An artificial intelligence technology hybrid-based knowledge data fusion method

Abstract

The application discloses a kind of knowledge data fusion methods based on artificial intelligence technology mixing, it is related to data processing technical field, including: obtaining complete knowledge data;Complete knowledge data is classified to form data fusion set;Get target accessory knowledge data;The innovation coefficient of target accessory knowledge data is calculated, and target accessory knowledge data sequence is obtained;Based on main fusion node and vice fusion node, target accessory knowledge data and subject knowledge data in data fusion set are fused.It is peeled off in the complete logic architecture in the data to be fused, complete knowledge data is classified, and target accessory knowledge data and subject knowledge data in data fusion set are fused, and the complete knowledge architecture is obtained, the subject knowledge data is used as the subject architecture of subsequent fusion, so that the fusion is more orderly, the situation of existing contradiction is handled, and the final fusion effect is more reasonable.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a knowledge data fusion method based on artificial intelligence technology. Background Technology

[0002] Different platforms or business domains possess their own data. With the development of data management and data infrastructure, there is a growing desire to integrate and connect data from multiple platforms and business domains. Knowledge graphs are a structured data representation method that can efficiently present the knowledge information contained within data. Achieving knowledge connectivity across multiple platforms and business domains through knowledge graphs can effectively improve the efficiency of data fusion, leading to enhanced business results and computational performance.

[0003] However, due to the uncertain types of knowledge data and their irregular arrangement, as well as the overlapping and other uncertainties, data fusion faces significant challenges, and the fusion effect is difficult to meet the requirements. Summary of the Invention

[0004] To address the aforementioned technical problems, a knowledge data fusion method based on artificial intelligence technology is provided. This technical solution solves the problems mentioned in the background section.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A knowledge data fusion method based on artificial intelligence technology includes:

[0007] The knowledge data awaiting fusion is aggregated into data to be fused, and the complete logical architecture is extracted from the data to be fused to obtain at least one complete knowledge data.

[0008] The complete knowledge data is classified to form at least one data fusion set, which consists of the main knowledge data and at least one auxiliary knowledge data.

[0009] Filter the supplementary knowledge data in the data fusion set to obtain at least one target supplementary knowledge data;

[0010] Calculate the innovation coefficient of the target-related knowledge data, and sort the target-related knowledge data in the data fusion set from largest to smallest according to the innovation coefficient to obtain the target-related knowledge data sequence;

[0011] Generate at least one main fusion node in the main knowledge data and at least one secondary fusion node in the target auxiliary knowledge data;

[0012] Based on the main fusion node and the secondary fusion node, the target affiliated knowledge data and the main knowledge data in the data fusion set are fused, and the fusion sequence is according to the target affiliated knowledge data sequence corresponding to the data fusion set.

[0013] Preferably, the stripping of the complete logical architecture in the data to be fused includes the following steps:

[0014] Based on big data, at least one knowledge data containing two knowledge points is obtained as sample knowledge data.

[0015] Half of the minimum value of the size of the sample knowledge data is taken as the preset value.

[0016] The data to be fused is segmented to obtain at least one local knowledge block, and the size of the local knowledge block is the preset value.

[0017] Based on big data, at least one application scenario of the local knowledge block is obtained, and at least one application feature is obtained by feature extraction.

[0018] The proportion of the number of coincident application features of the two local knowledge blocks to the total number of application features of the two local knowledge blocks is taken as the association degree of the two local knowledge blocks.

[0019] The association degree of the two sample knowledge data is calculated, and the maximum value of the association degree of the two sample knowledge data without knowledge point coincidence is taken as the association threshold.

[0020] When the association degree of the two local knowledge blocks exceeds the association threshold, an association path is connected between the two local knowledge blocks.

[0021] At least one complete knowledge data is formed, which satisfies that the complete knowledge data is composed of local knowledge blocks, any two local knowledge blocks in the complete knowledge data are connected by a single association path, the single association path is composed of at least one association path, and there is no association path between the local knowledge blocks of different complete knowledge data.

[0022] Preferably, the classification of the complete knowledge data to form at least one data fusion set includes the following steps:

[0023] The number of association paths connected by the local knowledge blocks in the complete knowledge data is counted as the characteristic value of the local knowledge block.

[0024] The local knowledge blocks in the complete knowledge data are sorted according to the characteristic value from large to small to obtain a local knowledge block sequence.

[0025] The local knowledge blocks are tested in sequence according to the sequence of the local knowledge blocks, and when the local knowledge blocks are deleted, the complete knowledge data is not conducted, and the local knowledge block is taken as a target local knowledge block, and the definition of conduction is that any two local knowledge blocks in the complete knowledge data are connected through a single connection path, otherwise, not conducted;

[0026] The target local knowledge blocks are summarized as main knowledge data, and the local knowledge blocks in the complete knowledge data except the main knowledge data are taken as subsidiary knowledge data.

[0027] Preferably, the subsidiary knowledge data in the data fusion set is screened to obtain at least one target subsidiary knowledge data, including the following steps:

[0028] After the main knowledge data is deleted from the complete knowledge data, at least one subsidiary knowledge data link is formed, which satisfies that any two subsidiary knowledge data in the subsidiary knowledge data link are connected through a single connection path, and there is no associated path between the subsidiary knowledge data in different subsidiary knowledge data links;

[0029] All combination possibilities of the two subsidiary knowledge data in the subsidiary knowledge data link are obtained to obtain at least one subsidiary knowledge data combination, and the two subsidiary knowledge data in the subsidiary knowledge data combination are taken as a first subsidiary knowledge data and a second subsidiary knowledge data respectively;

[0030] Based on big data, at least one first knowledge application event of the first subsidiary knowledge data is obtained, and first remaining knowledge used in the first knowledge application event is obtained, and the first remaining knowledge does not contain the first subsidiary knowledge data;

[0031] Based on big data, at least one second knowledge application event of the second subsidiary knowledge data is obtained, and second remaining knowledge used in the second knowledge application event is obtained, and the second remaining knowledge does not contain the second subsidiary knowledge data;

[0032] The second knowledge application event and the first knowledge application event in which the second remaining knowledge is consistent with the first remaining knowledge are established in a corresponding relationship, wherein the second knowledge application event and the first knowledge application event in which the second remaining knowledge is inconsistent with the first remaining knowledge do not have an overlapping relationship;

[0033] The second knowledge application event and the first knowledge application event are deduced and compared, when the second knowledge application event can be deduced from the first knowledge application event, the second knowledge application event is covered by the first knowledge application event, when the first knowledge application event can be deduced from the second knowledge application event, the first knowledge application event is covered by the second knowledge application event;

[0034] When the first knowledge application event of the first subsidiary knowledge data is covered by the second knowledge application event of the second subsidiary knowledge data, the first subsidiary knowledge data is deleted; when the second knowledge application event of the second subsidiary knowledge data is covered by the first knowledge application event of the first subsidiary knowledge data, the second subsidiary knowledge data is deleted.

[0035] The subsidiary knowledge data in the subsidiary knowledge data link after the pruning is taken as target subsidiary knowledge data.

[0036] Preferably, the deducing and comparing of the second knowledge application event and the first knowledge application event comprises the following steps:

[0037] Obtaining at least one sample application requirement, obtaining the completed part of the sample application requirement by the second knowledge application event as a second completed part, and obtaining the completed part of the sample application requirement by the first knowledge application event as a first completed part;

[0038] Merging the at least one second completed part to obtain a second total part, and merging the at least one first completed part to obtain a first total part;

[0039] Taking the intersection of the second total part and the first total part to obtain a characteristic part, taking the part different from the characteristic part in the second total part as a second target part, and taking the part different from the characteristic part in the first total part as a first target part;

[0040] Based on big data, obtaining at least one common sense requirement;

[0041] When the second target part is an empty set or the common sense requirement, the second knowledge application event can be deduced from the first knowledge application event; when the first target part is an empty set or the common sense requirement, the first knowledge application event can be deduced from the second knowledge application event.

[0042] Preferably, the obtaining of the at least one common sense requirement based on big data comprises the following steps:

[0043] Based on big data, obtaining at least one knowledge application requirement;

[0044] Forming a sample test personnel set, which is composed of personnel engaged in various occupations in a proportionate manner;

[0045] When the knowledge application requirements are all completed by the personnel in the sample test personnel set without consulting materials, the knowledge application requirements are taken as common sense requirements.

[0046] Preferably, the calculation of the innovation coefficient of the target subsidiary knowledge data comprises the following steps:

[0047] In the big data, the first occurrence time of the target subsidiary knowledge data is obtained as the target time;

[0048] The first processing result of at least one actual application requirement using the target subsidiary knowledge data is obtained, and the second processing result of the actual application requirement before the target time is obtained;

[0049] The first processing time consumption of the first processing result is counted, the second processing time consumption of the second processing result is counted, and the second processing time consumption is divided by the first processing time consumption to obtain a first coefficient;

[0050] The first completion degree of the first processing result to the actual application requirement is counted, the second completion degree of the second processing result to the actual application requirement is counted, and the first completion degree is divided by the second completion degree to obtain a second coefficient;

[0051] The first coefficient is multiplied by the second coefficient to obtain an innovation coefficient.

[0052] Preferably, the generating at least one main fusion node in the main knowledge data and at least one secondary fusion node in the target subsidiary knowledge data comprises the following steps:

[0053] The local knowledge blocks in the main knowledge data are segmented to obtain at least one main fusion node, and the knowledge in the main fusion node corresponds to the same entity;

[0054] The target subsidiary knowledge data is segmented to obtain at least one secondary fusion node, and the knowledge in the secondary fusion node corresponds to the same entity;

[0055] The main fusion node and the secondary fusion node corresponding to the same entity are established in a corresponding relationship.

[0056] Preferably, the fusing the target subsidiary knowledge data and the main knowledge data in the data fusion set comprises the following steps:

[0057] At least one main application event of the knowledge in the main fusion node is obtained, and at least one secondary application event of the knowledge in the secondary fusion node is obtained;

[0058] The main application event and the secondary application event are deduced and compared, when the main application event can be deduced from the secondary application event, the fusion result of the main fusion node and the secondary fusion node is the secondary fusion node, and when the secondary application event can be deduced from the main application event, the fusion result of the main fusion node and the fusion node is the main fusion node;

[0059] When the main application event and the secondary application event cannot be deduced from each other, the intersection of the knowledge in the main fusion node and the knowledge in the secondary fusion node is obtained to obtain a first fusion part;

[0060] The part different from the first fusion part in the knowledge in the main fusion node is taken as a main category part;

[0061] The part different from the first fusion part in the knowledge in the secondary fusion node is taken as a secondary category part;

[0062] The part with contradiction in the main category part and the secondary category part is taken as a main contradiction part and a secondary contradiction part respectively, the main contradiction part belongs to the main category part, and the secondary contradiction part belongs to the secondary category part;

[0063] The main category part and the secondary category part are fused, the secondary category part is deleted during the fusion, and a second fusion part is obtained;

[0064] The first fusion part and the second fusion part are summarized, and the fusion is completed.

[0065] Compared with the prior art, the present application has the advantages that:

[0066] By stripping the complete logical architecture in the data to be fused, classifying the complete knowledge data, screening the accessory knowledge data in the data fusion set, and fusing the target accessory knowledge data and the main knowledge data in the data fusion set, the complete knowledge architecture can be obtained by data stripping, the knowledge attribute can be judged according to the application of the knowledge, the data to be fused can be reduced, the workload of subsequent processing can be reduced, the main knowledge data can be used as the main architecture for subsequent fusion through the main knowledge data and the accessory knowledge data, the fusion is more orderly, the contradictory situation can be processed during the fusion through the fusion processing of the main fusion node and the secondary fusion node, the final fusion effect is more reasonable, and the actual demand can be met. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 It is a flowchart of the knowledge data fusion method based on the artificial intelligence technology mixing of the present application;

[0068] Figure 2 It is a flowchart of stripping the complete logical architecture in the data to be fused to obtain at least one complete knowledge data of the present application;

[0069] Figure 3 It is a flowchart of classifying the complete knowledge data to form at least one data fusion set of the present application;

[0070] Figure 4 It is a flowchart of screening the accessory knowledge data in the data fusion set to obtain at least one target accessory knowledge data of the present application;

[0071] Figure 5A flowchart for deducing and comparing the second knowledge application event and the first knowledge application event of the present application is shown in the figure;

[0072] Figure 6 A flowchart for obtaining at least one common sense demand based on big data of the present application is shown in the figure;

[0073] Figure 7 A flowchart for calculating the innovation coefficient of the target accessory knowledge data of the present application is shown in the figure;

[0074] Figure 8 A flowchart for generating at least one main fusion node in the main knowledge data and at least one secondary fusion node in the target accessory knowledge data of the present application is shown in the figure;

[0075] Figure 9 A flowchart for fusing the target accessory knowledge data and the main knowledge data in the data fusion set of the present application is shown in the figure. DETAILED DESCRIPTION

[0076] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be thought of by those skilled in the art.

[0077] Reference Figure 1 As shown, a knowledge data fusion method based on artificial intelligence technology mixing includes:

[0078] The knowledge data waiting for fusion is summarized as the to-be-fused data, and the complete logical architecture is stripped in the to-be-fused data to obtain at least one complete knowledge data;

[0079] The complete knowledge data is classified to form at least one data fusion set, and the data fusion set is composed of main knowledge data and at least one accessory knowledge data;

[0080] The accessory knowledge data in the data fusion set is screened to obtain at least one target accessory knowledge data;

[0081] The innovation coefficient of the target accessory knowledge data is calculated, the target accessory knowledge data in the data fusion set is sorted in descending order of the innovation coefficient, and a target accessory knowledge data sequence is obtained;

[0082] At least one main fusion node is generated in the main knowledge data, and at least one secondary fusion node is generated in the target accessory knowledge data;

[0083] Based on the main fusion node and the secondary fusion node, the target accessory knowledge data and the main knowledge data in the data fusion set are fused, and the fusion order is according to the target accessory knowledge data sequence corresponding to the data fusion set.

[0084] There are various data interleaving settings in the data to be fused, so when fusing, different types of data cannot be fused, and fusion needs to be performed in the same type of data, so data needs to be stripped in the data to be fused to obtain complete knowledge data, each complete knowledge data is the same type of data that can be fused, but the complete knowledge data involves a lot of knowledge, there may be approximate knowledge but the innovation degree is different, and there may be contradictions, if these situations are not processed, it is easy to cause the redundancy of the fused data to be too high, which has a great burden on the subsequent use, so in the present scheme, corresponding steps are set to process this situation.

[0085] Referring to Figure 2 The complete logical architecture is stripped in the data to be fused to obtain at least one complete knowledge data, including the following steps:

[0086] Based on big data, at least one knowledge data containing two knowledge points is obtained as sample knowledge data;

[0087] Half of the minimum value of the size of the sample knowledge data is taken as a preset value;

[0088] The data to be fused is segmented to obtain at least one local knowledge block, and the size of the local knowledge block is the preset value;

[0089] Based on big data, at least one application scenario of the local knowledge block is obtained, and at least one application feature is obtained by feature extraction on the at least one application scenario;

[0090] The proportion of the number of coinciding application features of two local knowledge blocks to the total number of application features of the two local knowledge blocks is taken as the association degree of the two local knowledge blocks;

[0091] The association degree of two sample knowledge data is calculated, and the maximum value of the association degree of two sample knowledge data without knowledge point coincidence is taken as an association threshold;

[0092] When the association degree of two local knowledge blocks exceeds the association threshold, an association path is connected between the two local knowledge blocks;

[0093] At least one complete knowledge data is formed, which satisfies that the complete knowledge data is composed of local knowledge blocks, any two local knowledge blocks in the complete knowledge data are connected through a single association path, the single association path is composed of at least one association path, and there is no association path between the local knowledge blocks of different complete knowledge data.

[0094] In the process of stripping, the local knowledge blocks are connected using the associated paths. When two local knowledge blocks are connected by an associated path, it indicates that there is a certain correlation. Therefore, any two local knowledge blocks in the complete knowledge data can be connected through multiple associated paths via other local knowledge blocks. Since the local knowledge blocks connected by the associated paths have a certain correlation, the local knowledge blocks in the complete knowledge data have a correlation. Since there is no associated path between the local knowledge blocks in different complete knowledge data, the local knowledge blocks in different complete knowledge data have no correlation. Therefore, each complete knowledge data is an independent whole, and targeted knowledge fusion can be performed in the complete knowledge data.

[0095] Referring to Figure 3 The classification of the complete knowledge data to form at least one data fusion set includes the following steps:

[0096] The number of associated paths connected by the local knowledge blocks in the complete knowledge data is counted as a characteristic value of the local knowledge blocks;

[0097] The local knowledge blocks in the complete knowledge data are sorted in descending order of the characteristic values to obtain a local knowledge block sequence;

[0098] The local knowledge blocks are tested for deletion in the order of the local knowledge block sequence. When the local knowledge blocks are deleted, the complete knowledge data is not connected, and the local knowledge blocks are taken as target local knowledge blocks. The definition of connection is that any two local knowledge blocks in the complete knowledge data are connected by a single associated path, otherwise, they are not connected.

[0099] The target local knowledge blocks are summarized as a main knowledge data, and the local knowledge blocks in the complete knowledge data other than the main knowledge data are taken as subsidiary knowledge data.

[0100] Since the complete knowledge data contains a large amount of data, a main architecture is needed for fusion. In this case, the main architecture is formed by the main knowledge data, and the subsidiary knowledge data is used as subsidiary fusion into the main architecture. The identification of the main knowledge data mainly depends on its importance in the complete knowledge data. The main knowledge data is composed of important local knowledge blocks. The identification of the important local knowledge blocks mainly depends on the associated paths connected to the local knowledge blocks. When the local knowledge blocks are removed, it will cause that two local knowledge blocks in the complete knowledge data cannot be connected by a single associated path. Because the single associated path originally used to connect the two local knowledge blocks passes through the removed local knowledge block. When the local knowledge block is removed, the single associated path is disconnected. Therefore, it indicates that the removed local knowledge block is more important, because it is not the marginal data in the correlation. Otherwise, its removal will not affect the connectivity of the complete knowledge data.

[0101] Referring to Figure 4 As shown in FIG. 1, the filtering of the auxiliary knowledge data in the data fusion set to obtain at least one target auxiliary knowledge data comprises the following steps:

[0102] After the main knowledge data is deleted from the complete knowledge data, at least one auxiliary knowledge data link is formed, which satisfies that any two auxiliary knowledge data in the auxiliary knowledge data link are connected through a single-link path, and there is no associated path between the auxiliary knowledge data in different auxiliary knowledge data links;

[0103] All combination possibilities of the two auxiliary knowledge data in the auxiliary knowledge data link are obtained to obtain at least one auxiliary knowledge data combination, and the two auxiliary knowledge data in the auxiliary knowledge data combination are respectively taken as a first auxiliary knowledge data and a second auxiliary knowledge data;

[0104] Based on big data, at least one first knowledge application event of the first auxiliary knowledge data is obtained, and first remaining knowledge used in the first knowledge application event is obtained, wherein the first remaining knowledge does not contain the first auxiliary knowledge data;

[0105] Based on big data, at least one second knowledge application event of the second auxiliary knowledge data is obtained, and second remaining knowledge used in the second knowledge application event is obtained, wherein the second remaining knowledge does not contain the second auxiliary knowledge data;

[0106] The second knowledge application event and the first knowledge application event in which the second remaining knowledge is consistent with the first remaining knowledge are established in a corresponding relationship, wherein the second knowledge application event and the first knowledge application event in which the second remaining knowledge is inconsistent with the first remaining knowledge do not have a covering relationship;

[0107] The second knowledge application event and the first knowledge application event are deduced and compared, when the second knowledge application event can be deduced from the first knowledge application event, the second knowledge application event is covered by the first knowledge application event, and when the first knowledge application event can be deduced from the second knowledge application event, the first knowledge application event is covered by the second knowledge application event;

[0108] When the first knowledge application event of the first auxiliary knowledge data is covered by the second knowledge application event of the second auxiliary knowledge data respectively, the first auxiliary knowledge data is deleted, and when the second knowledge application event of the second auxiliary knowledge data is covered by the first knowledge application event of the first auxiliary knowledge data, the second auxiliary knowledge data is deleted;

[0109] The auxiliary knowledge data in the auxiliary knowledge data link after the completion of the deletion and reduction are respectively taken as target auxiliary knowledge data.

[0110] There are many subsidiary knowledge data in the complete knowledge data, but there can be substitution between the subsidiary knowledge data, such as two subsidiary knowledge data are both for the same entity, but one is more advanced and one is more backward, the backward one is contained by the advanced one, for similar cases, only one of them needs to be retained, thereby, the subsequent fusion workload can be reduced.

[0111] With reference to Figure 5 The deducing comparison of the second knowledge application event and the first knowledge application event comprises the following steps:

[0112] Obtaining at least one sample application requirement, obtaining the completed part of the sample application requirement by the second knowledge application event as the second completed part, and obtaining the completed part of the sample application requirement by the first knowledge application event as the first completed part;

[0113] Merging at least one second completed part to obtain a second total part, and merging at least one first completed part to obtain a first total part;

[0114] Taking the intersection of the second total part and the first total part to obtain a feature part, taking the part different from the feature part in the second total part as a second target part, and taking the part different from the feature part in the first total part as a first target part;

[0115] Obtaining at least one common sense requirement based on big data;

[0116] When the second target part is an empty set or the common sense requirement, the second knowledge application event can be deduced from the first knowledge application event, and when the first target part is an empty set or the common sense requirement, the first knowledge application event can be deduced from the second knowledge application event.

[0117] The approximate knowledge is difficult to directly compare the relationship between each other when comparing, when two knowledge can be derived, then can be deleted, thus, the relationship of deduction can determine whether the two knowledge can be replaced by one, and because the deleted knowledge is the knowledge that can be preserved by deduction, thus, the fusion using the preserved knowledge is consistent with the fusion effect before deletion, and the principle of deduction is dependent on the comparison of the first knowledge application event of the first subsidiary knowledge data and the second knowledge application event of the second subsidiary knowledge data, when comparing, using sample application requirements to analyze the completion of the first knowledge application event and the second knowledge application event, when the second target part is empty set or common sense requirement, it is explained that the second target part can be completed by any person without assistance according to common sense, and the second target part is the difference part of the second total part different from the first total part, therefore, the second total part can be obtained by common sense and the first total part, thus, the deduction relationship of the second knowledge application event and the first knowledge application event can be determined, and the similar analysis is also applicable to the case that the first target part is empty set or common sense requirement.

[0118] Referring to Figure 6 As shown, based on big data, acquiring at least one common sense requirement comprises the following steps:

[0119] Based on big data, acquiring at least one knowledge application requirement;

[0120] Forming a sample test personnel set, which is composed of personnel engaged in various occupations in proportion;

[0121] When the knowledge application requirements are completed by the personnel in the sample test personnel set without consulting materials, the knowledge application requirements are taken as common sense requirements.

[0122] Referring to Figure 7 As shown, calculating the innovation coefficient of the target subsidiary knowledge data comprises the following steps:

[0123] In big data, acquiring the time when the target subsidiary knowledge data first appears as the target time;

[0124] Acquiring the first processing result of at least one actual application requirement using the target subsidiary knowledge data, and acquiring the second processing result of the actual application requirement before the target time;

[0125] Statistically, the first processing time of the first processing result is statistically the second processing time of the second processing result, and the second processing time is divided by the first processing time to obtain the first coefficient;

[0126] The first completion degree of the first processing result to the actual application requirement is counted, the second completion degree of the second processing result to the actual application requirement is counted, the first completion degree is divided by the second completion degree, and the second coefficient is obtained;

[0127] The first coefficient is multiplied by the second coefficient, and the innovation coefficient is obtained.

[0128] The innovation coefficient actually describes the optimization degree of the target subsidiary knowledge data compared with previous data. The time and completion degree of whether the target subsidiary knowledge data is processed are mainly described. When the target subsidiary knowledge data is more optimized than the previous data, the completion degree to the actual application requirement is higher, and the processing time is shorter. Therefore, the first coefficient is obtained by dividing the time consumed by the second processing by the time consumed by the first processing, the second coefficient is obtained by dividing the first completion degree by the second completion degree, when the target subsidiary knowledge data is more optimized than the previous data, the first coefficient and the second coefficient will be larger, and thus the innovation coefficient will be larger. Therefore, the innovation coefficient is positively correlated with the optimization of the target subsidiary knowledge data, and thus the innovation coefficient can describe the optimization of the target subsidiary knowledge data.

[0129] Referring to Figure 8 As shown in the figure, generating at least one main fusion node in the main knowledge data and at least one secondary fusion node in the target subsidiary knowledge data includes the following steps:

[0130] The local knowledge block in the main knowledge data is segmented to obtain at least one main fusion node, and the knowledge in the main fusion node corresponds to the same entity;

[0131] The target subsidiary knowledge data is segmented to obtain at least one secondary fusion node, and the knowledge in the secondary fusion node corresponds to the same entity;

[0132] The main fusion node and the secondary fusion node corresponding to the same entity are established in a corresponding relationship.

[0133] Each knowledge corresponds to an entity, that is, the thing it describes, such as a physical equation, which must correspond to an entity, which may be a gas, a fluid or a solid, or an application object, which is the sum of all application objects of knowledge application.

[0134] Referring to Figure 9 As shown in the figure, the target subsidiary knowledge data and the main knowledge data in the data fusion set are fused, including the following steps:

[0135] At least one main application event of the knowledge in the main fusion node is obtained, and at least one secondary application event of the knowledge in the secondary fusion node is obtained;

[0136] The main application event and the secondary application event are deduced and compared, when the main application event can be deduced from the secondary application event, the fusion result of the main fusion node and the secondary fusion node is the secondary fusion node, when the secondary application event can be deduced from the main application event, the fusion result of the main fusion node and the fusion node is the main fusion node;

[0137] When the main application event and the secondary application event cannot be deduced from each other, the knowledge in the main fusion node and the knowledge in the secondary fusion node are intersected to obtain a first fusion part;

[0138] The part of the knowledge in the main fusion node that is different from the first fusion part is taken as a main category part;

[0139] The part of the knowledge in the secondary fusion node that is different from the first fusion part is taken as a secondary category part;

[0140] The parts of the main category part and the secondary category part that have contradictions are taken as a main contradiction part and a secondary contradiction part respectively, the main contradiction part belongs to the main category part, and the secondary contradiction part belongs to the secondary category part;

[0141] The main category part and the secondary category part are fused, and when the fusion is performed, the secondary category part is deleted to obtain a second fusion part;

[0142] The first fusion part and the second fusion part are summarized to complete the fusion.

[0143] When the fusion is performed, the approximate knowledge and the contradiction need to be considered, the approximate knowledge can be determined by deduction, and for the knowledge that cannot be deduced from each other, the part without contradiction can be merged, and for the contradictory part, the part corresponding to the knowledge in the main fusion node is adopted because the knowledge in the main fusion node is more important and has higher data reliability.

[0144] Further, the present scheme further proposes a storage medium having a computer readable program stored thereon, the computer readable program is called to execute the above-mentioned knowledge data fusion method based on artificial intelligence technology mixing.

[0145] It can be understood that the storage medium can be a magnetic medium, for example, a floppy disk, a hard disk, a magnetic tape; an optical medium, for example, a DVD; or a semiconductor medium, for example, a solid state disk (SSD) and the like.

[0146] In summary, the application has the advantages that: by stripping the complete logical architecture in the data to be fused, classifying the complete knowledge data, screening the subsidiary knowledge data in the data fusion set, and fusing the target subsidiary knowledge data and the main knowledge data in the data fusion set, the complete knowledge architecture can be obtained by data stripping, the knowledge attribute can be judged according to the application of the knowledge, the data to be fused can be reduced, the work load of subsequent processing is reduced, the main knowledge data can be used as the main architecture of subsequent fusion by the main knowledge data and the subsidiary knowledge data, the fusion is more orderly, the contradictory situation can be processed during the fusion by the main fusion node and the auxiliary fusion node, the final fusion effect is more reasonable, and the actual demand is met.

[0147] The basic principle, main features and advantages of the application are shown and described above. Those skilled in the art should understand that the application is not limited by the above examples, and the above examples and descriptions in the specification are only the principles of the application. Without departing from the spirit and scope of the application, various changes and improvements can be made to the application, and these changes and improvements all fall within the scope of the claimed application. The scope of protection of the application is defined by the appended claims and their equivalents.

Claims

1. A knowledge data fusion method based on artificial intelligence technology, characterized in that, include: The knowledge data awaiting fusion is aggregated into data to be fused, and the complete logical architecture is extracted from the data to be fused to obtain at least one complete knowledge data. The complete knowledge data is classified to form at least one data fusion set, which consists of the main knowledge data and at least one auxiliary knowledge data. Filter the supplementary knowledge data in the data fusion set to obtain at least one target supplementary knowledge data; Calculate the innovation coefficient of the target-related knowledge data, and sort the target-related knowledge data in the data fusion set from largest to smallest according to the innovation coefficient to obtain the target-related knowledge data sequence; Generate at least one main fusion node in the main knowledge data and at least one secondary fusion node in the target auxiliary knowledge data; Based on the main fusion node and the secondary fusion node, the target auxiliary knowledge data and the main knowledge data in the data fusion set are fused, and the fusion order is carried out according to the target auxiliary knowledge data sequence corresponding to the data fusion set. The calculation of the innovation coefficient of the target-related knowledge data includes the following steps: In big data, the time when the target's associated knowledge data first appears is used as the target time. Obtain a first processing result of at least one practical application requirement using target-related knowledge data, and obtain a second processing result of practical application requirements prior to the target time. The time spent on the first processing of the first processing result is calculated, the time spent on the second processing of the second processing result is calculated, and the time spent on the second processing is divided by the time spent on the first processing to obtain the first coefficient; The first degree of completion of the first processing result to the actual application requirements is calculated, and the second degree of completion of the second processing result to the actual application requirements is calculated. The first degree of completion is divided by the second degree of completion to obtain the second coefficient. Multiplying the first coefficient by the second coefficient yields the innovation coefficient; The process of fusing target-related knowledge data and main knowledge data in the data fusion set includes the following steps: Obtain at least one primary application event of knowledge from the primary fusion node, and obtain at least one secondary application event of knowledge from the secondary fusion node; By comparing and contrasting the main application event and the secondary application event, if the main application event can be derived from the secondary application event, then the fusion result of the main fusion node and the secondary fusion node is the secondary fusion node; if the secondary application event can be derived from the main application event, then the fusion result of the main fusion node and the fusion node is the main fusion node. When the main application event and the secondary application event cannot be derived from each other, the intersection of the knowledge in the main fusion node and the knowledge in the secondary fusion node is taken to obtain the first fusion part; The part of the knowledge in the main fusion node that differs from the first fusion part is taken as the main category part; The parts of the knowledge in the sub-fusion node that differ from the first fusion part are treated as sub-category parts; The contradictory parts in the main category and the subcategory are respectively regarded as the main contradictory part and the sub contradictory part. The main contradictory part belongs to the main category, and the sub contradictory part belongs to the subcategory. The main category and the subcategory are merged. During the merging process, the subcategory is deleted to obtain the second merged part. The first and second fusion parts are combined to complete the fusion.

2. The knowledge data fusion method based on artificial intelligence technology according to claim 1, characterized in that, The process of extracting the complete logical architecture from the data to be merged to obtain at least one complete knowledge data includes the following steps: Based on big data, acquire at least one knowledge data containing two knowledge points, and use them as sample knowledge data respectively. Use half the minimum size of the sample knowledge data as the preset value; The data to be fused is divided into segments to obtain at least one local knowledge block, and the size of the local knowledge block is a preset value. Based on big data, obtain at least one application scenario of a local knowledge block, extract features from at least one application scenario, and obtain at least one application feature; The degree of association between two local knowledge blocks is the ratio of the number of overlapping application features to the total number of application features in the two local knowledge blocks. Calculate the correlation between two sample knowledge data, and take the maximum correlation between two sample knowledge data that do not overlap in knowledge points as the correlation threshold; When the correlation between two local knowledge blocks exceeds the correlation threshold, a correlation path is connected between the two local knowledge blocks. To form at least one complete knowledge data, satisfying the following conditions: the complete knowledge data consists of local knowledge blocks, any two local knowledge blocks in the complete knowledge data are connected by a single link path, the single link path consists of at least one associated path, and there is no associated path between the local knowledge blocks of different complete knowledge data.

3. The knowledge data fusion method based on artificial intelligence technology according to claim 2, characterized in that, The process of classifying complete knowledge data to form at least one data fusion set includes the following steps: The number of associated paths connecting local knowledge blocks in the complete knowledge data is counted and used as the feature value of the local knowledge block; The local knowledge blocks in the complete knowledge data are sorted from largest to smallest according to their feature values ​​to obtain a sequence of local knowledge blocks; According to the order of the local knowledge block sequence, the local knowledge blocks are deleted and tested. When a local knowledge block is deleted, the complete knowledge data is not connected. Then the local knowledge block is taken as the target local knowledge block. The connection is defined as: any two local knowledge blocks in the complete knowledge data are connected by a single path. Otherwise, they are not connected. The target local knowledge blocks are aggregated into main knowledge data, and the local knowledge blocks other than the main knowledge data in the complete knowledge data are respectively regarded as auxiliary knowledge data.

4. The knowledge data fusion method based on artificial intelligence technology according to claim 3, characterized in that, The process of filtering the supplementary knowledge data in the data fusion set to obtain at least one target supplementary knowledge data includes the following steps: After deleting the main knowledge data from the complete knowledge data, at least one subordinate knowledge data link is formed, satisfying that: any two subordinate knowledge data in the subordinate knowledge data link are connected by a single link path, and there is no associated path between the subordinate knowledge data in different subordinate knowledge data links; Obtain all possible combinations of two auxiliary knowledge data in the auxiliary knowledge data link, and obtain at least one auxiliary knowledge data combination. The two auxiliary knowledge data in the auxiliary knowledge data combination are respectively used as the first auxiliary knowledge data and the second auxiliary knowledge data. Based on big data, at least one first knowledge application event is obtained from the first auxiliary knowledge data, and the first other knowledge used in the first knowledge application event is obtained, wherein the first other knowledge does not contain the first auxiliary knowledge data. Based on big data, at least one second knowledge application event is obtained to acquire second auxiliary knowledge data, and the second remaining knowledge used in the second knowledge application event is obtained, wherein the second remaining knowledge does not contain second auxiliary knowledge data. Establish a correspondence between the second knowledge application event and the first knowledge application event that are consistent with the second other knowledge and the first other knowledge, wherein the second knowledge application event and the first knowledge application event that are inconsistent with the second other knowledge do not have an overriding relationship; The second knowledge application event and the first knowledge application event are compared and deduced. If the second knowledge application event can be deduced from the first knowledge application event, then the second knowledge application event is covered by the first knowledge application event. If the first knowledge application event can be deduced from the second knowledge application event, then the first knowledge application event is covered by the second knowledge application event. When the first knowledge application event of the first auxiliary knowledge data is covered by the second knowledge application event of the second auxiliary knowledge data, the first auxiliary knowledge data is deleted; when the second knowledge application event of the second auxiliary knowledge data is covered by the first knowledge application event of the first auxiliary knowledge data, the second auxiliary knowledge data is deleted. The ancillary knowledge data in the ancillary knowledge data links that have been deleted are respectively regarded as target ancillary knowledge data.

5. The knowledge data fusion method based on artificial intelligence technology according to claim 4, characterized in that, The process of comparing and deducing the second knowledge application event with the first knowledge application event includes the following steps: Obtain at least one sample application requirement, obtain the completion part of the second knowledge application in relation to the sample application requirement as the second completion part, and obtain the completion part of the first knowledge application in relation to the sample application requirement as the first completion part. At least one second completed part is merged to obtain a second total part, and at least one first completed part is merged to obtain a first total part; The intersection of the second general part and the first general part is used to obtain the feature part. The part that is different between the second general part and the feature part is taken as the second target part, and the part that is different between the first general part and the feature part is taken as the first target part. Based on big data, obtain at least one common-sense requirement; When the second objective part is an empty set or a common-sense requirement, the second knowledge application event can be deduced from the first knowledge application event. When the first objective part is an empty set or a common-sense requirement, the first knowledge application event can be deduced from the second knowledge application event.

6. The knowledge data fusion method based on artificial intelligence technology according to claim 5, characterized in that, The process of obtaining at least one common-sense requirement based on big data includes the following steps: Based on big data, identify at least one knowledge application requirement; A sample test personnel set is formed, which is composed of people engaged in various professions in equal proportions; When all knowledge application requirements are met by the personnel in the sample test group without consulting any materials, the knowledge application requirements are considered common sense requirements.

7. The knowledge data fusion method based on artificial intelligence technology according to claim 1, characterized in that, The step of generating at least one primary fusion node in the main knowledge data and at least one secondary fusion node in the target auxiliary knowledge data includes the following steps: The local knowledge blocks in the main knowledge data are segmented to obtain at least one main fusion node, and the knowledge in the main fusion node corresponds to the same entity; The target-related knowledge data is segmented to obtain at least one sub-fusion node, and the knowledge in the sub-fusion node corresponds to the same entity; Establish a correspondence between the primary fusion node and the secondary fusion node that correspond to the same entity.

Images