CN-114463041-B - Intelligent early warning method, device, equipment and storage medium

CN114463041BCN 114463041 BCN114463041 BCN 114463041BCN-114463041-B

Abstract

The invention relates to the technical field of artificial intelligence, and discloses an intelligent early warning method, device, equipment and storage medium, wherein the method comprises the steps of acquiring an early warning configuration file corresponding to a user identifier and all put field data; the method comprises the steps of carrying out index calculation on all input field data to obtain a plurality of index data, carrying out matrix generation on all input field data and all index data to obtain a matrix to be processed, inputting an early warning configuration file and the matrix to be processed into an early warning hosting model, carrying out early warning condition detection on the matrix to be processed to obtain a detection result, judging whether an early warning item exists in the detection result, carrying out trend analysis and action recommendation on the matrix to be processed and all early warning items through an input action recommendation model to obtain an early warning result, and therefore realizing automatic detection of the early warning item through the early warning hosting model, carrying out automatic trend analysis and action recommendation, timely carrying out early warning and action, avoiding loss expansion and saving input cost.

Inventors

LIU YANG
YAN RENSHUANG

Assignees

深圳市东信时代信息技术有限公司
深圳市东信时代信息技术有限公司

Dates

Publication Date: 20260421
Application Date: 20211231
Priority Date: 20211231

Claims (7)

1. An intelligent early warning method is characterized by comprising the following steps: Acquiring an early warning configuration file and all put field data corresponding to a user identifier; performing index calculation on all the put field data to obtain a plurality of index data; generating matrixes for all the put field data and all the index data to obtain a matrix to be processed; Inputting the early warning configuration file and the matrix to be processed into an early warning hosting model, and detecting early warning conditions of the matrix to be processed through the early warning hosting model to obtain a detection result; judging whether an early warning item exists in the detection result; If any early warning item exists in the detection result, inputting the matrix to be processed and the early warning item in the detection result into a throwing action recommendation model, and carrying out trend analysis and action recommendation on the matrix to be processed and all the early warning items through the throwing action recommendation model to obtain an early warning result; before the early warning configuration file corresponding to the user identifier is obtained, the method comprises the following steps: Acquiring advertisement types and early warning requirements corresponding to the user identification; Performing early warning attribute keyword recognition on the early warning requirement to recognize an early warning attribute result; based on the early warning attribute result, carrying out context semantic recognition on the early warning requirement to obtain an attribute configuration result corresponding to the early warning attribute result; Generating the early warning configuration file corresponding to the user identifier according to the advertisement type and the attribute configuration result; and detecting the early warning condition of the matrix to be processed through the early warning hosting model to obtain a detection result, wherein the detection result comprises the following steps: performing threshold interpretation on the early warning configuration file to obtain a threshold parameter array; index prediction is carried out on the matrix to be processed through the early warning hosting model, and a prediction result is obtained; Based on the threshold parameter array, carrying out early warning condition detection on the prediction result to obtain the detection result; And correspondingly comparing each early warning parameter in the threshold parameter array with index data in the prediction result to obtain a comparison result, and judging whether each item in the comparison result accords with a corresponding condition rule or not, wherein the item which accords with the corresponding condition rule is determined to be an early warning item.
2. The intelligent early warning method according to claim 1, wherein the performing index calculation on all the put field data to obtain a plurality of index data includes: screening index attribute data from all the put field data based on a preset index rule; and carrying out index analysis on the index attribute data through each index model in the preset index rule to obtain the index data.
3. The intelligent early warning method according to claim 2, wherein the index attribute data includes a delivery amount attribute value, a click attribute value, a download attribute value, a cost attribute value, and a consumption attribute value, and wherein the performing each index analysis on the index attribute data by each index model in the preset index rule to obtain the index data includes: carrying out click rate analysis on the input quantity attribute value and the click attribute value through a click index model in the preset index rule to obtain click index data; performing conversion rate analysis on the input quantity attribute value, the click attribute value and the download attribute value through a conversion index model in the preset index rule to obtain conversion index data; carrying out consumption analysis on the cost attribute value and the consumption attribute value through a consumption index model in the preset index rule to obtain consumption index data; And recording the click index data, the conversion index data and the consumption index data as the index data.
4. The intelligent early warning method according to claim 1, wherein before inputting the matrix to be processed and the early warning items in the detection result into the action recommendation model, the method comprises: the method comprises the steps of acquiring a historical user data set, wherein the historical user data set comprises a historical user data matrix, and action labels and action execution results which are associated with the historical user data matrix; inputting the historical user data matrix into a deep learning model containing initial parameters; Extracting action features of the historical user data matrix through the deep learning model, and performing action prediction according to the extracted action features to obtain a predicted action category and a predicted action result; Based on a cross entropy loss algorithm, obtaining a first loss value according to the predicted action category and the action label associated with the historical user data matrix, and obtaining a second loss value according to the predicted action result and the action execution result associated with the historical user data matrix; weighting the first loss value and the second loss value to obtain a total loss value; And when the total loss value does not reach the convergence condition, iteratively updating initial parameters in the deep learning model, and executing the step of extracting action characteristics of the historical user data matrix through the deep learning model until the total loss value reaches the convergence condition, and recording the converged deep learning model as a delivery action recommendation model.
5. An intelligent early warning device, characterized by comprising: The acquisition module is used for acquiring the early warning configuration file and all the put field data corresponding to the user identifier; The calculation module is used for carrying out index calculation on all the put field data to obtain a plurality of index data; the generation module is used for generating matrixes of all the put field data and all the index data to obtain matrixes to be processed; the detection module is used for inputting the early warning configuration file and the matrix to be processed into an early warning hosting model, and detecting early warning conditions of the matrix to be processed through the early warning hosting model to obtain a detection result; The judging module is used for judging whether the detection result has an early warning item or not; The early warning module is used for inputting the matrix to be processed and early warning items in the detection result into a throwing action recommendation model if any early warning item exists in the detection result, and carrying out trend analysis and action recommendation on the matrix to be processed and all the early warning items through the throwing action recommendation model to obtain an early warning result; before the early warning configuration file corresponding to the user identifier is obtained, the method comprises the following steps: Acquiring advertisement types and early warning requirements corresponding to the user identification; Performing early warning attribute keyword recognition on the early warning requirement to recognize an early warning attribute result; based on the early warning attribute result, carrying out context semantic recognition on the early warning requirement to obtain an attribute configuration result corresponding to the early warning attribute result; Generating the early warning configuration file corresponding to the user identifier according to the advertisement type and the attribute configuration result; the detection module comprises: the interpretation unit is used for performing threshold interpretation on the early warning configuration file to obtain a threshold parameter array; The prediction unit is used for performing index prediction on the matrix to be processed through the early warning hosting model to obtain a prediction result; The detection unit is used for detecting the early warning condition of the prediction result based on the threshold parameter array to obtain the detection result; And correspondingly comparing each early warning parameter in the threshold parameter array with index data in the prediction result to obtain a comparison result, and judging whether each item in the comparison result accords with a corresponding condition rule or not, wherein the item which accords with the corresponding condition rule is determined to be an early warning item.
6. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the intelligent pre-warning method of any one of claims 1 to 4 when the computer program is executed by the processor.
7. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the intelligent pre-warning method of any one of claims 1 to 4.

Description

Intelligent early warning method, device, equipment and storage medium Technical Field The invention relates to the technical field of artificial intelligence, in particular to an intelligent early warning method, device, equipment and storage medium. Background At present, in the existing commercial popularization activities, advertisement delivery plays a very important role, and in the existing advertisement delivery, advertisement optimizers often monitor a plurality of advertisement delivery indexes in real time manually, calculate monitoring indexes of each client, and combine configuration requirements of each client to adjust the delivery strategy of each advertisement, and due to the hysteresis and subjectivity of manual decision, problems of overestimation of budget, missing of optimal bid time, non-ideal conversion effect and the like are easy to occur, and finally the cost of the client is wasted. Disclosure of Invention The intelligent early warning method, the intelligent early warning device, the computer equipment and the storage medium provided by the invention realize automatic trend analysis and action recommendation when early warning projects exist, recommend the throwing action, and timely make early warning and action, avoid loss expansion and save throwing cost. An intelligent early warning method, comprising: Acquiring an early warning configuration file and all put field data corresponding to a user identifier; performing index calculation on all the put field data to obtain a plurality of index data; generating matrixes for all the put field data and all the index data to obtain a matrix to be processed; Inputting the early warning configuration file and the matrix to be processed into an early warning hosting model, and detecting early warning conditions of the matrix to be processed through the early warning hosting model to obtain a detection result; judging whether an early warning item exists in the detection result; if any early warning item exists in the detection result, inputting the matrix to be processed and the early warning item in the detection result into a throwing action recommendation model, and carrying out trend analysis and action recommendation on the matrix to be processed and all the early warning items through the throwing action recommendation model to obtain an early warning result. An intelligent early warning apparatus, comprising: The acquisition module is used for acquiring the early warning configuration file and all the put field data corresponding to the user identifier; The calculation module is used for carrying out index calculation on all the put field data to obtain a plurality of index data; the generation module is used for generating matrixes of all the put field data and all the index data to obtain matrixes to be processed; the detection module is used for inputting the early warning configuration file and the matrix to be processed into an early warning hosting model, and detecting early warning conditions of the matrix to be processed through the early warning hosting model to obtain a detection result; The judging module is used for judging whether the detection result has an early warning item or not; and the early warning module is used for inputting the matrix to be processed and early warning items in the detection result into a throwing action recommendation model if any early warning item exists in the detection result, and carrying out trend analysis and action recommendation on the matrix to be processed and all the early warning items through the throwing action recommendation model to obtain an early warning result. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the intelligent pre-warning method described above when the computer program is executed. A computer readable storage medium storing a computer program which when executed by a processor implements the steps of the intelligent pre-warning method described above. The intelligent early warning method, the device, the computer equipment and the storage medium provided by the invention are characterized in that an early warning configuration file and all input field data corresponding to a user identifier are obtained, index calculation is carried out on all input field data to obtain a plurality of index data, matrix generation is carried out on all input field data and all index data to obtain a matrix to be processed, the early warning configuration file and the matrix to be processed are input into an early warning hosting model, early warning condition detection is carried out on the matrix to be processed through the early warning hosting model to obtain a detection result, whether early warning items exist in the detection result or not is judged, if any early warning item exists in the detection result, the early warning items in the matrix to