Personalized intelligent question answering system based on artificial intelligence

Through a personalized intelligent question-and-answer system based on artificial intelligence, combined with face photos and speech text recognition technology, customized answers and accurate answer outputs are achieved based on user characteristics, solving the personalized and semantic recognition problems of the existing system, and improving the accuracy and efficiency of question-and-answer.

CN120371976AActive Publication Date: 2025-07-25HEBEI LIJIAQIN TECHNOLOGY CO LTD

Patent Information

Application Number
CN202510521920.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-04-24
Publication Date
2025-07-25
Estimated Expiration
2045-04-24

AI Technical Summary

Technical Problem

The existing intelligent question-and-answer system lacks personalization and cannot provide customized answers based on user's personal preferences, history and other information. It is not accurate enough in semantic recognition of questions, resulting in the answers not meeting user needs.

Method used

Based on artificial intelligence, personalized intelligent question and answer system collects user's face photos through the legality judgment module, extracts target features and calculates similarity with high-dimensional features to judge user legitimacy; uses the speech text recognition module to output voice content text; uses the entity reference extraction module to perform PCA analysis, calculates the similarity between the correlation vector matrix between entity reference and the answer library, and outputs the most similar preset answer text.

Benefits of technology

It improves the personalization and accuracy of Q&A, enhances the scalability and flexibility of the system, reduces misunderstandings and ambiguities, and improves the efficiency of information acquisition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120371976A_ABST
    Figure CN120371976A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of intelligent questioning and answering, in particular to a personalized intelligent questioning and answering system based on artificial intelligence. The method comprises the following steps: activating a face photo acquisition system based on question voice, acquiring a face photo of a current user, extracting a target feature representing a face from the face photo, carrying out similarity calculation on the target feature and a high-dimensional feature stored in a question and answer client, and judging the legitimacy of the user based on the similarity; substituting the question voice of the legal user into the voice text recognition model, and outputting a voice content text; and calculating module similarity between a correlation vector matrix between the entity references and a relation matrix corresponding to the central entity reference of the answer library, and outputting a preset answer text corresponding to the central entity reference with the maximum module similarity, thereby improving personalization, accuracy and intelligence levels of questions and answers, improving information acquisition efficiency, and improving user experience. And the expandability and the flexibility of the system are enhanced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent question - answering, and particularly to a personalized intelligent question - answering system based on artificial intelligence. Background Art

[0002] In recent years, artificial intelligence technology has provided strong support for the development of intelligent question - answering systems. An intelligent question - answering system can use natural language processing technology to understand users' questions and quickly give accurate answers through means such as knowledge graphs and machine learning algorithms.

[0003] However, most of the existing intelligent question - answering systems have some problems. On the one hand, they often lack personalization and cannot provide customized answers based on information such as users' personal preferences and historical records. On the other hand, the semantic recognition of users' questions is not precise enough, resulting in the answers given by the question - answering system not meeting users' needs.

[0004] Therefore, there is an urgent need for a personalized intelligent question - answering system based on artificial intelligence to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a personalized intelligent question - answering system based on artificial intelligence, which solves the technical problems that the existing question - answering systems cannot provide customized answers according to information such as users' personal preferences and historical records, and have low question - answering accuracy and intelligence.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A personalized intelligent question - answering system based on artificial intelligence, the system includes: A legality judgment module, which is used to obtain the user's question voice, activate the face photo acquisition system based on the question voice and collect the face photo of the current user, extract the target features representing the face from the face photo, calculate the similarity between the target features and the high - dimensional features stored in the question - answering client, and judge the legality of the user based on the similarity; A voice - to - text recognition module, which is used to substitute the question voice of a legal user into the voice - to - text recognition model and output the voice content text; An entity reference extraction module, which is used to perform entity reference extraction on the voice content text, determine the relevance between entity references in the voice content text according to the association rules, perform PCA analysis on the related entity references, and obtain the correlation vector matrix between the related entity references; An answer text output module, which is used to calculate the module similarity between the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library, and output the preset answer text corresponding to the central entity reference with the maximum module similarity.

[0007] Further, the process of calculating the similarity between the target feature and the high-dimensional feature stored in the Q&A client specifically includes the following steps: Calculate the basic similarity of the feature vector values between the target feature and the high-dimensional feature: ; Among them, is the basic similarity, represents the i-th feature vector value of the target feature representing the human face, represents the i-th feature vector value of the high-dimensional feature vector stored in the Q&A client, n is the dimension size of the feature vector, and the high-dimensional feature stored in the Q&A client is the pre-extracted user face feature.

[0008] Further, the process of determining the legitimacy of the user based on the similarity specifically includes the following steps: Obtain the Q&A time of the user's last Q&A and the Q&A time of the current Q&A, calculate the time interval between the Q&A time of the last Q&A and the Q&A time of the current Q&A, and record this time interval as the target time interval; calculate the historical similarity corresponding to all the user's historical time intervals and the query historical time interval in turn; Number the historical time interval and the target time interval to obtain a numbered value. Using the numbered value as the X-axis and setting the historical similarity and similarity as the Y-axis, construct a rectangular coordinate system. Connect the adjacent similarities in the rectangular coordinate system to generate a similarity curve. Draw perpendicular lines from both ends of the similarity curve to the X-axis to obtain two starting and ending line segments. The closed figure is composed of the similarity curve, the two starting and ending line segments, and the X-axis. Calculate the total area of the closed figure; Judge whether the total area exceeds the preset threshold. If so, judge that the user is a legitimate user. If not, judge that the user is an illegal user and reject their access to the Q&A system.

[0009] Further, substituting the question voice of the legitimate user into the speech text recognition model and outputting the speech content text specifically includes the following steps: For the question voice data , is the sample set of the question voice data; The speech content recognition model consists of two parts: a speech content encoding network f and a speech content context network. Among them, the speech content encoder network f consists of five convolutional networks, and the output of the speech content encoder is a low-frequency feature , and the hidden space mapped by the encoder network is represented by Z. Among them, : ; Among them, m is the convolutional receptive field; The speech content context network is a convolutional neural network, whose structure is the same as that of the speech content encoding network f, mapping the hidden space into the context information space C, and the output of the context network is , where : ; where v is the convolutional receptive field; Loss function based on the speech content recognition model The speech content recognition model is trained with this loss function, and this loss function is expressed as follows: ; where is the sequence length of the question speech data sample set X, is the loss of noise contrast estimation, is the true probability of the speech data sample to be evaluated, is the learning parameter; Output the speech content text.

[0010] Furthermore, entity mention extraction from the speech content text specifically includes: entity mention extraction from the speech content text based on the BERT model.

[0011] Furthermore, performing PCA analysis on the associated entity mentions to obtain the correlation vector matrix between the associated entity mentions specifically includes the following process: (1) Form a matrix M from the associated entity mentions and calculate the covariance matrix S of the samples of matrix M; (2) Calculate the eigenvectors , …, and eigenvalues t = 1, 2, …, ; (3) Project the data into the space of the eigenvectors using the formula: ; where the value is the eigenvalue corresponding to the associated entity mention in the original sample; (4) Select R vectors and form the correlation vector matrix.

[0012] Furthermore, calculating the module similarity requirement between the correlation vector matrix between entity mentions and the relationship matrix corresponding to the central entity mention in the answer library specifically includes the following process: ; where is the module similarity, is the number of vectors of the correlation vector matrix between entity references, represents the correlation vector matrix and the adjacency matrix of the relationship matrix, , represents the node number of the adjacency matrix, represents the adjacency matrix and is the sum of all the edge numbers in represents the degree of represents the degree of represents the community where the node exists, represents the community where the node exists, is the conditional restriction coefficient. If the number of characters of the entity reference corresponding to the correlation vector matrix between entity references and the relationship matrix of the central entity reference in the answer library is different, then is 0. If the number of characters of the entity reference corresponding to the correlation vector matrix between entity references and the relationship matrix of the central entity reference in the answer library, then is 1.

[0013] Furthermore, the face photo is an RGB image.

[0014] Compared with the existing solutions, the beneficial effects achieved by the present invention are: The present invention activates the face photo acquisition system based on the question voice and acquires the face photo of the current user, extracts the target features representing the face from the face photo, calculates the similarity between the target features and the high-dimensional features stored in the Q&A client, and determines the legitimacy of the user based on the similarity; substitutes the question voice of the legitimate user into the speech text recognition model to output the speech content text; calculates the module similarity between the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library, and outputs the preset answer text corresponding to the central entity reference with the largest module similarity, improving the personalization, accuracy, and intelligence level of the Q&A, enhancing the information acquisition efficiency, and enhancing the scalability and flexibility of the system.

[0015] Using natural language processing and deep learning technologies, the system can more accurately understand the user's question intention, reducing misunderstandings and ambiguities. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0017] Figure 1 It is a system block diagram of a personalized intelligent Q&A system based on artificial intelligence according to an embodiment of the present invention; Figure 2 It is a working flowchart of the first personalized intelligent Q&A system based on artificial intelligence according to an embodiment of the present invention; Figure 3 It is a working flowchart of the second personalized intelligent Q&A system based on artificial intelligence according to an embodiment of the present invention. Specific embodiments

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0019] In addition, the described features, structures or characteristics may be combined in any suitable manner in one or more example embodiments. In the following description, many specific details are provided to give a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will realize that the technical solutions of the present disclosure can be practiced without one or more of the specific details, or other methods, components, steps, etc. can be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0020] This embodiment provides a personalized intelligent Q&A system based on artificial intelligence, Figure 1 which is a system block diagram of a personalized intelligent Q&A system based on artificial intelligence according to an embodiment of the present invention, as Figure 1 shown. The system includes: A legality judgment module, configured to obtain the user's question voice, activate a face photo acquisition system based on the question voice and acquire the face photo of the current user, extract the target features representing the face from the face photo, calculate the similarity between the target features and the high-dimensional features stored in the Q&A client, and judge the legality of the user based on the similarity; A voice text recognition module, configured to substitute the question voice of a legal user into a voice text recognition model and output the voice content text; An entity reference extraction module, configured to perform entity reference extraction on the voice content text, determine the relevance between entity references in the voice content text according to the association rules, and perform PCA analysis on the associated entity references to obtain the correlation vector matrix between the associated entity references; An answer text output module for calculating the module similarity between the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library, and outputting the preset answer text corresponding to the central entity reference with the maximum module similarity.

[0021] In summary, the present invention activates the face photo collection system based on the question voice and collects the face photo of the current user, extracts the target features representing the face from the face photo, calculates the similarity between the target features and the high-dimensional features stored in the Q&A client, and determines the legitimacy of the user based on the similarity; substitutes the question voice of the legitimate user into the speech text recognition model to output the speech content text; calculates the module similarity between the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library, and outputs the preset answer text corresponding to the central entity reference with the maximum module similarity, improving the personalization, accuracy, and intelligence level of the Q&A, enhancing the information acquisition efficiency, and enhancing the scalability and flexibility of the system.

[0022] It should be noted that the face photo collection system usually consists of a front-end collection device, a data transmission module, a data processing center, and a user interaction interface, etc., and the face photo it collects is an RGB picture.

[0023] In some embodiments, calculating the similarity between the target features and the high-dimensional features stored in the Q&A client specifically includes the following process: Calculate the basic similarity of the feature vector values between the target features and the high-dimensional features: ; Wherein, is the basic similarity, represents the i-th feature vector value of the target features representing the face, represents the i-th feature vector value of the high-dimensional feature vector stored in the Q&A client, n is the dimension size of the feature vector, and the high-dimensional features stored in the Q&A client are the pre-extracted user face features.

[0024] Furthermore, Figure 2 is the flowchart of the operation of the first personalized intelligent Q&A system based on artificial intelligence in the embodiments of the present invention. As Figure 2 shown, determining the legitimacy of the user based on the similarity specifically includes the following process: Step S201: Obtain the Q&A moment of the user's last Q&A and the Q&A moment of the current Q&A, calculate the time interval between the Q&A moment of the last Q&A and the Q&A moment of the current Q&A, and record this time interval as the target time interval; calculate the historical similarities corresponding to all the historical time intervals of the user and the query historical time intervals in sequence; Step S202: Number the historical time interval and the target time interval to obtain a numbered value. Using the numbered value as the X-axis and setting the historical similarity and the similarity as the Y-axis, construct a rectangular coordinate system. Connect the adjacent similarities in the rectangular coordinate system to generate a similarity curve. Draw perpendicular lines from both ends of the similarity curve to the X-axis to obtain two start and end line segments. A closed figure is formed by the similarity curve, the two start and end line segments, and the X-axis. Calculate the total area of the closed figure; Step S203: Determine whether the total area exceeds a preset threshold. If so, determine that the user is a legitimate user. If not, determine that the user is an illegal user and reject their access to the Q&A system.

[0025] In some embodiments, substituting the question voice of a legitimate user into the voice text recognition model and outputting the voice content text specifically includes the following process: For the question voice data , is the question voice data sample set; The voice content recognition model consists of two parts: a voice content encoding network f and a voice content context network. Among them, the voice content encoder network f consists of five convolutional networks, and the output of the voice content encoder is a low-frequency feature . The hidden space mapped by the encoder network is represented by Z, where, : ; where m is the convolutional receptive field; The voice content context network is a convolutional neural network, and its structure is the same as that of the voice content encoding network f. It maps the hidden space into a context information space C, and the output of the context network is , where, : ; where v is the convolutional receptive field; Based on the loss function of the voice content recognition model train the voice content recognition model. This loss function is expressed as follows: ; where, is the sequence length of the question voice data sample set X, is the loss of noise contrast estimation, is the true probability of the voice data sample to be evaluated, is the learning parameter; Output the voice content text.

[0026] In some embodiments, the entity reference extraction from the speech content text based on the BERT (Bidirectional Encoder Representations from Transformers) model specifically includes the following processes: Text preprocessing: Preprocess the transformed text, including word segmentation, stop word removal, punctuation processing, etc. This step is to enable the BERT model to better understand and process the text.

[0027] BERT model loading and fine-tuning: Load the pre-trained BERT model and fine-tune it according to the entity reference extraction task. During the fine-tuning process, a dataset with entity reference annotations can be used for training to enable the BERT model to learn specific patterns of entity reference extraction.

[0028] Entity reference extraction: Use the fine-tuned BERT model to perform entity reference extraction on the text. The BERT model will output the vector representations of each word or phrase and their semantic relationships. By comparing these vector representations and semantic relationships, the entity references in the text can be identified.

[0029] In some embodiments, Figure 3 is the flowchart of the second personalized intelligent Q&A system based on artificial intelligence according to the embodiments of the present invention. As Figure 3 shown, performing PCA analysis on the associated entity references to obtain the correlation vector matrix between the associated entity references specifically includes the following processes: Step S301: Form a matrix M from the associated entity references and calculate the covariance matrix S of the samples of matrix M; Step S302: Calculate the eigenvectors and eigenvalues of the covariance matrix S; where the eigenvectors are , …, , and the eigenvalues t = 1, 2, …, ; Step S303: Project the data into the space of the eigenvectors; Using the formula: ; where, the value is the eigenvalue corresponding to the associated entity reference in the original sample; Step S304: Select R vectors and form a correlation vector matrix.

[0030] In some embodiments, the module similarity requirement between the correlation vector matrix of computational entity references and the relationship matrix corresponding to the central entity reference in the answer library specifically includes the following process: ; Among them, is the module similarity, is the number of vectors of the correlation vector matrix between entity references, represents the adjacency matrix of the correlation vector matrix and the relationship matrix, , represents the node number of the adjacency matrix, represents the adjacency matrix the sum of all the edge numbers in, represents 's degree, represents 's degree, represents the community where the node exists, represents the community where the node exists, is the conditional restriction coefficient. If the number of entity reference characters of the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library is different, then is 0. If the number of entity reference characters of the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library, then is 1.

[0031] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more collections of available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0032] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraint requirements of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0033] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0034] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only for some logical function divisions, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0035] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0036] As described above, only the specific implementation manners of this application are provided, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. An artificial intelligence-based personalized intelligent question-answering system, characterized in that, The system includes: A legitimacy judgment module, which is used to obtain the user's question voice, activate the face photo collection system based on the question voice and collect the face photo of the current user, extract the target features representing the face from the face photo, calculate the similarity between the target features and the high-dimensional features stored in the Q&A client, and judge the legitimacy of the user based on the similarity; A voice text recognition module, which is used to substitute the question voice of a legitimate user into the voice text recognition model and output the voice content text; An entity reference extraction module, which is used to perform entity reference extraction on the voice content text, determine the relevance between entity references in the voice content text according to the association rules, perform PCA analysis on the relevant entity references, and obtain the correlation vector matrix between the relevant entity references; An answer text output module, which is used to calculate the module similarity between the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library, and output the preset answer text corresponding to the central entity reference with the largest module similarity.

2. The personalized intelligent Q&A system based on artificial intelligence according to claim 1, characterized in that, Calculating the similarity between the target features and the high-dimensional features stored in the Q&A client specifically includes the following process: Calculating the basic similarity of the feature vector values between the target features and the high-dimensional features: ; Among them, is the basic similarity, represents the i-th eigenvector value of the target feature characterizing the human face, represents the i-th eigenvector value of the high-dimensional eigenvector saved in the Q&A client, where n is the dimensionality size of the eigenvector, and the high-dimensional feature saved in the Q&A client is the pre-extracted user face feature.

3. The personalized intelligent Q&A system based on artificial intelligence according to claim 2, wherein Judging the legitimacy of the user based on the similarity specifically includes the following process: Obtaining the Q&A moment of the user's previous Q&A and the Q&A moment of the current Q&A, calculating the time interval between the Q&A moment of the previous Q&A and the Q&A moment of the current Q&A, and recording the time interval as the target time interval; successively calculating all the historical time intervals of the user and the historical similarities corresponding to the query historical time intervals; Numbering the historical time interval and the target time interval to obtain a number value, constructing a rectangular coordinate system with the number value as the X-axis and setting the historical similarity and similarity as the Y-axis, connecting the adjacent similarities in the rectangular coordinate system to generate a similarity curve, drawing perpendicular lines from both ends of the similarity curve to the X-axis to obtain two starting and ending line segments, and forming a closed figure by the similarity curve, the two starting and ending line segments, and the X-axis, and calculating the total area of the closed figure; Judging whether the total area exceeds a preset threshold. If so, judging the user as a legitimate user; if not, judging the user as an illegal user and rejecting their access to the Q&A system.

4. The personalized intelligent Q&A system based on artificial intelligence according to claim 1, wherein Substituting the question voice of a legitimate user into the voice text recognition model and outputting the voice content text specifically includes the following process: For the problem voice data , is the sample set of the problem voice data; The speech content recognition model consists of two parts: a speech content encoding network f and a speech content context network. Among them, the speech content encoder network f is composed of five convolutional networks, and the output of the speech content encoder is a low-frequency feature , and the hidden space mapped by the encoder network is represented by Z, where :[[]]END]] ; where m is the convolutional receptive field; The speech content context network is a convolutional neural network, whose structure is the same as that of the speech content encoding network f, mapping the hidden space into the context information space C, and the output of the context network is , where :[[]]END]] ; where v is the convolutional receptive field; Loss function of the speech content recognition model Train the speech content recognition model, and the loss function is expressed as follows: ; Among them, is the sequence length of the problem voice data sample set X, is the loss of noise contrast estimation, is the true probability of the voice data sample to be evaluated, is the learning parameter; Outputting the voice content text.

5. The personalized intelligent Q&A system based on artificial intelligence according to claim 1, wherein Performing entity reference extraction on the voice content text specifically includes: performing entity reference extraction on the voice content text based on the BERT model.

6. The personalized intelligent Q&A system based on artificial intelligence according to claim 1, characterized in that Performing PCA analysis on the relevant entity references to obtain the correlation vector matrix between the relevant entity references specifically includes the following process: (1) Forming a matrix M from the relevant entity references and calculating the sample covariance matrix S of the matrix M; (2) Calculate the eigenvectors of the covariance matrix S and … and eigenvalues t = 1, 2, … ; (3) Projecting the data into the space of the eigenvectors, using the formula: ; Among them, the value is the eigenvalue corresponding to the associated entity reference in the original sample; (4) Select R vectors and form a correlation vector matrix.

7. The personalized intelligent question-answering system based on artificial intelligence according to claim 1, characterized in that, Calculating the requirements for the module similarity between the correlation vector matrix between entity references and the relationship matrix corresponding to the central entity reference in the answer library specifically includes the following process: ; Among them, is the module similarity, is the number of vectors of the correlation vector matrix between entity references, represents the correlation vector matrix and the adjacency matrix of the relationship matrix, , represents the node number of the adjacency matrix, represents the adjacency matrix the sum of all the edge numbers in, represents the degree of, represents the degree of, represents the node existing community, represents the node existing community, is the conditional restriction coefficient. If the number of characters of the entity references corresponding to the correlation vector matrix between entity references and the relationship matrix of the central entity reference in the answer library is different, then is 0. If the number of characters of the entity references corresponding to the correlation vector matrix between entity references and the relationship matrix of the central entity reference in the answer library, then is 1.

8. The personalized intelligent Q&A system based on artificial intelligence according to claim 1, characterized in that, The face photo is an RGB image.

Citation Information

Patent Citations

  • Entity linking method for Chinese knowledge graph question-answering system

    CN111563149A

  • Voice interaction method, interaction device and server

    CN112053688A

  • Knowledge base question and answer entity linking method and system based on similarity

    CN112100356A

  • Identity recognition method, device and equipment for video question and answer process and storage medium

    CN113032758A

  • User identity verification method and device

    CN113515643A

Cited By

  • Video monitoring intelligent early warning and communication method and system based on deep learning

    CN120876914A