Search

CN-121983343-A - Infectious disease transmission risk real-time prediction system based on multi-source data fusion

CN121983343ACN 121983343 ACN121983343 ACN 121983343ACN-121983343-A

Abstract

The invention relates to the technical field of knowledge fusion, in particular to a real-time infectious disease transmission risk prediction system based on multi-source data fusion. According to the invention, a case report source identification code, an administrative division code, a space-time unit identification, a case state identification and a symptom classification identification are written into a knowledge registration unit, a space-time knowledge boundary is constructed to form a change track, after a state participation area is gathered, regional case distribution is presented in a fusion knowledge structure, the case state change direction among the regions forms a corresponding relation in continuous space-time units and is converted into propagation associated knowledge, the evolution sequence of the regional knowledge structure is mapped into propagation risk prediction content, so that risk grades and change trends are continuously generated based on the knowledge relation, the influence of single-time data fluctuation on prediction expression is avoided, and the consistency and the interpretability of information in propagation risk prediction are improved.

Inventors

  • ZHU YUNFENG
  • MING HONGYAN
  • LI ZHAOHUI
  • LIU SHENGYING

Assignees

  • 广州国际旅行卫生保健中心(广州海关口岸门诊部)

Dates

Publication Date
20260505
Application Date
20260114

Claims (10)

  1. 1. An infectious disease transmission risk real-time prediction system based on multi-source data fusion, which is characterized by comprising: The source knowledge registration module acquires a case reporting source identification code, an administrative division code, a space-time unit identification, a case state identification and a symptom classification identification, gathers the case state identification and the symptom classification identification according to the space-time unit identification and writes the case state identification and the symptom classification identification into a corresponding knowledge registration unit to generate a case source space-time knowledge unit; The trusted knowledge correction module judges whether case state identifiers and symptom classification identifiers of different case reporting sources under the same time space unit identifier form a consistent structure or not based on the case source time space knowledge unit, records deviation and accumulates when structural separation occurs, and adjusts the trusted knowledge identifier state according to the accumulated result to generate a cross-source trusted knowledge identifier set; the regional knowledge aggregation module invokes the cross-source trusted knowledge identification set to screen case source space-time knowledge units corresponding to the trusted identification state, and the regional knowledge aggregation module forms a regional case fusion knowledge structure according to the regional identification collection; and the association knowledge constraint module judges whether the case state change directions of adjacent areas under continuous space-time unit marks are consistent according to the regional case fusion knowledge structure, records stable corresponding relations and generates regional transmission association knowledge relations.
  2. 2. The infectious disease propagation risk real-time prediction system based on multi-source data fusion according to claim 1, wherein the case source spatiotemporal knowledge unit comprises case report source identification information, spatial location identification information and case state information, the cross-source trusted knowledge identification set is a data source credibility degree, a data consistency identification and a credible state change record, the regional case fusion knowledge structure comprises regional case distribution characteristics, regional case number characteristics and regional case space coverage, and the regional propagation association knowledge relationship is a regional propagation direction relationship, a regional propagation continuity characteristic and a regional propagation stability identification.
  3. 3. The infectious disease transmission risk real-time prediction system based on multi-source data fusion according to claim 1, wherein the case source spatiotemporal knowledge unit formed in the source knowledge registration module uses a combination relationship of case reporting source identification, spatiotemporal unit identification and case semantic state identification as a basic structural unit; the cross-source trusted knowledge identification set generated by the trusted knowledge correction module is used for describing the trusted relation state of the case report source in different time and space knowledge scenes; the regional case fusion knowledge structure formed in the regional knowledge convergence module is used for reconstructing a regional level knowledge expression structure with a regional identifier as a center by separating from a single case reporting source view angle and reconstructing a plurality of case source space-time knowledge units; And the region formed in the associated knowledge constraint module propagates the associated knowledge relationship and is configured to define a region case fusion knowledge structure.
  4. 4. The multi-source data fusion-based real-time infection disease transmission risk prediction system of claim 1, wherein the source knowledge registration module comprises: The source identifier acquisition submodule acquires a case report source identifier code, administrative division codes, space-time unit identifiers, case state identifiers and symptom classification identifiers, acquires the non-empty count and the sampling count calculation ratio of five types of identifiers in each case record, and acquires element preparation rate; the state symptom collecting sub-module calls case state identification and symptom classification identification according to the element preparation rate and collects according to the time space unit identification, calculates each recorded state difference value and classification difference value in the collection and forms a summation item to obtain collection consistency; And the knowledge unit writing submodule calls the collection consistency and the space-time unit identification, screens the collection consistency threshold value, writes the screening item into the corresponding knowledge registration unit, forms a source identification index sequence and a division index sequence, and generates a case source space-time knowledge unit.
  5. 5. The multi-source data fusion based real-time infection disease transmission risk prediction system of claim 1, wherein the trusted knowledge correction module comprises: The cross-source structure judging submodule judges case state identifiers and symptom classification identifiers corresponding to reporting sources of different cases under the same time-space unit identifier based on the case source time-space knowledge unit, collects state identifier sequences and symptom classification sequences of the reporting sources in the same time-space unit, carries out item-by-item corresponding comparison, judges whether a consistent structure is formed or not according to a consistency benchmark of identifier values, and generates a structure judging value; The deviation accumulation evaluation sub-module calls a reporting source identifier pair which does not form a consistent structure according to the structure discrimination value, detects a state identifier difference value and a symptom classification difference value, calculates the absolute quantity of the difference value, and accumulates according to a time space unit identifier to obtain a deviation accumulation quantity; and the trusted identification adjustment submodule invokes the deviation accumulation amount, judges a trusted identification adjustment threshold, adjusts the trusted identification state of the reporting source according to a judgment result to form a cross-source corresponding relation set, and generates a cross-source trusted knowledge identification set.
  6. 6. The multi-source data fusion-based real-time infection disease transmission risk prediction system of claim 1, wherein the regional knowledge gathering module comprises: the trusted unit screening submodule calls the case source space-time knowledge unit corresponding to the cross-source trusted knowledge identification set screening trusted identification state, acquires the trusted identification state sequence and the source identification index sequence to compare item by item, screens reserved items according to the trusted state value standard, calculates reserved proportion and acquires the trusted screening rate; the regional identification gathering submodule screens the reserved case source space-time knowledge units according to the credible screening rate, acquires administrative division codes as regional identification and carries out consistency comparison, gathers item sets according to the regional identification and counts each regional item to obtain regional gathering quantity; and the regional fusion index construction submodule calls a regional collection item set according to the regional collection quantity, acquires a space-time unit identification sequence and a symptom classification sequence, calculates the corresponding count quantity, performs merging operation with the regional collection quantity, and generates a regional case fusion knowledge structure.
  7. 7. The multi-source data fusion-based real-time infection disease transmission risk prediction system of claim 1, wherein the associated knowledge constraint module comprises: The adjacent region judging submodule judges the adjacent region relation according to the region case fusion knowledge structure, collects a region identification sequence, performs item-by-item comparison on the region identifications according to a space adjacent reference, screens and forms an adjacent region corresponding comparison sequence, and acquires the number of the region adjacent pairs; based on the number of the adjacent pairs of the regions, the change direction consistency submodule calls a continuous space-time unit identification sequence corresponding to the adjacent regions and collects a case state change sequence, calculates a state change difference value of the adjacent regions in the adjacent space-time units and compares change direction symbols to obtain a direction consistency degree; And the stable relation generation submodule calls the direction consistency degree, screens the consistency judgment threshold value, accumulates adjacent area relation indexes corresponding to the records, and generates an area propagation association knowledge relation.
  8. 8. The multi-source data fusion based real-time infection disease transmission risk prediction system of claim 1, further comprising: the prediction knowledge generation module maps the change sequence of the regional case fusion knowledge structure under the continuous space-time unit identification aiming at the regional transmission association knowledge relationship to generate regional transmission risk real-time prediction knowledge; the regional spreading risk real-time prediction knowledge comprises regional spreading risk grades, regional spreading change trends and regional spreading risk real-time prediction results.
  9. 9. The infectious disease transmission risk real-time prediction system based on multi-source data fusion according to claim 8, wherein the regional transmission risk real-time prediction knowledge output by the prediction knowledge generation module is expressed in the form of a knowledge unit or a knowledge sequence; The regional propagation risk real-time prediction knowledge is generated based on inheritance and evolution of case source space-time knowledge units, cross-source trusted knowledge identification sets, regional case fusion knowledge structures and regional propagation association knowledge relations.
  10. 10. The multi-source data fusion based real-time infection disease transmission risk prediction system of claim 8, wherein the predictive knowledge generation module comprises: The change sequence mapping submodule transfers the association knowledge relation to the region, calls a region case fusion knowledge structure and collects continuous space-time unit identification sequences, compares case state sequencing positions in adjacent space-time units to form a front sequence and a rear sequence of bit difference values, gathers the sequence of the sequence bit difference values to generate a change mapping index, and obtains a change mapping degree; Based on the change mapping degree, the propagation sequence quantization submodule calls a case state change sequence of the adjacent region under the continuous space-time unit identification, collects a region association index sequence, calculates a change sequence difference value merging absolute quantity of each region in the adjacent space-time unit, and obtains propagation sequence strength; And the risk prediction identification generation submodule calls the propagation sequence intensity to filter against a risk judgment threshold value, accumulates region association indexes corresponding to the filtering items to form a time sequence mark set, and generates region propagation risk real-time prediction knowledge.

Description

Infectious disease transmission risk real-time prediction system based on multi-source data fusion Technical Field The invention relates to the technical field of knowledge fusion, in particular to a real-time infectious disease transmission risk prediction system based on multi-source data fusion. Background The technical field of knowledge fusion comprises related technical content oriented to the cooperative utilization of multiple types of information in a complex application scene. The core of the technical field is to perform unified expression association organization and collaborative use on data with different sources, different structures and different time and scale, and generally relates to semantic alignment data time sequence arrangement of data acquisition source labeling data and construction of cross-source information association rules. By systematically organizing multi-source data in a semantic layer and a time layer of a content layer, information existing in a scattered manner can be jointly used under the same technical system, so that continuous description and analysis of complex object states are supported, and the technical contents such as data source management information association logic construction, fusion result output and the like are integrally covered in the field. The infectious disease transmission risk real-time prediction system based on multi-source data fusion refers to a technical theme for unified integration and expression of related information about infectious disease transmission. The technical matters aimed at by the patent theme comprise collection and association of case report data population mobile record public place access record environment monitoring data and medical resource use records, semantic corresponding relation setting is carried out on data items through time mark unification processing on data of different sources, and the data items are combined according to preset association rules to form a continuous data set which can be used for describing the infectious disease transmission state. The system generally adopts a data source registration mode to determine the acquisition paths of various data, and performs synchronous arrangement and joint presentation on related data by correspondingly matching the occurrence time and place of the case with the activity track of the personnel and combining the regional environment index with the medical contact record. The multi-source data is mainly synchronously arranged and presented in parallel in the existing operation, the difference of data sources is mainly distinguished through labeling, the source credible state is lack of structural expression along with time change, so that cross-source deviation can only be corrected through manual check, case data on the area level still exists in a static set after being assembled, the inter-area propagation relationship is maintained by depending on a preset matching rule, the propagation direction change in continuous space-time units is difficult to precipitate into a stable relationship, when the individual reporting source has abnormal fluctuation in a short time, abnormal information still can participate in area combination and influence propagation state description, cross-area propagation chains are easy to generate associated fracture or repeated superposition under frequent scenes of population flow or place access record change, and the reliability degree of propagation risk prediction in terms of time continuity and expression consistency is weakened. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides a real-time infectious disease transmission risk prediction system based on multi-source data fusion. In order to achieve the purpose, the invention adopts the following technical scheme that the infectious disease transmission risk real-time prediction system based on multi-source data fusion comprises: The source knowledge registration module acquires a case reporting source identification code, an administrative division code, a space-time unit identification, a case state identification and a symptom classification identification, gathers the case state identification and the symptom classification identification according to the space-time unit identification and writes the case state identification and the symptom classification identification into a corresponding knowledge registration unit to generate a case source space-time knowledge unit; The trusted knowledge correction module judges whether case state identifiers and symptom classification identifiers of different case reporting sources under the same time space unit identifier form a consistent structure or not based on the case source time space knowledge unit, records deviation and accumulates when structural separation occurs, and adjusts the trusted knowledge identifier state according to the accumulated result to generate a cross-