Persistent memory construction method, device and storage medium based on knowledge graph

Abstract

The present application discloses a method, device and storage medium for constructing persistent memory based on a knowledge graph, and belongs to the field of artificial intelligence technology. The method includes: using the user's identity information as a feature node to construct a main partition of the user knowledge graph corresponding to the user; using the user's behavior information as a feature node to construct a behavior sub-partition of the user knowledge graph; based on the user knowledge graph, obtaining and outputting the response result corresponding to the information to be responded to in the response database. The present application analyzes the user's behavior habits and other characteristics based on the user's identity information and behavior information, and constructs a user knowledge graph. Through the user knowledge graph, the present application can generate customized response results according to different users, meet the personalized needs of users, and optimize the user experience.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, device and storage medium for constructing persistent memory based on a knowledge graph. Background Art

[0002] With the popularization and development of artificial intelligence technology, smart butlers have also been widely used due to their convenience and intelligence. Smart butlers can provide corresponding services according to user needs through artificial intelligence technology, helping users save time, worry and effort in handling various matters.

[0003] In the related art, when the smart butler receives the same question, it will output the same answer based on the database. However, in actual applications, different users have different personalized needs based on their own behavior habits. The answers to the questions of the smart butler are too homogeneous and cannot make targeted responses according to different users. This results in the inability of the smart butler in the related art to meet the personalized needs of users.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide a method for constructing persistent memory based on a knowledge graph, aiming to solve the technical problem that the answers to questions from smart butlers are too homogeneous and cannot meet the personalized needs of users.

[0006] To achieve the above objectives, the present application provides a method for constructing a persistent memory based on a knowledge graph, and the method for constructing a persistent memory based on a knowledge graph comprises the following steps:

[0007] Using the user's identity information as a feature node, constructing a primary partition of the user knowledge graph corresponding to the user;

[0008] Taking the user's behavior information as a feature node, constructing a behavior sub-partition of the user knowledge graph;

[0009] Based on the user knowledge graph, a response result corresponding to the information to be responded to is obtained and output in a response database.

[0010] Optionally, the step of acquiring and outputting a response result corresponding to the information to be responded to in a response database based on the user knowledge graph includes:

[0011] Determining keywords corresponding to the information to be responded to according to the information to be responded to;

[0012] Matching the keyword with the behavior sub-partition to determine a target sub-partition;

[0013] According to the main partition and the target sub-partition, corresponding response data is acquired in a response database, and the response result is output according to the response data.

[0014] Optionally, the step of acquiring corresponding response data in the response data according to the primary partition and the target sub-partition, and outputting the response result according to the response data includes:

[0015] Matching the primary partition and the target sub-partition with the response data in the response database;

[0016] According to the matching results, target response data is determined;

[0017] Determine the target timing information corresponding to the target response data, and splice the target response data according to the target timing information to generate and output the response result.

[0018] Optionally, after the step of acquiring corresponding response data from the response data according to the primary partition and the target sub-partition, and outputting the response result according to the response data, the method further includes:

[0019] Determine target behavior information based on user behavior records;

[0020] According to the target behavior information, the target behavior information is compared with the response result to determine a distinguishing behavior feature;

[0021] The behavior sub-partition corresponding to the distinguishing behavior feature is determined, and the user knowledge graph is updated according to the distinguishing behavior feature and the target behavior information.

[0022] Optionally, after the step of acquiring and outputting a response result corresponding to the information to be responded to in the response database based on the user knowledge graph, the method further includes:

[0023] Determine a corresponding execution action according to the response result;

[0024] Determine the functional component corresponding to the execution action, and execute the control action of the functional component.

[0025] Optionally, the step of using the user's behavior information as a feature node to construct a behavior sub-partition of the user knowledge graph includes:

[0026] Determining an action record of the user in the behavior information, and determining a behavior attribute value of the action record;

[0027] Using the action records as feature nodes, and connecting the feature nodes according to the timing information of the action records;

[0028] The behavior attribute value is associated with the corresponding feature node to generate the behavior sub-partition.

[0029] Optionally, before the step of acquiring and outputting response data corresponding to the information to be responded to in the response database based on the user knowledge graph, the step further includes:

[0030] When the voice information and / or image information of the user is acquired, extracting voice information features and / or image information features;

[0031] Matching the voice information feature and / or the image information feature with the identity information;

[0032] According to the matching results, the user knowledge graph is determined.

[0033] Optionally, the step of acquiring and outputting a response result corresponding to the information to be responded to in a response database based on the user knowledge graph includes:

[0034] Constructing a user profile according to the entity data in the main partition and the behavior sub-partition in the user knowledge graph;

[0035] Determine a corresponding user feature classification according to the user portrait, and determine response data in the response database according to the information to be responded to;

[0036] Determining the matching degree between the user portrait and the response data according to the satisfaction degree of the historical response results corresponding to the user feature classification;

[0037] According to the matching degree, target response data in the response data is selected, and the response result is output according to the target response data.

[0038] In addition, to achieve the above-mentioned purpose, the present application also provides a knowledge graph-based persistent memory construction device, the knowledge graph-based persistent memory construction device comprising: a memory, a processor, and a knowledge graph-based persistent memory construction program stored on the memory and executable on the processor, the knowledge graph-based persistent memory construction program being configured to implement the steps of the knowledge graph-based persistent memory construction method as described above.

[0039] In addition, to achieve the above-mentioned purpose, the present application also provides a storage medium, on which a persistent memory construction program based on a knowledge graph is stored. When the persistent memory construction program based on the knowledge graph is executed by a processor, the steps of the persistent memory construction method based on the knowledge graph as described above are implemented.

[0040] This application can analyze the user's identity information and behavior habits, and build the main partition of the user knowledge graph based on the user's identity information, and build the behavior sub-partition of the user knowledge graph based on the user's behavior information. Through the user knowledge graph corresponding to the user identity, this application can provide users with targeted services based on the user's identity information and behavior habits to meet the user's personalized needs and optimize the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of the first embodiment of the method for constructing persistent memory based on knowledge graph in this application;

[0042] Figure 2 This is a flow chart of the second embodiment of the method for constructing persistent memory based on knowledge graph of the present application;

[0043] Figure 3 This is a flow chart of the third embodiment of the method for constructing persistent memory based on knowledge graph of the present application;

[0044] Figure 4 This is a flow chart of the fourth embodiment of the method for constructing persistent memory based on knowledge graph of the present application;

[0045] Figure 5 It is a structural diagram of a knowledge graph-based persistent memory construction device in the hardware operating environment involved in the embodiment of the present application.

[0046] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0047] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0048] This application can build a user knowledge graph based on the user's identity information and behavior information. According to the user knowledge graph, according to the user's different identity characteristics and behavior habits, corresponding response results are generated to provide users with customized services to meet their personalized needs.

[0049] In order to better understand the above technical solution, exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0050] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0051] Embodiment 1

[0052] Smart butlers usually generate models based on training and testing results, and output corresponding responses through the model according to the user's pending response information. However, different users may have different personalized needs in the same scenario due to differences in behavioral habits, age, gender and other characteristics. In order to meet the personalized needs of users, the embodiment of the present application provides a method for constructing persistent memory based on knowledge graphs, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the method for constructing persistent memory based on knowledge graph in this application.

[0053] In this embodiment, the method for constructing persistent memory based on knowledge graph includes:

[0054] Step S10: Using the user's identity information as a feature node, constructing a primary partition of the user knowledge graph corresponding to the user;

[0055] In this embodiment, the smart butler is a personal smart technology assistant that can help users improve business management and quality of life. It can be applied to smart terminals in scenarios such as smart homes, unmanned supermarkets, and smart car terminals. It can meet the personalized needs of users by connecting and controlling various functional components of the corresponding scenarios. The smart butler can build a corresponding user knowledge graph for each user and form persistent memory information of the corresponding user. After receiving the user's pending response information, the smart butler can identify and extract the user's identity information and determine the corresponding user knowledge graph.

[0056] In specific implementation, the smart butler can construct the main partition of the user knowledge graph through the identity information entered by the user. When receiving the user's registration information, the smart butler can output the corresponding user registration interface on the display screen of the smart terminal, and obtain the user's identity information based on the user registration interface. Among them, the identity information includes but is not limited to user name, appearance, age, height, gender and other data. The smart butler can choose to establish a feature node based on one of the data such as user name or user appearance as entity data, and associate age, height, gender and other data with the user name as attribute values ​​to construct the main partition of the user knowledge graph, and initially form memory information about the user.

[0057] Optionally, the smart butler can also construct the main partition of the user knowledge graph according to the user registration instruction sent by the management terminal. The smart butler can obtain the identity information and behavior information of the user to be registered from the user registration instruction, and construct the main partition of the user knowledge graph according to the identity information of the user to be registered.

[0058] It should be noted that the knowledge graph is a method of representing and organizing knowledge in a graphical structure. It is constructed through entities and relations. Entities represent objects in the real world, and relations represent the connections between entities. Smart Butler can store user knowledge graphs of multiple users at the same time to form persistent memory information for multiple users. Based on different user knowledge graphs, Smart Butler can distinguish different users and analyze users' personalized needs based on user knowledge graphs.

[0059] Furthermore, the smart butler can also use the intimacy between the user and other users as an edge relationship to connect the root feature node of the user's user knowledge graph with the root feature node of other user knowledge graphs. Based on the edge relationship, the smart butler can determine the user's information access rights to other users' user knowledge graphs.

[0060] Optionally, the smart butler can also determine the attribute value of the root feature node of the user knowledge graph based on the received information such as user hobbies and habits, so as to facilitate the analysis of user behavior habits.

[0061] For example, when the smart butler is applied to an unmanned supermarket, the appearance data of the user entering the unmanned supermarket can be obtained through the supermarket's video recording equipment or snapshot equipment. The smart butler can match the appearance data with the user's identity information in the user knowledge graph, determine the user knowledge graph corresponding to the user, and provide the user with corresponding services. In addition, when the user knowledge graph of the user does not exist in the knowledge graph database, the smart butler can construct the user knowledge graph of the user based on the user's appearance data as entity data.

[0062] Step S20: Using the user's behavior information as a feature node to construct a behavior sub-partition of the user knowledge graph;

[0063] In this embodiment, the smart butler can analyze the user's habits, hobbies and other characteristics based on the acquired user behavior information, and construct the behavior sub-partition of the user knowledge graph. The user behavior information can be the user's historical behavior information or the historical interaction record between the user and the smart butler. The smart butler can generate the behavior sub-partition based on the time series information corresponding to the user behavior information, or based on the number of executions of different actions in the user behavior information.

[0064] As an optional implementation, the smart butler can determine the action performed by the user to complete the corresponding behavior based on the action record in the user behavior information. The smart butler can determine the target timing information of the action based on the preset rules for constructing the behavior sub-partition and the user's identity information. According to the target timing information, the smart butler will connect the feature nodes according to the target timing information. At the same time, the smart butler will associate the behavior attribute value of the user behavior information with the corresponding feature node to generate a behavior sub-partition.

[0065] In specific implementation, the smart butler can perform time series analysis on the user knowledge graph by introducing the time dimension in the behavior sub-partition in the construction of the user knowledge graph. By analyzing the changes and trends of the user knowledge graph based on time series, the user's behavior habits can be analyzed more accurately, making the response results more in line with the user's personalized needs.

[0066] As another optional implementation, the smart butler can also count the number of executions of each action in the action record based on the action record in the user behavior information, and use the number of executions of each action as an attribute value to associate it with the feature nodes constructed by each action record to generate a behavioral sub-partition of the user knowledge graph.

[0067] Exemplarily, after obtaining user behavior information, the intelligent terminal can use the action records in the user behavior information as entity information to construct feature nodes in a tree-like manner. The intelligent terminal can determine the attribute values ​​corresponding to different feature nodes in the tree-like knowledge graph based on the number of executions corresponding to the user's action records. The intelligent terminal can determine the data flow direction of the user knowledge graph based on the attribute values ​​of the feature nodes and based on preset rules.

[0068] Step S30: Based on the user knowledge graph, obtain and output the response result corresponding to the information to be responded to in the response database.

[0069] In this embodiment, the user can interact with the smart butler through voice, text, etc. After obtaining the interactive information input by the user, the smart butler will convert the interactive information into text information, and determine the corresponding information to be responded to based on the text information. The smart butler can obtain the response data in the response database based on the entity data corresponding to the information to be responded to in the user knowledge graph, or it can build a user portrait based on the user knowledge graph, and obtain the response data in the response database based on the user feature classification of the user portrait.

[0070] As an optional implementation, after obtaining the information to be responded to, the smart butler can extract keywords from the information to be responded to, and determine the target sub-partition by matching the keywords with the behavior sub-partition in the user knowledge graph. According to the identity information and target sub-partition corresponding to the main partition in the user knowledge graph, the smart butler can obtain the corresponding response data in the response database, and generate and output the corresponding response result according to the response data.

[0071] Furthermore, the smart butler will match the user's identity information and the target sub-partition with the response data in the response database to determine the target response data. After obtaining the target response data, the smart butler can determine the target timing information corresponding to the target response data based on the timing information of the characteristic nodes in the target sub-partition, or obtain the target timing information associated with the target response data in the response database. The smart butler will splice the target response data according to the target timing information, generate and output the response result.

[0072] As another optional implementation, the smart butler can build a user portrait based on the entity data in the main partition and behavior sub-partition in the user knowledge graph. Based on the user portrait, the smart butler can determine the corresponding user feature classification. After obtaining the user's pending response information, the smart butler will determine the response data in the response database based on the pending response information, and determine the matching degree between the user portrait and the response data based on the satisfaction of the historical response results corresponding to the user feature classification. The smart butler can select the target response data with the highest matching degree with the user portrait in the response data, and output the response result based on the target response data.

[0073] It should be noted that user profiling is a method of classification based on user characteristics and preferences. It divides users into different groups by collecting and analyzing user behavior habits, interests, hobbies, and lifestyle characteristics in order to better understand user needs and behaviors. Smart Butler can determine user types based on user profiling and output targeted response results.

[0074] For example, the smart butler can be applied to a smart car terminal. After receiving the pending response data for playing music, the smart butler can determine the user's preference for different types of music based on the user's user profile, and select and play the corresponding music. In addition, after receiving the pending response data for planning a route to the destination, the smart butler can determine the user's preference for road type based on the user profile, and plan a route for the user based on the current road information.

[0075] The embodiment of the present application constructs a user knowledge graph based on the user's identity information and the user's behavior information to form a persistent memory of the user, and after receiving the user's response data, outputs a customized response result based on the user knowledge graph to meet the user's personalized needs.

[0076] Embodiment 2

[0077] Based on the same inventive concept, the present application also provides a second embodiment, referring to Figure 2 , Figure 2 This is a flow chart of the second embodiment of the method for constructing persistent memory based on knowledge graph in this application.

[0078] In this embodiment, after obtaining and outputting the response result corresponding to the information to be responded to in the response database based on the user knowledge graph as described in step S30, the method further includes:

[0079] Step S31: Determine target behavior information according to the user behavior record;

[0080] Step S32: comparing the target behavior information with the response result to determine distinguishing behavior features;

[0081] Step S33: Determine the behavior sub-partition corresponding to the distinguishing behavior feature, and update the user knowledge graph according to the distinguishing behavior feature and the target behavior information.

[0082] In this embodiment, the smart butler can record the user's behavior during the user's use and generate a user behavior record. The smart butler can record the interaction process with the user and can also obtain the user's behavior video through the camera connected to the smart terminal.

[0083] Specifically, after outputting the response result, the smart butler can obtain the user behavior record based on the response result through the smart terminal, and determine the user's target behavior information based on the user behavior record. The smart butler can determine the distinguishing behavior features of the target behavior information and the user's knowledge graph, as well as the behavior sub-partition corresponding to the distinguishing behavior features, and update the distinguishing behavior features to the behavior sub-partition. The smart butler can modify the attribute value of the feature node in the behavior sub-partition based on the distinguishing behavior features, and can also add or replace the feature nodes in the behavior sub-partition based on the distinguishing behavior features.

[0084] Optionally, the user can also evaluate the response results of the smart housekeeper. The smart housekeeper can record the user evaluation and modify the user knowledge graph based on the user evaluation. In addition, the smart housekeeper can also select the response data with higher user evaluation in the response database based on historical user evaluations.

[0085] For example, the user can interact with the smart butler after the smart butler constructs the main partition of the user knowledge graph based on the user's identity information. The smart butler can gradually construct the behavioral sub-partitions of the user knowledge graph through the interaction information with the user and the user's satisfaction with the response results to improve the user knowledge graph.

[0086] The embodiments of the present application record the user's target behavior information during the interaction with the user, match the target behavior information with the user knowledge graph, and update the user knowledge graph based on the target behavior information, so as to update the user's status in time and analyze the user's behavior habits more accurately.

[0087] Since the system introduced in the second embodiment of the present application is a system used to implement the method of the first embodiment of the present application, based on the method introduced in the first embodiment of the present application, the person skilled in the art can understand the specific structure and deformation of the system, so it is not repeated here. All systems used in the method of the first embodiment of the present application belong to the scope of protection of this application.

[0088] Embodiment 3

[0089] Based on the same inventive concept, the present application also provides a third embodiment, referring to Figure 3 , Figure 3 This is a flow chart of the third embodiment of the method for constructing persistent memory based on knowledge graph in this application.

[0090] In this embodiment, after obtaining and outputting the response result corresponding to the information to be responded to in the response database based on the user knowledge graph as described in step S30, the method further includes:

[0091] Step S34: Determine the corresponding execution action according to the response result;

[0092] Step S35: Determine the functional component corresponding to the execution action, and execute the control action of the functional component.

[0093] In this embodiment, the smart butler can determine the user's needs based on the user's pending response information to determine the response result. After determining the response result, the smart butler can split the corresponding execution action based on the response result and determine the functional component corresponding to the execution action. Through the smart terminal, the smart butler can execute the control action of the corresponding functional component in the application scenario to complete the response to the user's pending response information.

[0094] Exemplarily, when the smart housekeeper is applied to the scenario of smart home, the control action of the corresponding furniture can be determined according to the user's pending response information. The smart housekeeper can send the corresponding action execution instruction to the corresponding furniture through the message queue telemetry transmission (MQTT, MessageQueuing Telemetry Transport) protocol. For example, when the smart housekeeper receives the user's pending response information to lower the air conditioner temperature, the smart housekeeper can determine the user's preference for different indoor temperatures based on the user's knowledge graph. The smart housekeeper can calculate the most suitable air conditioning inlet temperature by obtaining the outdoor temperature and outdoor air humidity, and control the air conditioner to modify the operating parameters. It can also send the target indoor temperature and operating parameter modification instructions to the corresponding air conditioner through the MQTT protocol according to the user's preference for different indoor temperatures.

[0095] In an embodiment of the present application, after determining a response result based on the personalized needs of the user, the smart butler can control the corresponding functional components in the current application scenario to implement corresponding functions according to the execution action corresponding to the response result to meet the personalized needs of the user.

[0096] Since the system introduced in the third embodiment of the present application is a system used to implement the method of the first embodiment of the present application, based on the method introduced in the first embodiment of the present application, the person skilled in the art can understand the specific structure and deformation of the system, so it is not repeated here. All systems used in the method of the first embodiment of the present application belong to the scope of protection of the present application.

[0097] Embodiment 4

[0098] Based on the same inventive concept, the present application also provides a fourth embodiment, referring to Figure 4 , Figure 4 This is a flow chart of the fourth embodiment of the method for constructing persistent memory based on knowledge graph in this application.

[0099] In this embodiment, before obtaining and outputting the response result corresponding to the to-be-responded information in the response database based on the user knowledge graph as described in step S30, the method further includes:

[0100] Step S36: when the voice information and / or image information of the user is acquired, extracting voice information features and / or image information features;

[0101] In this embodiment, the smart butler can identify the user's identity through a variety of methods. The smart butler can obtain the user's voice information and / or image information based on the camera and / or microphone connected to the smart terminal, or determine the user's identity through the user's logged-in account.

[0102] As an optional implementation, when the smart butler identifies the user through the user's voice information and / or image information, it will first capture the user's image information and / or voice information through the camera and / or microphone connected to the smart terminal. When the user's image information is obtained, the smart butler can extract the image information features through image convolution and other methods, and / or, when the user's voice information is obtained, the smart butler can identify the timbre in the voice information to determine the user's voice information features.

[0103] For example, when the smart housekeeper detects the presence of a person nearby through the smart terminal, it can obtain the video captured by the camera according to the smart terminal, and select the image information containing the corresponding appearance information in the video. The smart housekeeper can extract the image features of the image information and obtain the corresponding image information features.

[0104] As another optional implementation, the smart butler can also determine the corresponding user knowledge graph based on the currently logged in user. The user can log in through fingerprint, face recognition or key, and the smart butler can determine the user's identity information and user knowledge graph based on the information received during the login process.

[0105] Step S37: matching the voice information feature and / or the image information feature with the identity information;

[0106] Step S38: Determine the user knowledge graph based on the matching results.

[0107] In this embodiment, when the smart butler extracts the user's language information features and / or image information features, the smart butler can match the language information features and / or image information features with the user identity information in the user knowledge graph, and determine the most suitable user identity information based on the matching results, and determine the user knowledge graph corresponding to the user identity information.

[0108] Exemplarily, when the smart butler receives a response message in the form of user voice, it will first extract the voice information features of the response message, and based on the voice information features, traverse all user knowledge graphs to determine the user knowledge graph whose timbre attribute value in the user's identity information matches the voice information features.

[0109] In the embodiment of the present application, the smart butler can identify the identity information of the current user through voice or image, and determine the corresponding user knowledge graph. Through the user knowledge graph, the smart butler can output a response result that better meets the user's expectations.

[0110] Since the system introduced in the fourth embodiment of the present application is a system used to implement the method of the first embodiment of the present application, based on the method introduced in the first embodiment of the present application, the person skilled in the art can understand the specific structure and deformation of the system, so it is not repeated here. All systems used in the method of the first embodiment of the present application belong to the scope of protection of this application.

[0111] Embodiment 5

[0112] Reference Figure 5 , Figure 5 A schematic diagram of the structure of a device for building persistent memory based on a knowledge graph in the hardware operating environment involved in the embodiment of the present application.

[0113] like Figure 5 As shown, the knowledge graph-based persistent memory construction device may include: a processor 1001, such as a core processor (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (WIreless-FIdelity, WI-FI) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0114] Those skilled in the art will understand that Figure 5 The structure shown in does not constitute a limitation on the device for building persistent memory based on the knowledge graph, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0115] like Figure 5 As shown, the memory 1005 as a storage medium may include an operating system, a data storage module, a network communication module, a user interface module, and a persistent memory construction program based on a knowledge graph.

[0116] exist Figure 5In the persistent memory construction device based on the knowledge graph shown, the network interface 1004 is mainly used for data communication with other devices; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the persistent memory construction device based on the knowledge graph of the present application can be set in the persistent memory construction device based on the knowledge graph, and the persistent memory construction device based on the knowledge graph calls the persistent memory construction program based on the knowledge graph stored in the memory 1005 through the processor 1001, and performs the following steps:

[0117] Using the user's identity information as a feature node, constructing a primary partition of the user knowledge graph corresponding to the user;

[0118] Taking the user's behavior information as a feature node, constructing a behavior sub-partition of the user knowledge graph;

[0119] Based on the user knowledge graph, a response result corresponding to the information to be responded to is obtained and output in a response database.

[0120] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0121] Determining keywords corresponding to the information to be responded to according to the information to be responded to;

[0122] Matching the keyword with the behavior sub-partition to determine a target sub-partition;

[0123] According to the main partition and the target sub-partition, corresponding response data is acquired in a response database, and the response result is output according to the response data.

[0124] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0125] Matching the primary partition and the target sub-partition with the response data in the response database;

[0126] According to the matching results, target response data is determined;

[0127] Determine the target timing information corresponding to the target response data, and splice the target response data according to the target timing information to generate and output the response result.

[0128] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0129] Determine target behavior information based on user behavior records;

[0130] According to the target behavior information, the target behavior information is compared with the response result to determine a distinguishing behavior feature;

[0131] The behavior sub-partition corresponding to the distinguishing behavior feature is determined, and the user knowledge graph is updated according to the distinguishing behavior feature and the target behavior information.

[0132] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0133] Determine a corresponding execution action according to the response result;

[0134] Determine the functional component corresponding to the execution action, and execute the control action of the functional component.

[0135] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0136] Determining an action record of the user in the behavior information, and determining a behavior attribute value of the action record;

[0137] Using the action records as feature nodes, and connecting the feature nodes according to the timing information of the action records;

[0138] The behavior attribute value is associated with the corresponding feature node to generate the behavior sub-partition.

[0139] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0140] When the voice information and / or image information of the user is acquired, extracting voice information features and / or image information features;

[0141] Matching the voice information feature and / or the image information feature with the identity information;

[0142] According to the matching results, the user knowledge graph is determined.

[0143] Furthermore, the knowledge graph-based persistent memory construction device calls the knowledge graph-based persistent memory construction program stored in the memory 1005 through the processor 1001, and further performs the following steps:

[0144] Constructing a user profile according to the entity data in the main partition and the behavior sub-partition in the user knowledge graph;

[0145] Determine a corresponding user feature classification according to the user portrait, and determine response data in the response database according to the information to be responded to;

[0146] Determining the matching degree between the user portrait and the response data according to the satisfaction degree of the historical response results corresponding to the user feature classification;

[0147] According to the matching degree, target response data in the response data is selected, and the response result is output according to the target response data.

[0148] In addition, the present application also provides a computer-readable storage medium, which stores the knowledge graph-based persistent memory construction program, and the knowledge graph-based persistent memory construction program can also be executed by a processor to implement the steps of each embodiment of the above-mentioned knowledge graph-based persistent memory construction method.

[0149] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0150] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0151] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0152] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0153] It should be noted that in the claims, any reference signs placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The present application may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0154] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0155] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for constructing persistent memory based on knowledge graph, characterized in that: Applied to the intelligent housekeeper, the method for constructing persistent memory based on the knowledge graph includes the following steps: Using the user's identity information as a feature node, constructing a primary partition of the user knowledge graph corresponding to the user; Acquire the behavior information of the user, determine the action record performed by the user to complete the corresponding behavior in the behavior information, and determine the behavior attribute value of the action record; Taking the action record as a feature node, connecting the feature nodes according to the time sequence information of the action record, and associating the behavior attribute value with the corresponding feature node to generate a behavior sub-partition of the user knowledge graph; Based on the user knowledge graph, obtaining and outputting a response result corresponding to the information to be responded to in a response database; Determine a corresponding execution action according to the response result, determine a functional component corresponding to the execution action, and execute a control action of the functional component; Obtaining, through a smart terminal, a user behavior record of the user based on the response result, and determining target behavior information of the user according to the user behavior record, wherein the user behavior record includes a video recording of the user's behavior and an interaction process of the user; According to the target behavior information, the target behavior information is compared with the response result to determine a distinguishing behavior feature; Determine the behavior sub-partition corresponding to the distinguishing behavior feature, and update the user knowledge graph according to the distinguishing behavior feature and the target behavior information; The step of acquiring and outputting a response result corresponding to the information to be responded to in a response database based on the user knowledge graph includes: According to the information to be responded to, determine the keyword corresponding to the information to be responded to; match the keyword with the behavior sub-partition to determine the target sub-partition; according to the main partition and the target sub-partition, obtain the corresponding response data in the response database, and output the response result according to the response data; The step of acquiring corresponding response data in a response database according to the primary partition and the target sub-partition, and outputting the response result according to the response data comprises: Match the identity information corresponding to the main partition in the user knowledge graph and the target sub-partition with the response data in the response database; determine the target response data based on the matching result; determine the target timing information corresponding to the target response data based on the timing information of the feature node in the target sub-partition, and splice the target response data based on the target timing information to generate and output the response result.

2. The method for constructing persistent memory based on knowledge graph according to claim 1, characterized in that: Before the step of acquiring and outputting the response result corresponding to the information to be responded to in the response database based on the user knowledge graph, the step further includes: When the voice information and / or image information of the user is acquired, extracting voice information features and / or image information features; Matching the voice information feature and / or the image information feature with the identity information; According to the matching results, the user knowledge graph is determined.

3. The method for constructing persistent memory based on knowledge graph according to claim 1, characterized in that: The step of acquiring and outputting a response result corresponding to the information to be responded to in the response database based on the user knowledge graph includes: Constructing a user profile according to the entity data in the main partition and the behavior sub-partition in the user knowledge graph; Determine a corresponding user feature classification according to the user portrait, and determine response data in the response database according to the information to be responded to; Determining the matching degree between the user portrait and the response data according to the satisfaction degree of the historical response results corresponding to the user feature classification; According to the matching degree, target response data in the response data is selected, and the response result is output according to the target response data.

4. A persistent memory construction device based on knowledge graph, characterized in that: The knowledge graph-based persistent memory construction device includes: a memory, a processor, and a knowledge graph-based persistent memory construction program stored on the memory and executable on the processor, wherein the knowledge graph-based persistent memory construction program is configured to implement the steps of the knowledge graph-based persistent memory construction method as described in any one of claims 1 to 3.

5. A storage medium, characterized in that: The storage medium stores a knowledge graph-based persistent memory construction program, which, when executed by a processor, implements the steps of the knowledge graph-based persistent memory construction method as described in any one of claims 1 to 3.

Images