memory storage space by effectively integrating pairs of interaction data and associated data related to the same or similar interaction events. In one embodiment of the present disclosure, the electronic device (100) may effectively retrieve related data when generating a response by merging pairs where the similarity is greater than or equal to a threshold value.

[0153] FIG. 8 is a reference diagram for explaining the operation of merging pairs of interaction data and association data according to one embodiment of the present disclosure.

[0154] For example, if the electronic device (100) satisfies a specified condition, it can identify pairs among the pairs (810) whose similarity is greater than or equal to a threshold value. For example, the similarity between the pair of third interaction data and third association data and the pair of fourth interaction data and fourth association data may be greater than or equal to a threshold value.

[0155] The electronic device (100) can merge (820) pairs of which similarity is greater than or equal to a threshold value. For example, the electronic device (100) can obtain an updated pair of third interaction data and third association data (830) by merging a pair of third interaction data and third association data with a pair of fourth interaction data and fourth association data.

[0156] FIG. 9 is a reference diagram for further explaining the operation of determining similarity between pairs of interaction data and association data according to one embodiment of the present disclosure.

[0157] In one embodiment of the present disclosure, the electronic device (100) determines (910) the similarity between pairs of interaction data and association data, and can merge pairs where the determined similarity is greater than or equal to a threshold value. In one embodiment of the present disclosure, the electronic device (100) can merge pairs where at least one of the similarity between interaction data or the similarity between association data is greater than or equal to a threshold value.

[0158] According to one embodiment of the present disclosure, an electronic device (100) can determine (912) the similarity between pairs (920, 950) of interaction data and association data based on interaction data (930, 960). In one embodiment of the present disclosure, the electronic device (100) can determine (912) the similarity between pairs based on at least one of the content of an interaction event (932, 962), the time of occurrence of an interaction event (934, 964), or the location information of the electronic device at the time of occurrence of an interaction event (936, 966). The electronic device (100) can merge pairs (920, 950) where the similarity is greater than or equal to a threshold value. For example, the electronic device (100) can merge pairs for the same interaction event occurring at the same location at regular intervals. For example, the electronic device (100) can merge pairs for an interaction event where an exercise application is run at home at 2:00 every Sunday.

[0159] In one embodiment of the present disclosure, an electronic device (100) can identify the content of an interaction event (932, 962) included in the interaction data in each pair of interaction data and associated data (920, 950). For example, the electronic device (100) can identify at least one of a topic, keyword, or action performing a function of an agent application included in the interaction data. The electronic device (100) can merge pairs in which the similarity of the content of the interaction event (932, 962) is greater than or equal to a threshold value.

[0160] In one embodiment of the present disclosure, the electronic device (100) can identify at least one of the time of occurrence of an interaction event included in the interaction data (934, 964) or the location information of the electronic device at the time of the interaction event (936, 966) in each pair of interaction data and associated data (920, 950). The electronic device (100) can identify interaction events that occur within a threshold time interval (e.g., 10 seconds, 1 minute) or at a fixed time interval (e.g., 1 hour interval, 1 week interval). For example, the electronic device (100) can merge pairs of interaction events for clicking the "Play Music" button of an agent application twice within 10 seconds, and keep pairs of interaction events for clicking the "Play Music" button after 10 minutes as separate pairs.

[0161] In one embodiment of the present disclosure, an electronic device (100) can identify an interaction event occurring at a designated location or a nearby location. The electronic device (100) can merge pairs (920, 950) in which the similarity of at least one of the time of occurrence (934, 964) or location information (936, 966) of the interaction event is greater than or equal to a threshold value.

[0162] In one embodiment, the electronic device (100) can identify association data (940, 970) in each pair of interaction data and association data (920, 950). In one embodiment, the electronic device (100) can merge the pairs (920, 950) based on at least one of fact information (942, 972), usage history information (944, 974), or reaction information (946, 976). In one embodiment, the electronic device (100) can merge the pairs (920, 950) in which the similarity of at least one of the fact information (942, 972), usage history information (944, 974), or reaction information (946, 976) is greater than or equal to a threshold value.

[0163] For example, the electronic device (100) can merge pairs (920, 950) based on the judgment that pairs (920, 950) of identical or similar factual information are interaction events regarding related topics. For example, the electronic device (100) can merge pairs (920, 950) where the similarity of factual information (942, 972) is greater than or equal to a threshold value. For example, if a user inputs "The capital of Korea is Busan," the electronic device (100) can recognize through a communication interface that the information entered by the user (i.e., interaction event) differs from the actual fact (Seoul is the capital). The electronic device (100) can identify whether input (i.e., interaction event) containing the same error has been repeated previously. The electronic device (100) can merge and store pairs (920, 950) containing similar errors (e.g., errors regarding the capital of Korea or errors regarding the capital).

[0164] For example, the electronic device (100) can merge pairs (920, 950) of interaction data and related data based on usage history information (944, 974). The usage history information (944, 974) is data indicating how often and in what pattern a user uses a specific function or information, and based on this, pairs (920, 950) of similar behavioral patterns of the user can be integrated. The electronic device (100) can merge related pairs (920, 950) by identifying whether the user has repeatedly performed similar tasks in a specific agent application based on the usage history information (944, 974). For example, if the user requests "set an alarm for 3 PM" multiple times or repeatedly performs a command such as "check my schedule," the electronic device (100) recognizes that the repeated interaction event is a habitual behavior of the user and can merge related pairs (920, 950).

[0165] For example, the electronic device (100) can merge pairs (920, 950) of interaction data and associated data based on user reaction information (946, 976). The electronic device (100) can group data with similar emotional reactions from the user or merge data showing the same reaction into one. For example, if there is a specific feature to which the user shows a positive reaction such as "This feature is really good!" or "It is really convenient," the electronic device (100) can merge and store pairs (920, 950) of interaction events containing multiple positive feedback for the specific feature. If negative feedback such as "This feature is too inconvenient" or "This doesn't work" is repeatedly entered for the same feature, the electronic device (100) can merge and store pairs (920, 950) of interaction events containing multiple negative feedback.

[0166] For example, the electronic device (100) can merge pairs (920, 950) even when a user performs repetitive actions showing a strong positive or negative reaction to a specific interaction. For example, if a user writes a comment saying "I love this music so much!" or presses a "Like" button and then plays the same song repeatedly, the electronic device (100) can merge pairs (920, 950) for interaction events containing multiple positive feedbacks. For example, the electronic device (100) can manage this as a single data such as "the user really likes the song."

[0167] In one embodiment of the present disclosure, the electronic device (100) determines (910) the similarity between pairs of interaction data and associated data, and can connect pairs in which the determined similarity is greater than or equal to a threshold value. In one embodiment of the present disclosure, the electronic device (100) can connect pairs in which at least one of the similarity between interaction data or the similarity between associated data is greater than or equal to a threshold value. Since the method by which the electronic device (100) determines the similarity to connect pairs overlaps with the method by which the electronic device (100) determines the similarity to merge pairs described above, the redundant description is omitted.

[0168] FIG. 10 is a flowchart illustrating a method for an electronic device (100) to output response information according to one embodiment of the present disclosure.

[0169] Referring to FIG. 10, the method of the electronic device (100) outputting response information may include operations 1010 to 1040. In one embodiment of the present disclosure, operations 1010 to 1040 may be executed by at least one processor included in the electronic device (100). The method of the electronic device (100) managing memory is not limited to that illustrated in FIG. 10, and in one or more embodiments, operations not illustrated in FIG. 10 may be further included, or some operations may be omitted.

[0170] Referring to operation 1010, the electronic device (100) can obtain actions or conversations input by a user through an agent application. For example, the electronic device (100) can obtain at least one of a user's actions operating the agent application or conversations communicating with the agent application. Referring to operation 1010, the electronic device (100) can obtain new interaction events through the agent application.

[0171] Referring to operation 1020, the electronic device (100) may obtain from memory pairs of interaction data and associated data related to an action or conversation input by a user, and priority scores. Each priority score may be a priority score corresponding to each pair. In one embodiment of the present disclosure, the electronic device (100) may identify pairs of stored interaction data and associated data that have a high priority score (or correspond to a high priority score).

[0172] Referring to operation 1030, the electronic device (100) can generate response information for an action or conversation input by a user based on a priority score. For example, the electronic device (100) can obtain response information more suitable for the user (e.g., response information reflecting the user's characteristics) by generating response information using a pair of interaction data and association data corresponding to a high priority score. For example, through operation 1030, the electronic device (100) can obtain response information reflecting at least one of the user's thoughts, the user's recent interests, or content to which the user shows a strong reaction.

[0173] For example, an agent application mounted on an electronic device (100) can receive input from a user saying, "What should I eat for dinner tonight?" The electronic device (100) can generate a response recommending a menu based on a priority score, the user's meal history over a recent period (e.g., one month), or food data to which the user has responded positively.

[0174] Referring to operation 1040, the electronic device (100) may output response information. In one embodiment of the present disclosure, the electronic device (100) may output response information in the form of text, voice, or image, but is not limited to the disclosed embodiment. The form of the response information may be determined based on at least one of a user's request or the purpose of an agent application.

[0175] In one embodiment of the present disclosure, the electronic device (100) can generate and output user-customized response information. The electronic device (100) can generate response information that reflects the user's preferences that change over time.

[0176] FIG. 11 is a reference diagram illustrating an example in which an electronic device (100) outputs response information through an agent application according to one embodiment of the present disclosure.

[0177] In one embodiment of the present disclosure, an electronic device (100) may obtain a conversation from a user on February 2, 2025, through an agent application, with the content “It’s today T_T, bottom of the 10th inning, 2 outs, final day” (1110). In one embodiment of the present disclosure, the electronic device (100) may transmit (or send) (1122) the action or conversation obtained through the agent application to an artificial intelligence model (1120).

[0178] In one embodiment of the present disclosure, an electronic device (100) may use an artificial intelligence model (1120) to generate a query to find information related to the content of an acquired action or conversation in memory. For example, the electronic device (100) may identify at least one of the contents of a user's conversation (e.g., 'bottom of the 10th inning, 2 outs', 'today', 'final game', etc.) as a keyword. The electronic device (100) may generate a query based on at least one keyword. The electronic device (100) may use the query to trigger (1124) memory (or database (1130)). In one embodiment of the present disclosure, the trigger (1124) may include not only the action of the artificial intelligence model (1120) querying the database (1230), but also the action of executing a specific search process within the database (1230).

[0179] In one embodiment of the present disclosure, an electronic device (100) may provide (1126) to an artificial intelligence model (1120) pairs of interaction data and associated data stored in memory (or, database (1130)) and priority scores assigned to the pairs. The electronic device (100) may provide (1126) to an artificial intelligence model (1120) pairs of interaction data and associated data related to input user actions or conversations and priority scores assigned to the pairs using a query.

[0180] For example, the electronic device (100) can use an artificial intelligence model (1120) to identify interaction data and association data related to the content of the conversation stored in memory (e.g., related to baseball or drama) and priority scores assigned to each pair of interaction data and association data based on keywords included in the user's conversation (e.g., '2 outs in the bottom of the 10th inning', 'today', 'final episode', etc.).

[0181] In one embodiment of the present disclosure, an electronic device (100) may use an artificial intelligence model (1120) to identify pairs of interaction data and associated data that are of high importance or recently relevant based on a priority score. For example, as previously described with reference to FIGS. 1 through 10, the priority score may be determined based on at least one of a predefined criterion, a user's interests that change over time, and a user's individual characteristics.

[0182] The priority score corresponding to pairs of interaction data and associated data related to recent users and interests may be higher than the priority score corresponding to pairs of interaction data and associated data not related to recent users and interests. The priority score corresponding to pairs of interaction data and associated data suitable for the user's individual characteristics (e.g., differing from actual facts, or having high response intensity, positive or negative degree) may be higher than the priority score corresponding to pairs of interaction data and associated data not suitable for the user's individual characteristics.

[0183] For example, if a user has repeatedly conversed with the agent application about the ending of the drama "Bottom of the 10th Inning, 2 Outs" during the past month but has not conversed about baseball, the priority score (1160) corresponding to pairs of interaction data and associated data related to the drama "Bottom of the 10th Inning, 2 Outs" based on usage history information may be higher than the priority score (1150) corresponding to pairs of interaction data and associated data related to Korean professional baseball. Among the priority scores related to the drama "Bottom of the 10th Inning, 2 Outs", the second score (1164) determined based on usage history information may be higher than the second score (1154) among the priority scores related to Korean professional baseball.

[0184] For example, if a user reacted to the agent application with high intensity regarding the drama's bottom of the 10th inning, 2 outs, but reacted with low intensity regarding Korean professional baseball, the priority score (1160) corresponding to pairs of interaction data and associated data related to the drama's bottom of the 10th inning, 2 outs, based on the reaction information, may be higher than the priority score (1150) corresponding to pairs of interaction data and associated data related to Korean professional baseball. Among the priority scores, the third score (1166) determined based on the reaction information may be higher than the third score (1156) among the priority scores related to Korean professional baseball. The first score (1152, 1162) may be determined by a predefined criterion. Since the method for determining priority scores for pairs has been explained in detail with reference to FIGS. 5a and 5b, a redundant explanation will be omitted.

[0185] The electronic device (100) can generate response information based on the identified pairs. The electronic device (100) can output response information (1128) through an agent application. For example, the electronic device (100) can generate response information based on interaction data included in the pair corresponding to the high priority score among the pairs queried in the database (1130) using an artificial intelligence model (1120).

[0186] For example, the electronic device (100) may decide to provide a notification regarding the final broadcast of the drama "2 outs in the 10th inning" based on interaction data for the drama "2 outs in the 10th inning" included in a pair of interaction data and associated data with a high priority score stored in memory in response to a user's conversation. The electronic device (100) may generate a result that reflects the user's characteristics (usage history information, reaction information, etc.) by generating and outputting customized response information such as "I will let you know so you don't miss the final broadcast of the 2 outs in the 10th inning" (1140). The electronic device (100) may output the response information to the user in the form of voice or text through an agent application.

[0187] FIG. 12 is a block diagram illustrating the configuration of an electronic device (100) according to one embodiment of the present disclosure.

[0188] According to one embodiment of the present disclosure, an electronic device (100) may include a communication interface (110), a memory (120), and a processor (130).

[0189] The communication interface (110) can communicate with at least one electronic device. Here, 'communication' may mean the operation of transmitting and / or receiving data, signals, requests, and / or commands, etc. The communication interface (110) can perform wired or wireless communication with at least one electronic device.

[0190] For example, the communication interface (110) may include at least one of a communication module, a communication circuit, a communication device, an input / output port, and an input / output plug for performing wired or wireless communication with at least one electronic device. For example, the communication interface (110) may include at least one wireless communication module, a wireless communication circuit, or a wireless communication device for performing wireless communication with at least one electronic device.

[0191] For example, the communication interface (110) may include a short-range communication module, such as an IR (infrared) communication module, capable of receiving control commands from a remote controller located at a short distance. In this case, the communication interface (110) may receive control signals from the remote controller.

[0192] For example, the communication interface (110) may include at least one communication module that performs communication according to wireless communication standards such as Bluetooth, Wi-Fi, BLE (Bluetooth Low Energy), NFC (Near Field Communication), RFID (Radio Frequency Identification), Wi-Fi Direct, UWB, or ZIGBEE. Alternatively, the communication interface (110) may further include a communication module that performs communication with a server to support long-distance communication according to long-distance communication standards. For example, the communication interface (110) may include a communication module that performs communication through a network for internet communication. Additionally, the communication interface (110) may include a communication module that performs communication through a communication network according to communication standards such as 3G, 4G, 5G and / or 6G. For example, the electronic device (100) may communicate with at least one of an external database, an external server, or an external trusted database through the communication interface (110).

[0193] The memory (120) can store a program for processing and controlling the processor (130), and can store data that is input to or output from the electronic device (100). Additionally, the memory (120) can store data necessary for the operation of the electronic device (100).

[0194] The memory (120) may include at least one type of storage medium among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), RAM (Random Access Memory), SRAM (Static Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic disk, and optical disk.

[0195] The memory (120) may not exist separately but may be configured to be included in the processor (130). The memory (120) may be composed of volatile memory, non-volatile memory, or a combination of volatile and non-volatile memory. The memory (120) may store a program or at least one instruction for performing operations according to the embodiments described below. The memory (120) may provide stored data to the processor (130) upon the request of the processor (130).

[0196] The memory (120) may include an agent application (122). The agent application (122) may be installed on the electronic device (100). For example, the agent application (122) may be installed on the electronic device (100) from the beginning, or it may be installed by user operation or by the operation of the electronic device (100).

[0197] The memory (120) may include a database (124). The database (124) may store the pairs of interaction data and association data described above, and priority scores corresponding to the pairs of interaction data and association data. The database (124) is a space where data is stored.

[0198] The memory (120) may include an artificial intelligence model (126). The artificial intelligence model (126) may represent an algorithm or a set of algorithms for managing pairs of interaction data and association data.

[0199] The artificial intelligence model (126) can determine priority scores for pairs of interaction data and association data. The artificial intelligence model (126) can manage pairs of interaction data and association data based on priority scores. The artificial intelligence model (126) can delete at least one of the pairs based on priority scores. The artificial intelligence model (126) can merge pairs where the similarity is greater than or equal to a threshold value.

[0200] The artificial intelligence model (126) may represent a neural network model. For example, the artificial intelligence model (126) may represent an algorithm or a set of algorithms for implementing artificial intelligence technology. Artificial intelligence technology (hereinafter referred to as 'AI technology') is a technology that obtains a desired result by performing operations through a neural network to process input data, such as analysis and / or classification. The neural network receives input data, performs the aforementioned analysis and / or classification operations, and outputs result data.

[0201] Examples of artificial intelligence models (126) include, but are not limited to, CNN (Convolutional Neural Network), DNN (Deep Neural Network), RNN (Recurrent Neural Network), RBM (Restricted Boltzmann Machine), DBN (Deep Belief Network), BRDNN (Bidirectional Recurrent Deep Neural Network), and Deep Q-Networks. Additionally, neural networks can be subdivided. For example, a CNN neural network can be subdivided into a DCNN (Deep Convolutional Neural Network) or a Capsnet neural network.

[0202] The processor (130) controls the overall operation of the electronic device (100). The processor (130) is configured to control a series of processes to enable the electronic device (100) to operate according to the embodiments described below, and may be composed of one or more processors. One or more processors included in the processor (130) may be circuitry such as a System on Chip (SoC) or an Integrated Circuit (IC).

[0203] One or more processors included in the processor (130) may be general-purpose processors such as a CPU (Central Processing Unit), MPU (Micro Processor Unit), AP (Application Processor), DSP (Digital Signal Processor), graphics-dedicated processors such as a GPU (Graphic Processing Unit) or VPU (Vision Processing Unit), artificial intelligence-dedicated processors such as an NPU (Neural Processing Unit), or communication-dedicated processors such as a CP (Communication Processor). If one or more processors included in the processor (130) are artificial intelligence-dedicated processors, said artificial intelligence-dedicated processors may be designed with a hardware structure specialized for processing a specific artificial intelligence model. The processor (130) may be implemented as a single-core processor or as a multi-core processor.

[0204] The processor (130) can write data to memory (120) or read data stored in memory (120), and in particular, can process data according to a predefined operation rule or artificial intelligence model by executing a program or at least one instruction stored in memory (120). Accordingly, the processor (130) can perform operations described in subsequent embodiments, and operations described as being performed by the electronic device (100) or detailed elements included in the electronic device (100) in subsequent embodiments can be seen as being performed by the processor (130) unless otherwise specified.

[0205] For example, the processor (130) may perform the function of the electronic device (100) described in the present disclosure by individually or collectively executing at least one instruction stored in the memory (120). Alternatively, according to one embodiment, the electronic device (100) may perform the function described in the present disclosure by individually or collectively executing at least one instruction stored in the memory (120) by the processor (130).

[0206] In particular, the processor (130) can process data according to a predefined operation rule or artificial intelligence model by executing a program or at least one instruction stored in memory (120). Accordingly, the processor (130) can perform operations described in subsequent embodiments, and operations described as being performed by the electronic device (100) or detailed components included in the electronic device (100) in subsequent embodiments can be seen as being performed by the processor (130) unless otherwise specified.

[0207] A method of operation of an electronic device (100) according to one embodiment of the present disclosure may include the step of acquiring pairs of interaction data and associated data related to each interaction event in response to the occurrence of an interaction event between an agent application mounted on the electronic device (100) and a user. A method of operation of the electronic device (100) may include the step of storing the pairs of interaction data and associated data related to each interaction event in memory. A method of operation of the electronic device (100) may include the step of determining a priority score for the pairs of interaction data and associated data related to each interaction event using an artificial intelligence model. A method of operation of the electronic device (100) may include the step of storing the priority score in memory. A method of operation of the electronic device (100) may include the step of managing the pairs of interaction data and associated data stored in memory based on the priority score when the capacity of the memory exceeds a threshold ratio.

[0208] The method of operation of the electronic device (100) may include the step of obtaining an action or conversation input by a user through an agent application. The method of operation of the electronic device (100) may include the step of obtaining from memory pairs of interaction data and associated data related to the action or conversation input by the user, and a priority score. The method of operation of the electronic device (100) may include the step of generating response information for the action or conversation input by the user based on the priority score. The method of operation of the electronic device (100) may include the step of outputting the response information through an agent application.

[0209] According to one embodiment of the present disclosure, a method of operation of an electronic device (100) may include the step of deleting some pairs based on a priority score. A method of operation of the electronic device (100) may include the step of merging pairs whose similarity is greater than or equal to a threshold value based on satisfying a specified condition for merging pairs.

[0210] According to one embodiment of the present disclosure, a method of operating an electronic device (100) may include the step of updating a priority score stored in memory based on satisfying an update condition indicating a condition for updating a priority score. According to one embodiment of the present disclosure, the update condition may include an idle state of the electronic device.

[0211] According to one embodiment of the present disclosure, a method of operating an electronic device (100) may include the step of obtaining associated data comprising at least one of fact information representing actual facts corresponding to information included in an interaction event, user usage history information regarding a subject or agent application related to the interaction event, and user response information related to the interaction data.

[0212] A method of operation of an electronic device (100) according to one embodiment of the present disclosure may include the step of determining a priority score based on at least one of a predefined standard, a user's interest that changes over time, and individual characteristics of the user.

[0213] A method of operation of an electronic device (100) according to one embodiment of the present disclosure may include the step of determining a priority score by applying different weights to each of a predefined standard, a user's interest, and a user's individual characteristics.

[0214] A method of operation of an electronic device (100) according to one embodiment of the present disclosure may include the step of determining a high priority score for a pair containing an interaction event that has a high similarity to the most recently acquired interaction data among pairs stored in memory.

[0215] A method of operation of an electronic device (100) according to one embodiment of the present disclosure may include a step of determining a priority score based on factual information and user response information included in associated data. According to one embodiment of the present disclosure, the interaction data may include at least one of the content of an interaction event, the time of occurrence of an interaction event, or location information of the electronic device at the time of occurrence of the interaction event.

[0216] A method of operation of an electronic device (100) according to one embodiment of the present disclosure may include a step of summarizing interaction data related to each interaction event through an artificial intelligence model. A method according to one embodiment of the present disclosure may include a step of storing pairs of summarized interaction data and associated data corresponding to the summarized interaction data in memory.

[0217] An electronic device (100) according to one embodiment of the present disclosure may include a communication interface (110); a memory (120) for storing at least one instruction; and at least one processor (130) including a circuit device. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) may acquire pairs of interaction data and associated data related to each interaction event in response to the occurrence of an interaction event between an agent application mounted on the electronic device (100) and a user. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) may store pairs of interaction data and associated data related to each interaction event in the memory (120). By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can determine a priority score for pairs of interaction data and associated data associated with each interaction event using an artificial intelligence model. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can manage pairs of interaction data and associated data stored in memory (120) based on the priority score when the capacity of memory (120) exceeds a threshold ratio.

[0218] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) may delete some pairs based on priority scores. According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) may merge pairs whose similarity is greater than or equal to a threshold value based on satisfying specified conditions for merging pairs.

[0219] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can obtain an action or conversation input by a user through an agent application. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can obtain from memory pairs of interaction data and associated data related to the action or conversation input by the user, and a priority score. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can generate response information for the action or conversation input by the user based on the priority score. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can output the response information through an agent application.

[0220] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can update a priority score stored in memory (120) based on satisfying an update condition indicating a condition for updating a priority score. According to one embodiment of the present disclosure, the update condition may include an idle state of the electronic device.

[0221] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can obtain the associated data including at least one of fact information representing actual facts corresponding to information included in an interaction event, a subject related to the interaction event or user usage history information regarding the agent application, and user response information related to the interaction data.

[0222] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can determine a priority score based on at least one of a predefined criterion, a user's interest that changes over time, and the user's individual characteristics.

[0223] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can determine a priority score by applying different weights to each of the predefined criteria, user interests, and individual characteristics of the user.

[0224] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can determine a high priority score for a pair containing an interaction event that has a high similarity to the most recently acquired interaction data among the pairs stored in memory (120).

[0225] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can determine a priority score based on factual information and user response information included in associated data. According to one embodiment of the present disclosure, the interaction data may include at least one of the content of an interaction event, the time of occurrence of an interaction event, or location information of the electronic device at the time of occurrence of the interaction event.

[0226] According to one embodiment of the present disclosure, by executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can summarize interaction data associated with each interaction event through an artificial intelligence model. By executing at least one instruction individually or collectively by at least one processor (130), the electronic device (100) can store pairs of summarized interaction data and associated data corresponding to the summarized interaction data in memory (120).

[0227] According to one embodiment of the present disclosure, a computer-readable recording medium may be provided on which a program for performing any one of the methods described above on a computer is recorded.

[0228] A device-readable storage medium may be provided in the form of a non-transitory storage medium. Here, 'non-transitory storage medium' simply means that it is a tangible device and does not contain a signal (e.g., electromagnetic waves), and the term does not distinguish between cases where data is stored semi-permanently and cases where it is stored temporarily. For example, a 'non-transitory storage medium' may include a buffer in which data is stored temporarily.

[0229] According to one embodiment, the method according to the various embodiments disclosed herein may be provided by being included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)), or distributed online (e.g., download or upload) through an application store or directly between two user devices (e.g., smartphones). In the case of online distribution, at least a portion of the computer program product (e.g., downloadable app) may be temporarily stored or temporarily created on a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

Claims

1. In a method of operating an electronic device, A step (210) of acquiring pairs of interaction data associated with each interaction event and associated data associated with said interaction data in response to an interaction event occurring between an agent application mounted on the electronic device and a user; A step (220) of storing pairs of the interaction data and the associated data related to each of the above interaction events in memory; A step (230) of determining priority scores for pairs of the interaction data and the association data related to each of the interaction events using an artificial intelligence model; A method comprising the step (240, 250) of managing pairs of the interaction data and the association data stored in the memory based on the priority score when the capacity of the memory exceeds a threshold ratio.

2. In Paragraph 1, A step of obtaining an action or conversation input from the user through the agent application; A step of obtaining from the memory pairs of the interaction data and the associated data related to actions or conversations input by the user, and the priority score; A step of generating response information for an action or conversation input by the user based on the above priority score; and A method comprising the step of outputting the above response information through the above agent application.

3. In Paragraph 1 or 2, The step of managing pairs of the interaction data and the association data stored in the memory is: A method comprising the step of deleting at least one of the pairs based on the above priority score.

4. In any one of paragraphs 1 through 3, A method comprising further a step of merging pairs whose similarity is greater than or equal to a threshold value, based on satisfying specified conditions for merging the above pairs.

5. In any one of claims 1 to 4, the step of obtaining the pairs is, A method comprising the step of obtaining associated data including at least one of fact information representing actual facts corresponding to information included in the interaction event, information on the user's usage history regarding the subject related to the interaction event or the agent application, and information on the user's response related to the interaction data.

6. In any one of paragraphs 1 through 5, The step of determining the above priority score is, A method comprising the step of updating the priority score stored in the memory based on satisfying an update condition indicating a condition for updating the priority score.

7. In Paragraph 6, The above update condition includes the idle state of the electronic device, a method.

8. In any one of paragraphs 1 through 7, The step of determining the above priority score is, A method comprising the step of determining a priority score based on at least one of a predefined standard, the user's interests that change over time, and the user's individual characteristics.

9. In Paragraph 8, The step of determining the above priority score is, A method comprising the step of determining the priority score by applying different weights to each of the above-defined criteria, the user's interests, and the user's individual characteristics.

10. In any one of paragraphs 1 to 9, the step of determining the priority score is, A method for determining a high priority score for a pair containing an interaction event that has a high similarity to the most recently acquired interaction data among the pairs stored in the memory.

11. In any one of paragraphs 1 to 10, the step of determining the priority score is: A method comprising the step of determining the priority score based on factual information and user response information included in the above-mentioned associated data.

12. In any one of paragraphs 1 through 11, A method in which the interaction data comprises at least one of the content of the interaction event, the time of occurrence of the interaction event, or the location information of the electronic device at the time of occurrence of the interaction event.

13. In any one of paragraphs 1 through 12, The step of storing pairs of the interaction data and the associated data related to each of the above interaction events in memory is: A step of summarizing the interaction data related to each of the interaction events through the artificial intelligence model; and A method comprising the step of storing in memory pairs of the summarized interaction data and the associated data corresponding to the summarized interaction data.

14. In an electronic device (100), Communication interface (110); Memory (120) for storing at least one instruction; and It includes at least one processor (130) including a circuit device, and By having the above at least one instruction executed individually or collectively by the above at least one processor (130), the electronic device (100) In response to the occurrence of an interaction event between an agent application mounted on the electronic device (100) and a user, pairs of interaction data associated with each interaction event and associated data associated with said interaction data are obtained. Using an artificial intelligence model, priority scores are determined for pairs of the interaction data and the association data associated with each of the above interaction events, and The pairs of the interaction data and the associated data related to each of the above interaction events and the priority score are stored in the memory (120), and An electronic device (100) that manages pairs of the interaction data and the association data stored in the memory (120) based on the priority score when the capacity of the memory (120) exceeds a threshold ratio.

15. A computer-readable recording medium having a program recorded thereon for performing the method of any one of paragraphs 1 through 13 on a computer.

Images