A method and system for automatic review of error correction questions

Abstract

Embodiments of the present invention provide an automatic review method and device for error correction questions, which belong to the field of image processing. The method includes: obtaining single-point answer images corresponding to each written answer on the answer medium to be read; using a preset image detection method and / or based on a preset recognition model to perform image recognition on each of the acquired single-point answer images to obtain a recognition result Obtain the corresponding review feature according to the identification information, and review each written answer based on the preset review model according to the review feature. The method of this specific embodiment can overcome the technical problem of inaccurate image recognition of the answer due to the complexity of the answering mode of the error correction question, thereby having the beneficial effect of quickly and accurately reviewing the error correction question.

Description

technical field [0001] The present invention relates to the technical field of image information processing, and more specifically, to a method and system for automatic review of error correction questions. Background technique [0002] In contemporary language exams, (short essay) correction questions belong to the objective examination of subjective test questions. It not only examines the basic knowledge of all aspects of the respondent, but also examines the comprehensive ability of the respondent to use the language. In terms of setting the questions, the questions examine the examinee's ability to comprehensively use the corresponding language in the discourse from the three perspectives of morphology, syntax and writing logic; in terms of details, the tense, voice, predicate verb, noun, article, adjective , conjunctions, pronouns, and the succession of sentences are examined. The respondent answers the questions by adding, deleting or modifying the words in the test...

AI Technical Summary

Problems solved by technology

[0004] However, for the error correction questions, it is limited by the complexity and diversity of handwriting in the answers of the answerers (such as the handwritten mark "∧" of "increase" will appear in various handwriting styles due to different answerers), so the error correction questions have so far still Requires manual review by correctors

Especially in large-scale examinations, such as the high school entrance examination and language level examination, the manual review method is time-consuming and labor-intensive, which brings great work pressure to the reviewers; in addition, manual review is often easily affected by the subjectivity of the reviewers. impact, making the review results unfair

Images