Abnormal article detection method and device

Abstract

The invention discloses an abnormal article detection method and device, and relates to the field of warehouse logistics. A specific embodiment of the method comprises the following steps: collecting name information, category information and logistics attribute information of an in-warehouse article; according to the in-warehouse article name information and the in-warehouse article category information, performing classification judgment on the in-warehouse articles based on a preset text classification model, and determining the in-warehouse articles meeting a preset abnormal condition as first abnormal articles; taking the articles left after the first abnormal article is removed from the in-warehouse articles as to-be-detected articles; and screening out a second abnormal article from the to-be-detected articles by using a preset anomaly detection algorithm according to the category information of the to-be-detected articles and the logistics attribute information of the to-be-detected articles. According to the embodiment, active anomaly detection can be carried out according to various data information of the article, and the method can adapt to various sample conditions, so that the detection result is accurate.

Description

technical field [0001] The invention relates to the field of warehousing and logistics, in particular to a method and device for detecting abnormal items. Background technique [0002] In the field of warehousing and logistics, logistics billing is calculated based on the weight and volume of the waybill, and the weight and volume of the waybill includes the weight and volume of items and consumables, so the correctness of the weight and volume of items has an important impact on billing. At present, the methods for abnormal monitoring of item weight and volume mainly include the following methods: (1) use the method of experience or historical data statistics to monitor the items at the time of storage based on a certain threshold; (2) based on statistical methods, find out Outliers; (3) According to the comparison and verification between different systems, abnormal prompts are given. [0003] In the course of realizing the present invention, the inventor finds that there...

AI Technical Summary

Problems solved by technology

[0004] For (1) the threshold setting method, if the difference in weight and volume of items contained in the same category is too large, a reasonable threshold cannot be set; and this method cannot monitor weak abnormal changes; this detection method cannot be adapted to a variety of sample situations, and Not accurate enough;

[0005] For (2) the method of statistical learning, the requirements for the sample data are high, and the samples need to conform to a certain distribution; and it is easy to detect some items with a small sample size but relatively extreme weight and volume as abnormal items; this detection method cannot adapt In a variety of sample situations, it is not accurate enough and prone to misreporting;

[0006] For (3) the method based on the comparison of different systems, this method is an after-the-fact behavior and cannot actively discover abnormalities, and the cause of the inconsistency may also be due to the transmission or storage of data streams between systems; this detection method is not timely and accurate enough

Images