Automated Text Objective Question Scoring via Image Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automatic scoring systems for text objective questions rely heavily on human resources and are prone to subjective errors due to variations in teacher scoring styles, mood, and mental status, making them inefficient and inaccurate.
Innovation Solution
An intelligent scoring method and system that acquires an answer image, segments it to identify the answer string, calculates identification confidence using statistical models and acoustical models, and determines the answer's correctness based on predefined confidence thresholds, enabling fully automated scoring of text objective questions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual scoring by teachers is used for text objective questions, then scoring can be performed with existing systems, but human resources are consumed and subjective errors occur
Solution Approach 1:
The patent replaces the mechanical system of manual teacher scoring with an automated image recognition and text processing system. The system captures images of answer sheets, performs OCR to extract handwritten text, and automatically compares it with standard answers using confidence threshold evaluation, thereby eliminating human involvement in the scoring process while maintaining or improving accuracy.
Solution Approach 2:
The scoring system performs self-service by automatically evaluating answers without requiring teacher intervention. The system independently completes the entire scoring workflow including image processing, text recognition, answer comparison, and score generation, making the scoring process autonomous and eliminating dependency on human scorers.
2Measurement precision
If teachers are gathered for centralized scoring training and testing, then scoring consistency can be improved, but significant human resources and time are consumed
Solution Approach 1:
The patent replaces the time-consuming centralized training process with an automated system that inherently maintains scoring consistency through standardized algorithms. The system uses consistent image processing, text recognition, and comparison algorithms for all evaluations, eliminating the need for repeated training sessions to ensure uniform scoring standards across different evaluators.
Solution Approach 2:
The system changes the parameters of scoring evaluation by using configurable confidence thresholds and adjustable matching criteria that can be optimized for different question types and difficulty levels. This allows the system to adapt to various scoring requirements without requiring retraining of human teachers, as the parameters can be modified through software configuration.
3Productivity
If automated scoring is implemented for text objective questions, then human resource consumption is reduced, but identification accuracy must be maintained
Solution Approach 1:
The patent implements feedback mechanisms in the form of confidence threshold evaluation, where the system calculates the confidence level of each recognized answer and compares it against predefined thresholds. This feedback loop allows the system to identify low-confidence recognitions and handle them appropriately, thereby maintaining high identification accuracy while processing large volumes of answers efficiently.
Solution Approach 2:
The system performs preliminary actions by pre-processing images, pre-segmenting answer regions, and pre-configuring confidence thresholds based on question characteristics. This preliminary preparation enables the main scoring process to run efficiently with high accuracy, as the system has already optimized the data and parameters before the actual recognition and comparison operations.
Data Source
AI summary
An intelligent scoring method and system for a text objective question, the method comprising: acquiring an answer image of a text objective question (101); segmenting the answer image to obtain one or more segmentation results of an answer string to be identified (102); determining whether any of the segmentation results has the same number of characters as the standard answer (103); if no, the answer is determined to be wrong (106); otherwise, calculating identification confidence of the segmentation result having the same number of words as the standard answer, and/or calculating the identification confidence of respective characters in the segmentation result having the same number of words as the standard answer (104); determining whether the answer is correct according to the calculated identification confidence (105). The method can automatically score text objective questions, thus reducing consumption of human resource, and improving scoring efficiency and accuracy.