WO2022218186A1 - Procédé et appareil pour générer un graphe de connaissances personnalisé, et dispositif informatique - Google Patents

Procédé et appareil pour générer un graphe de connaissances personnalisé, et dispositif informatique Download PDF

Info

Publication number
WO2022218186A1
WO2022218186A1 PCT/CN2022/085187 CN2022085187W WO2022218186A1 WO 2022218186 A1 WO2022218186 A1 WO 2022218186A1 CN 2022085187 W CN2022085187 W CN 2022085187W WO 2022218186 A1 WO2022218186 A1 WO 2022218186A1
Authority
WO
WIPO (PCT)
Prior art keywords
knowledge
user
corpus
knowledge points
points
Prior art date
Application number
PCT/CN2022/085187
Other languages
English (en)
Chinese (zh)
Inventor
刘倩
曾涛
王照栋
冯璐
张清
Original Assignee
京东科技控股股份有限公司
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 京东科技控股股份有限公司 filed Critical 京东科技控股股份有限公司
Publication of WO2022218186A1 publication Critical patent/WO2022218186A1/fr

Links

Images

Classifications

    • 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
    • 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
    • G06F16/335—Filtering based on additional data, e.g. user or group profiles

Definitions

  • the present disclosure relates to the technical field of artificial intelligence, and in particular, to a method, device and computer equipment for generating a personalized knowledge graph.
  • fragmented knowledge there is a lot of fragmented knowledge on the Internet, and a complete knowledge system has not been formed. It is very inconvenient for users to find information from fragmented knowledge. At the same time, fragmented knowledge is not conducive to the overall improvement of users. In addition, if the fragmented knowledge is formed into a knowledge system, it requires professionals to spend a lot of time and energy to realize it. Therefore, if users want to use the established knowledge system, they need to pay expensive fees.
  • the present disclosure aims to solve one of the technical problems in the related art at least to a certain extent.
  • the embodiment of the first aspect of the present disclosure provides a method for generating a personalized knowledge graph, including:
  • each knowledge point corresponding to the user is determined;
  • a knowledge graph corresponding to the user is generated according to the association relationship between the various knowledge points and the core information corresponding to each of the knowledge points.
  • the embodiment of the second aspect of the present disclosure provides an apparatus for generating a personalized knowledge graph, including:
  • the first acquisition module is used to acquire the user's basic information and interest tags
  • a first determining module configured to determine each knowledge point corresponding to the user according to the user's basic information and interest tags
  • a second determining module configured to analyze the respective knowledge points to determine the association relationship between the respective knowledge points
  • the second acquisition module is used to acquire the corpus associated with each of the knowledge points from the multimodal corpus
  • a first generation module configured to fuse the corpus associated with each of the knowledge points to generate core information corresponding to each of the knowledge points
  • the second generation module is configured to generate a knowledge graph corresponding to the user according to the association relationship between the various knowledge points and the core information corresponding to each of the knowledge points.
  • Embodiments of the third aspect of the present disclosure provide a computer device, including: a memory, a processor, and a computer program stored in the memory and running on the processor, when the processor executes the program, the implementation of the present disclosure
  • the method for generating a personalized knowledge graph proposed by the embodiment of the first aspect
  • Embodiments of the fourth aspect of the present disclosure provide a non-transitory computer-readable storage medium that stores a computer program, and when the computer program is executed by a processor, realizes the personalized knowledge graph according to the embodiments of the first aspect of the present disclosure. Generate method.
  • the embodiment of the fifth aspect of the present disclosure provides a computer program product.
  • an instruction processor in the computer program product is executed, the method for generating a personalized knowledge graph proposed by the embodiment of the first aspect of the present disclosure is executed.
  • the method, device, computer equipment and storage medium for generating a personalized knowledge graph can first obtain the user's basic information and interest tags, and then determine each knowledge point corresponding to the user according to the user's basic information and interest tags. Then analyze each knowledge point to determine the relationship between each knowledge point, and then obtain the corpus associated with each knowledge point from the multimodal corpus, and then fuse the corpus associated with each knowledge point. , so as to generate the core information corresponding to each knowledge point, so that the knowledge map corresponding to the user can be generated according to the relationship between the various knowledge points and the core information corresponding to each knowledge point.
  • each knowledge point corresponding to the user can be determined according to the user's basic information and interest tags, and then the association relationship between each knowledge point can be obtained, and then the obtained corpus can be fused to obtain the core information corresponding to each knowledge point. , so that a knowledge graph corresponding to the user can be generated.
  • the knowledge graph combines the relevant information of the user, so it has strong pertinence, can better meet the user's needs, greatly save the user's time, and reduce the user's economic burden. , which improves efficiency.
  • FIG. 1 is a schematic flowchart of a method for generating a personalized knowledge graph according to an embodiment of the present disclosure
  • FIG. 2 is a schematic flowchart of a method for generating a personalized knowledge graph according to another embodiment of the present disclosure
  • FIG. 3 is a schematic flowchart of a method for generating a personalized knowledge graph according to another embodiment of the present disclosure
  • FIG. 4 is a schematic structural diagram of an apparatus for generating a personalized knowledge graph provided by an embodiment of the present disclosure
  • FIG. 5 shows a block diagram of an exemplary computer device suitable for use in implementing embodiments of the present disclosure.
  • FIG. 1 is a schematic flowchart of a method for generating a personalized knowledge graph according to an embodiment of the present disclosure.
  • the embodiment of the present disclosure is exemplified in that the method for generating a personalized knowledge graph is configured in an apparatus for generating a personalized knowledge graph, and the apparatus for generating a personalized knowledge graph can be applied to any computer equipment, so that the computer equipment The generation function of the personalized knowledge graph can be performed.
  • the computer equipment can be a personal computer (Personal Computer, PC for short), a cloud device, a mobile device, etc.
  • the mobile device can be, for example, a mobile phone, a tablet computer, a personal digital assistant, a wearable device, a vehicle-mounted device, etc. with various operating systems, Hardware devices for touch screens and/or display screens.
  • the device for generating the personalized knowledge graph is simply referred to as a "generating device”.
  • the method for generating a personalized knowledge graph may include the following steps S101 to S106.
  • Step 101 Obtain basic information and interest tags of the user.
  • the basic information of the user may be: age, education, occupation and other information
  • the interest tags may be interest tags such as "fitness”, “health”, “cooking”, etc.
  • the basic information and interest tags of the user may be determined according to the historical behavior data of the user.
  • user A often searches for "funds", “stocks” and other content, and it can be determined that the interest tag corresponding to the user can be “finance”.
  • user B often searches for "fitness”, “health care at 40", “how to exercise at 40”, etc., it can be determined that the basic information of user B can be "40 years old", and the interest tag can be "fitness", " nurture”.
  • the basic information and interest tags may also be determined according to the information filled in by the user on the personalized learning interface.
  • user A can fill in information such as age, occupation, hobby, and specialty in the personalized learning interface, so that the corresponding basic information and interest tags can be determined according to the information filled in by user A.
  • Step 102 Determine each knowledge point corresponding to the user according to the user's basic information and interest tags.
  • the knowledge point may be information such as book name, time information, and keywords related to the user's interest tag, which may be in the form of words, sentences, and the like.
  • user B's basic information is "35 years old, market manager", and the interest tags are "finance” and "running".
  • the knowledge points corresponding to "finance” can be: fund fixed investment, income ratio, fund size, etc., “running”
  • the corresponding knowledge points can be: correct running posture, running hazards, etc., so that it can be determined that the knowledge points corresponding to user B can be fund fixed investment, income ratio, fund size, correct running posture, running hazards, etc.
  • Step 103 Analyze each knowledge point to determine the association relationship between each knowledge point.
  • a data mining engine such as a superpixel algorithm can be used to analyze and mine each knowledge point, so as to determine the relationship between each knowledge point.
  • entity relationship extraction can also be performed on each knowledge point to determine the entity relationship between each knowledge point.
  • knowledge point 1 can be "fund”, and entity relationship extraction is performed on it, and the result can be: fund.
  • Knowledge point 2 can be "insurance fund is a classification of funds”, and the entity relationship extraction is performed on it, and the obtained results can be: insurance fund, fund, classification.
  • Knowledge point 3 can be "an investment fund is a kind of fund”. Extract the entity relationship, and the result can be: "investment fund, fund”. Therefore, it can be determined that knowledge point 1 is a large level, knowledge points 2 and 3 are the same small level subordinate to knowledge point 1, and knowledge points 2 and 3 are in a parallel relationship.
  • the knowledge structure classification method can also be used to process each knowledge point to determine the structural relationship between each knowledge point.
  • knowledge point 1 is "IELTS test learning materials”
  • knowledge point 2 is "IELTS test writing video explanation”.
  • the "IELTS test learning materials” and “IELTS test writing video explanation” can be analyzed. It can be determined that knowledge point 1 is a large level, and knowledge point 2 is a small level subordinate to knowledge point 1.
  • the structural relationship between the determined knowledge points can also be verified, for example, by using a root cause analysis method, etc., so as to further ensure the accuracy of the determined structural relationship between the knowledge points.
  • Step 104 From the multimodal corpus, acquire a corpus set associated with each knowledge point.
  • the multimodal corpus can be a corpus that is generated in advance and can contain various modalities, such as text, notes, web links, audio materials, video materials, news broadcasts, on-site interviews and other modal corpora. .
  • the knowledge point is: fund fixed investment
  • find the content related to it from the multimodal corpus can be: the type of fund A is hybrid, the fund's historical performance in the past year, the fund's Analysis indicators include anti-risk fluctuations, investment cost performance and so on.
  • Step 105 fuse the corpus associated with each knowledge point to generate core information corresponding to each knowledge point.
  • the corpus set may include the corpus associated with the knowledge point, and may also include the recommendation degree corresponding to each corpus.
  • each corpus can be fused according to the keywords corresponding to it, so as to generate the core information corresponding to the knowledge point.
  • the model can also be used to fuse each corpus in the corpus, such as using a multi-document summarization model, etc., to generate core information corresponding to knowledge points.
  • the top N corpora with the highest recommendation degree in the corpus can also be fused to generate the core information corresponding to the knowledge point.
  • the first 10 corpora are fused and so on.
  • a threshold may be set in advance, and each corpus with a recommendation degree greater than the threshold may be fused to generate the core information corresponding to the knowledge point.
  • Step 106 Generate a knowledge graph corresponding to the user according to the association relationship between each knowledge point and the core information corresponding to each knowledge point.
  • knowledge point 1 corresponding to user A is "Fund's anti-risk volatility index”
  • knowledge point 2 is "fund's investment cost performance index” both of which are fund analysis indicators, which are in a parallel relationship, and then analyze the knowledge points. It is determined that the core information corresponding to knowledge point 1 can be "Using funds to resist risk fluctuations to select funds", and the core information corresponding to knowledge point 2 can be "Using funds to invest in cost-effective selection of funds”.
  • the user's association relationship generates a knowledge graph corresponding to the user.
  • the generated knowledge graph combines the user's basic information and interest tags, which is more in line with the user's needs and easier for the user to understand, so it is a more targeted and personalized knowledge graph.
  • the user's basic information and interest tags may be obtained first, then each knowledge point corresponding to the user is determined according to the user's basic information and interest tags, and then each knowledge point is analyzed to determine the relationship between each knowledge point.
  • the corpus associated with each knowledge point can be obtained from the multimodal corpus, and then the corpus associated with each knowledge point can be fused to generate the core information corresponding to each knowledge point.
  • the relationship between each knowledge point and the core information corresponding to each knowledge point generate a knowledge graph corresponding to the user.
  • each knowledge point corresponding to the user can be determined according to the user's basic information and interest tags, and then the association relationship between each knowledge point can be obtained, and then the obtained corpus can be fused to obtain the core information corresponding to each knowledge point. , so that the knowledge graph corresponding to the user can be generated.
  • the knowledge graph combines the relevant information of the user, so it has strong pertinence, can better meet the user's needs, greatly save the user's time, reduce the user's economic burden, and improve the efficiency.
  • each reference user may be determined first according to the user's basic information and interest tags, and then each knowledge point corresponding to the user may be determined according to each reference user. The above process will be further described below with reference to FIG. 2 .
  • FIG. 2 is a schematic flowchart of a method for generating a personalized knowledge graph according to an embodiment of the present disclosure. As shown in FIG. 2, the method for generating the personalized knowledge graph may include the following steps S201 to S208.
  • step 201 basic information and interest tags of the user are acquired.
  • Step 202 Determine each reference user according to the user's basic information and interest tags.
  • the reference users may be reference users set in advance according to basic information, interest tags, and their corresponding relationships.
  • one reference user may be determined, or multiple reference users may also be determined.
  • Step 203 Determine each knowledge point corresponding to the user according to the knowledge point in each knowledge graph corresponding to each reference user.
  • the knowledge points in each knowledge graph corresponding to each reference user may be fused according to keywords, so as to determine each knowledge point corresponding to the user.
  • reference user 1 For example, according to the user's basic information and interest tags, it is determined that there are three reference users, namely reference user 1, reference user 2, and reference user 3.
  • the knowledge points corresponding to reference user 1 are: A and B, and reference user 2 corresponds to
  • the knowledge points of user 3 are C and D, and the knowledge points corresponding to reference user 3 are B and E.
  • they can be fused to determine the knowledge points corresponding to users: A, B, C, D, and E respectively.
  • knowledge points in each knowledge graph corresponding to each reference user may also be screened first, and then each knowledge point that satisfies the condition is determined as each knowledge point corresponding to the user.
  • each reference user the knowledge points in the corresponding knowledge graphs, and the important value of each knowledge point can be determined.
  • Each knowledge point of the threshold is determined as each knowledge point corresponding to the user.
  • each knowledge point corresponding to the user may also be determined according to the corresponding relationship between each knowledge point and the interest tag and/or basic information.
  • each knowledge point and the interest tag and/or basic information may be set in advance.
  • the interest label is "IELTS test”
  • the knowledge points "IELTS test skills” and “IELTS test materials” correspond to the interest label, so it can be determined that the knowledge point corresponding to the user can be "IELTS test skills”. ", "IELTS test data”.
  • the basic information is "female, 20 years old, fitness”, and the corresponding knowledge point may be "fitness method” and the like.
  • the interest label can be "fund”
  • the basic information can be "37 years old, senior engineer”
  • the corresponding knowledge point can be "fund type", "fund selection”, etc., so that it can be determined that the knowledge point corresponding to the user can be " Fund Type” and "Fund Selection”.
  • Step 204 Analyze each knowledge point to determine the association relationship between each knowledge point.
  • Step 205 Match each knowledge point with each corpus in the multimodal corpus to obtain each candidate corpus containing each knowledge point.
  • the multimodal corpus can be generated in advance, which can contain various modal corpora, such as text, notes, web links, audio materials, video materials, news broadcasts, live interviews, and so on.
  • the degree of matching between each knowledge point and each corpus in the multimodal corpus can be determined according to the semantic similarity between each knowledge point and each corpus in the multimodal corpus.
  • the top N corpora with high matching degree can be selected as candidate predictions.
  • the top 5 corpora with the highest matching degree may be determined as candidate corpora
  • the top 11 corpora with the highest matching degree may be determined as candidate corpus, and the like.
  • a certain threshold may also be set in advance, and each corpus containing the same knowledge point whose matching degree is greater than the threshold may be determined as candidate corpora.
  • Step 206 using the reading comprehension model to process each candidate corpus to determine the target corpus associated with the knowledge point contained in each candidate corpus.
  • each candidate corpus and the corresponding knowledge point can be input into the reading comprehension model, and then the reading comprehension model can output the starting position and end position of the corpus containing the knowledge point in each candidate corpus, so that the knowledge point associated with the knowledge point can be determined. target corpus.
  • each candidate corpus containing each knowledge point can be obtained, and then the reading comprehension model is used to process each candidate corpus.
  • the target corpus associated with the knowledge point contained in each candidate corpus is determined, and the determined target corpus is more accurate and can better meet user needs.
  • Step 207 fuse the corpus associated with each knowledge point to generate core information corresponding to each knowledge point.
  • Step 208 Generate a knowledge graph corresponding to the user according to the association relationship between each knowledge point and the core information corresponding to each knowledge point.
  • each reference user may be determined first according to the acquired basic information and interest tags of the user, and then each knowledge point corresponding to the user may be determined according to the knowledge points in each knowledge graph corresponding to each reference user, and then Analyze each knowledge point to determine the relationship between each knowledge point. Match each knowledge point with each corpus in the multimodal corpus to obtain each candidate corpus containing each knowledge point, and then use the reading comprehension model to process each candidate corpus to determine each candidate corpus. The target corpus associated with knowledge points contained in . Then, the corpus associated with each knowledge point is fused to generate the core information corresponding to each knowledge point, so that the knowledge corresponding to the user can be generated according to the relationship between each knowledge point and the core information corresponding to each knowledge point. Atlas.
  • each reference user can be determined first according to the user's basic information and interest tags, and then each knowledge point corresponding to the user can be determined according to each reference user, and then the target can be determined by matching and processing each knowledge point. corpus, so that the knowledge graph corresponding to the user can be generated.
  • the knowledge graph combines the relevant information of the user, so it has strong pertinence, can better meet the user's needs, greatly save the user's time, improve the efficiency, and reduce the user's economic burden as much as possible.
  • the knowledge graph can be used to obtain relevant information.
  • the generating device can first determine the user information, and then push the relevant information to the user according to the knowledge graph matching the user information, so that the user can obtain the information that meets the needs more conveniently.
  • FIG. 3 describes the above process in detail.
  • FIG. 3 is a schematic flowchart of a method for generating a personalized knowledge graph according to an embodiment of the present disclosure. As shown in FIG. 3 , the method for generating the personalized knowledge graph may include the following steps S301 to S305.
  • Step 301 Receive an information acquisition request, where the acquisition request includes an identifier of the user.
  • the information acquisition request may also include a relevant request of the user, for example, it may be "know about the fund”, “what is the fund's scheduled investment”, and so on.
  • the user identification may be content that uniquely characterizes the user's identity, such as the user's name, account number, mobile phone number, and other information.
  • Step 302 Acquire a knowledge graph according to the user's identification, wherein the knowledge graph includes a plurality of interrelated knowledge points.
  • the knowledge graph may be a pre-generated knowledge graph, which may be generated with reference to the methods for generating a personalized knowledge graph in the foregoing embodiments.
  • the current user can be determined by using the user identifier, so as to determine the knowledge graph matching the user.
  • Step 303 returning the knowledge point corresponding to the root node in the knowledge graph to obtain the first feedback information.
  • the knowledge point corresponding to the root node in the knowledge graph can be directly returned to the user.
  • the received information acquisition request may include "what is a fund scheduled investment”.
  • the generating device can directly push to the user the knowledge point corresponding to the root node of "what is fund fixed investment” in the knowledge graph "fund fixed investment is the abbreviation of regular fixed investment fund”, and then according to the user's operation, obtain the user's first Feedback.
  • a question and answer sentence associated with the knowledge point corresponding to the root node can also be generated and returned.
  • the generating device can first push related sentences such as "have you ever bought a fund” and "whether the fund is risky" to the user, and then can obtain the user's first feedback information according to the user's answer to the related question-and-answer sentence.
  • Step 304 obtain the target first-level child node to be returned from each first-level child node associated with the root node.
  • the first feedback information may be parsed, and then the target first-level child node may be determined according to the degree of matching between the first feedback information and each first-level child node.
  • the matching degree between the first feedback information and each first-level child node may be determined according to the semantic similarity, and then the first-level child node with the highest matching degree is determined as the target first-level child node.
  • the first-level child node with the highest correlation between each first-level child node and the first feedback information may also be determined as the target first-level child node.
  • Step 305 Return to the target first-level child node to obtain second feedback information, and obtain the target second-level child node to be returned from each second-level child node associated with the target first-level child node according to the second feedback information, until the target second-level child node is returned. The traversal of each knowledge point in the knowledge graph is completed.
  • the knowledge point corresponding to the target first-level child node can be directly returned to the user, and then the second feedback information is obtained according to the user's operation.
  • the question-and-answer sentence associated with the target first-level child node may also be returned to the user, and then the second feedback information is obtained according to the user's answer content.
  • the second feedback information may be parsed, and then the target second-level child node may be determined according to the degree of matching between the second feedback information and each second-level child node associated with the target first-level child node.
  • the root node A corresponds to three first-level child nodes A1, A2, and A3. You can first push the question and answer statement associated with the root node A to the user, and then filter the three first-level child nodes according to the first feedback information returned by the user, and determine that A2 is the target first-level child node. Push the question and answer statement associated with A2 to the user.
  • A2 corresponds to two second-level child nodes A21 and A22, and then according to the obtained second feedback information, the two second-level child nodes A21 and A22 are screened, and it can be determined that A22 is the target second-level child node. A22 is pushed to the user. A22 does not have a third-level child node, and then the content of A21 to be returned to the user can be determined according to the third feedback information of the user. A21 does not have a third-level child node, and then the content corresponding to the A1 node can be returned to the user according to the user's feedback information.
  • A1 corresponds to a second-level child node A11, which can determine the content corresponding to the node A11 returned to the user according to the feedback information of the user.
  • A11 has no three-level child node, so the content of A3 returned to the user can be determined according to the user's feedback information.
  • A3 has no child nodes.
  • the above process of determining the target first-level child nodes and the target second-level child nodes can be repeated to determine the target child nodes at all levels in the knowledge graph, and push the corresponding target child nodes at all levels to the user.
  • knowledge points The knowledge points in the knowledge graph are more in line with the needs of users, and can improve the efficiency of users to obtain information and the utilization of information.
  • the information acquisition request is received first, and then the knowledge graph is obtained according to the user ID, and the knowledge point corresponding to the root node in the knowledge graph can be returned to obtain the first feedback information, and then according to the first feedback information, from Obtain the target first-level child node to be returned from each first-level child node associated with the root node, and then return to the target first-level child node to obtain the second feedback information, and according to the second feedback information, from the target first-level child node.
  • the target second-level child node to be returned is obtained from each second-level child node, until the traversal of each knowledge point in the knowledge graph is completed.
  • the present disclosure also provides an apparatus for generating a personalized knowledge graph.
  • FIG. 4 is a schematic structural diagram of an apparatus for generating a personalized knowledge graph according to an embodiment of the present disclosure.
  • the apparatus 100 for generating a personalized knowledge graph may include: a first obtaining module 110 , a first determining module 120 , a second determining module 130 , a second obtaining module 140 , a first generating module 150 , and a second determining module 130 .
  • Generation module 160 may include: a first obtaining module 110 , a first determining module 120 , a second determining module 130 , a second obtaining module 140 , a first generating module 150 , and a second determining module 130 .
  • the first obtaining module 110 is used to obtain the basic information and interest tags of the user.
  • the first determining module 120 is configured to determine each knowledge point corresponding to the user according to the user's basic information and interest tags.
  • the second determining module 130 is configured to analyze the respective knowledge points to determine the association relationship between the respective knowledge points.
  • the second obtaining module 140 is configured to obtain a corpus set associated with each of the knowledge points from the multimodal corpus.
  • the first generating module 150 is configured to fuse the corpus associated with each of the knowledge points to generate core information corresponding to each of the knowledge points.
  • the second generating module 160 is configured to generate a knowledge graph corresponding to the user according to the association relationship between the various knowledge points and the core information corresponding to each of the knowledge points.
  • the second obtaining module 140 is specifically configured to match each of the knowledge points with respective corpora in the multimodal corpus, so as to obtain each candidate corpus containing each of the knowledge points ; Process each candidate corpus using a reading comprehension model to determine the target corpus associated with the knowledge point contained in each candidate corpus.
  • the second determining module 130 is specifically configured to perform entity relationship extraction on each of the knowledge points to determine the entity relationship between the various knowledge points; and use the knowledge structure classification method to perform entity relationship extraction on each of the knowledge points Processing is performed to determine the structural relationship among the various knowledge points.
  • the first determining module 120 is specifically configured to determine each knowledge point corresponding to the user according to the corresponding relationship between each knowledge point and an interest tag and/or basic information.
  • the first determining module 120 is specifically configured to determine each reference user according to the basic information and interest tags of the user; determine the reference user according to the knowledge points in each knowledge graph corresponding to each reference user Each knowledge point corresponding to the user.
  • the first determining module 120 is further configured to receive an information acquisition request, where the acquisition request includes a user identifier.
  • the first obtaining module 110 is further configured to obtain a knowledge graph according to the user identification, wherein the knowledge graph includes a plurality of interrelated knowledge points.
  • the first obtaining module 110 is further configured to return the knowledge point corresponding to the root node in the knowledge graph to obtain the first feedback information.
  • the first obtaining module 110 is further configured to obtain the target first-level child node to be returned from each first-level child node associated with the root node according to the first feedback information.
  • the first obtaining module 110 is further configured to return the target first-level child node to obtain second feedback information, and according to the second feedback information, from each second-level child node associated with the target first-level child node Obtain the target secondary child node to be returned, until the traversal of each knowledge point in the knowledge graph is completed.
  • the first obtaining module 110 is specifically configured to generate and return a question and answer sentence associated with the knowledge point corresponding to the root node.
  • the device for generating a personalized knowledge graph can first obtain the basic information and interest tags of the user, and then determine each knowledge point corresponding to the user according to the user's basic information and interest tags, and then analyze each knowledge point. To determine the relationship between each knowledge point, then you can obtain the corpus associated with each knowledge point from the multimodal corpus, and then fuse the corpus associated with each knowledge point to generate the corresponding knowledge point. Therefore, a knowledge graph corresponding to the user can be generated according to the relationship between each knowledge point and the core information corresponding to each knowledge point.
  • each knowledge point corresponding to the user can be determined according to the user's basic information and interest tags, and then the association relationship between each knowledge point can be obtained, and then the obtained corpus can be fused to obtain the core information corresponding to each knowledge point. , so that a knowledge graph corresponding to the user can be generated.
  • the knowledge graph combines the relevant information of the user, so it has strong pertinence, can better meet the user's needs, greatly save the user's time, and reduce the user's economic burden. , which improves efficiency.
  • the present disclosure also proposes a computer device, including: a memory, a processor, and a computer program stored in the memory and running on the processor.
  • a computer program stored in the memory and running on the processor.
  • the present disclosure also provides a non-transitory computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the method for generating a personalized knowledge graph as proposed in the foregoing embodiments of the present disclosure is implemented.
  • the present disclosure also proposes a computer program product, when the instruction processor in the computer program product executes, executes the method for generating a personalized knowledge graph as proposed in the foregoing embodiments of the present disclosure.
  • FIG. 5 shows a block diagram of an exemplary computer device suitable for use in implementing embodiments of the present disclosure.
  • the computer device 12 shown in FIG. 5 is only an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present disclosure.
  • computer device 12 takes the form of a general-purpose computing device.
  • Components of computer device 12 may include, but are not limited to, one or more processors or processing units 16 , system memory 28 , and a bus 18 connecting various system components including system memory 28 and processing unit 16 .
  • Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of a variety of bus structures.
  • these architectures include, but are not limited to, Industry Standard Architecture (hereinafter referred to as: ISA) bus, Micro Channel Architecture (hereinafter referred to as: MAC) bus, enhanced ISA bus, video electronics Standards Association (Video Electronics Standards Association; hereinafter referred to as: VESA) local bus and Peripheral Component Interconnection (Peripheral Component Interconnection; hereinafter referred to as: PCI) bus.
  • ISA Industry Standard Architecture
  • MAC Micro Channel Architecture
  • VESA Video Electronics Standards Association
  • PCI Peripheral Component Interconnection
  • Computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer device 12, including both volatile and nonvolatile media, removable and non-removable media.
  • the memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (Random Access Memory; hereinafter: RAM) 30 and/or cache memory 32 .
  • Computer device 12 may further include other removable/non-removable, volatile/non-volatile computer system storage media.
  • storage system 34 may be used to read and write to non-removable, non-volatile magnetic media (not shown in FIG. 5, commonly referred to as a "hard disk drive").
  • a magnetic disk drive for reading and writing to removable non-volatile magnetic disks (eg, "floppy disks") and removable non-volatile optical disks (eg, compact disk read only memory) may be provided.
  • Memory 28 may include at least one program product having a set (eg, at least one) of program modules configured to perform the functions of various embodiments of the present disclosure.
  • a program/utility 40 having a set (at least one) of program modules 42, which may be stored, for example, in memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data , each or some combination of these examples may include an implementation of a network environment.
  • Program modules 42 generally perform the functions and/or methods of the embodiments described in this disclosure.
  • Computer device 12 may also communicate with one or more external devices 14 (eg, keyboard, pointing device, display 24, etc.), may also communicate with one or more devices that enable a user to interact with computer device 12, and/or communicate with Any device (eg, network card, modem, etc.) that enables the computer device 12 to communicate with one or more other computing devices. Such communication may take place through input/output (I/O) interface 22 .
  • the computer device 12 can also communicate with one or more networks (such as a local area network (Local Area Network; hereinafter referred to as: LAN), a wide area network (Wide Area Network; hereinafter referred to as: WAN) and/or a public network, such as the Internet, through the network adapter 20 ) communication.
  • networks such as a local area network (Local Area Network; hereinafter referred to as: LAN), a wide area network (Wide Area Network; hereinafter referred to as: WAN) and/or a public network, such as the Internet, through the network
  • network adapter 20 communicates with other modules of computer device 12 via bus 18 .
  • bus 18 It should be understood that, although not shown, other hardware and/or software modules may be used in conjunction with computer device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives and data backup storage systems.
  • the processing unit 16 executes various functional applications and data processing by running the programs stored in the system memory 28 , for example, implements the methods mentioned in the foregoing embodiments.
  • the basic information and interest tags of the user can be obtained first, then each knowledge point corresponding to the user can be determined according to the user's basic information and interest tags, and then each knowledge point can be analyzed to determine the relationship between each knowledge point.
  • the corpus associated with each knowledge point can be obtained from the multimodal corpus, and then the corpus associated with each knowledge point can be fused to generate the core information corresponding to each knowledge point.
  • the relationship between each knowledge point and the core information corresponding to each knowledge point generate a knowledge graph corresponding to the user.
  • each knowledge point corresponding to the user can be determined according to the user's basic information and interest tags, and then the association relationship between each knowledge point can be obtained, and then the obtained corpus can be fused to obtain the core information corresponding to each knowledge point. , so that a knowledge graph corresponding to the user can be generated.
  • the knowledge graph combines the relevant information of the user, so it has strong pertinence, can better meet the user's needs, greatly save the user's time, and reduce the user's economic burden. , which improves efficiency.
  • first and second are only used for descriptive purposes, and should not be construed as indicating or implying relative importance or implying the number of indicated technical features. Thus, a feature delimited with “first”, “second” may expressly or implicitly include at least one of that feature.
  • plurality means at least two, such as two, three, etc., unless expressly and specifically defined otherwise.
  • a "computer-readable medium” can be any device that can contain, store, communicate, propagate, or transport the program for use by or in conjunction with an instruction execution system, apparatus, or apparatus.
  • computer readable media include the following: electrical connections with one or more wiring (electronic devices), portable computer disk cartridges (magnetic devices), random access memory (RAM), Read Only Memory (ROM), Erasable Editable Read Only Memory (EPROM or Flash Memory), Fiber Optic Devices, and Portable Compact Disc Read Only Memory (CDROM).
  • the computer readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, followed by editing, interpretation, or other suitable medium as necessary process to obtain the program electronically and then store it in computer memory.
  • portions of the present disclosure may be implemented in hardware, software, firmware, or a combination thereof.
  • various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system.
  • a suitable instruction execution system For example, if implemented in hardware as in another embodiment, it can be implemented by any one of the following techniques known in the art, or a combination thereof: discrete with logic gates for implementing logic functions on data signals Logic circuits, application specific integrated circuits with suitable combinational logic gates, Programmable Gate Arrays (PGA), Field Programmable Gate Arrays (FPGA), etc.
  • each functional unit in each embodiment of the present disclosure may be integrated into one processing module, or each unit may exist physically alone, or two or more units may be integrated into one module.
  • the above-mentioned integrated modules can be implemented in the form of hardware, and can also be implemented in the form of software function modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium.
  • the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, and the like.

Abstract

La présente divulgation concerne un procédé et un appareil pour générer un graphe de connaissances personnalisé, et un dispositif informatique, et se rapporte en particulier au domaine technique de l'intelligence artificielle et analogues. Le procédé consiste à : acquérir des informations de base et des balises d'intérêt d'un utilisateur ; déterminer, en fonction des informations de base et des balises d'intérêt de l'utilisateur, des points de connaissance correspondant à l'utilisateur ; analyser les points de connaissance afin de déterminer la relation d'association entre les points de connaissance ; acquérir des ensembles de corpus associés à tous les points de connaissance à partir d'une base de données multimodale de corpus ; fusionner les ensembles de corpus associés à tous les points de connaissance pour générer des informations de noyau correspondant à chaque point de connaissance ; et générer, en fonction de la relation d'association entre les points de connaissance et les informations de noyau correspondant à chaque point de connaissance, un graphe de connaissances correspondant à l'utilisateur.
PCT/CN2022/085187 2021-04-15 2022-04-02 Procédé et appareil pour générer un graphe de connaissances personnalisé, et dispositif informatique WO2022218186A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110406534.4A CN113806550A (zh) 2021-04-15 2021-04-15 个性化知识图谱的生成方法、装置及计算机设备
CN202110406534.4 2021-04-15

Publications (1)

Publication Number Publication Date
WO2022218186A1 true WO2022218186A1 (fr) 2022-10-20

Family

ID=78892931

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/085187 WO2022218186A1 (fr) 2021-04-15 2022-04-02 Procédé et appareil pour générer un graphe de connaissances personnalisé, et dispositif informatique

Country Status (2)

Country Link
CN (1) CN113806550A (fr)
WO (1) WO2022218186A1 (fr)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116054910A (zh) * 2022-12-20 2023-05-02 中国人民解放军63819部队 基于知识图谱构建的地球站设备故障分析及装置
CN116342340A (zh) * 2023-03-31 2023-06-27 上海毅学堂智能科技有限公司 基于多版本教材知识图谱实现的个性化教育系统及方法
CN116757709A (zh) * 2023-08-22 2023-09-15 南京海关工业产品检测中心 一种基于知识图谱的铜精矿进口风险分析方法及系统
CN116955589A (zh) * 2023-09-19 2023-10-27 山东山大鸥玛软件股份有限公司 一种基于教材知识图谱的智能命题方法、系统、命题终端及存储介质
CN117172978A (zh) * 2023-11-02 2023-12-05 北京国电通网络技术有限公司 学习路径信息生成方法、装置、电子设备和介质
CN117672027A (zh) * 2024-02-01 2024-03-08 青岛培诺教育科技股份有限公司 一种vr教学方法、装置、设备及介质
CN116054910B (zh) * 2022-12-20 2024-05-14 中国人民解放军63819部队 基于知识图谱构建的地球站设备故障分析及装置

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113806550A (zh) * 2021-04-15 2021-12-17 京东科技控股股份有限公司 个性化知识图谱的生成方法、装置及计算机设备
CN114519131B (zh) * 2021-12-29 2023-07-25 航天科工网络信息发展有限公司 一种面向异构资源的知识融合处理方法和装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109902187A (zh) * 2019-03-21 2019-06-18 广东小天才科技有限公司 一种特征知识图谱的构建方法及装置、终端设备
CN111046194A (zh) * 2019-12-31 2020-04-21 重庆和贯科技有限公司 构建多模态教学知识图谱的方法
CN112331201A (zh) * 2020-11-03 2021-02-05 珠海格力电器股份有限公司 语音的交互方法和装置、存储介质、电子装置
CN112506945A (zh) * 2020-12-03 2021-03-16 华中师范大学 基于知识图谱的自适应导学方法及系统
CN113806550A (zh) * 2021-04-15 2021-12-17 京东科技控股股份有限公司 个性化知识图谱的生成方法、装置及计算机设备

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170185670A1 (en) * 2015-12-28 2017-06-29 Google Inc. Generating labels for images associated with a user
CN109271525A (zh) * 2018-08-08 2019-01-25 北京百度网讯科技有限公司 用于生成知识图谱的方法、装置、设备以及计算机可读存储介质
CN110046811B (zh) * 2019-04-03 2021-08-20 上海松鼠课堂人工智能科技有限公司 适合自适应学习的知识点追根溯源方法
CN112395403A (zh) * 2020-11-30 2021-02-23 广东国粒教育技术有限公司 一种基于知识图谱的问答方法、系统、电子设备及介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109902187A (zh) * 2019-03-21 2019-06-18 广东小天才科技有限公司 一种特征知识图谱的构建方法及装置、终端设备
CN111046194A (zh) * 2019-12-31 2020-04-21 重庆和贯科技有限公司 构建多模态教学知识图谱的方法
CN112331201A (zh) * 2020-11-03 2021-02-05 珠海格力电器股份有限公司 语音的交互方法和装置、存储介质、电子装置
CN112506945A (zh) * 2020-12-03 2021-03-16 华中师范大学 基于知识图谱的自适应导学方法及系统
CN113806550A (zh) * 2021-04-15 2021-12-17 京东科技控股股份有限公司 个性化知识图谱的生成方法、装置及计算机设备

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116054910A (zh) * 2022-12-20 2023-05-02 中国人民解放军63819部队 基于知识图谱构建的地球站设备故障分析及装置
CN116054910B (zh) * 2022-12-20 2024-05-14 中国人民解放军63819部队 基于知识图谱构建的地球站设备故障分析及装置
CN116342340A (zh) * 2023-03-31 2023-06-27 上海毅学堂智能科技有限公司 基于多版本教材知识图谱实现的个性化教育系统及方法
CN116342340B (zh) * 2023-03-31 2023-10-17 上海毅学堂智能科技有限公司 基于多版本教材知识图谱实现的个性化教育系统及方法
CN116757709A (zh) * 2023-08-22 2023-09-15 南京海关工业产品检测中心 一种基于知识图谱的铜精矿进口风险分析方法及系统
CN116757709B (zh) * 2023-08-22 2023-11-14 南京海关工业产品检测中心 一种基于知识图谱的铜精矿进口风险分析方法及系统
CN116955589A (zh) * 2023-09-19 2023-10-27 山东山大鸥玛软件股份有限公司 一种基于教材知识图谱的智能命题方法、系统、命题终端及存储介质
CN116955589B (zh) * 2023-09-19 2024-01-30 山东山大鸥玛软件股份有限公司 一种基于教材知识图谱的智能命题方法、系统、命题终端及存储介质
CN117172978A (zh) * 2023-11-02 2023-12-05 北京国电通网络技术有限公司 学习路径信息生成方法、装置、电子设备和介质
CN117172978B (zh) * 2023-11-02 2024-02-02 北京国电通网络技术有限公司 学习路径信息生成方法、装置、电子设备和介质
CN117672027A (zh) * 2024-02-01 2024-03-08 青岛培诺教育科技股份有限公司 一种vr教学方法、装置、设备及介质
CN117672027B (zh) * 2024-02-01 2024-04-30 青岛培诺教育科技股份有限公司 一种vr教学方法、装置、设备及介质

Also Published As

Publication number Publication date
CN113806550A (zh) 2021-12-17

Similar Documents

Publication Publication Date Title
WO2022218186A1 (fr) Procédé et appareil pour générer un graphe de connaissances personnalisé, et dispositif informatique
Hofmann et al. Text mining and visualization: Case studies using open-source tools
US10831762B2 (en) Extracting and denoising concept mentions using distributed representations of concepts
US11797607B2 (en) Method and apparatus for constructing quality evaluation model, device and storage medium
Hussain et al. Approximation of COSMIC functional size to support early effort estimation in Agile
US9442917B2 (en) Detecting semantic errors in text using ontology-based extraction rules
US10650094B2 (en) Predicting style breaches within textual content
CN110276023B (zh) Poi变迁事件发现方法、装置、计算设备和介质
Ahasanuzzaman et al. CAPS: a supervised technique for classifying Stack Overflow posts concerning API issues
CN109325201A (zh) 实体关系数据的生成方法、装置、设备及存储介质
Li et al. A policy-based process mining framework: mining business policy texts for discovering process models
AU2022223275B2 (en) Auditing citations in a textual document
CN113157899B (zh) 一种大数据画像分析方法、服务器及可读存储介质
Dawood et al. Integrating IFC and NLP for automating change request validations
CN111656453A (zh) 用于信息提取的层次实体识别和语义建模框架
CN113704415B (zh) 医学文本的向量表示生成方法和装置
CN108319586B (zh) 一种信息提取规则的生成和语义解析方法及装置
US20210390258A1 (en) Systems and methods for identification of repetitive language in document using linguistic analysis and correction thereof
CN113704420A (zh) 文本中的角色识别方法、装置、电子设备及存储介质
AU2019290658B2 (en) Systems and methods for identifying and linking events in structured proceedings
Ashfaq et al. An intelligent analytics approach to minimize complexity in ambiguous software requirements
Zhao et al. Natural language query for technical knowledge graph navigation
Yang et al. Leveraging Large Language Models for Concept Graph Recovery and Question Answering in NLP Education
CN114416174A (zh) 基于元数据的模型重构方法、装置、电子设备及存储介质
CN116402058A (zh) 文本识别方法、装置、计算机设备和存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22787413

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE