Big data information analysis method and system based on artificial intelligence

Abstract

The application relates to the field of artificial intelligence, in particular to a big data information analysis method and system based on artificial intelligence, which comprises the following steps: acquiring a jump section direction vector and classifying label nodes with consistent directions, splitting a path lacking a mapping relationship to form an interrupted section, identifying nodes with continuous directions as bridge section entrances, rearranging path sections to generate a sliding connection sequence, and combining semantic fields to sort behavior trajectories to generate a focused analysis set; in the application, the jump section direction is classified according to node order and channel number, the direction aggregation boundary is determined, the semantic field is combined with the path position to identify an interruption, the positioning accuracy of the node sequence structure and the rationality of the paragraph division are improved, the bridge section connection is based on direction consistency screening, the path continuation logic is maintained, the behavior trajectory and the aggregation number are sorted according to the sequence and the level, the behavior aggregation sequence of the semantic chain is constructed, and the coherence of path identification in the channel aggregation area and the stability of node attribution are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a big data information analysis method and system based on artificial intelligence. BACKGROUND

[0002] The technical field of artificial intelligence includes the ability to intelligently reason, understand data and make decisions on complex problems based on machine learning, deep learning, natural language processing, knowledge graph, neural network modeling and expert systems. The core content of this technology field is to use mathematical models with self-adaptive and generalization capabilities to automatically learn and infer structured or unstructured data, thereby forming an information processing process with perception, understanding and feedback capabilities. Artificial intelligence systems are widely used in voice recognition, image recognition, intelligent recommendation, intelligent search, intelligent decision support and other fields. The key steps include sample training, model establishment, parameter optimization, data input analysis and output response control.

[0003] Among them, the big data information analysis method based on artificial intelligence refers to the use of neural network learning mechanism and graph neural structure modeling mechanism, based on the structured log data and semi-structured text data to construct a unified feature vector set, through the convolution conversion function to compress the dimension and extract the local features of the multi-dimensional data, combined with the feature focusing strategy based on attention mechanism, to identify the key indicators and analyze the logical sequence pattern of the input data, and to complete the inductive classification of abnormal behavior and potential rules through the probability inference method. This scheme uses the detailed records and time series in the historical data set as the basis, uses the graph construction mechanism to generate the node attribute relationship matrix, and generates the result label set according to the construction principle of the clustering classification tree.

[0004] The existing artificial intelligence big data information analysis method relies on the feature vector construction method to model the nodes globally, and it is difficult to realize the differentiated identification of the connection characteristics at the path direction mutation point. The path order cannot preserve the direction continuity of the segment in the multi-direction intersection scene. In the processing of semantic fields, it mainly depends on the abstract representation of the model hidden layer, and lacks an explicit synchronization mechanism for the order of the semantic fields within the path segment, resulting in that the semantic chain break points in some behavior paths cannot be identified. The path segment migration does not rely on the connection conditions between the actual segments to extend the structure, and often uses node similarity or adjacency probability in the graph structure to predict the connection, which lacks logical judgment of the connection direction, causing the lack of direction continuity basis in the path relay construction. The trajectory number and cluster number are only classified according to the label feature matching, without considering their position order and hierarchical proportion in the path structure, which may cause classification distortion of the same semantics but different trajectory distribution. In the channel aggregation path, there is a lack of behavior chain regularization method controlled by the aggregation level, which makes it difficult to avoid the semantic redundancy and node segment intersection in the path behavior merging, affecting the integration accuracy of the structure output. SUMMARY

[0005] In order to solve the technical problems existing in the prior art, the embodiment of the present application provides a big data information analysis method based on artificial intelligence, comprising the following steps:

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: a big data information analysis method based on artificial intelligence, comprising the following steps:

[0007] S1: obtaining a jump point direction vector in a trajectory, comparing a channel number to which a direction mutation point belongs with a path number and an adjacent jump point order, merging label nodes according to direction consistency, and obtaining a direction jump aggregation label group graph;

[0008] S2: based on the identified node path in the direction jump aggregation label group graph, reading a semantic trajectory field and a path number according to the label node in the aggregation section, splitting a path section without an upstream and downstream mapping relationship, and obtaining a semantic interruption structure mapping fragment set;

[0009] S3: based on the broken node in the semantic interruption structure mapping fragment set, analyzing a front and rear jump section connection number and a direction order, marking an interruption section with consistent direction as a bridge section entrance, and obtaining a bridge section entrance path node structure graph;

[0010] S4: based on the bridge section connection relationship in the bridge section entrance path node structure graph, comparing a jump direction sequence, reordering path sections with consistent offset and continuous levels, and obtaining a path sliding structure connected sequence graph;

[0011] S5: based on the label node group in the path sliding structure connected sequence graph, analyzing a semantic expansion structure in a channel aggregation area, merging continuous trajectories and semantic distribution into a behavior fragment, and obtaining a semantic behavior focusing analysis set.

[0012] As a further scheme of the present application, the direction jump aggregation label group graph comprises a jump point direction vector sequence, a channel number order mapping set, a direction convergence label node group, and a path structure branch section set, the semantic interruption structure mapping fragment set comprises a path section number sequence, a semantic field broken node set, a non-continuous mapping path index, and a structure interruption node classification result, the bridge section entrance path node structure graph comprises a jump section connection number index, a path direction sequence set, a connectable path fragment set, and a bridge section entrance node identification set, the path sliding structure connected sequence graph comprises a jump direction offset ordering group, a structure level continuous node group, a path linkage sliding section number sequence, and a channel sequence mapping table, and the semantic behavior focusing analysis set comprises a behavior trajectory aggregation sequence, a cluster number classification table, a path level relationship set, and a semantic mode structure item.

[0013] As a further scheme of the present application, the specific steps of S1 are:

[0014] S101: Obtain the node set of the jump structure in the track, extract the connection order between the direction vector of the node and the adjacent node, analyze the label node direction structure difference in the direction mutation area by combining the arrangement trend of the direction vector and the node jump position sequence, and obtain the jump point direction offset relationship;

[0015] S102: Based on the jump point direction offset relationship, extract the channel number and path number of the label node where the mutation point is located, call the channel number arrangement order and the number position of the jump point in the path, filter the channel node sequence with consistent direction offset trend and integrate the sequence, and obtain the channel label order structure;

[0016] S103: Based on the channel label order structure, number the path number and channel structure sequence to which the label node belongs, collect the label combination with convergent direction in the same path and perform path structure splitting, branch and cluster the obtained label node sequence, and obtain the direction jump aggregation label group graph.

[0017] As a further scheme of the present application, the specific steps of S2 are:

[0018] S201: Based on the node path identified in the direction jump aggregation label group graph, extract the label node number in the aggregation segment, read the corresponding semantic track label field and path paragraph number, pair and associate the semantic fields according to the path number order, and obtain the semantic field path mapping relationship;

[0019] S202: Call the semantic field path mapping relationship, filter the path number sequence to which the semantic field does not continuously correspond, identify the label node with interrupted semantic corresponding relationship, extract the identification value of the interrupted node according to the path number order and classify and organize, and obtain the mapping structure interrupted node sequence;

[0020] S203: Based on the mapping structure interrupted node sequence, extract the path number and paragraph position of the contained node, mark the label path where the interrupt point is located according to the path number segmentation, match and merge the corresponding path paragraph number and the interrupted node number, and obtain the semantic interruption structure mapping fragment set.

[0021] As a further scheme of the present application, the specific steps of S3 are:

[0022] S301: Based on the path breaking node in the semantic interruption structure mapping fragment set, extract the front and rear jump segment connection number and path direction order corresponding to the node, arrange the jump segment connection structure according to the path number, and obtain the path jump segment direction order by sequentially sorting the jump segment number and comparing the path direction.

[0023] S302: calling the path hop segment direction sequence, extracting the arrangement order of adjacent hop segments in the path direction sequence, calculating the frequency of repeated occurrence of the path direction in the hop segment sequence, marking the node number combination with consistent repetition frequency, and obtaining the direction sequence coincidence position;

[0024] S303: based on the direction sequence coincidence position, screening the node pair with direction continuation characteristics, extracting the path number and entrance node position, classifying and integrating the hop segment path number in the path order, and obtaining the bridge segment entrance path node structure diagram.

[0025] As a further scheme of the application, the calculation formula of the frequency of repeated occurrence of the path direction in the hop segment sequence is specifically:

[0026]

[0027] Among them, F rep represents the frequency of repeated occurrence of the path direction in the hop segment sequence, D z represents the cumulative occurrence number of the path direction z in the hop segment sequence, represents the average of the path direction occurrence number, G z represents the difference between the average of the path direction occurrence number and the average of the path direction occurrence number, V z represents the number of path segments corresponding to the direction z in the hop segment sequence, R z represents the number of hop segments covered by the path segment distribution range of the direction z, L z represents the number of hop segment label nodes of the path segment where the direction z is located, and u represents the total number of path direction categories in the hop segment sequence.

[0028] As a further scheme of the application, the specific steps of S4 are:

[0029] S401: based on the bridge segment connection node in the bridge segment entrance path node structure diagram, extracting the jump sequence between adjacent nodes, collecting the start and end node numbers and jump direction of the path segment, and combining according to the path number order to obtain the path direction sequence combination;

[0030] S402: calling the path direction sequence combination, extracting the direction identifier and structure level value of the path segment, calculating the occurrence frequency of the direction identifier in the path sequence, screening the path number with consistent frequency and same structure level, and obtaining the direction structure frequency combination path;

[0031] S403: based on the direction structure frequency combination path, extracting the jump node number sequence corresponding to the classified path segment, aligning the node number and path number position according to the direction frequency classification order, and obtaining the path sliding structure connection sequence diagram.

[0032] As a further scheme of the present application, the calculation formula of the frequency of occurrence of the direction identifier in the path sequence is specifically:

[0033]

[0034] wherein, represents the frequency of occurrence of the i-th type of direction identifier in the path sequence, n i represents the total number of occurrences of the i-th type of direction identifier in the current path direction sequence, w ij represents the direction weight value of the i-th type of direction identifier in the j-th occurrence, l ij represents the structural level value of the i-th type of direction identifier in the j-th occurrence path segment, m ij represents the path occupation density of the structural unit connected by the i-th type of direction identifier in the j-th occurrence path segment, s ij represents the number of adjacent direction identifiers connected to the i-th type of direction identifier at the position of the j-th occurrence path segment.

[0035] As a further scheme of the present application, the specific steps of S5 are:

[0036] S501: Based on the label node group in the path sliding structure connected sequence graph, the corresponding channel aggregation section number is extracted, the semantic field information of the label node in the aggregation section is collected, the node sequence with path extension relationship between fields is identified, the path is merged according to the structural continuity, and the semantic expansion structure node is obtained;

[0037] S502: The semantic expansion structure node is called, the behavior trajectory number, the cluster number and the corresponding aggregation level value of the associated node in the path segment are extracted, the corresponding structural features between the trajectory number and the cluster number are analyzed, the label nodes with verified number relationship are classified and integrated, and the behavior trajectory cluster association sequence value is obtained;

[0038] S503: Based on the behavior trajectory cluster association sequence value, the label node number group with consistent structure level in the path segment is extracted, the behavior trajectory and the node number are combined and associated, the node behavior with associated features in the channel aggregation area is regularized, and the semantic behavior focusing analysis set is obtained.

[0039] A big data information analysis system based on artificial intelligence, comprising:

[0040] The direction aggregation point identification module obtains the direction vector of the jump structure node, extracts the channel number and the path number of the mutation point label node, combines the direction vector sequence and the channel arrangement position, combines the nodes with continuous directions, aggregates the path numbers according to the jump point order, and obtains the direction jump aggregation label group atlas;

[0041] The semantic mapping broken link module reads the semantic label field and path segment number of the aggregated segment, and merges the path numbers of the nodes with discontinuous upstream and downstream semantic fields, to obtain a set of semantic interruption structure mapping segments based on the direction jump aggregated label group graph;

[0042] The channel connection recombination module extracts the jump segment connection number and channel direction, compares the connection node position and direction sequence, and groups the jump segments with structural connection, to obtain a bridge segment entry path node structure graph based on the broken nodes in the set of semantic interruption structure mapping segments.

[0043] The sliding structure linkage module extracts the jump sequence to form a path direction sequence based on the connection nodes in the bridge segment entry path node structure graph, calls the path offset trend and structure level value, combines the path segments with consistent structures, constructs a continuous jump channel, and obtains a path sliding structure interconnection sequence graph.

[0044] The behavior characteristic focusing module identifies the semantic expansion structure of the channel aggregation section based on the label nodes in the path sliding structure interconnection sequence graph, calls the behavior trajectory field, group relationship and aggregation level, and induces the node groups with coinciding groups and consistent structures, to obtain a semantic behavior focusing analysis set.

[0045] Compared with the prior art, the advantages and positive effects of the present application are:

[0046] In the present application, the direction aggregation boundary is determined by classifying the jump segment direction according to the node sequence and channel number, the interruption is identified by combining the semantic field with the path position, the positioning accuracy of the node sequence structure and the rationality of the paragraph division are improved, the bridge segment connection is screened based on the direction consistency, the path continuation logic is maintained, the behavior trajectory and aggregation number are regularized in sequence and level, the behavior aggregation sequence of the semantic chain is constructed, and the coherence of the path recognition in the channel aggregation area and the stability of the node attribution are enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0048] Figure 1 The step flowchart of the present application is shown in the figure.

[0049] Figure 2 The system module diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0050] The technical solutions in the present application will be described below with reference to the drawings.

[0051] In the embodiments of the present application, the words such as "example", "for example" are used to represent an example, illustration, or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0052] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times, and it should be pointed out that the meanings expressed are consistent when the distinction is not emphasized. "Of", "corresponding" and "corresponding" can be used interchangeably at times, and it should be pointed out that the meanings expressed are consistent when the distinction is not emphasized.

[0053] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1, and the meanings expressed are consistent when the distinction is not emphasized.

[0054] In order to make the technical problems, technical schemes and advantages to be solved by the present application more clear, the following will be described in detail in combination with the drawings and specific embodiments.

[0055] Please refer to Figure 1 The embodiments of the present application provide a big data information analysis method based on artificial intelligence, comprising the following steps:

[0056] S1: Extracting the jump point direction vector in the jump structure in the trajectory, extracting the channel number and path number to which the path direction mutation point belongs, sequentially associating according to the arrangement position of the jump points between channels, grouping and integrating the label nodes with adjacent channel direction convergence relationship, dividing the label sequence in the aggregation result into multiple branch paragraphs according to the path structure, and obtaining a direction jump aggregation label group graph;

[0057] S2: Based on the identified node path in the direction jump aggregation label group graph, reading the semantic trajectory annotation field and continuous path paragraph number of all label nodes in the aggregation paragraph, segmenting the label path whose semantic field does not appear continuous mapping between upstream and downstream nodes, and integrating the node path at the mapping interruption into the chain break structure sequence, to obtain a semantic interruption structure mapping segment set;

[0058] S3: Based on the semantic interruption structure mapping fragment set, the path breaking node in the path segment set is extracted, the connection node number and path direction sequence of the jumping segment before and after the breaking segment are extracted, the position of the connection structure between the path segments is compared, the discontinuous fragments with the connection characteristics are screened combined with the path direction, the path segments meeting the connection conditions are classified into the bridge segment entrance set, and the bridge segment entrance path node structure diagram is obtained;

[0059] S4: Based on the bridge segment entrance path node structure diagram, all bridge segment connection nodes are extracted, the jumping sequence between adjacent nodes is formed to form a path direction sequence, the label paths with the same direction offset trend in the sequence are sorted, the path segments with the same offset direction and consistent structure level are recombined into a linkage sliding structure path, and a path sliding structure connection sequence diagram is obtained.

[0060] S5: Based on each label node group in the path sliding structure connection sequence diagram, the semantic expansion structure in the channel aggregation section is identified, the behavior trajectory, group relationship and path aggregation level of the label nodes in the aggregation area are integrated and induced, the label behavior mode with analysis value is aggregated into a readable structure item, and a semantic behavior focus analysis set is obtained.

[0061] The direction jump aggregation label group atlas includes a jump point direction vector sequence, a channel number order mapping set, a direction convergent label node group, and a path structure branch paragraph set. The semantic interruption structure mapping fragment set includes a path paragraph number sequence, a semantic field break node set, a non-continuous mapping path index, and a structure interruption node classification result. The bridge segment entrance path node structure diagram includes a jump segment connection number index, a path direction sequence set, an connectable path segment set, and a bridge segment entrance node identification set. The path sliding structure connection sequence diagram includes a jump direction offset sorting group, a structure level continuous node group, a path linkage sliding segment number sequence, and a channel sequence mapping table. The semantic behavior focus analysis set includes a behavior trajectory aggregation sequence, a group number classification table, a path level relationship set, and a semantic mode structure item.

[0062] The specific steps of S1 are:

[0063] S101: Obtain the node set of the jump structure in the trajectory, extract the direction vector of the node and the connection sequence between adjacent nodes, combine the arrangement trend of the direction vector and the node jump position sequence, analyze the direction structure difference of the label nodes in the direction mutation area, and obtain the jump point direction offset relationship.

[0064] The spatial position coordinates and time label values of each node in the jump trajectory are obtained, the nodes are arranged in time sequence, the position change vector between each two adjacent nodes is recorded, the coordinate difference vector between the current node and the subsequent node is recorded as the direction vector, the node set is labeled according to the angle trend of the direction change, the time difference sequence of the jump points is obtained, the node whose jump frequency in a unit time period is more than 3 times is marked as a frequent jump node, the jump section position whose direction change amplitude exceeds 45 degrees is identified according to the turning angle of the direction vector of the frequent jump node, the channel number and path number of the label where the jump section is located are obtained, the channel index set where the jump point is located in the trajectory is formed, the time sequence of the label appearing in the multiple channel paths is recorded according to the arrangement order of the front and rear jump points in the label path number, and the angle size formed by the turning position of the direction vector is arranged from large to small, and the angle change trend of the direction vector before and after the jump section is analyzed according to the time sequence, the position node where the direction change sequence appears trend reversal in different time periods is calculated, the jump point group whose direction continuously rises or falls is identified, whether the angle difference of the direction change between adjacent jump points reaches the change reference interval is judged, the reference interval is set to 【30 degrees, 90 degrees】, if the direction angle difference before and after the jump point is located in the range, it is classified into the direction mutation area, and then the jump sequence number of the nodes in each channel is obtained, all the direction mutation sections in the whole path are found, and the angle offset value between the position of the jump point in each channel and the path trunk is compared, if the angle offset value is greater than the set direction offset judgment reference value 45 degrees, the node is marked as a direction offset node, the front and rear jump section label identification of the label where each group of direction offset nodes belongs in the path graph is further read, whether a continuous path mapping relationship is formed in the jump sequence is compared, if the jump section path jumps over the intermediate sequence section and is directly connected, it is recorded as a jump section path jump, and is mapped to each label index according to the jump section number, finally, the direction change nodes, path numbers, jump sequences and angle offset value sets that meet the above conditions are marked as a mutation jump point group, and the jump point direction offset relationship is obtained.

[0065] S102: Based on the jump point direction offset relationship, the channel number and path number of the label node where the mutation point is located are extracted, the channel number arrangement order and the number position of the jump point in the path are called, the channel node sequence with consistent direction offset trend is screened and integrated in sequence, and the channel label order structure is obtained.

[0066] Read the label node information corresponding to each mutation point from the offset node set, take the channel number and path number of the label node as the basic index field, construct a sequential list combining the relative position number of the node in the path sequence where the label hop point is located, take the channel number arrangement order as reference to the path timestamp, sort the labels to which the hop point belongs in time order, if there are multiple label nodes corresponding to the same channel number in different paths, rearrange them in ascending order of path number, analyze the direction offset vector between the hops, read the angle change value of the direction vector in each group of consecutive hops, if the signs of adjacent angle change values remain the same, define that the group of hop points has the same direction offset trend, filter by performing a difference operation on the hop direction difference sequence, record the number of consecutive segments with positive or negative difference value changes, if the number of consecutive segments exceeds two and the angle change amplitude does not undergo sign reversal, mark it as a consistent direction trend segment, set the angle change difference value as the judgment basis, use the angle difference threshold interval 【20 degrees, 60 degrees】 as the screening standard, if the hop direction difference value always stays within this interval, consider that the direction change is smooth and consistent, on this basis, merge all hops that meet the consistent direction trend relationship in channel number order, while marking the start and end label numbers and the original path number of each segment, then integrate the label node position order between the hops in the path, so that the position number of each group of labels in the channel path remains monotonically increasing, if the hop number is in reverse order or the hop boundary is crossed, exclude the segment combination from participating in integration, as an example: if labels T1 to T5 belong to channel C7, the path numbers are P21, P22, P23, P24, P25 in turn, the direction offset angle is 35°, 40°, 38°, 36°, the direction trend does not change, and the number is arranged in time order, then combine T1-T5 into a consistent trend channel segment, repeat the above processing in multiple such combinations, finally connect all label node sets that pass the direction trend screening in number order to form a stable and continuous label sequence chain, and generate the channel label order structure.

[0067] S103: Based on the channel label order structure, segment the path number and channel structure sequence to which the label node belongs, combine and split the path structure of the labels with the same direction trend in the same path, divide and cluster the obtained label node sequence branches, and obtain the direction jump aggregation label group atlas;

[0068] The path number and the channel number are combined to form a double-layer path index group, the label nodes in the index group are preliminarily segmented in ascending order of path number, in the label sequence in each segment, the jump direction vector is called and the difference value between adjacent nodes is analyzed, the node sequence with a difference value in the interval of 20 degrees to 50 degrees and in the same direction is regarded as a direction convergent segment, and then the consistency of the label channel number and the path number is checked, the label combination with the same path number and direction convergence is marked, a sequence record table is generated for each combination, the table records the channel, path and position index of the label in the combination, and then the sequence record table is combined to split the jump segment with adjacent path segments but a mutation of more than 50 degrees in the direction difference, the mutation nodes are taken as path structure dividing points, an independent path segment number is generated for each dividing point, the path segments before and after the dividing point are divided into independent data segments, after the structure splitting is completed, the direction density value of each path segment after the path splitting is calculated, the direction density value is obtained by the ratio of the sum of all jump angles in the path segment to the number of jump segments, and the direction density value is taken as a reference for branch division of the label node, the segment with a direction density value between 30 degrees and 40 degrees is marked as a same-direction clustering candidate segment, the index numbers of the label nodes in multiple candidate segments are grouped, adjacent path numbers and direction consistent candidate segments are combined to form a label direction aggregation candidate set, the candidate set is renumbered in the order of the labels in the original trajectory, and the label groups after renumbering are divided according to the intersection of the channel numbers, if two label groups have path number coincidence, direction consistency and intersection label quantity not less than 3, they are determined as the same clustering set, finally all label sets meeting the above conditions are output as the direction cluster boundary of the jump path, and the direction jump aggregation label group atlas is obtained.

[0069] The specific steps of S2 are:

[0070] S201: Based on the node path identified in the direction jump aggregation label group atlas, the label node number in the aggregation segment is extracted, the corresponding semantic trajectory label field and path paragraph number are read, the semantic field is paired and associated according to the path number sequence, and the semantic field path mapping relationship is obtained;

[0071] First, read the start and end node number of each aggregation segment, write the number sequence to the index list, perform path positioning operation on each number item and call its label number and corresponding channel number synchronously, then read the semantic trajectory annotation field from the semantic information table corresponding to the label number, which represents the behavior meaning assumed by the label node in different stages. The semantic field is recorded in the form of scalar encoding, such as "behavior start", "state keep", "event alternate" and other state segment labels. The record form is a set of behavior state code and corresponding time period, for example: label T024 has annotation sequence

(B1, 0-15), (B2, 15-35)

[0072] S202: Call semantic field path mapping relationship, filter path number sequence with non-continuous semantic field, identify label nodes with broken semantic correspondence, extract the identification value of the broken node in order of path number and classify, get the sequence of broken nodes in the mapping structure;

[0073] The semantic field records corresponding to each path are arranged in ascending order of path number, the semantic field index and its position in the path number sequence are read path by path, and string comparison operation is performed on the semantic field contents between the two paths. The comparison method adopts consistency check of field value, that is, when the semantic fields corresponding to path Pi and path Pi+1 are inconsistent, and the field does not appear on any other node in the path sequence, it is determined that there is a semantic interruption between the paths, and the scanning is continued in the increasing direction of path number. For each group of path number combinations with discontinuous semantic fields, a number difference sequence is formed, and the path segment with a number difference greater than 1 is regarded as a structural fracture segment. The path number difference is recorded as a fracture span index. The fracture span allowed range is set to 1 to 3. If the number jump of a path segment exceeds 3 from the previous segment, it is marked as an abnormal interruption fragment, and the label node number corresponding to the interruption position of the fragment is obtained from the path sequence. The label node number format is Txxx, where xxx is the sequential encoding, for example, T016 represents the 16th label node. The channel number to which the label node belongs and its corresponding sequential position in the channel are recorded, and are stored in a triple format, that is, (T016, C03, Pos=12), where C03 is the channel number, and Pos is the sequential position in the channel. All label nodes that meet the interruption judgment condition are written into an interruption identification list. It is also necessary to judge whether the label node appears in the semantic field multiplexing set during the screening process. If a node appears under two path numbers and its semantic field value remains unchanged, it is not included in the interruption set and is excluded from the interruption identification list. After the screening is completed, all path interruption label numbers are classified and arranged, and the mapping relationship between the position of each interruption label node and the corresponding path number is output in ascending order of path number. The output format is P number→T number sequence, for example, P12→T034, T037, which means that T034 and T037 are two semantic interruption label nodes in the 12th path. Finally, a complete path interruption position index is formed, and a mapping structure interruption node sequence is obtained.

[0074] S203: Based on the mapping structure interruption node sequence, the path number and paragraph position of the contained nodes are extracted, the label path where the interruption point is marked is divided according to the path number, the corresponding path paragraph number and interruption node number are matched and merged, and a semantic interruption structure mapping fragment set is obtained.

[0075] The path number and node identification value corresponding to the reading interruption node are read, the path number sequence is arranged in ascending order, and then the channel number and path paragraph number of each interruption node are taken as references to determine the positioning interval of the node in the whole path structure. It is identified whether the path numbers and channel numbers of the adjacent label nodes before and after the node are consistent in the interval. If the path numbers of more than three consecutive nodes do not increase or the channel numbers jump, the number of the interruption node in the paragraph is regarded as a path segmentation point. A segmentation index table is generated for all segmentation points. The starting label node number, ending label node number and corresponding path number range of each paragraph are recorded in the segmentation index table. Then, all label nodes in the path of the paragraph where the interruption node is located are aggregated. Through the matching operation of the label node number and the interruption node number in the path paragraph, the path paragraph and the interruption node are merged and judged. The judgment standard is: if the interruption node number is located in the front 30% or the back 30% of the paragraph path, the paragraph is marked as an interruption tendency paragraph. The proportion division standard is set according to the average length of the whole path. For example: if the path number P21 contains T011 to T021, a total of 11 nodes, and the interruption node is T013, the position of T013 in the sequence is the 3rd, accounting for 27.3%, which meets the interruption tendency marking condition. Therefore, T011 to T021 are all included in the interruption segment aggregation set. On this basis, the node marked as an interruption tendency paragraph in all paragraphs is matched again. If the same interruption node number appears among multiple paragraphs, these paragraphs are classified into the same semantic segmentation set. Each semantic segmentation set is sorted by path number for the second time. The same numbered paragraphs are merged and the paragraph number label index is updated. Finally, all structure fragment sets aggregated by path and interruption node number are output, and the semantic interruption structure mapping fragment set is obtained.

[0076] The specific steps of S3 are:

[0077] S301: Based on the path breaking node in the semantic interruption structure mapping fragment set, the front and back jump segment connection numbers and path direction sequence corresponding to the node are extracted. The jump segment connection structure is arranged according to the path number. The jump segment number is sequentially sorted and compared with the path direction, and the path jump segment direction sequence is obtained.

[0078] First, read the label number of each broken node, and call its jump connection relationship in the original path sequence. The jump connection relationship refers to the jump pair between adjacent label nodes, where one node is the starting point of the jump, and the subsequent node is the end point. After reading the jump connection pair, extract the path number information of each jump, combine the order of the nodes before and after the jump in the path, and judge whether the jump connection is forward jump, that is, whether the node number is monotonically increasing. Mark the jump pairs that do not meet the monotonically increasing condition as reverse jumps, and mark the jump path direction as "reverse order". Then, according to the path number, all jump connection relationships are summarized, the path number is taken as the primary key for aggregation, and a jump set indexed by path number is formed. In the jump set of each path, the jump numbers are sorted in ascending order. Compare the order of each jump in the path with its connection direction. If the jump connection order is consistent with the path direction, record the direction consistent identifier as "+1"; if not, record the direction consistent identifier as "-1". Take path P012 as an example. If its jumps are T005→T006, T006→T007, and T007→T006 in turn, the starting node number is higher than the ending node number in the third jump, the direction is reversed, and the direction marker is "-1". The direction identifier sequence of all jumps in each path is counted, and a direction identifier arrangement vector is formed for each path. Then, the direction vectors of all paths are sorted in ascending order according to the path number to form a direction sorting table. The direction sorting table records the overall consistency of the jump connection order and direction in each path. According to the number of direction identifier changes, the direction stability interval is divided. Paths with a direction change frequency of less than or equal to 1 are marked as direction continuous segments, and paths with a direction change frequency of more than 2 are marked as direction variable segments. The direction stability judgment interval is set as: 【0, 1】 stable, 【2, 3】 interrupted, 【4, ∞】 unstable. If the jump order in path P045 is T010→T011, T011→T010, and T010→T011 in turn, the direction changes 2 times, and is marked as a direction interrupted segment. Finally, the jump direction identifier results of each path are output and classified to obtain the path jump direction sorting.

[0079] S302: Call the path jump direction sorting, extract the arrangement order of adjacent jumps in the path direction sequence, calculate the repeated appearance frequency of the path direction in the jump sequence, mark the jump node number combination with consistent repeated frequency, and obtain the direction sequence coincidence position;

[0080] The calculation formula of the repeated appearance frequency of the path direction in the jump sequence is as follows:

[0081]

[0082] Where F rep represents the repeated appearance frequency of the path direction in the jump sequence, D z represents the cumulative appearance number of path direction z in the jump sequence, Mean of the number of occurrences of the path direction, G z Mean square sum of the difference between the position numbers of adjacent segments in the path in the direction z, V z Number of path segments corresponding to the direction z in the segment sequence, R z Number of segments covered by the distribution range of the path segment in the direction z, L z Number of segment label nodes of the path segment in the direction z, u represents the total number of path directions in the segment sequence.

[0083] Assume:

[0084] The total number of sample paths is 500, and after sequence extraction and segment extraction, a total of 5 valid path directions are identified, i.e. u = 5;

[0085] The following statistical values are obtained in actual monitoring:

[0086] D1 = 12, D2 = 15, D3 = 10, D4 = 8, D5 = 14;

[0087] The mean of the cumulative number of occurrences of the five path directions is represented, and the calculation process is as follows:

[0088]

[0089] The direction 1 number group is T101→T104→T107, the difference is 3 and 3, and the mean square sum is G1 = 3 2 + 3 2 = 18;

[0090] Similarly, G2 = 20, G3 = 10, G4 = 8, G5 = 16;

[0091] The number is obtained by extracting the number of path segments where the segment is located:

[0092] V1 = 4, V2 = 3, V3 = 2, V4 = 2, V5 = 3

[0093] The total number of segments in the path segment that appear in the direction is obtained, respectively:

[0094] R1 = 10, R2 = 9, R3 = 7, R4 = 6, R5 = 8

[0095] Through node statistics, we get:

[0096] L1 = 30, L2 = 28, L3 = 20, L4 = 18, L5 = 25

[0097] The intermediate value is calculated as follows:

[0098] The first term:

[0099] Secondly:

[0100] Thirdly:

[0101] Fourthly:

[0102] Fifthly:

[0103] F rep ≈5.66+9.45+7.97+5.64+9.15=37.87;

[0104] The results show that the repeated distribution of the path direction in the jump sequence has a bias trend, and the overall frequency strength of the direction is high, which provides a quantitative basis for marking the overlapping position of the direction sequence.

[0105] S303: Based on the overlapping position of the direction sequence, the node pair with the direction continuation characteristic is screened, the path number and the entrance node position are extracted, the jump path number is classified and integrated according to the path order, and the bridge section entrance path node structure diagram is obtained;

[0106] First, all hop connection pairs are extracted from the path hop segment direction sorting result, each connection pair is represented as a combination of start and end node numbers, and its connection direction identifier is recorded. The direction identifier uses the symbol "+1" to represent the positive direction, and "-1" to represent the reverse direction. After obtaining all the hop connection information, the connection pairs are grouped according to the path number. The hops with the same path number are grouped into the same group. The values of the connection direction identifiers of adjacent hops in each group are compared. If the connection direction identifiers of three consecutive groups of hops are consistent and the hop numbers are strictly increasing, then the segment is marked as a direction continuation segment. The marking condition is that the end node number of each hop in the hop sequence is equal to the start node number of the next hop minus one. In the example, if the hop sequence of path P05 is T03→T04, T04→T05, T05→T06, and the direction identifier is "+1", then the segment satisfies the direction continuation characteristic. T03 is taken as the entry node of the direction segment, and its path number P05 and entry node position number T03 are recorded. Then the next hop T06→T05 is checked, and its direction identifier is "-1", which is considered as the termination of the continuation segment. Continue to traverse the next hop that satisfies the continuation condition, and select all node pairs that satisfy the continuation criterion. Record the start node number, path number and direction state of each pair. Arrange the position index of all entry nodes in the path. All start nodes are considered as potential entry points of the bridge segment. Combine the path number and direction state corresponding to each entry node, and combine it with the adjacent hop number to form the bridge segment hop section structure. The hop section structure is indexed by path number. All hop path numbers are classified by direction. The hop chains with consistent direction are grouped into the same hop section. The entry node, termination node and direction sequence combination of each hop section are recorded in the number mapping table. The number mapping table takes the path number as the key, and the start hop number and direction sequence as the value. Finally, the path number, node start point and connection direction relationship corresponding to all bridge entry pairs are marked and summarized to obtain the bridge entry path node structure diagram.

[0107] The specific steps of S4 are:

[0108] S401: Based on the bridge connection nodes in the bridge entry path node structure diagram, extract the jump sequence between adjacent nodes, collect the start and end node numbers and jump direction of the path segment, and combine them according to the path number sequence to obtain the path direction sequence combination;

[0109] First, all the nodes identified as the entrance are screened from the figure, and the position number of each entrance node in its path segment is recorded. Then, the jump relationship between each entrance node and its adjacent node is obtained. By traversing the path number list, a jump pair is established between each pair of adjacent nodes. The structure of the jump pair is "start number-end number". The jump direction is judged. The judgment standard is whether the start number is less than the end number. If it is true, the jump direction is marked as "forward". Otherwise, it is "reverse". The identification value of the jump direction is recorded and paired with the path number to generate a paired group with the structure of "path number-jump direction". Then, the arrangement order of the jump direction in each path is read. The jump segments with the same direction are combined together as a direction segment. The start and end nodes of the direction segment are used as the start and end nodes of the segment. The start number, end number, path number, and direction state of each segment are recorded in the structured path table. If there are repeated jump segments or reverse jump segments, the path segment is not counted in the valid direction segment set. A fault tolerance mechanism is set during the selection of valid segments, which allows 1 direction change in the path to be excluded from the exclusion condition. If the number of direction changes is ≥2, the path segment is excluded and does not participate in subsequent combination. For example, if the jump segment sequence in path P014 is T06→T07 (positive), T07→T08 (positive), and T08→T07 (negative), the number of direction changes is 1, which is still counted in the valid segment. The start and end nodes of the segment are T06 and T08. After arranging all valid direction segments in ascending order of path number, the adjacent path segments are compared in order. If the end and start numbers of two path segments are adjacent and the direction states are consistent, they are merged into a combined segment. Then, the identification values of each combined segment are used to generate a direction sequence mapping index table. The table records the path number, start and end node numbers, direction identification, and jump order number in the segment. Finally, the path direction sequence combination is obtained.

[0110] S402: Call the path direction sequence combination, extract the direction identification and structure level value of the path segment, calculate the frequency of the direction identification in the path sequence, select the path numbers with the same frequency and the same structure level, and obtain the direction structure frequency combination path.

[0111] The calculation formula of the frequency of the direction identification in the path sequence is as follows:

[0112]

[0113] wherein, represents the frequency of the i-th type of direction identification in the path sequence, n i represents the total number of occurrences of the i-th type of direction identification in the current path direction sequence, w ij represents the direction weight value of the i-th type of direction identification in the j-th occurrence, l ijrepresents the structural level value of the path segment in the jth occurrence of the i-type direction identifier, m ij represents the path occupation density of the structural unit connected by the path segment in the jth occurrence of the i-type direction identifier, s ij represents the number of adjacent direction identifiers connected to the position of the path segment in the jth occurrence of the i-type direction identifier;

[0114] Assume:

[0115] First occurrence (j = 1): w i1 = 1.2, l i1 = 3, m i1 = 2.5, s i1 = 2;

[0116] Second occurrence (j = 2): w i2 = 1.0, l i2 = 2, m i2 = 2.0, s i2 = 1;

[0117] Third occurrence (j = 3): w i3 = 0.8, l i3 = 4, m i3 = 3.0, s i3 = 3;

[0118] First time:

[0119] Second time:

[0120] Third time:

[0121] ∑ = 0.56 + 0 + 1.93 = 2.49;

[0122]

[0123] Total: 2.49 + 1.732 = 4.222;

[0124]

[0125] The result shows that the frequency of the i-type direction identifier in the path sequence is 1.407, which reflects the distribution characteristics of the direction identifier in the path sequence and can be used for subsequent path optimization and analysis.

[0126] S403: Based on the direction structure frequency combination path, extract the jump node number sequence corresponding to the classified path segment, align the node number and path number position according to the direction frequency classification order, and obtain the path sliding structure connection sequence diagram;

[0127] First, all path segment sets classified into the same direction section are obtained from the path direction sequence combination, the path number corresponding to each path segment in the set and the paragraph start and end node number are recorded, then the jump node number in each path segment is extracted according to the jump segment connection order, the complete node number sequence in the jump segment is constructed, all node numbers are classified into the path number according to the path segment to form a number group, and the jump direction frequency of each group of node numbers is counted, the occurrence frequency of the direction identifiers "+1" and "-1" is counted respectively, if the proportion of a direction identifier in a path segment is more than 80%, the direction frequency classification value of the path segment is marked as the direction state corresponding to the direction identifier, the rule is fixed value interval judgment, which does not allow fuzzy expression, and the direction consistency of the node number is screened through the direction frequency classification value, only the node numbers with the same direction state are reserved for the next step processing, then the node number sequence screened is sorted in ascending order of path number, and the sorting result is aligned with the order in the original path segment number sequence, the alignment standard is that the order of the node number should be consistent with the order of the path number, if the node number jump and the path number repetition phenomenon occur, the group of nodes is marked as "sequence conflict group", and the group of numbers is excluded from the path direction sequence and no longer participates in the construction of the connected graph, example explanation: if the jump segment node numbers in path P011 are T102, T103 and T104, the direction frequency is "+1", then it is classified into the direction consistent group, if the direction frequency in P012 is "-1" and the node numbers are T101, T100 and T099, then the sorting should be T099, T100 and T101, and there is no jump after aligning the original path number in ascending order, which is considered as an effective group, all effective node groups are generated according to the path number sequence to generate a connected index table, in which the relative position of the node number and the path attribution number are marked, and finally the continuous node sequence graph is drawn based on the index table and the direction connection state is marked to obtain the path slip structure connected sequence graph.

[0128] The specific steps of S5 are:

[0129] S501: Based on the labeled node group in the path slip structure connected sequence graph, the corresponding channel aggregation section number is extracted, the semantic field information of the labeled node in the aggregation section is collected, the node sequence with path extension relationship between fields is identified, the path is merged according to the structural continuity, and the semantic expansion structure node is obtained.

[0130] First, identify the path segment number to which each label node belongs and its position index in the path, arrange all path segment numbers in ascending order and generate the corresponding channel aggregation section number index list, then read the label node number in each aggregation section in turn and call the corresponding semantic field information, the semantic field is represented in the form of text encoding, such as "enter", "keep", "switch", "exit", etc., all semantic field information is archived according to the aggregation section number, then the archived semantic field group is sequentially arranged, it is judged whether the values of the continuous label nodes in the semantic field constitute a semantic process chain, the semantic process chain is defined as at least three continuous nodes whose semantic fields constitute a logical sequence, for example: T102 is "enter", T103 is "keep", and T104 is "exit", which form an effective semantic chain, if T105 is "enter", T106 is "keep", and T107 is still "keep", it does not meet the "exit" end condition, and is considered as an incomplete chain and is rejected, the node group that meets the semantic sequence relationship is further executed for path extension relationship analysis, it is judged whether the label nodes in the group exist in the original path sequence, if the label numbers in the semantic continuous chain are also continuous, it is considered as a continuous path sequence, if the number interval exceeds two or more jumps, it is marked as a non-continuous segment, the continuous and non-continuous path segments are marked as S (sequence) and B (break) types respectively, and are stored in two path buffer tables respectively, the next aggregation section is continuously scanned, and the above operation is repeated, after completion, all S type path segments are merged, and the path segment range index is established according to the start and end label node numbers, then the B type segments are connected back to their corresponding original path numbers, and it is checked whether they can be merged into S type segments by forming a head-tail connection with other B type segments, the connection condition is: the end node semantic field of two B segments is "keep" or "switch", and the start node is "enter", then it is judged that it has path extension relationship, the B type segment is merged into the last S type path segment, after the merging judgment of all aggregation sections, the continuous semantic path segments in the merging result are finally recorded as the extension result, and the semantic extension structure node is obtained.

[0131] S502: Call the semantic extension structure node, extract the behavior trajectory number, cluster number and belonging aggregation level value of the associated node in the path segment, analyze the corresponding structure characteristics between the trajectory number and the cluster number, classify and integrate the label nodes whose number relationship has been verified, and obtain the behavior trajectory cluster association sequence;

[0132] Firstly, the number of each extension node in the path segment and its belonging label information are read, the correspondence between the path segment number and the label number is recorded, and the behavior trajectory number to which each label node belongs in the path segment is obtained, the behavior trajectory number represents the time positioning and the jump path in the path evolution of the label, then the corresponding cluster number of each node is read, that is, the index classification position of the label in the label cluster, and the corresponding aggregation level value of the label is extracted, the aggregation level value represents the level of the label node in the clustering structure, which is divided into three fixed segments, such as L1 for top layer, L2 for middle layer, and L3 for bottom layer, the trajectory number, the cluster number and the level value are combined to form a three-element index group, one-to-one relationship verification is performed on all three-element groups, the label nodes with one-to-one correspondence between the trajectory number and the cluster number in the same path segment are marked, and it is judged whether the corresponding mode satisfies the cluster continuity, the cluster continuity is defined as the cluster number being continuously increased or unchanged in sequence, if the label trajectory number is T101, T102 and T103, and the cluster number is G02, G02 and G03, it is regarded as a continuous cluster sequence, if it is G02, G04 and G03, it is regarded as a non-continuous cluster sequence and is rejected, further, the trajectory number position and the aggregation level value of the label node marked as a continuous cluster are recorded, the node groups with the same aggregation level in the same trajectory segment are screened, and they are merged according to the cluster number, the merging rule is that the label nodes with continuous trajectory number, consistent cluster number and consistent aggregation level are grouped into one group, each group forms an independent number set, the set format is {T101, G02, L2}, {T102, G02, L2} and {T103, G03, L2}, the first two items are grouped into one group, and the last item is not included in the group because the cluster numbers are different, finally, all label node number sequences satisfying the cluster number consistency and the behavior trajectory continuity are uniformly output, and the behavior trajectory cluster association sequence value is obtained.

[0133] S503: Based on the behavior trajectory cluster association sequence, the label node number groups with consistent structure level in the path segment are extracted, the behavior trajectory and the node number are combined and associated, the node behavior with associated characteristics in the channel aggregation area is regularized, and a semantic behavior focusing analysis set is obtained.

[0134] First, each label node in the sequence is screened, the path segment number and the corresponding aggregation level value of the node are read, and the label nodes are grouped with the aggregation level as the primary key. Nodes with the same aggregation level after grouping are extracted as a structural consistency set. Then, the behavior trajectory number of each label node in the structural consistency set is read, and the trajectory number represents the time sequence activity identification of the node in the path segment. Subsequently, the trajectory number and the label number are combined in ascending order, and the path segment number is used as an auxiliary index to form a trajectory-label pair. The position alignment operation is performed on all trajectory-label pairs to determine whether the trajectory number is monotonically increasing. If the number jump does not exceed 2, it is retained, otherwise the trajectory pair is excluded and not included in the subsequent integration. At the same time, it is determined whether the trajectory numbers in each aggregation level group overlap. If the same trajectory number appears between two discontinuous nodes, it is considered as a behavior split fragment, its path number is recorded and marked as a conflict, and it does not participate in the regularization operation. After completing the trajectory number cleaning, all trajectory numbers are bound with the node numbers to form a behavior association group, which is grouped into a channel aggregation section according to the path segment number. Each aggregation section in the behavior association group is taken as a basic behavior unit, and the semantic field value and channel number of each node number in the unit are read. It is determined whether the same trajectory number is distributed in multiple channel numbers. If there is a cross-channel jump but the semantic field remains consistent, it is marked as "cross-channel semantic continuity". The nodes of this type are reorganized and sorted according to the channel number and are uniformly included in the regularization process. Finally, the behavior label set after regularization according to the four elements of path, channel, level and semantic is output, and the semantic behavior focusing analysis set is obtained.

[0135] Please refer to Figure 2 An artificial intelligence-based big data information analysis system, comprising:

[0136] The direction aggregation point recognition module obtains the direction vector of the jump structure node, extracts the channel number and path number of the mutation point label node, combines the direction vector sequence and the channel arrangement position, combines the nodes with continuous directions, aggregates the path numbers in the order of jump points, and obtains the direction jump aggregation label group atlas;

[0137] The semantic mapping breakage module reads the semantic annotation field and path segment number of the labels in the aggregation section based on the node path in the direction jump aggregation label group atlas, merges the path numbers of the nodes with discontinuous upstream and downstream semantic fields, and obtains the semantic interruption structure mapping fragment set;

[0138] The pathway connection reorganization module extracts the jump section connection number and pathway direction based on the broken nodes in the semantic interruption structure mapping fragment set, compares the connection node position and direction sequence, groups the jump sections with structural connection, and obtains the bridge section entry path node structure diagram;

[0139] The sliding structure linkage module extracts a jump sequence to form a path direction sequence based on the connection nodes in the bridge section entry path node structure diagram, calls a path offset trend and a structure level value, combines path sections with consistent structures, constructs a continuous jump channel, and obtains a path sliding structure interconnection sequence diagram;

[0140] The behavior characteristic focusing module identifies semantic expansion structures of a channel aggregation section based on a label node in the path sliding structure interconnection sequence diagram, calls a behavior trajectory field, a cluster relationship, and an aggregation level, induces a node group with coinciding clusters and consistent structures, and obtains a semantic behavior focusing analysis set.

[0141] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A big data information analysis method based on artificial intelligence, characterized by, Comprise the following steps: S1: Obtain the jump structure in the trajectory, analyze the jump direction vector, compare the channel number of the direction mutation point with the path number and the adjacent jump point order, merge the label nodes according to the direction consistency, and obtain the direction jump aggregation label group atlas; S2: Based on the identified node path in the direction jump aggregation label group atlas, according to the label node in the aggregation section, read the semantic trajectory field and the path number, split the path section without upstream and downstream mapping relationship, and obtain the semantic interruption structure mapping segment set; S3: Based on the broken node in the semantic interruption structure mapping segment set, analyze the connection number and direction order of the front and rear jump section, mark the discontinuous section with consistent direction as the bridge section entrance, and obtain the bridge section entrance path node structure diagram; S4: Based on the bridge section entrance path node structure diagram, compare the jump direction sequence, reorder the path section with consistent offset and continuous level, and obtain the path sliding structure connected sequence diagram; S5: Based on the label node group in the path sliding structure connected sequence diagram, analyze the semantic expansion structure in the channel aggregation area, combine the continuous trajectory and semantic distribution into behavior segment, and obtain the semantic behavior focusing analysis set; The specific steps of S4 are: S401: Based on the bridge section connection node in the bridge section entrance path node structure diagram, extract the jump sequence between adjacent nodes, collect the start and end node number and jump direction of the path section, combine according to the path number sequence, and obtain the path direction sequence combination; S402: Call the path direction sequence combination, extract the direction identifier and structure level value of the path section, calculate the frequency of the direction identifier in the path sequence, filter the path numbers with consistent frequency and same structure level, and obtain the direction structure frequency combination path; S403: Based on the direction structure frequency combination path, extract the jump node number sequence corresponding to the classified path section, align the node number and path number position according to the direction frequency classification order, and obtain the path sliding structure connected sequence diagram; The specific steps of S5 are: S501: Based on the label node group in the path sliding structure connected sequence diagram, extract the corresponding channel aggregation section number, collect the semantic field information of the label node in the aggregation section, identify the node sequence with path extension relationship between fields, merge the path according to the structure continuity, and obtain the semantic expansion structure node; S502: Call the semantic expansion structure node, extract the behavior trajectory number, cluster number and corresponding aggregation level value of the associated node in the path section, analyze the corresponding structure characteristics between the trajectory number and the cluster number, classify and integrate the label nodes with verified number relationship, and obtain the behavior trajectory cluster association sequence value; S503: Based on the behavior trajectory cluster association sequence value, extract the label node number group with consistent structure level in the path section, combine the behavior trajectory and node number, and regularize the node behavior with association characteristics in the channel aggregation area, and obtain the semantic behavior focusing analysis set. 2.The big data information analysis method based on artificial intelligence according to claim 1, wherein, The direction jump aggregation label group atlas includes a jump point direction vector sequence, a channel number order mapping set, a direction convergence label node group, and a path structure branch paragraph set. The semantic interruption structure mapping fragment set includes a path paragraph number sequence, a semantic field break node set, a non-continuous mapping path index, and a structure interruption node classification result. The bridge section entrance path node structure diagram includes a jump section connection number index, a passage direction sequence set, a connectable path fragment set, and a bridge section entrance node identification set. The path slip structure interconnection sequence diagram includes a jump direction offset ordering group, a structure level continuous node group, a path linkage slip section number sequence, and a channel sequence mapping table. The semantic behavior focus analysis set includes a behavior trajectory aggregation sequence, a cluster number classification table, a path level relationship set, and a semantic mode structure item. 3.The big data information analysis method based on artificial intelligence according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the node set of the jump structure in the trajectory, extract the connectivity order between the direction vector of the node and the adjacent node, combine the arrangement trend of the direction vector and the node jump position sequence, analyze the direction structure difference of the label node in the direction mutation area, and obtain the jump point direction offset relationship; S102: Based on the jump point direction offset relationship, extract the channel number and path number of the label node where the mutation point is located, call the channel number arrangement order and the number position of the jump point in the path, screen the channel node sequence with consistent direction offset trend and integrate the sequence, and obtain the channel label order structure; S103: Based on the channel label order structure, segment the path number and channel structure sequence to which the label node belongs, collect the label combination with consistent direction in the same path and perform path structure splitting, branch and classify the obtained label node sequence, and obtain the direction jump aggregation label group atlas. 4.The big data information analysis method based on artificial intelligence according to claim 1, wherein, The specific steps of S2 are as follows: S201: Based on the identified node path in the direction jump aggregation label group atlas, extract the label node number in the aggregation section, read the corresponding semantic trajectory label field and path paragraph number, pair and associate the semantic fields according to the path number order, and obtain the semantic field path mapping relationship; S202: Call the semantic field path mapping relationship, screen the path number sequence with non-continuous corresponding semantic fields, identify the label node with interrupted semantic corresponding relationship, extract the identification value of the interruption node according to the path number order and classify and arrange, and obtain the mapping structure interruption node sequence; S203: Based on the mapping structure interruption node sequence, extract the path number and paragraph position of the contained node, mark the label path where the interruption point is located according to the path number segmentation, match and merge the corresponding path paragraph number and interruption node number, and obtain the semantic interruption structure mapping fragment set. 5.The big data information analysis method based on artificial intelligence according to claim 1, wherein, The specific steps of S3 are as follows: S301: Based on the path breaking node in the semantic interruption structure mapping fragment set, extract the front and rear jump section connection number and passage direction order corresponding to the node, arrange the jump section connection structure according to the path number, and sort the jump section number and compare the passage direction, and obtain the path jump section direction order; S302: Call the path hop segment direction sequence, extract the adjacent hop segment in the path direction sequence, calculate the frequency of repeated occurrence of the path direction in the hop segment sequence, mark the node number combination with the same frequency, and obtain the direction sequence coincidence position; S303: Based on the direction sequence coincidence position, filter the node pairs with direction continuation characteristics, extract the path number and entrance node position, and integrate the hop path number in the path order to obtain the bridge segment entrance path node structure diagram. 6.The big data information analysis method based on artificial intelligence according to claim 5, characterized in that, The calculation formula of the frequency of repeated occurrence of the path direction in the hop segment sequence is: ； wherein, the frequency of repetition of the direction of passage in the sequence of hops, the direction of passage the cumulative number of occurrences in the sequence of hops, the mean of the number of occurrences of the direction of passage, the direction the sum of the squares of the differences between the position numbers of adjacent hops in the path, the direction the number of path segments in the sequence of hops, the direction the number of hops covered by the range of path segments, the direction the number of hop label nodes in the path segment, the total number of types of direction of passage in the sequence of hops. 7.The big data information analysis method based on artificial intelligence according to claim 1, wherein, The calculation formula of the frequency of repeated occurrence of the path direction in the hop segment sequence is: ； wherein, representing the frequency of occurrence of the class direction identifier in the path sequence, representing the total number of occurrences of the class direction identifier in the current path direction sequence, representing the direction weight value of the class direction identifier in the occurrence, representing the structure level value of the path segment in the occurrence of the class direction identifier, representing the path occupancy density of the structure unit connected by the path segment in the occurrence of the class direction identifier, representing the number of adjacent direction identifiers connected to the path segment position in the occurrence of the class direction identifier.

8. An artificial intelligence-based big data information analysis system, characterized by, The system is used to realize the artificial intelligence-based big data information analysis method of any one of claims 1-7, and the system comprises: The direction aggregation point recognition module obtains the direction vector of the jump structure node, extracts the channel number and path number of the mutation point label node, combines the direction vector sequence and the channel arrangement position, combines the nodes with continuous directions, aggregates the path numbers in the order of jump points, and obtains the direction jump aggregation label group atlas; The semantic mapping broken link module reads the semantic annotation field and path segment number of the aggregated segment based on the node path in the direction jump aggregation label group atlas, merges the path numbers of the nodes with discontinuous upstream and downstream semantic fields, and obtains a semantic interruption structure mapping fragment set; The path connection recombination module extracts the hop segment connection number and path direction based on the broken nodes in the semantic interruption structure mapping fragment set, compares the connection node position and direction sequence, groups the hop segments with structural connection, and obtains a bridge segment entrance path node structure diagram; The sliding structure linkage module extracts the path direction sequence formed by the jump sequence based on the connection nodes in the bridge segment entrance path node structure diagram, calls the path offset trend and structure level value, combines the path segments with consistent structures, constructs a continuous jump path, and obtains a path sliding structure connection sequence diagram; The behavior characteristic focusing module identifies the semantic expansion structure of the channel aggregation section based on the label nodes in the path sliding structure connection sequence diagram, calls the behavior trajectory field, group relationship and aggregation level, and induces the node group with consistent group coincidence and structure, to obtain a semantic behavior focusing analysis set.

Images