CN115358395A - Knowledge graph updating method and device, storage medium and electronic device - Google Patents

Knowledge graph updating method and device, storage medium and electronic device Download PDF

Info

Publication number
CN115358395A
CN115358395A CN202210911171.4A CN202210911171A CN115358395A CN 115358395 A CN115358395 A CN 115358395A CN 202210911171 A CN202210911171 A CN 202210911171A CN 115358395 A CN115358395 A CN 115358395A
Authority
CN
China
Prior art keywords
question
answer text
knowledge
knowledge graph
target
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210911171.4A
Other languages
Chinese (zh)
Inventor
邓邱伟
窦方正
刘朝振
张旭
区波
翟建光
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Qingdao Haier Technology Co Ltd
Qingdao Haier Intelligent Home Appliance Technology Co Ltd
Haier Smart Home Co Ltd
Original Assignee
Qingdao Haier Technology Co Ltd
Qingdao Haier Intelligent Home Appliance Technology Co Ltd
Haier Smart Home Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Qingdao Haier Technology Co Ltd, Qingdao Haier Intelligent Home Appliance Technology Co Ltd, Haier Smart Home Co Ltd filed Critical Qingdao Haier Technology Co Ltd
Priority to CN202210911171.4A priority Critical patent/CN115358395A/en
Publication of CN115358395A publication Critical patent/CN115358395A/en
Priority to PCT/CN2022/135124 priority patent/WO2024021407A1/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/332—Query formulation
    • G06F16/3329—Natural language query formulation or dialogue systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
    • G06F16/367—Ontology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Human Computer Interaction (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The application discloses a method and a device for updating a knowledge graph, a storage medium and an electronic device, which relate to the technical field of smart families, wherein the method for updating the knowledge graph comprises the following steps: acquiring an external field knowledge document under the condition that the first knowledge graph cannot respond to the first question and answer text corresponding to the target object; inputting the first question and answer text and the knowledge document of the external field into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, wherein the machine reading understanding model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises: the system comprises a question and answer text, a preset external field knowledge document and an answer text corresponding to the question and answer text and the preset external field knowledge document; and performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.

Description

Knowledge graph updating method and device, storage medium and electronic device
Technical Field
The application relates to the technical field of smart homes, in particular to a knowledge graph updating method and device, a storage medium and an electronic device.
Background
In the related art, in a question-answering system supported by a domain knowledge graph, question-answering feedback of the whole question-answering system to all knowledge in the existing domain cannot be realized through a single knowledge graph, for example, in a question-answering scene of intelligent household equipment, a user cannot remember various information such as how to use the equipment, notes, various parameters and maintenance methods, or search for the equipment from paper specifications one by one, each piece of equipment in a future family is networked, the question-answering system can solve the application problem, but with the continuous increase of the equipment, equipment specification files can grow along with the increase of the equipment, and the construction cost and the iteration updating cost of the corresponding knowledge graph are also increased continuously. Therefore, the traditional intelligent question-answering system based on the artificial design field knowledge map library cannot meet the increasing user requirements, particularly in the scenes of high-frequency equipment iterative upgrade and new product release, the online period of the knowledge map library is long, the real-time maintenance is needed, the maintenance cost is high, the field knowledge map can be started to consume a large amount of resources for real-time updating when new knowledge appears, the updating efficiency is low, the question-answering support can not be carried out on the user during the updating period, and the normal use of the user on the map is influenced.
Therefore, an effective solution is not provided for the problems that the update efficiency of the knowledge graph is low, the knowledge graph cannot timely feed back the question and answer of the target object to effectively respond in the update process, and the like in the related technology.
Disclosure of Invention
The embodiment of the application provides a method and a device for updating a knowledge graph, a storage medium and an electronic device, which are used for at least solving the problems that the updating efficiency of the knowledge graph is low, the knowledge graph cannot timely feed back the question and answer of a target object to effectively respond in the updating process and the like in the related technology.
According to an embodiment of the present application, there is provided a knowledge-graph updating method, including: acquiring an external field knowledge document under the condition that the first knowledge graph cannot respond to the first question and answer text corresponding to the target object; inputting the first question and answer text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, wherein the machine reading understanding model is trained by machine learning by using multiple groups of data, and each group of data in the multiple groups of data comprises: the method comprises the steps of obtaining a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document; and performing question-answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question-answer text to obtain a second knowledge graph.
In an exemplary embodiment, before content updating the first knowledge-graph based on the target answer text and the first question-answer text, the method further comprises: counting the occurrence times of the first question and answer text; determining to update the first knowledge graph on line under the condition that the occurrence times are greater than or equal to a first preset threshold value; and determining to perform offline updating on the first knowledge graph under the condition that the occurrence times are smaller than a first preset threshold value.
In an exemplary embodiment, the first knowledge-graph is content-updated based on the target answer text and the first question-answer text, the method further comprising: determining available space resources corresponding to a storage space used for map updating by the first knowledge map; under the condition that the available space resources are smaller than or equal to a first preset resource threshold value and the data traffic corresponding to the first question and answer text is larger than a preset traffic threshold value, determining that only a target answer text corresponding to the first question and answer text is stored in the available space resources, and updating the content of the first knowledge graph by using the target answer text; and under the condition that the available space resources are larger than a first preset resource threshold value and the data traffic corresponding to the first question and answer text is smaller than or equal to a preset traffic threshold value, determining that the first question and answer text, the target answer text and the external domain knowledge document are correspondingly stored in the available space resources, and updating the content of the first knowledge graph by using the first question and answer text, the target answer text and the external domain knowledge document.
In an exemplary embodiment, the first knowledge-graph is content-updated based on the target answer text and the first question-answer text, the method further comprising: acquiring a preset problem category rule; determining question categories of the first question and answer text output by the machine reading understanding model and corresponding frequencies of the first question and answer text in each question category through the question category rules; taking the frequency as weight, and calculating an updated value corresponding to the first question and answer text; and under the condition that a plurality of first question and answer texts exist at the same time, determining that the content of the first knowledge graph is updated by using the target first question and answer text with the maximum updating value and the target answer text corresponding to the target first question and answer text.
In an exemplary embodiment, calculating an updated value corresponding to the first question-and-answer text by using the frequency as a weight includes: determining a target heat value of the question category corresponding to the first question and answer text based on heat values respectively corresponding to different question categories in the first knowledge graph; and multiplying the target heat value by the frequency, and taking a value corresponding to a multiplication result as an updated value corresponding to the first question-answering text.
In an exemplary embodiment, before the first question and answer text and the external domain knowledge document are input into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, the method further includes: acquiring a third knowledge graph corresponding to new equipment knowledge under the condition that the new equipment knowledge needs to be added in the first knowledge graph; adding the third knowledge-graph to the external domain knowledge document.
In an exemplary embodiment, after performing question and answer feedback to the target object according to the target answer text, and performing content update on the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph, the method further includes: replacing a first knowledge graph which cannot respond to a first question-answer text corresponding to the target object by using the second knowledge graph as a current knowledge graph; identifying the first question and answer text to obtain a query text for querying in a map, wherein the identifying is used for indicating an intention entity determining the first question and answer text and attribute information corresponding to the first question and answer text; and inputting the query text into the current knowledge graph to determine a query result, wherein the query result is used for indicating whether the updated graph effectively supports the response to the first question and answer text.
According to another embodiment of the present application, there is also provided a knowledge-graph updating apparatus, including: the acquisition module is used for acquiring an external field knowledge document under the condition that the first knowledge graph cannot respond to the first question-answer text corresponding to the target object; an answer module, configured to input the first question-and-answer text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question-and-answer text, where the machine reading understanding model is trained through machine learning by using multiple sets of data, and each set of data in the multiple sets of data includes: the system comprises a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document; and the updating module is used for performing question and answer feedback on the target object according to the target answer text and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
In an exemplary embodiment, the apparatus further includes: the statistic module is used for counting the occurrence times of the first question and answer text; determining to perform online updating on the first knowledge graph under the condition that the occurrence number is greater than or equal to a first preset threshold value; and determining to perform offline updating on the first knowledge graph under the condition that the occurrence times are smaller than a first preset threshold value.
In an exemplary embodiment, the apparatus further includes: the storage module is used for determining available space resources corresponding to a storage space used by the first knowledge graph for graph updating; under the condition that the available space resources are smaller than or equal to a first preset resource threshold value and the data traffic corresponding to the first question and answer text is larger than a preset traffic threshold value, determining that only a target answer text corresponding to the first question and answer text is stored in the available space resources, and updating the content of the first knowledge graph by using the target answer text; and under the condition that the available space resources are larger than a first preset resource threshold value and the data traffic corresponding to the first question and answer text is smaller than or equal to a preset traffic threshold value, determining that the first question and answer text, the target answer text and the external domain knowledge document are correspondingly stored in the available space resources, and updating the content of the first knowledge graph by using the first question and answer text, the target answer text and the external domain knowledge document.
In an exemplary embodiment, the apparatus further includes: the category module is used for acquiring a preset problem category rule; determining question categories of the first question and answer text output by the machine reading understanding model and corresponding frequencies of the first question and answer text in each question category through the question category rules; taking the frequency as weight, and calculating an updated value corresponding to the first question and answer text; and under the condition that a plurality of first question and answer texts exist at the same time, determining that the content of the first knowledge graph is updated by using the target first question and answer text with the maximum updating value and the target answer text corresponding to the target first question and answer text.
In an exemplary embodiment, the category module is further configured to determine a target heat value of the question category corresponding to the first question and answer text based on heat values respectively corresponding to different question categories in the first knowledge graph; and multiplying the target heat value by the frequency, and taking a value corresponding to a multiplication result as an updated value corresponding to the first question-answering text.
In an exemplary embodiment, the apparatus further comprises: the map module is used for acquiring a third knowledge map corresponding to new equipment knowledge under the condition that the new equipment knowledge needs to be added in the first knowledge map; adding the third knowledge-graph to the external domain knowledge document.
In an exemplary embodiment, the apparatus further includes: the determining module is used for replacing a first knowledge graph which cannot respond to a first question-answer text corresponding to the target object by using the second knowledge graph as a current knowledge graph; identifying the first question and answer text to obtain a query text for querying in a map, wherein the identification is used for indicating an intention entity determining the first question and answer text and attribute information corresponding to the first question and answer text; inputting the query text into the current knowledge graph, and determining a query result, wherein the query result is used for indicating whether the updated graph effectively supports the response to the first question-answering text.
According to another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium, in which a computer program is stored, where the computer program is configured to execute the above-mentioned method for updating a knowledge-graph when running.
According to another aspect of the embodiments of the present application, there is also provided an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-mentioned method for updating a knowledge graph through the computer program.
In the embodiment of the application, under the condition that the first knowledge graph cannot respond to the first question and answer text corresponding to the target object, the knowledge document of the external field is obtained; inputting the first question and answer text and the knowledge document of the external field into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, wherein the machine reading understanding model is trained by machine learning by using a plurality of groups of data, and each group of data in the plurality of groups of data comprises: the system comprises a question and answer text, a preset external field knowledge document and an answer text corresponding to the question and answer text and the preset external field knowledge document; performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph; by adopting the technical scheme, the problems that the knowledge graph updating efficiency is low, the knowledge graph cannot timely feed back the question and answer of the target object to effectively respond in the updating process and the like in the related technology are solved, the first question and answer text of the target object is processed and responded through the first knowledge graph, the trained machine reading understanding model is used for processing the first question and answer text and the knowledge document in the external field under the condition that the first knowledge graph cannot support and respond to the first question and answer text, the target answer text corresponding to the first question and answer text is obtained, the question and answer feedback is carried out on the target object by using the target answer text, the first knowledge graph is updated to be the second knowledge graph containing new knowledge, the fact that the knowledge graph updating does not conflict with the use of the target object is guaranteed, and the knowledge graph updating efficiency is improved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application.
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a diagram illustrating a hardware environment of a knowledge graph updating method according to an embodiment of the present application;
FIG. 2 is a flow diagram of a method for knowledge-graph update according to an embodiment of the present application;
FIG. 3 is a block diagram of a question-answering system architecture according to an embodiment of the present application;
FIG. 4 is a block diagram of an apparatus for knowledge-graph update according to an embodiment of the present application;
fig. 5 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It should be understood that the data so used may be interchanged under appropriate circumstances such that embodiments of the application described herein may be implemented in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to one aspect of the embodiment of the application, a knowledge graph updating method is provided. The knowledge graph updating method is widely applied to full-house intelligent digital control application scenes such as Smart homes (Smart Home), intelligent homes, intelligent household equipment ecology, intelligent house (Intellignee House) ecology and the like. Alternatively, in the present embodiment, the above-described method for updating a knowledge graph may be applied to a hardware environment formed by the terminal device 102 and the server 104 as shown in fig. 1. As shown in fig. 1, the server 104 is connected to the terminal device 102 through a network, and may be configured to provide a service (e.g., an application service) for the terminal or a client installed on the terminal, set a database on the server or independent of the server, and provide a data storage service for the server 104, and configure a cloud computing and/or edge computing service on the server or independent of the server, and provide a data operation service for the server 104.
The network may include, but is not limited to, at least one of: wired networks, wireless networks. The wired network may include, but is not limited to, at least one of: wide area networks, metropolitan area networks, local area networks, which may include, but are not limited to, at least one of the following: WIFI (Wireless Fidelity ), bluetooth. Terminal equipment 102 can be but not limited to be PC, the cell-phone, the panel computer, intelligent air conditioner, intelligent cigarette machine, intelligent refrigerator, intelligent oven, intelligent kitchen range, intelligent washing machine, intelligent water heater, intelligent washing equipment, intelligent dish washer, intelligent projection equipment, intelligent TV, intelligent clothes hanger, intelligent (window) curtain, intelligence audio-visual, smart jack, intelligent stereo set, intelligent audio amplifier, intelligent new trend equipment, intelligent kitchen guarding equipment, intelligent bathroom equipment, intelligence robot of sweeping the floor, intelligence robot of wiping the window, intelligence robot of mopping the ground, intelligent air purification equipment, intelligent steam ager, intelligent microwave oven, intelligent kitchen is precious, intelligent clarifier, intelligent water dispenser, intelligent lock etc..
In this embodiment, a method for updating a knowledge graph is provided, and is applied to the computer terminal, and fig. 2 is a flowchart of a method for updating a knowledge graph according to an embodiment of the present application, where the flowchart includes the following steps:
step S202, under the condition that the first knowledge graph can not respond to the first question-answer text corresponding to the target object, acquiring an external field knowledge document;
optionally, when the first knowledge-map is a map related to the device function description, the corresponding external domain knowledge document is a description document of which the description domain is various devices. That is, the above-mentioned external domain knowledge documents may be understood as various documents in the domain corresponding to the knowledge graph, and the document may be obtained by actively inputting the target object, or by actively collecting documents from the network according to the knowledge domain corresponding to the current first knowledge graph, which is not limited in this application.
Step S204, inputting the first question-answering text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question-answering text, wherein the machine reading understanding model is trained by machine learning using multiple sets of data, and each set of data in the multiple sets of data includes: the method comprises the steps of obtaining a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document;
step S206, performing question and answer feedback to the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
Through the steps, under the condition that the first knowledge graph cannot respond to the first question-answer text corresponding to the target object, the knowledge document of the external field is obtained; inputting the first question-answer text and the knowledge document of the external field into a machine reading understanding model to obtain a target answer text corresponding to the first question-answer text, wherein the machine reading understanding model is trained by machine learning by using multiple groups of data, and each group of data in the multiple groups of data comprises: the system comprises a question and answer text, a preset external field knowledge document and an answer text corresponding to the question and answer text and the preset external field knowledge document; performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph; by adopting the technical scheme, the problems that the knowledge graph updating efficiency is low, the knowledge graph cannot timely feed back the question and answer of the target object to effectively respond in the updating process and the like in the related technology are solved, the first question and answer text of the target object is processed and responded through the first knowledge graph, the trained machine reading understanding model is used for processing the first question and answer text and the knowledge document in the external field under the condition that the first knowledge graph cannot support and respond to the first question and answer text, the target answer text corresponding to the first question and answer text is obtained, the question and answer feedback is carried out on the target object by using the target answer text, the first knowledge graph is updated to be the second knowledge graph containing new knowledge, the fact that the knowledge graph updating does not conflict with the use of the target object is guaranteed, and the knowledge graph updating efficiency is improved.
In an exemplary embodiment, before content updating the first knowledge-graph based on the target answer text and the first question-answer text, the method further comprises: counting the occurrence times of the first question and answer text; determining to update the first knowledge graph on line under the condition that the occurrence times are greater than or equal to a first preset threshold value; and determining to perform offline updating on the first knowledge graph under the condition that the occurrence times are smaller than a first preset threshold value.
In other words, when the first knowledge graph is updated, the updating can be performed according to the choice, combined use or condition of the requirement, for example, the question-answering efficiency is higher when the online updating is performed on the high-frequency user question (corresponding to the first question-answering text) and the knowledge (corresponding to the target answer text) is updated to the knowledge graph, and the offline updating can be only used when the online updating is performed only twice on the low-frequency user question. In addition, when the online updating has an influence on the effect of the question-answering system, a compromise strategy can be omitted or adopted.
According to the embodiment, the requirement frequency of the target object for the answer text corresponding to the first question-answer text can be identified based on the occurrence condition of the first question-answer text, so that the time length of an answer period can be determined based on the requirement frequency, the answer text corresponding to the first question-answer text with high requirement frequency is updated in the knowledge graph in real time, the knowledge graph can be enabled to realize quick response to the target object in a short time, when the requirement frequency is low, the answer text corresponding to the corresponding first question-answer text can be temporarily cached or stored, when the target object does not use the knowledge graph for information query, the knowledge graph is updated by using the stored answer text, the use experience of the target object is guaranteed, meanwhile, the updating efficiency of the knowledge graph corresponding to the idle time is improved, and the knowledge graph can cover a larger information range as much as possible when the target object uses.
In an exemplary embodiment, the content of the first knowledge-graph is updated based on the target answer text and the first question-answer text, and the method further comprises: determining available space resources corresponding to a storage space used for map updating by the first knowledge map; under the condition that the available space resources are smaller than or equal to a first preset resource threshold value and the data traffic corresponding to the first question and answer text is larger than a preset traffic threshold value, determining that only a target answer text corresponding to the first question and answer text is stored in the available space resources, and updating the content of the first knowledge graph by using the target answer text; and under the condition that the available space resources are larger than a first preset resource threshold value and the data traffic corresponding to the first question and answer text is smaller than or equal to a preset traffic threshold value, determining that the first question and answer text, the target answer text and the external domain knowledge document are correspondingly stored in the available space resources, and updating the content of the first knowledge graph by using the first question and answer text, the target answer text and the external domain knowledge document.
By the embodiment, when the method for updating the knowledge graph is applied to the corresponding equipment or application platform, how to update and use the data generated in the method for updating the knowledge graph can be determined by determining the data traffic size and the corresponding space resources which can be supported by the equipment or application platform, so that the normal operation function corresponding to the equipment or application platform is not affected while the updating effect is ensured, and the accuracy of the question and answer function corresponding to the equipment or application platform is improved by the aid of the data.
In an exemplary embodiment, the first knowledge-graph is content-updated based on the target answer text and the first question-answer text, the method further comprising: acquiring a preset problem category rule; determining question categories of the first question and answer text output by the machine reading understanding model and corresponding frequencies of the first question and answer text in each question category through the question category rules; taking the frequency as weight, and calculating an updated value corresponding to the first question and answer text; and under the condition that a plurality of first question and answer texts exist at the same time, determining that the content of the first knowledge graph is updated by using the target first question and answer text with the maximum updating value and the target answer text corresponding to the target first question and answer text.
It can be understood that when the above method is applied to the relevant question-answering system, further setting of updating can be performed according to the flow of the question-answering system, and optionally, from the viewpoint of storage policy, the method can be regarded as three methods: the first is a total storage strategy, which is suitable for the conditions of small flow and sufficient storage resources, stores all input and output of a user, and fully utilizes the information when updating a knowledge graph offline; secondly, only storing the knowledge which can not be covered by the knowledge map and needs to be read by a machine to understand the answer, and only considering the unknown or unmatched knowledge during off-line updating, the method is suitable for the conditions of large flow and insufficient storage resources; third, not only are the questions that the machine reads to understand the answers stored, but the frequency of all categories of questions is also stored, the former for the latter as auxiliary information, such as: the detection knowledge graph can be regarded as the representation of importance, and the detection knowledge graph is updated with emphasis as the weight.
In an exemplary embodiment, calculating an updated value corresponding to the first question-and-answer text by using the frequency as a weight includes: determining a target heat value of the question category corresponding to the first question and answer text based on heat values respectively corresponding to different question categories in the first knowledge graph; and multiplying the target heat value by the frequency, and taking a value corresponding to a multiplication result as an updated value corresponding to the first question-answering text.
It can be understood that, because the types of the questions included in the first knowledge graph are various, some questions can primarily support question answering through the data of the knowledge graph, and the questions can be supported only by updating the knowledge graph when the questions are answered in detail, so that in order to avoid data abnormality or use abnormality in the knowledge graph caused by frequent updating, the data sets corresponding to the question answering texts to be updated can be arranged by determining the updating values corresponding to the first question answering texts, and the data sets corresponding to the first question answering texts with the largest updating values and the highest requirements are sequentially updated from the first question answering texts with the largest updating values, so that the knowledge graph is completely updated.
In an exemplary embodiment, before inputting the first question and answer text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, the method further includes: acquiring a third knowledge graph corresponding to new equipment knowledge under the condition that the new equipment knowledge needs to be added in the first knowledge graph; adding the third knowledge-graph to the external domain knowledge document.
It can be understood that when a third knowledge graph exists to update content to the first knowledge graph, the expanded equipment supporting range of the first knowledge graph is described, at this time, in order to improve the updating efficiency, the third knowledge graph is added to the external domain knowledge document, when the target object uses data corresponding to the third knowledge graph, the first knowledge graph is updated, or when the target object does not use the first knowledge graph, the third knowledge graph is used to update the first knowledge graph offline.
In an exemplary embodiment, after performing question and answer feedback to the target object according to the target answer text, and performing content update on the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph, the method further includes: replacing a first knowledge graph which cannot respond to a first question and answer text corresponding to a target object by using the second knowledge graph as a current knowledge graph; identifying the first question and answer text to obtain a query text for querying in a map, wherein the identification is used for indicating an intention entity determining the first question and answer text and attribute information corresponding to the first question and answer text; inputting the query text into the current knowledge graph, and determining a query result, wherein the query result is used for indicating whether the updated graph effectively supports the response to the first question-answering text.
When the knowledge graph is used, the intention of the target object and the entity attribute requirements need to be determined, so that after the knowledge graph is updated, whether the current updated knowledge graph can effectively support the question and answer before corresponding to the current updated knowledge graph is determined through the question and answer test, the accuracy of the knowledge graph is further verified, and the question and answer of the target object can be accurately fed back after the knowledge graph is used.
In order to better understand the process of the method for updating the knowledge graph, the following describes a flow of the method for updating the knowledge graph with reference to an optional embodiment, but the flow is not limited to the technical solution of the embodiment of the present application.
A common implementation method of machine question answering is Knowledge Base Question Answering (KBQA), which is based on Knowledge Base (KB) or Knowledge Graph (KG), and the principle is to construct a human knowledge system understandable by a machine, and then align Natural Language Understanding (NLU) with the knowledge base to find out relevant answers. However, to construct a knowledge base (map) in a complex domain requires large-scale domain data and multiple subdivision steps, from knowledge acquisition, knowledge mining, knowledge storage to knowledge inference, and a large amount of resource cost is consumed on the link, including but not limited to a large amount of labor cost for data and model making, hardware cost for knowledge storage, and cost for computing resources used for training various models and inference on the link. In addition, under the scene of the intelligent hardware specification based on the traditional KBQA, the upgrading of each device, the new product release of each type of device and the like all need to fuse the latest device knowledge into the KB or the KG, so that the frequent updating online operation of the KBQA system is brought.
Based on the analysis, the existing two methods are a method for constructing a relatively complete knowledge graph in advance from the perspective of knowledge graph construction and updating the knowledge graph as required, and a method for improving the matching degree between user input and the knowledge graph through some technologies. The former has the problem of long construction and update periods, which leads to the lengthening of the on-line period of the system, and has the defects of high construction and maintenance costs, low timeliness and the like. The latter does not solve the problem of the former, but can also improve the question and answer accuracy to a certain extent, but the method does not introduce a large amount of domain knowledge beyond the acquired knowledge graph.
As an optional embodiment, the invention designs a knowledge graph-based question-answering system architecture assisted by a machine reading understanding technology, which not only realizes question-answering of knowledge not covered in a knowledge graph by the machine reading understanding technology, but also integrates new knowledge into the knowledge graph based on a machine reading understanding result to realize online updating of the knowledge graph. The framework can still ensure the question and answer effect on the premise of not updating the knowledge graph offline, the offline updating of the knowledge graph can select proper time as required, and the input and machine reading comprehension answers during question and answer of a user can be used as additional auxiliary information.
Optionally, fig. 3 is a block diagram of a structure of a question-answering system architecture according to an embodiment of the present application, where the question-answering system architecture includes: an input module 32, a machine reading understanding module 34, a knowledge graph module 36, an update module 38, a storage feedback module 40;
optionally, the input module 32 includes user input and external domain knowledge, where the user input is a text input by the user, and the external domain knowledge is various documents in the domain, such as instruction documents of various devices in the instruction domain. It should be noted that when the user inputs the knowledge graph module 36 to perform question and answer query or inference, the intention and the concerned information of the user need to be analyzed and extracted in a refined manner through the intention and entity attribute identification technology. The input and the input processing are recorded through the storage module and used as system feedback information, and the knowledge graph can be assisted to update and upgrade in the future.
Optionally, a machine reading understanding module 34, which may introduce various machine reading understanding technologies based on deep learning, reinforcement learning, etc., such as graph neural network, interactive machine reading understanding, etc. The input of the module is user question text and external domain knowledge document, the output is answer A corresponding to the user input question, and the module can be expressed by formula as follows: a = MRC (I, D); wherein, I = w 1 ,w 2 ,…,w m ，D＝{d 1 ,d 2 ,…,d n MRC represents a type of machine reading understanding model which can give answer texts needed by a user by inputting a user question I and an external domain knowledge document set D.
Optionally, the update module 38 includes two parts, i.e. online update and offline update, and the function modules can be selected, used jointly or conditionally according to requirements. For example, if online update affects the system effect, the system effect may be discarded, or a compromise strategy is adopted, online update is performed on a high-frequency user question, knowledge is updated to a knowledge graph, the question-answering efficiency is higher, and for a low-frequency question, only one-time or two-time question may appear, and only offline update may be used. The online update of the knowledge graph can adopt various existing methods, but the principle needing to be ensured is that the system application is not influenced, so that the sub-graph update or the distributed update is feasible in two directions.
Optionally, the feedback module 40 is stored, and the module may be designed in detail according to the flow of the question answering system, such as a storage scheme, a storage device selection, and the like. From the storage policy point of view, three methods can be seen. The first is a total storage strategy, which is suitable for the conditions of small flow and sufficient storage resources, stores all the input of a user, and fully utilizes the information when updating a knowledge graph offline; secondly, only storing the knowledge which can not be covered by the knowledge map and needs to be read by a machine to understand the answer, and only considering the unknown or unmatched knowledge during off-line updating, the method is suitable for the conditions of large flow and insufficient storage resources; thirdly, not only the questions that the machine reads to understand the answers are stored, but also the frequency of all categories of questions, the former for which the latter as auxiliary information, such as tokens that can be considered important, as weights, have an important point to update the detection knowledge graph.
In summary, in the optional example of the present invention, by using the MRC technology, the external knowledge or unsupervised knowledge in the field can be introduced into the system without updating the knowledge graph, so as to improve the user question and answer experience and meet the requirement that the knowledge graph cannot cover the user question and answer in the scene. Meanwhile, the external knowledge can be updated to the knowledge graph on line through machine reading understanding and related technologies, and the knowledge graph can be used as auxiliary information when being updated off line through a storage strategy.
As an alternative example, in practical applications, when a new device is put on the market and enters a user's home, the knowledge graph needs to be updated once for all links, even with the aid of a knowledge graph automation production platform, and such long link construction or update results in accumulation of resource and time costs and gradual performance degradation. This phenomenon is particularly serious in the scenario of a relatively high-frequency iterative update of the device specification question-answering system. The MRC technology can take the knowledge of new equipment as external knowledge of the field, does not need to extract structured information, and can support the question and answer requirements of users by importing knowledge maps. And online and offline branch updating can ensure that the knowledge coverage rate of the knowledge graph is continuously improved, and high-efficiency question answering under the input of the same user is met.
In the embodiment, external knowledge can be acquired through a machine reading and understanding (MRC) technology under the condition that the knowledge graph cannot acquire the question and answer of the user, the knowledge graph is updated on line by using the acquired external knowledge, and in addition, new knowledge can be recorded and used as auxiliary information when the knowledge graph is updated off line. Therefore, the domain knowledge graph does not need to consume a large amount of resources to construct a super-complete domain knowledge graph at the beginning, is relatively complete, and even can support basic services, and a relatively complete domain knowledge graph is finally realized through continuous iteration of the process.
Through the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but the former is a better implementation mode in many cases. Based on such understanding, the technical solutions of the present application or portions thereof that contribute to the prior art may be embodied in the form of a software product, where the computer software product is stored in a storage medium (such as a ROM/RAM, a magnetic disk, and an optical disk), and includes several instructions for enabling a terminal device (which may be a mobile phone, a computer, a server, or a network device) to execute the method of the embodiments of the present application.
Fig. 4 is a block diagram of a knowledge-graph updating apparatus according to an embodiment of the present application. As shown in fig. 4, includes:
the obtaining module 42 is configured to obtain an external domain knowledge document when the first knowledge graph cannot respond to the first question-answer text corresponding to the target object;
an answer module 44, configured to input the first question and answer text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, where the machine reading understanding model is trained through machine learning by using multiple sets of data, and each set of data in the multiple sets of data includes: the method comprises the steps of obtaining a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document;
and an updating module 46, configured to perform question and answer feedback on the target object according to the target answer text, and perform content updating on the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
By the device, under the condition that the first knowledge graph cannot respond to the first question and answer text corresponding to the target object, the knowledge document of the external field is obtained; inputting the first question-answer text and the knowledge document of the external field into a machine reading understanding model to obtain a target answer text corresponding to the first question-answer text, wherein the machine reading understanding model is trained by machine learning by using multiple groups of data, and each group of data in the multiple groups of data comprises: the system comprises a question and answer text, a preset external field knowledge document and an answer text corresponding to the question and answer text and the preset external field knowledge document; performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph; by adopting the technical scheme, the problems that the knowledge graph updating efficiency is low, the knowledge graph cannot timely feed back the question and answer of the target object to effectively respond in the updating process and the like in the related technology are solved, the first question and answer text of the target object is processed and responded through the first knowledge graph, the trained machine reading understanding model is used for processing the first question and answer text and the knowledge document in the external field under the condition that the first knowledge graph cannot support and respond to the first question and answer text, the target answer text corresponding to the first question and answer text is obtained, the question and answer feedback is carried out on the target object by using the target answer text, the first knowledge graph is updated to be the second knowledge graph containing new knowledge, the fact that the knowledge graph updating does not conflict with the use of the target object is guaranteed, and the knowledge graph updating efficiency is improved.
In an exemplary embodiment, the apparatus further includes: the statistic module is used for counting the occurrence times of the first question and answer text; determining to update the first knowledge graph on line under the condition that the occurrence times are greater than or equal to a first preset threshold value; and determining to perform offline updating on the first knowledge graph under the condition that the occurrence times are smaller than a first preset threshold value.
In an exemplary embodiment, the apparatus further includes: the storage module is used for determining available space resources corresponding to a storage space used by the first knowledge graph for graph updating; under the condition that the available space resources are smaller than or equal to a first preset resource threshold value and the data traffic corresponding to the first question and answer text is larger than a preset traffic threshold value, determining that only a target answer text corresponding to the first question and answer text is stored in the available space resources, and updating the content of the first knowledge graph by using the target answer text; and under the condition that the available space resources are larger than a first preset resource threshold value and the data traffic corresponding to the first question and answer text is smaller than or equal to a preset traffic threshold value, determining that the first question and answer text, the target answer text and the external domain knowledge document are correspondingly stored in the available space resources, and updating the content of the first knowledge graph by using the first question and answer text, the target answer text and the external domain knowledge document.
In an exemplary embodiment, the apparatus further includes: the category module is used for acquiring a preset problem category rule; determining question categories of the first question and answer text output by the machine reading understanding model and corresponding frequencies of the first question and answer text in each question category through the question category rules; taking the frequency as weight, and calculating an updated value corresponding to the first question and answer text; and under the condition that a plurality of first question and answer texts exist at the same time, determining that the content of the first knowledge graph is updated by using the target first question and answer text with the maximum updating value and the target answer text corresponding to the target first question and answer text.
In an exemplary embodiment, the category module is further configured to determine a target heat value of the question category corresponding to the first question and answer text based on heat values corresponding to different question categories in the first knowledge graph; and multiplying the target heat value by the frequency, and taking a value corresponding to a multiplication result as an updated value corresponding to the first question-answering text.
In an exemplary embodiment, the apparatus further comprises: the map module is used for acquiring a third knowledge map corresponding to new equipment knowledge under the condition that the new equipment knowledge needs to be added in the first knowledge map; adding the third knowledge-graph to the external domain knowledge document.
In an exemplary embodiment, the apparatus further comprises: the determining module is used for replacing a first knowledge graph which cannot respond to a first question and answer text corresponding to the target object by using the second knowledge graph as a current knowledge graph; identifying the first question and answer text to obtain a query text for querying in a map, wherein the identification is used for indicating an intention entity determining the first question and answer text and attribute information corresponding to the first question and answer text; inputting the query text into the current knowledge graph, and determining a query result, wherein the query result is used for indicating whether the updated graph effectively supports the response to the first question-answering text.
Embodiments of the present application further provide a storage medium including a stored program, where the program executes the method of any one of the above.
Alternatively, in this embodiment, the storage medium may be configured to store program codes for performing the following steps:
s1, acquiring an external field knowledge document under the condition that a first knowledge graph cannot respond to a first question and answer text corresponding to a target object;
s2, inputting the first question and answer text and the external field knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, wherein the machine reading understanding model is trained by machine learning by using multiple groups of data, and each group of data in the multiple groups of data comprises: the system comprises a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document;
and S3, performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
Embodiments of the present application further provide an electronic device, comprising a memory in which a computer program is stored and a processor configured to execute the computer program to perform the steps in any of the method embodiments described above.
Optionally, the electronic apparatus may further include a transmission device and an input/output device, wherein the transmission device is connected to the processor, and the input/output device is connected to the processor.
Optionally, as shown in fig. 5, the electronic device includes a memory 702 and a processor 704, the memory 702 stores a computer program, and the processor 704 is configured to execute the steps in any of the above method embodiments through the computer program.
Optionally, in this embodiment, the electronic apparatus may be located in at least one network device of a plurality of network devices of a computer network.
Optionally, in this embodiment, the processor may be configured to execute the following steps by a computer program:
s1, acquiring an external field knowledge document under the condition that a first knowledge graph cannot respond to a first question and answer text corresponding to a target object;
s2, inputting the first question and answer text and the external field knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, wherein the machine reading understanding model is trained by machine learning by using multiple groups of data, and each group of data in the multiple groups of data comprises: the method comprises the steps of obtaining a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document;
and S3, performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
Alternatively, it can be understood by those skilled in the art that the structure shown in fig. 5 is only an illustration, and the electronic device may also be a terminal device such as a smart phone (e.g., an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, a Mobile Internet Device (MID), a PAD, and the like. Fig. 5 is a diagram illustrating a structure of the electronic device. For example, the electronic device may also include more or fewer components (e.g., network interfaces, etc.) than shown in FIG. 5, or have a different configuration than shown in FIG. 5.
The memory 702 may be used to store software programs and modules, such as program instructions/modules corresponding to the communication connection method and apparatus in the embodiments of the present disclosure, and the processor 704 executes various functional applications and data processing by running the software programs and modules stored in the memory 702, so as to implement the communication connection method described above. The memory 702 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 702 can further include memory located remotely from the processor 704, which can be connected to the terminal over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof. As an example, as shown in fig. 5, the memory 702 may include, but is not limited to, the obtaining module 42, the answering module 44, and the updating module 46 in the communication connection device. In addition, other module units in the communication connection device may also be included, but are not limited to, and are not described in detail in this example.
Optionally, the transmitting device 706 is used for receiving or sending data via a network. Examples of the network may include a wired network and a wireless network. In one example, the transmission device 706 includes a Network adapter (NIC) that can be connected to a router via a Network cable and other Network devices to communicate with the internet or a local area Network. In one example, the transmitting device 1106 is a Radio Frequency (RF) module used for communicating with the internet in a wireless manner.
In addition, the electronic device further includes: a display 708 for displaying the knowledge-graph; and a connection bus 710 for connecting the respective module components in the electronic apparatus.
Optionally, in this embodiment, the storage medium may include but is not limited to: various media capable of storing program codes, such as a usb disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic disk, or an optical disk.
Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation manners, and this embodiment is not described herein again.
It will be apparent to those skilled in the art that the modules or steps of the present application described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.
The foregoing is only a preferred embodiment of the present application and it should be noted that those skilled in the art can make several improvements and modifications without departing from the principle of the present application, and these improvements and modifications should also be considered as the protection scope of the present application.

Claims (10)

1. A method for updating a knowledge graph, comprising:
acquiring an external field knowledge document under the condition that the first knowledge graph cannot respond to the first question and answer text corresponding to the target object;
inputting the first question and answer text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, wherein the machine reading understanding model is trained by machine learning by using multiple groups of data, and each group of data in the multiple groups of data comprises: the system comprises a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document;
and performing question and answer feedback on the target object according to the target answer text, and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
2. The method for updating a knowledge-graph of claim 1, wherein before content updating the first knowledge-graph based on the target answer text and the first question-answer text, the method further comprises:
counting the occurrence times of the first question and answer text;
determining to update the first knowledge graph on line under the condition that the occurrence times are greater than or equal to a first preset threshold value;
and determining to perform offline updating on the first knowledge graph under the condition that the occurrence times are smaller than a first preset threshold value.
3. The method for updating a knowledge-graph according to claim 1, wherein the first knowledge-graph is updated in content based on the target answer text and the first question and answer text, and the method further comprises:
determining available space resources corresponding to a storage space used for map updating by the first knowledge map;
under the condition that the available space resources are smaller than or equal to a first preset resource threshold value and the data traffic corresponding to the first question and answer text is larger than a preset traffic threshold value, determining that only a target answer text corresponding to the first question and answer text is stored in the available space resources, and updating the content of the first knowledge graph by using the target answer text;
and under the condition that the available space resources are larger than a first preset resource threshold value and the data traffic corresponding to the first question and answer text is smaller than or equal to a preset traffic threshold value, determining that the first question and answer text, the target answer text and the external domain knowledge document are correspondingly stored in the available space resources, and updating the content of the first knowledge graph by using the first question and answer text, the target answer text and the external domain knowledge document.
4. The method for updating a knowledge-graph according to claim 1, wherein the first knowledge-graph is updated in content based on the target answer text and the first question-answer text, and the method further comprises:
acquiring a preset problem category rule;
determining question categories of the first question and answer text output by the machine reading understanding model and corresponding frequencies of the first question and answer text in each question category through the question category rules;
taking the frequency as weight, and calculating an updated value corresponding to the first question and answer text;
and under the condition that a plurality of first question and answer texts exist at the same time, determining that the content of the first knowledge graph is updated by using the target first question and answer text with the maximum updating value and the target answer text corresponding to the target first question and answer text.
5. The method for updating a knowledge graph according to claim 4, wherein calculating the updated value corresponding to the first question and answer text by using the frequency as a weight comprises:
determining a target heat value of the question category corresponding to the first question and answer text based on heat values respectively corresponding to different question categories in the first knowledge graph;
and multiplying the target heat value by the frequency, and taking a value corresponding to a multiplication result as an updated value corresponding to the first question-answering text.
6. The method for updating a knowledge graph according to claim 1, wherein before the first question and answer text and the external domain knowledge document are input into a machine reading understanding model to obtain a target answer text corresponding to the first question and answer text, the method further comprises:
acquiring a third knowledge graph corresponding to new equipment knowledge under the condition that the new equipment knowledge needs to be added in the first knowledge graph;
adding the third knowledge-graph to the external domain knowledge document.
7. The method for updating a knowledge graph according to claim 1, wherein after performing question and answer feedback to the target object according to the target answer text and performing content update on the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph, the method further comprises:
replacing a first knowledge graph which cannot respond to a first question-answer text corresponding to the target object by using the second knowledge graph as a current knowledge graph;
identifying the first question and answer text to obtain a query text for querying in a map, wherein the identification is used for indicating an intention entity determining the first question and answer text and attribute information corresponding to the first question and answer text;
and inputting the query text into the current knowledge graph to determine a query result, wherein the query result is used for indicating whether the updated graph effectively supports the response to the first question and answer text.
8. An apparatus for knowledge-graph updating, comprising:
the acquisition module is used for acquiring an external field knowledge document under the condition that the first knowledge graph cannot respond to the first question-answer text corresponding to the target object;
an answer module, configured to input the first question-and-answer text and the external domain knowledge document into a machine reading understanding model to obtain a target answer text corresponding to the first question-and-answer text, where the machine reading understanding model is trained through machine learning by using multiple sets of data, and each set of data in the multiple sets of data includes: the method comprises the steps of obtaining a question and answer text, a preset external domain knowledge document and an answer text corresponding to the question and answer text and the preset external domain knowledge document;
and the updating module is used for performing question and answer feedback on the target object according to the target answer text and updating the content of the first knowledge graph based on the target answer text and the first question and answer text to obtain a second knowledge graph.
9. A computer-readable storage medium, comprising a stored program, wherein the program when executed performs the method of any of claims 1 to 7.
10. An electronic device comprising a memory and a processor, characterized in that the memory has stored therein a computer program, the processor being arranged to execute the method of any of claims 1 to 7 by means of the computer program.
CN202210911171.4A 2022-07-29 2022-07-29 Knowledge graph updating method and device, storage medium and electronic device Pending CN115358395A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202210911171.4A CN115358395A (en) 2022-07-29 2022-07-29 Knowledge graph updating method and device, storage medium and electronic device
PCT/CN2022/135124 WO2024021407A1 (en) 2022-07-29 2022-11-29 Knowledge graph updating method and apparatus, and storage medium and electronic apparatus

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210911171.4A CN115358395A (en) 2022-07-29 2022-07-29 Knowledge graph updating method and device, storage medium and electronic device

Publications (1)

Publication Number Publication Date
CN115358395A true CN115358395A (en) 2022-11-18

Family

ID=84032539

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210911171.4A Pending CN115358395A (en) 2022-07-29 2022-07-29 Knowledge graph updating method and device, storage medium and electronic device

Country Status (2)

Country Link
CN (1) CN115358395A (en)
WO (1) WO2024021407A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024021407A1 (en) * 2022-07-29 2024-02-01 青岛海尔科技有限公司 Knowledge graph updating method and apparatus, and storage medium and electronic apparatus
CN117725423A (en) * 2024-02-18 2024-03-19 青岛海尔科技有限公司 Method and device for generating feedback information based on large model

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9996525B2 (en) * 2016-10-19 2018-06-12 International Business Machines Corporation System and method for supplementing a question answering system with mixed-language source documents
CN110390003A (en) * 2019-06-19 2019-10-29 北京百度网讯科技有限公司 Question and answer processing method and system, computer equipment and readable medium based on medical treatment
CN113220895B (en) * 2021-04-23 2024-02-02 北京大数医达科技有限公司 Information processing method and device based on reinforcement learning and terminal equipment
CN114780697A (en) * 2022-04-18 2022-07-22 国网安徽省电力有限公司电力科学研究院 Intelligent question-answering system based on knowledge graph
CN115358395A (en) * 2022-07-29 2022-11-18 青岛海尔科技有限公司 Knowledge graph updating method and device, storage medium and electronic device

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2024021407A1 (en) * 2022-07-29 2024-02-01 青岛海尔科技有限公司 Knowledge graph updating method and apparatus, and storage medium and electronic apparatus
CN117725423A (en) * 2024-02-18 2024-03-19 青岛海尔科技有限公司 Method and device for generating feedback information based on large model
CN117725423B (en) * 2024-02-18 2024-05-24 青岛海尔科技有限公司 Method and device for generating feedback information based on large model

Also Published As

Publication number Publication date
WO2024021407A1 (en) 2024-02-01

Similar Documents

Publication Publication Date Title
CN115358395A (en) Knowledge graph updating method and device, storage medium and electronic device
CN106874522A (en) Information recommendation method, device, storage medium and processor
CN115269774A (en) Text intention recognition method and device, storage medium and electronic device
CN114676689A (en) Sentence text recognition method and device, storage medium and electronic device
CN115877726A (en) Control method of intelligent household equipment, computer equipment and storage medium
CN114493028A (en) Method and device for establishing prediction model, storage medium and electronic device
CN113741930A (en) Application upgrading method and device, electronic equipment and computer readable storage medium
CN116757402A (en) Service policy determining method and device, storage medium and electronic device
CN115345225A (en) Method and device for determining recommended scene, storage medium and electronic device
CN114925158A (en) Sentence text intention recognition method and device, storage medium and electronic device
CN110874639A (en) Method and device for acquiring operation information
CN109872722B (en) Voice interaction method and device, storage medium and air conditioner
CN114943273A (en) Data processing method, storage medium, and computer terminal
CN110992148A (en) Information pushing method and device
CN114268650A (en) Configuration method and system for establishing communication connection
CN110232393B (en) Data processing method and device, storage medium and electronic device
CN115796185A (en) Semantic intention determination method and device, storage medium and electronic device
CN115481317A (en) Recommendation method for workbench scene, storage medium and electronic device
CN115174383B (en) Equipment updating method, storage medium and electronic device
CN115526230A (en) Behavior habit recognition method and device, storage medium and electronic device
CN112163709B (en) Method and device for electricity utilization promotion, storage medium, and electronic device
CN115314332B (en) Equipment theme setting method and device, storage medium and electronic device
CN115174296B (en) Equipment function access method and device, storage medium and electronic device
CN117744970A (en) Personnel and post matching method and device, storage medium and electronic device
CN118692447A (en) Voice reply test method and device for intelligent household appliance

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination