Unstructured test question data labeling method and device and storage medium

Abstract

The invention discloses an unstructured test question data labeling method and device and a storage medium. The method comprises the steps: extracting multiple pieces of unlabeled data from a test question data set according to a self-service sampling method to be preprocessed; inputting the plurality of pieces of pre-processed unlabeled data into a deep learning network, and inputting labeled data into the deep learning network for correction; respectively inputting an output result of the deep learning network into at least two different types of base classifiers for ensemble learning, wherein each base classifier comprises a plurality of weak learners and a strong learner of the same type; and constructing a transition probability matrix for the output data in all the base classifiers,solving the transition probability matrix, and generating a labeling result. According to the method, the labeling accuracy is improved, the labeling problem of test questions, test papers and other unstructured data is effectively solved, on the basis, the automatic storage problem of unstructured texts can be solved, and a large amount of manual labor can be saved.

Description

technical field [0001] The invention relates to the technical field of educational informatization, in particular to a labeling method, equipment and storage medium for unstructured test question data. Background technique [0002] With the rapid development of artificial intelligence, more and more data needs to be labeled. For example, if we want an artificial intelligence robot to be able to identify whether a photo is a giant panda, then we first need to collect a batch of photos containing giant pandas, and then manually label these photos, and the content of the label is which photos are giant pandas photos, which photos are not photos of giant pandas, and then feed the marked content and photo data to the artificial intelligence program for deep learning. When the artificial intelligence robot completes the deep learning, when a new photo appears At this time, this program can complete intelligent identification, which one is a photo of a giant panda and which one is...

AI Technical Summary

Problems solved by technology

[0009] (2) A lot of manual labor, error-prone, and high cost

However, the text of the test paper requires marking where is the question, where is the question, where is the option, and where is the answer. The length of the text is long, and there is no better solution at present.

Images