A Dynamic Monitoring Algorithm for Text Data Risk Signals Based on the Coupled DWT-IFLOF

The composition elements of text data are complex, and the risk signals between multiple elements accumulate and superimpose. Compared with actual signals, this leads to significant monitoring deviations in the distribution of risk signals and risk value intervals. Therefore, a dynamic monitoring algorithm for text data risk signals based on the coupled DWT-IFLOF has been developed. Segmenting the original text data and utilizing the distribution of key information on iForest to generate risk signals covering all elements of the text data; Extracting feature parameters of signals through DWT and using them as benchmarks for determining the specific node risk status in risk signals; After dividing the benchmark neighborhood, calculate the relationship between its LOFnor and feature values, and output the risk value of the corresponding node. The test results show that the design method can accurately monitor the distribution of risk signals based on the conversion of text data signals, without deviation from the actual signal risk value range.