Traffic event classification method and system based on social media data and medium

Abstract

The invention discloses a traffic event classification method and system based on social media data, and a medium. The method comprises the steps of crawling traffic event data from a preset social media platform and constructing a data set; performing text coding on the preprocessed data set and dividing the preprocessed data set into a training set, a verification set and a test set; inputting the training set and the verification set into a traffic event classification model for model training and effect evaluation to obtain an optimal hyper-parameter, wherein the traffic event classification model comprises a pre-training model; and carrying out multi-group cross validation on the traffic event classification model based on the optimal hyper-parameter, then carrying out traffic classification prediction on the test set, and confirming an optimal traffic event classification result in multiple groups of prediction results through a voting algorithm. According to the method, massive traffic event information is automatically collected on the social media platform to serve as a data source, and training and effect verification are performed on the traffic event classification model established on the basis of the pre-training model, so that efficient and accurate traffic event intelligent classification can be realized by the model.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a traffic event classification method, system and medium based on social media data. Background technique [0002] With the development of informatization, social media platforms have been widely used in daily life. Their fast information dissemination, timely information updates, and long information retention time make them an important source of data in the field of artificial intelligence. Among them, social media traffic information refers to the content related to traffic incidents published by traffic-related departments or individual users on social media platforms. Through data mining of massive social media traffic information, traffic conditions can be fed back to facilitate residents’ travel planning. [0003] At present, the detection of traffic incidents mostly adopts machine learning models, or general deep learning models. The machine learning model is ...

AI Technical Summary

Problems solved by technology

The machine learning model is fast and interpretable, but its accuracy is relatively low; the general deep learning model relies too much on the data set. When the training set is insufficient, overfitting will occur, resulting in poor online performance of the model. However, a sufficient training set requires a lot of manpower, so the current automatic detection of traffic incidents still has the problem of low accuracy and efficiency.

Images