Ayush KaulThakur College of Engineering and Technology, Mumbai Sharad BharadiaThakur College of Engineering and Technology, Mumbai Prince SinhaThakur College of Engineering and Technology, Mumbai. Say no to plagiarism. Get a tailor-made essay on "Why Violent Video Games Shouldn't Be Banned"? Get Original Essay In this modern era, when the world is moving towards automation, there is a need for automation in the answer evaluation system. Currently, online answer evaluation is available for MCQ based questions, so theoretical answer evaluation is hectic for the controller. The teacher manually checks the answer and accepts the votes. The current system requires more manpower and time to evaluate the response. This project is an application based on evaluating responses using machine learning. The project is specially developed to reduce labor and time consumption. Since in manual response evaluation, manpower and time consumption are much greater. Furthermore, in the manual system, it is possible that the marks assigned to two same answers are different. This application system provides automatic answer rating based on the keyword provided to the application in the form of dataset by the user which will provide fair distribution of votes and reduce time and manpower. Keywords: OCR, backpropagation algorithm, ReLU, ANN Introduction (Title 1) Manual evaluation of responses is a very tedious task. Manual inspection is a time-consuming process and also requires a lot of manpower. Furthermore, the card controller is unable to assign equal scores. Therefore, our system will evaluate the answer based on some keywords and labor will also be saved. Simply scan the sheet and then, based on the keyword contained in the answer, the system will provide the score for the question based on the dataset present. Furthermore, with this system, the error in evaluating the marks relating to the specific question will be reduced. Therefore, our system will evaluate the answer based on some keywords and labor will also be saved. Just scan the sheet then the system will divide the answer using OCR[3], based on the keyword in the answer the system will provide the score to the question based on the dataset present[4]. A question of this type is needed which provides easy evaluation of the answer and can provide suitable scores. Furthermore, this application will help various colleges, universities and coaching institutes to evaluate the answer in less time and with less manpower. Checking the answers requires high concentration due to the large amount of time which often leads to errors. Automating this task will increase the efficiency of response evaluation at scale. After a brief discussion, it was understood that the answer sheet is evaluated keeping in mind some key words that the moderators look for the answer while evaluating an answer. Our proposed algorithm will take keywords as input. These keywords will be provided by the subject matter expert. Our proposed algorithm will match these keywords to the detected words extracted from the answer sheet using a supervised learning algorithm. The learning phase of the model will require a handwritten dataset for English language alphabets. These datasets are available online in various formats for you to use to train your model. The machine learning model used in the proposed algorithm consists of neural networks with multiple hidden layers. The model calculates the error using the backpropagation algorithm. The network weights are updated in thedirection opposite to the partial differentiation of the error with respect to the weighted input to the neuron in a particular layer. The activation function used for the model is ReLU (rectified linear unit) which calculates as:f( x)=max(0,x)Here the variable x is an input to the function. The proposed algorithm will also consider the response length as a parameter for response evaluation. The ideal length of the answer will be taken as input by the teacher. Research Paper “An Approach to Evaluation of Subjective Questions for Online Examination System” by Sheeba Praveen, Assistant Professor, CSE Department, Integral University, Lucknow, UP, India. In recent years we have seen a number of government and semi-government exams move online, for example [IBPS Common Written Examination (CWE)]. This system or any other similar system represents a resource saving advantage. However, we have observed that these systems only accommodate multiple choice questions and there are no plans to extend these systems to subjective questions. Our goal is to design an algorithm for automatically evaluating the descriptive response of a single sentence. The contribution presents an approach to verify the level of learning of the student/learner, evaluating the relevant descriptive exam answer sheets. Representing the descriptive response as a graph and comparing it to the standard response are the key steps in our approach. B Vanni, M. shyni, and R. Deepalakshmi, “High-precision optical character recognition algorithms using ANN learning array” in Proc. 2014 IEEE International Conference on Circuit, Power and Computing Technologies ( ICCPCT), International Conference 2014. Optical character recognition refers to the process of translating handwritten or printed text into a format understood by machines for the purposes of editing, searching and indexing. The performance of current OCR illustrates and explains actual imaging errors and defects in recognition with illustrated examples. This paper aims to create an application interface for OCR using artificial neural network as the backend to achieve a high recognition accuracy rate. The proposed algorithm using the concept of neural network provides a high accuracy rate in character recognition. The proposed approach is implemented and tested on an isolated character database consisting of English characters, digits and special keyboard characters. Proposed Methodology This project is an application for automated evaluation of responses using the matching keyword from a dataset based on a machine learning algorithm. There are some applications available, but they are different from this one and use different methodologies. Some available applications only evaluate MCQs (multiple choice questions) and not the subjective question[1]. To use this application you just need to scan the answer to that question, then the system will split the answer keyword using OCR [3]. Based on the keywords written in the answer and the keywords in the dataset, the application will provide scores between 1 and 5. Steps to evaluate the answer Provide an answer sheet to the system in jpeg (.jpg) format provide keywords , maximum score and minimum length required for the answer. The system will separate the words from the given answer the given words will be stored in the .csv file the length of the answer will be calculated by counting the words in the CSV file checks the percentage of matching keywords checks the percentage of words written compared to the minimum length check the percentage of votes awarded for the date 0=.