基于文本相似度的主观题自动评阅系统的设计与实现外文翻译资料

英语原文共 5 页，剩余内容已隐藏，支付完成后下载完整资料

毕业设计(论文)

外文资料翻译

说明：

封面之后放英文原文；英文原文之后放译文。

原文：

Abstract— Every year educational institutes conduct various examinations, which include institutional and non-institutional competitive exams. Now a days online tests andexaminations are becoming popular to reduce the burden of the examination evaluation process. The online exams include either objective or multiple-choice questions. Nevertheless, the exams include only objective or multiple-choice questions. However, subjective-based questions and answers are not involved due to the evaluation process complexity and efficiency of the evaluation process. An automatic answer checker application that checks the written answers and marks the weightage similar to a human being is more helpful in the current modern era is necessary. Hence, the software applications built to check subjective answers may be more useful for allocating marks to the user after verifying the answer for oon;ine examination.This type of tool/application/system has the challenge of having an abundant resource database, including questions, corresponding answers, and the marks allocated to the corresponding answers. At the same time system need to verify the answers provided by the users by checking the template answers and the answers provided by the user. However, Artificial Intelligence is required to identify the core element of the answers while allocating marks. Hence, an Artificial Intelligence-based answer verifier is proposed to do the job of examiner/evaluator automatically. As a result of this artificial intelligence -based answer verifier, the evaluators time and energy can be conserved, with improved work efficiency.

Keywords- Subjective answer verifier; Answer verifier; Similarity check ;Artificial Intelligence.

I.INTRODUCTION

The Online Examination is beneficial to users as in the present day, and the online exams are based on objective questions and exams are getting digitized all over the world. In this scenario, exam questions can even be based on subjective answers. Meaning that the traditional pen-paper based tests are replaced by computer-based tests that have proven to be both: (i)more consistent in allocating marks and (ii)faster than teachers correcting papers [1,11,12]. The traditional exam usually consisted of subjective answers, which were not the best way of grading the students perception of the subject. Because sometimes, examiners get bored by checking many answer sheets, and there may be an increase in the false evaluation.

Hence, the Artificial Intelligence-based Answer Verifier is required to grade the student after he/she has solved the question paper.

However, the system reduces the workload by automating the manual checking process also. An automatic subjective questions marking is a key technology in the network test system. To solve this problem, Artificial Intelligence (AI) based Answer Verifier (AV) is required to analyze like a gradinh teacher and think while receiving subjective questions [3]. In various institutes, the results are declared

after time because teachers take a long time to evaluate the subjective papers. As a bundle of answers, booklets need to be evaluated and each booklet may contain the answer in a different manner, which requires a longer duration. Hence, then AI-based Answer Verifier can come into role [4], based on the nearness theory of fuzzy mathematics [2]. It does the automatic scoring of subjective questions through specific reference values. In the investigation process, a decision-based scoring algorithm is identified as an efficient one [6]. The system based on the AI will save time and effort of humans. The online examination makes an interesting working environment by giving faster access. The existing system mainly calculates the score based on one or two parameters like keywords, or synonyms. But, coming to the 'Artificial Intelligence-based Answer Verifier', it calculates the score of the student by combining various parameters such as keywords, question specific things along with the proper grammar, which in terms provides a more accurate score. The artificial intelligence-based approach will continue to grow, and ultimately provide a full breadth of services and benefits to the students/teachers to have an efficient grading system.

II. LITERATURE SURVEY

Organizations/educational institutes always depend on the grading system through examinations. However, most of the examinations are objective. These systems or any other such system are more advantageous in terms of saving resources but failed to include subjective questions [1, 9, 10]. This paper attempted to evaluate the descriptive answer. The evaluation is done through graphical comparison with a standard answer.

A subjective answer verifier was proposed [2] by allotting the marks according to the percentage of accuracy present in the answer for different users providing three different answers. The system should have a database that includes questions, corresponding answers and the marks allocated to the corresponding answers. At the same time system need to

verify the answers provided by the users by checking the template answers and the answers provided by the user. However, Artificial Intelligence is required to identify the core element of the answers while allocating marks.

The system used a part-to-speech tagger to recognize the user answers. The answers were purely ranked based on the keyword similarities to heuristic metrics. The application achieved 70% efficiency as it was not able to consider mathematical formulas, brief description, examples and problems

剩余内容已隐藏，支付完成后下载完整资料

资料编号：[260420]，资料为PDF文档或Word文档，PDF文档可免费转换为Word