A multi-source heterogeneous information fusion method and system

Abstract

The present application provides a multi-source heterogeneous information fusion method and system, which can configure a first neural network according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, and the first neural network is used to solve the data error caused by the fusion. Obviously, when the heterogeneous communication to be processed is processed by the first neural network to obtain the corrected heterogeneous communication, the data error in the corrected heterogeneous communication can be avoided or reduced, and the performance of the corrected heterogeneous communication can be improved. The description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication can be used to adjust the processing performance of the second neural network. In this way, when the first neural network is configured according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, the processing performance of the second neural network can also be improved.

Description

Technical Field

[0001] The present application relates to the technical field of data fusion, and in particular to a multi-source heterogeneous information fusion method and system. Background Art

[0002] With the continuous development of information fusion technology, the amount of communication information is increasing, and it is necessary to fuse the communication information. This can effectively reduce the situation where the data processing terminal is paralyzed due to overload caused by excessive communication information.

[0003] However, in the process of merging communication information, different communication information may be merged, which may lead to communication information errors. Summary of the invention

[0004] In view of this, the present application provides a multi-source heterogeneous information fusion method and system.

[0005] In a first aspect, a multi-source heterogeneous information fusion method is provided, the method comprising:

[0006] Obtaining heterogeneous communication information to be processed;

[0007] Transmitting the to-be-processed heterogeneous communication information to the first neural network;

[0008] The first neural network processes the heterogeneous communication information to be processed to obtain a modified heterogeneous communication corresponding to the heterogeneous communication information to be processed;

[0009] The target heterogeneous communication corresponding to the modified heterogeneous communication is passed to the second neural network, so that the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target heterogeneous communication; wherein the first neural network is configured according to the description of the pre-identification reference heterogeneous communication and the description of the modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, and the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed are the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed for the same sequence of heterogeneous communications.

[0010] Furthermore, the configuration process of the first neural network includes:

[0011] Transmitting the heterogeneous communication information of the reference heterogeneous communication before identification and the reference heterogeneous communication to be processed into the first neural network;

[0012] Acquiring, by the first neural network, a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed;

[0013] Determining an information loss vector corresponding to the modified reference heterogeneous communication according to the description of the reference heterogeneous communication before identification and the description of the modified reference heterogeneous communication;

[0014] configuring the first neural network according to the information loss vector to obtain a configured first neural network;

[0015] The heterogeneous communication information of the reference heterogeneous communication to be processed includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents a compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; the obtaining, by the first neural network, the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed includes:

[0016] Acquire a global communication information queue according to the global communication information;

[0017] The reference heterogeneous communication to be processed and the global communication information queue are processed by the first neural network to obtain a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed.

[0018] Further, the description of the pre-identification reference heterogeneous communication is a description vector obtained by transforming a key vector of the pre-identification reference heterogeneous communication;

[0019] The description of the modified reference heterogeneous communication is a description vector obtained by transforming a key vector of the modified reference heterogeneous communication;

[0020] Alternatively, when the second neural network performs fusion processing for characterizing multi-dimensional analysis based on the target description of the target heterogeneous communication, the description of the pre-identification reference heterogeneous communication is the target description of the pre-identification reference heterogeneous communication, and the description of the corrected reference heterogeneous communication is the target description of the corrected reference heterogeneous communication.

[0021] Furthermore, the method further comprises:

[0022] Acquiring, by a third neural network, a description of the pre-identification reference heterogeneous communication;

[0023] The description of the modified reference heterogeneous communication is obtained through the third neural network; wherein the third neural network includes the description screening method of the second neural network.

[0024] Further, the heterogeneous communication information to be processed includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents a compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; the first neural network processes the heterogeneous communication information to be processed to obtain the modified heterogeneous communication corresponding to the heterogeneous communication information to be processed, including:

[0025] Acquire a global communication information queue according to the global communication information;

[0026] Integrating the to-be-processed heterogeneous communications and the global communication information queue to obtain an integrated queue;

[0027] Processing the integrated queue to obtain modified heterogeneous communication;

[0028] The heterogeneous communication information to be processed includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents a compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; the first neural network processes the heterogeneous communication information to be processed to obtain the modified heterogeneous communication corresponding to the heterogeneous communication information to be processed, including:

[0029] Acquire a global communication information queue according to the global communication information;

[0030] Processing the to-be-processed heterogeneous communication by using the first sub-parameter of the first neural network to obtain original processed heterogeneous communication;

[0031] The global communication information queue is processed by the second sub-parameter of the first neural network to obtain an importance information queue, wherein each attribute range of the importance information queue corresponds to an importance information;

[0032] For each key content of the original processing heterogeneous communication, determine the attribute range corresponding to the key content from the importance information queue, and modify the key vector of the key content according to the importance information of the attribute range to obtain a modified key vector;

[0033] Corrected heterogeneous communications are obtained based on the corrected key vectors of each key content.

[0034] In a second aspect, a multi-source heterogeneous information fusion system is provided, including a data acquisition terminal and a data processing terminal, wherein the data acquisition terminal is communicatively connected to the data processing terminal, and the data processing terminal is specifically used for:

[0035] Obtaining heterogeneous communication information to be processed;

[0036] Transmitting the to-be-processed heterogeneous communication information to the first neural network;

[0037] The first neural network processes the heterogeneous communication information to be processed to obtain a modified heterogeneous communication corresponding to the heterogeneous communication information to be processed;

[0038] The target heterogeneous communication corresponding to the modified heterogeneous communication is passed to the second neural network, so that the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target heterogeneous communication; wherein the first neural network is configured according to the description of the pre-identification reference heterogeneous communication and the description of the modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, and the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed are the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed for the same sequence of heterogeneous communications.

[0039] Furthermore, the data processing terminal is specifically used for:

[0040] Transmitting the heterogeneous communication information of the reference heterogeneous communication before identification and the reference heterogeneous communication to be processed into the first neural network;

[0041] Acquiring, by the first neural network, a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed;

[0042] Determining an information loss vector corresponding to the modified reference heterogeneous communication according to the description of the reference heterogeneous communication before identification and the description of the modified reference heterogeneous communication;

[0043] configuring the first neural network according to the information loss vector to obtain a configured first neural network;

[0044] The heterogeneous communication information of the reference heterogeneous communication to be processed includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents a compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; the obtaining, by the first neural network, the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed includes:

[0045] Acquire a global communication information queue according to the global communication information;

[0046] The reference heterogeneous communication to be processed and the global communication information queue are processed by the first neural network to obtain a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed.

[0047] Furthermore, the data processing terminal is specifically used for:

[0048] The description of the pre-identification reference heterogeneous communication is a description vector obtained by transforming a key vector of the pre-identification reference heterogeneous communication;

[0049] The description of the modified reference heterogeneous communication is a description vector obtained by transforming a key vector of the modified reference heterogeneous communication;

[0050] Alternatively, when the second neural network performs fusion processing for characterizing multi-dimensional analysis based on the target description of the target heterogeneous communication, the description of the pre-identification reference heterogeneous communication is the target description of the pre-identification reference heterogeneous communication, and the description of the corrected reference heterogeneous communication is the target description of the corrected reference heterogeneous communication.

[0051] Furthermore, the data processing terminal is specifically used for:

[0052] Acquiring, by a third neural network, a description of the pre-identification reference heterogeneous communication;

[0053] The description of the modified reference heterogeneous communication is obtained through the third neural network; wherein the third neural network includes the description screening method of the second neural network.

[0054] Furthermore, the data processing terminal is specifically used for:

[0055] Acquire a global communication information queue according to the global communication information;

[0056] Integrating the to-be-processed heterogeneous communications and the global communication information queue to obtain an integrated queue;

[0057] Processing the integrated queue to obtain modified heterogeneous communication;

[0058] Wherein, the data processing terminal is specifically used for:

[0059] Acquire a global communication information queue according to the global communication information;

[0060] Processing the to-be-processed heterogeneous communication by using the first sub-parameter of the first neural network to obtain original processed heterogeneous communication;

[0061] The global communication information queue is processed by the second sub-parameter of the first neural network to obtain an importance information queue, wherein each attribute range of the importance information queue corresponds to an importance information;

[0062] For each key content of the original processing heterogeneous communication, determine the attribute range corresponding to the key content from the importance information queue, and modify the key vector of the key content according to the importance information of the attribute range to obtain a modified key vector;

[0063] Corrected heterogeneous communications are obtained based on the corrected key vectors of each key content.

[0064] A multi-source heterogeneous information fusion method and system provided in an embodiment of the present application can configure a first neural network according to a description of a reference heterogeneous communication before identification and a description of a corrected reference heterogeneous communication, and the first neural network is used to solve the data error caused by the fusion. Obviously, when the heterogeneous communication to be processed is processed by the first neural network to obtain a corrected heterogeneous communication, the data error in the corrected heterogeneous communication can be avoided or reduced, and the performance of the corrected heterogeneous communication can be improved. The description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication can be used to adjust the processing performance of the second neural network. In this way, when the first neural network is configured according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, the processing performance of the second neural network can also be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0066] Figure 1 A flowchart of a multi-source heterogeneous information fusion method provided in an embodiment of the present application.

[0067] Figure 2 A block diagram of a multi-source heterogeneous information fusion device provided in an embodiment of the present application.

[0068] Figure 3 An architectural diagram of a multi-source heterogeneous information fusion system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0069] In order to better understand the above technical scheme, the technical scheme of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical scheme of the present application, rather than limitations on the technical scheme of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.

[0070] See also Figure 1 , shows a multi-source heterogeneous information fusion method, which may include the technical solutions described in the following steps 100-400.

[0071] Step 100: Obtain heterogeneous communication information to be processed.

[0072] Exemplarily, the heterogeneous communication information to be processed is used to represent network communication information.

[0073] Step 200: passing the heterogeneous communication information to be processed to the first neural network.

[0074] Step 300: The first neural network processes the heterogeneous communication information to be processed to obtain modified heterogeneous communication corresponding to the heterogeneous communication information to be processed.

[0075] Step 400, the target heterogeneous communication corresponding to the modified heterogeneous communication is transmitted to the second neural network, so that the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target heterogeneous communication.

[0076] Exemplarily, the first neural network is configured based on a description of a pre-identification reference heterogeneous communication and a description of a modified reference heterogeneous communication corresponding to a reference heterogeneous communication to be processed, wherein the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed are pre-identification reference heterogeneous communication and reference heterogeneous communication to be processed for the same sequence of heterogeneous communications.

[0077] It can be understood that when executing the technical solution described in the above steps 100-400, the first neural network can be configured according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, and the first neural network is used to solve the data error caused by fusion. Obviously, when the heterogeneous communication to be processed is processed by the first neural network to obtain the corrected heterogeneous communication, the data error in the corrected heterogeneous communication can be avoided or alleviated, and the performance of the corrected heterogeneous communication can be improved. The description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication can be used to adjust the processing performance of the second neural network. In this way, when the first neural network is configured according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, the processing performance of the second neural network can also be improved.

[0078] In an alternative embodiment, the inventors discovered that in the configuration process of the first neural network, there is a problem of inaccurate calculation, making it difficult to accurately obtain the configured first neural network. In order to improve the above technical problem, the steps of the configuration process of the first neural network described in step 200 may specifically include the technical solutions described in the following steps q1 to q4.

[0079] Step q1, the heterogeneous communication information of the reference heterogeneous communication before identification and the reference heterogeneous communication to be processed is transmitted to the first neural network.

[0080] Step q2: the first neural network obtains a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed.

[0081] Step q3: determining an information loss vector corresponding to the modified reference heterogeneous communication according to the description of the reference heterogeneous communication before identification and the description of the modified reference heterogeneous communication.

[0082] Step q4, configuring the first neural network according to the information loss vector to obtain a configured first neural network.

[0083] It can be understood that when executing the technical solution described in the above steps q1 to q4, the problem of inaccurate calculation is improved during the configuration process of the first neural network, so that the configured first neural network can be accurately obtained.

[0084] In an alternative embodiment, the inventors found that the heterogeneous communication information of the reference heterogeneous communication to be processed includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents the compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; when the first neural network obtains the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, there is a problem of inaccurate global communication information queue, which makes it difficult to accurately correct. In order to improve the above technical problems, the heterogeneous communication information of the reference heterogeneous communication to be processed described in step q2 includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents the compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; the step of obtaining the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed by the first neural network can specifically include the technical solutions described in the following steps q21 and q22.

[0085] Step q21, obtaining a global communication information queue according to the global communication information.

[0086] Step q22, processing the reference heterogeneous communication to be processed and the global communication information queue through the first neural network to obtain a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed

[0087] It can be understood that when executing the technical solution described in the above steps q21 and q22, the heterogeneous communication information of the reference heterogeneous communication to be processed includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents the compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; when the first neural network obtains the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, the problem of inaccurate global communication information queuing is improved, so that it can be accurately corrected.

[0088] In an alternative embodiment, the technical solution described in the following steps w1 to w3 may be included.

[0089] Step w1, the description of the pre-identification reference heterogeneous communication is a description vector obtained by transforming a key vector of the pre-identification reference heterogeneous communication.

[0090] Step w2, the description of the modified reference heterogeneous communication is a description vector obtained by transforming the key vector of the modified reference heterogeneous communication.

[0091] Step w3, or, when the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target description of the target heterogeneous communication, the description of the pre-identification reference heterogeneous communication is the target description of the pre-identification reference heterogeneous communication, and the description of the corrected reference heterogeneous communication is the target description of the corrected reference heterogeneous communication.

[0092] It can be understood that when executing the technical solution described in the above steps w1 to w3, the target description can be accurately obtained through accurate judgment.

[0093] Based on the above foundation, the technical solutions described in steps e1 and e2 may also be included.

[0094] Step e1, obtaining the description of the pre-identification reference heterogeneous communication through a third neural network.

[0095] Step e2, obtaining the description of the modified reference heterogeneous communication through the third neural network; wherein the third neural network includes the description screening method of the second neural network.

[0096] It can be understood that when executing the technical solutions described in the above steps e1 and e2, the accuracy of the description of the modified reference heterogeneous communication is improved by accurately identifying the description of the previous reference heterogeneous communication.

[0097] In a possible embodiment, the inventors found that the heterogeneous communication information to be processed includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents a compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; when the heterogeneous communication information to be processed is processed by the first neural network, there is a problem of inaccurate integration of the heterogeneous communication to be processed and the global communication information queue, so that it is difficult to accurately obtain the modified heterogeneous communication corresponding to the heterogeneous communication information to be processed, In order to improve the above technical problems, the heterogeneous communication information to be processed described in step 300 includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents the compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; the step of processing the heterogeneous communication information to be processed by the first neural network to obtain the corrected heterogeneous communication corresponding to the heterogeneous communication information to be processed can include the following steps, specifically including the technical solutions described in the following steps r1-r3.

[0098] Step r1, obtaining a global communication information queue according to the global communication information.

[0099] Step r2, integrating the to-be-processed heterogeneous communications and the global communication information queue to obtain an integrated queue.

[0100] Step r3, processing the integrated queue to obtain modified heterogeneous communication.

[0101] It can be understood that when executing the technical solution described in the above steps r1 to r3, the heterogeneous communication information to be processed includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents a compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; when the heterogeneous communication information to be processed is processed by the first neural network, the problem of inaccurate integration of the heterogeneous communication to be processed and the global communication information queue is improved, so that the corrected heterogeneous communication corresponding to the heterogeneous communication information to be processed can be accurately obtained.

[0102] In an alternative embodiment, the inventors found that the heterogeneous communication information to be processed includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents a compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; when the first neural network processes the heterogeneous communication information to be processed, there is a problem that each attribute range of the importance information queue corresponds to an inaccurate importance information, so that it is difficult to accurately obtain the modified information corresponding to the heterogeneous communication information to be processed. Positive heterogeneous communication, in order to improve the above technical problems, the heterogeneous communication information to be processed described in step 300 includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents the compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; the step of processing the heterogeneous communication information to be processed by the first neural network to obtain the corrected heterogeneous communication corresponding to the heterogeneous communication information to be processed may specifically include the technical solutions described in the following steps t1-t5.

[0103] Step t1, obtaining a global communication information queue according to the global communication information.

[0104] Step t2: Processing the to-be-processed heterogeneous communication by using the first sub-parameter of the first neural network to obtain the original processed heterogeneous communication.

[0105] Step t3, processing the global communication information queue by the second sub-parameter of the first neural network to obtain an importance information queue, wherein each attribute range of the importance information queue corresponds to an importance information.

[0106] Step t4, for each key content of the original processing heterogeneous communication, determine the attribute range corresponding to the key content from the importance information queue, and modify the key vector of the key content according to the importance information of the attribute range to obtain a modified key vector.

[0107] Step t5, obtaining the modified heterogeneous communication according to the modified key vectors of each key content.

[0108] It can be understood that when executing the technical solution described in the above steps t1 to t5, the heterogeneous communication information to be processed includes the heterogeneous communication to be processed and the global communication information corresponding to the heterogeneous communication to be processed; the global communication information represents a compensatory description of the heterogeneous communication to be processed relative to the heterogeneous communication before identification, and the heterogeneous communication before identification and the heterogeneous communication to be processed are the heterogeneous communication before identification and the heterogeneous communication to be processed for the same sequence of heterogeneous communications; when the heterogeneous communication information to be processed is processed by the first neural network, the problem of inaccurate importance information corresponding to each attribute range of the importance information queue is improved, so that the corrected heterogeneous communication corresponding to the heterogeneous communication information to be processed can be accurately obtained.

[0109] In a possible embodiment, the inventors have discovered that when the global communication information queue is processed by the second sub-parameter of the first neural network, there is a problem of inaccurate error vectors in various attribute ranges, making it difficult to accurately obtain the importance information queue. In order to improve the above technical problem, the step of processing the global communication information queue by the second sub-parameter of the first neural network to obtain the importance information queue as described in step t3 may specifically include the technical solutions described in the following steps t31 to t33.

[0110] Step t31, processing the global communication information queue by the second sub-parameter of the first neural network to obtain the error vector of each attribute range in the global communication information queue.

[0111] Step t32: for each attribute range, determine the importance information of the attribute range according to the error vector of the attribute range; wherein, the higher the error vector of the attribute range, the greater the importance information of the attribute range.

[0112] Step t33, generating the importance information queue according to the importance information of each attribute range.

[0113] It can be understood that when executing the technical solution described in the above steps t31 to t33, when the global communication information queue is processed by the second sub-parameter of the first neural network, the problem of inaccurate error vectors in each attribute range is improved, so that the importance information queue can be accurately obtained.

[0114] In a possible embodiment, the inventors found that when the heterogeneous communication information to be processed is transmitted to the first neural network, there is a problem of feedback error, which makes it difficult to accurately transmit the heterogeneous communication information to be processed to the first neural network. In order to improve the above technical problem, the step of transmitting the heterogeneous communication information to be processed to the first neural network described in step 200 can specifically include the technical solutions described in the following steps y1-y3.

[0115] Step y1, obtaining the correlation degree of the heterogeneous communication information to be processed.

[0116] Step y2: determining whether to perform feedback processing on the heterogeneous communication information to be processed according to the correlation degree.

[0117] Step y3: If yes, the heterogeneous communication information to be processed is transmitted to the first neural network.

[0118] It can be understood that when executing the technical solution described in the above steps y1 to y3, when the heterogeneous communication information to be processed is transmitted to the first neural network, the problem of feedback error is improved, so that the heterogeneous communication information to be processed can be accurately transmitted to the first neural network.

[0119] In a possible embodiment, the inventors discovered that when the target heterogeneous communication corresponding to the modified heterogeneous communication is transmitted to the second neural network, there is a problem that the modified heterogeneous communication is inaccurately determined as the target heterogeneous communication, making it difficult to accurately transmit the target heterogeneous communication corresponding to the modified heterogeneous communication to the second neural network. In order to improve the above technical problem, the step of transmitting the target heterogeneous communication corresponding to the modified heterogeneous communication to the second neural network described in step 400 may specifically include the technical solutions described in the following steps a1 to a3.

[0120] Step a1: determining the modified heterogeneous communication as the target heterogeneous communication.

[0121] Step a2, alternatively, post-processing the modified heterogeneous communication to obtain enhanced heterogeneous communication corresponding to the modified heterogeneous communication, and determining the enhanced heterogeneous communication as the target heterogeneous communication.

[0122] Step a3, transmitting the target heterogeneous communication to the second neural network.

[0123] It can be understood that when executing the technical solution described in the above steps a1 to a3, when the target heterogeneous communication corresponding to the corrected heterogeneous communication is transmitted to the second neural network, the problem of inaccurate determination of the corrected heterogeneous communication as the target heterogeneous communication is improved, so that the target heterogeneous communication corresponding to the corrected heterogeneous communication can be accurately transmitted to the second neural network.

[0124] Based on the above, please refer to Figure 2 , a multi-source heterogeneous information fusion device 200 is provided, which is applied to a data processing terminal, and the device includes:

[0125] The information acquisition module 210 is used to acquire the heterogeneous communication information to be processed;

[0126] An information input module 220, used for inputting the to-be-processed heterogeneous communication information into the first neural network;

[0127] A communication correction module 230, configured to process the heterogeneous communication information to be processed by the first neural network to obtain a corrected heterogeneous communication corresponding to the heterogeneous communication information to be processed;

[0128] The communication fusion module 240 is used to pass the target heterogeneous communication corresponding to the modified heterogeneous communication to the second neural network, so that the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target heterogeneous communication; wherein the first neural network is configured according to the description of the pre-identification reference heterogeneous communication and the description of the modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, and the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed are the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed for the same sequence of heterogeneous communications.

[0129] Based on the above, please refer to Figure 3 , shows a multi-source heterogeneous information fusion system 300, including a processor 310 and a memory 320 that communicate with each other, and the processor 310 is used to read and execute a computer program from the memory 320 to implement the above method.

[0130] Based on the above, a computer-readable storage medium is also provided, on which a computer program stored implements the above method when running.

[0131] In summary, based on the above scheme, the first neural network can be configured according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, and the first neural network is used to solve the data error caused by fusion. Obviously, when the heterogeneous communication to be processed is processed by the first neural network to obtain the corrected heterogeneous communication, the data error in the corrected heterogeneous communication can be avoided or alleviated, and the performance of the corrected heterogeneous communication can be improved. The description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication can be used to adjust the processing performance of the second neural network. In this way, when the first neural network is configured according to the description of the reference heterogeneous communication before identification and the description of the corrected reference heterogeneous communication, the processing performance of the second neural network can also be improved.

[0132] It should be understood that the system and its modules shown above can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or a dedicated design hardware. Those skilled in the art will understand that the above methods and systems can be implemented using computer executable instructions and / or included in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The system and its modules of the present application can not only be implemented by hardware circuits such as ultra-large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but can also be implemented by software such as executed by various types of processors, and can also be implemented by a combination of the above hardware circuits and software (e.g., firmware).

[0133] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects that may be produced may be any one or a combination of the above, or any other beneficial effects that may be obtained.

[0134] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of the present application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements and amendments to the present application. Such modifications, improvements and amendments are suggested in the present application, so such modifications, improvements and amendments still belong to the spirit and scope of the exemplary embodiments of the present application.

[0135] At the same time, the present application uses specific words to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be appropriately combined.

[0136] In addition, it will be appreciated by those skilled in the art that various aspects of the present application may be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any new and useful improvements thereto. Accordingly, various aspects of the present application may be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of the present application may be represented as a computer product located in one or more computer-readable media, which includes computer-readable program code.

[0137] A computer storage medium may include a propagated data signal containing computer program code, for example, in baseband or as part of a carrier wave. The propagated signal may be in a variety of forms, including electromagnetic, optical, etc., or a suitable combination. A computer storage medium may be any computer-readable medium other than a computer-readable storage medium, which can be connected to an instruction execution system, device or apparatus to communicate, propagate or transmit the program for use. The program code on the computer storage medium may be transmitted via any suitable medium, including radio, cable, fiber optic cable, RF, or similar media, or any combination of the above media.

[0138] The computer program coding required for the operation of each part of the application can be written in any one or more programming languages, including object-oriented programming languages ​​such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., conventional procedural programming languages ​​such as C language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages ​​such as Python, Ruby and Groovy, or other programming languages, etc. The program coding can be run completely on the user's computer, or run on the user's computer as an independent software package, or run partly on the user's computer and partly on the remote computer, or run completely on the remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as software as a service (SaaS).

[0139] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this application, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this application. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this application. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0140] Similarly, it should be noted that in order to simplify the description of the disclosure of this application and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this application, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the object of this application are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.

[0141] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used for the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers allow adaptive changes. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which can be changed according to the required characteristics of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining the digits. Although the numerical domains and parameters used to confirm the breadth of the range in some embodiments of the present application are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0142] Each patent, patent application, patent application disclosure, and other materials, such as articles, books, instructions, publications, documents, etc., cited in this application are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the content of this application, documents that limit the broadest scope of the claims of this application (currently or later attached to this application) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the attached materials of this application are inconsistent or conflicting with the content described in this application, the descriptions, definitions, and / or use of terms in this application shall prevail.

[0143] Finally, it should be understood that the embodiments described in this application are only used to illustrate the principles of the embodiments of the present application. Other variations may also fall within the scope of the present application. Therefore, as an example and not a limitation, the alternative configurations of the embodiments of the present application may be considered to be consistent with the teachings of the present application. Accordingly, the embodiments of the present application are not limited to the embodiments explicitly introduced and described in the present application.

[0144] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A multi-source heterogeneous information fusion method, characterized in that: The method comprises: Obtaining heterogeneous communication information to be processed; Transmitting the to-be-processed heterogeneous communication information to the first neural network; The first neural network processes the heterogeneous communication information to be processed to obtain a modified heterogeneous communication corresponding to the heterogeneous communication information to be processed; The target heterogeneous communication corresponding to the modified heterogeneous communication is passed to the second neural network, so that the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target heterogeneous communication; wherein the first neural network is configured according to the description of the pre-identification reference heterogeneous communication and the description of the modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, and the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed are the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed for the same sequence heterogeneous communication; The configuration process of the first neural network includes: Transmitting the heterogeneous communication information of the reference heterogeneous communication before identification and the reference heterogeneous communication to be processed into the first neural network; Acquiring, by the first neural network, a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed; Determining an information loss vector corresponding to the modified reference heterogeneous communication according to the description of the reference heterogeneous communication before identification and the description of the modified reference heterogeneous communication; configuring the first neural network according to the information loss vector to obtain a configured first neural network; The heterogeneous communication information of the reference heterogeneous communication to be processed includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents a compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; the obtaining, by the first neural network, the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed includes: Acquire a global communication information queue according to the global communication information; Processing the reference heterogeneous communication to be processed and the global communication information queue by the first neural network to obtain a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed; The description of the reference heterogeneous communication before identification is a description vector obtained by transforming a key vector of the reference heterogeneous communication before identification; The description of the modified reference heterogeneous communication is a description vector obtained by transforming a key vector of the modified reference heterogeneous communication; Alternatively, when the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target description of the target heterogeneous communication, the description of the pre-identification reference heterogeneous communication is the target description of the pre-identification reference heterogeneous communication, and the description of the modified reference heterogeneous communication is the target description of the modified reference heterogeneous communication; Wherein, the method further comprises: Acquiring, by a third neural network, a description of the pre-identification reference heterogeneous communication; The description of the modified reference heterogeneous communication is obtained through the third neural network; wherein the third neural network includes the description screening method of the second neural network.

2. The multi-source heterogeneous information fusion method according to claim 1 is characterized by: The to-be-processed heterogeneous communication information includes the to-be-processed heterogeneous communication and the global communication information corresponding to the to-be-processed heterogeneous communication; The global communication information represents a compensatory description of the to-be-processed heterogeneous communication relative to the pre-identification heterogeneous communication, wherein the pre-identification heterogeneous communication and the to-be-processed heterogeneous communication are the pre-identification heterogeneous communication and the to-be-processed heterogeneous communication for the same sequence of heterogeneous communications; The first neural network processes the heterogeneous communication information to be processed to obtain a modified heterogeneous communication corresponding to the heterogeneous communication information to be processed, including: Acquire a global communication information queue according to the global communication information; Integrating the to-be-processed heterogeneous communications and the global communication information queue to obtain an integrated queue; The integrated queue is processed to obtain a modified heterogeneous communication.

3. The multi-source heterogeneous information fusion method according to claim 1, characterized in that: The to-be-processed heterogeneous communication information includes the to-be-processed heterogeneous communication and the global communication information corresponding to the to-be-processed heterogeneous communication; The global communication information represents a compensatory description of the to-be-processed heterogeneous communication relative to the pre-identification heterogeneous communication, wherein the pre-identification heterogeneous communication and the to-be-processed heterogeneous communication are the pre-identification heterogeneous communication and the to-be-processed heterogeneous communication for the same sequence of heterogeneous communications; The first neural network processes the heterogeneous communication information to be processed to obtain a modified heterogeneous communication corresponding to the heterogeneous communication information to be processed, including: Acquire a global communication information queue according to the global communication information; Processing the to-be-processed heterogeneous communication by using the first sub-parameter of the first neural network to obtain original processed heterogeneous communication; The global communication information queue is processed by the second sub-parameter of the first neural network to obtain an importance information queue, wherein each attribute range of the importance information queue corresponds to an importance information; For each key content of the original processing heterogeneous communication, determine the attribute range corresponding to the key content from the importance information queue, and modify the key vector of the key content according to the importance information of the attribute range to obtain a modified key vector; Corrected heterogeneous communications are obtained based on the corrected key vectors of each key content.

4. A multi-source heterogeneous information fusion system, characterized in that: It includes a data acquisition terminal and a data processing terminal, the data acquisition terminal and the data processing terminal are communicatively connected, and the data processing terminal is specifically used for: Obtaining heterogeneous communication information to be processed; Transmitting the to-be-processed heterogeneous communication information to the first neural network; The first neural network processes the heterogeneous communication information to be processed to obtain a modified heterogeneous communication corresponding to the heterogeneous communication information to be processed; The target heterogeneous communication corresponding to the modified heterogeneous communication is passed to the second neural network, so that the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target heterogeneous communication; wherein the first neural network is configured according to the description of the pre-identification reference heterogeneous communication and the description of the modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed, and the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed are the pre-identification reference heterogeneous communication and the reference heterogeneous communication to be processed for the same sequence heterogeneous communication; The data processing terminal is further used for: Transmitting the heterogeneous communication information of the reference heterogeneous communication before identification and the reference heterogeneous communication to be processed into the first neural network; Acquiring, by the first neural network, a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed; Determining an information loss vector corresponding to the modified reference heterogeneous communication according to the description of the reference heterogeneous communication before identification and the description of the modified reference heterogeneous communication; configuring the first neural network according to the information loss vector to obtain a configured first neural network; The heterogeneous communication information of the reference heterogeneous communication to be processed includes the reference heterogeneous communication to be processed and the global communication information corresponding to the reference heterogeneous communication to be processed; the global communication information represents a compensatory description of the reference heterogeneous communication to be processed relative to the reference heterogeneous communication before identification; the obtaining, by the first neural network, the corrected reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed includes: Acquire a global communication information queue according to the global communication information; Processing the reference heterogeneous communication to be processed and the global communication information queue by the first neural network to obtain a modified reference heterogeneous communication corresponding to the reference heterogeneous communication to be processed; The data processing terminal is further used for: The description of the pre-identification reference heterogeneous communication is a description vector obtained by transforming a key vector of the pre-identification reference heterogeneous communication; The description of the modified reference heterogeneous communication is a description vector obtained by transforming a key vector of the modified reference heterogeneous communication; Alternatively, when the second neural network performs fusion processing for characterizing multi-dimensional analysis according to the target description of the target heterogeneous communication, the description of the pre-identification reference heterogeneous communication is the target description of the pre-identification reference heterogeneous communication, and the description of the modified reference heterogeneous communication is the target description of the modified reference heterogeneous communication; The data processing terminal is further used for: Acquiring, by a third neural network, a description of the pre-identification reference heterogeneous communication; The description of the modified reference heterogeneous communication is obtained through the third neural network; wherein the third neural network includes the description screening method of the second neural network.

5. The multi-source heterogeneous information fusion system according to claim 4 is characterized in that: The data processing terminal is also specifically used for: Acquire a global communication information queue according to the global communication information; Integrating the to-be-processed heterogeneous communications and the global communication information queue to obtain an integrated queue; The integrated queue is processed to obtain a modified heterogeneous communication.

6. The multi-source heterogeneous information fusion system according to claim 4 is characterized by: The data processing terminal is also specifically used for: Acquire a global communication information queue according to the global communication information; Processing the to-be-processed heterogeneous communication by using the first sub-parameter of the first neural network to obtain original processed heterogeneous communication; The global communication information queue is processed by the second sub-parameter of the first neural network to obtain an importance information queue, wherein each attribute range of the importance information queue corresponds to an importance information; For each key content of the original processing heterogeneous communication, determine the attribute range corresponding to the key content from the importance information queue, and modify the key vector of the key content according to the importance information of the attribute range to obtain a modified key vector; Corrected heterogeneous communications are obtained based on the corrected key vectors of each key content.

Images