International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)

ISSN: 0976-1353 Volume 29 Issue 3 – APRIL 2022.

Automatic Paper Corrector Using

NLP

MR.MK.NAGARAJAN R RAJA SUBRAMANIAN AKSHAYA S

Department of Computer Science Department of Computer Science Department of Computer Science
and Engineering, and Engineering, and Engineering,
Kalasalingam Academy of Kalasalingam Academy of Kalasalingam Academy of
Research Education, Research Education, Research Education,
Virudhunagar, Tamil Nadu, India Virudhunagar, Tamil Nadu, India Virudhunagar, Tamil Nadu, India

BASIREDDY SREYA
Department of Computer Science and Engineering,
Kalasalingam Academy of Research Education,
Virudhunagar, Tamil Nadu, India

ABSTRACT:This paper proposes the idea of a security and reducing human labour. Regular
"automated paper corrector."This is designed to language handling (NLP) is an area of semantics,
check subjective responses in an online exam and software engineering, and man-made reasoning
assign marks to the user after the answer has been that looks at how PCs associate with human
verified. The mechanism necessitates that you language, explicitly how to build PCs that can
save the original response. To reduce manual cycle and examines a lot of normal language
labour, an automatic paper corrector has been data.The objective is to foster a PC that can "get"
developed. Everything has gone virtual as a result the items in archives, remembering language
of the Covid19 epidemic. It is difficult to virtually subtleties for setting. From that point onward, the
correct papers during online exams. So, using this framework can separate right information and bits
Automatic paper corrector, students can submit of knowledge from the articles, as well as order
their answers online, and it will correct the paper and sort out them.
and display the necessary marks. This paper will
II.PROBLEM STATEMENT
explain the design and features of automatic
paper correcting. In the existing system the exams are done
manually. If the system is not provided the right
I. INTRODUCTION
inputs, it will produce incorrect outputs. Data
It is made up of statistical natural language security is lacking. More manpower is necessary
processing libraries and programmes. The to complete the task. Time can be saved by using
libraries support tokenization, classification, an automatic paper corrector. Under the current
parsing, stemming, tagging, and sematic system, exams are currently completed manually.
reasoning. A machine can understand human
language using this set of techniques.The
proposed solution can eliminate all of the flaws in
the current system while still providing acceptable
6
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 29 Issue 3 – APRIL 2022.
III.PROPOSED WORK 3.3 FILTERING STOP WORDS, STEMMING
It is made of libraries and programmes for Stop words are terms that you don't want to read,
statistical natural language processing. therefore they're removed from your text as it's
Tokenization, classification, parsing, stemming, being processed. Stop words like 'in,' 'is,' and 'an'
tagging, and sematic reasoning are all possible are frequently employed as stop words because
with the libraries. A machine can understand they offer little to the text on their own. Stemming
human language using this set of techniques. The is a text processing technique that involves words
proposed solution can eliminate all of the flaws in to their root, or the most fundamental piece of the
the current system while still providing acceptable word. The names "making a difference" and
security and reducing human labour. "aide," for instance, have similar starting points.
Stemming permits you to focus on the
3.1 TOKENIZING
fundamental importance of a word as opposed to
Tokenizing text allows you to rapidly break it the complexities.
down into individual words or sentences.This will
enable you to work with smaller chunks of text
that are still cohesive and clear when removed
from their context.On a word-by-word basis,
tokenization:The building blocks of natural
language are words.They're the tiniest unit of
meaning that can nonetheless be comprehended
on their own.It's the first stage in converting
unstructured data into structured data that can be
more easily analysed.You'll tokenize words by
word and sentences by sentence while analysing
text.The following are the advantages of both
types of tokenization:You can detect common
words by tokenizing your text word by word. If
you look at a bunch of job ads, for example, you'll
Fig 2.USE CASE DIAGRAM
find that the word "Python" appears quite
frequently. This could suggest a significant 3.4 TKINTER PROGRAMMING
demand for Python.
Tkinter is Python's standard GUI library.
Whenever Python is used associated with Tkinter,
3.2BLOCK DIAGRAM:
it's fast and straightforward to create graphical
UIs. Tkinter accompanies a helpful thing. Tk GUI
tool stash's arranged connection point. Making a
GUI application with Tkinter's straightforward.
you must simply follow the means beneath. • In
your task, incorporate the Tkinter module. • For
the GUI programming, make the essential
window. • Add a minimum of one amongst the
previously mentioned gadgets to the GUI
program. • To react to each event that the user has
Fig 1: FLOW PROCESS FOR OUTPUT triggered, enter the main event loop.

7
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 29 Issue 3 – APRIL 2022.
IV HARDWARE REQUIREMENT CONCLUSION
Processor – i3, This project might be expanded by scanning the
Hard Disk – 5 GB, text and then converting it to legible text using
Memory – 4GB RAM 4.3. optical recognition. OCR (Optical Character
Recognition) is a software technology that
4.1 SOFTWARE REQUIREMENT recognises text (written or printed) within an
1. Software tools related to NLP: Python image file or physical document (such as a
Tensor Flow ,PyTorch ,Hugging Faces scanned document) and turns it into machine-
Transformers, Frontend ,Basic GUI tools, Spacy readable text for data processing. Text recognition
Streamlit or Gradio. is another term for it. optical The most important
element of OCR is connecting to a scanner and
scanning the document. The amount of variables
to consider while building OCR software is
minimised because scanning the paper
standardises the inputs. This phase also improves
the efficiency of the overall process by ensuring
that the document is perfectly aligned and sized.
Students can use this Automatic paper corrector to
enter their answers online, and the application will
Fig 3:OVERVIEW OF NLP[process of correct the paper and provide the necessary marks.
tokenizing] A computer programme that assesses and scores
written responses in the same way that a human
V. ADVANTAGES would. This software application verifies
subjective responses in an online exam and
1. Examiners become bored when they have to assigns marks to the user when the user's answer
check a large number of answer sheets; therefore, is validated.
the system reduces their workload by accurately
REFERENCES:
automating the manual checking process.
2.The system computes the score and displays the [1] Song, Shengli, Haitao Huang, and Tongxiao
findings immediately. Ruan. "Abstractive text summarization using
LSTM-CNN based deep learning." Multimedia
3. It eliminates human errors that are frequent Tools and Applications 78.1 (2019):857-875.
during manual checks.
[2] R. Mar ́ın, P. Sanz, O. Coltell, J. Inesta, F.
4.The technology generates a completely unbiased
outcome. Barber, and D. Corella, “Student-teacher
communication directed to computer-based
5.As a result, the technology eliminates human learning environments,” Displays, vol. 17, no. 3,
effort while saving time and resources. pp. 167–178, 2019.

[3] .Shirwandkar, Nikhil S., and Samidha

Kulkarni. "Extractive text summarization
usingdeep learning." 2018 Fourth International
Conference on Computing Communication

8
International Journal of Emerging Technology in Computer Science & Electronics (IJETCSE)
ISSN: 0976-1353 Volume 29 Issue 3 – APRIL 2022.
Control and Automation (ICCUBEA). IEEE,
2018.

[4] Standards for Web Applications on

Mobile:current state androadmap
https://www.w3.org/2015/01/mobile-web-app-
state/

[5] .R. M. McHugh, C. G. Horner, J. B. Colditz,

and T. L. Wallace, “Bridges and barriers
adolescent perceptions of student–teacher
relationships,” Urban Education, vol. 48, no. 1,
pp. 9– 43, 2013.

[6] Nenkova, Ani, and Kathleen McKeown. "A

survey of text summarization techniques."Mining
text data. Springer, Boston, MA, 2012. 43-76.

[7] Mihalcea, Rada, Courtney Corley, and Carlo

Strapparava. "Corpus-based and knowledgebased
measures of text semantic similarity." Aaai. Vol.
6. No. 2006. 2006.

[8] Proceedings of IEEE 24th International

Symposium on Fault- Tolerant Computing, J.D.
BrightComputer Science Department, Johns
Hopkins University, Baltimore, MD, USA-G.P.
Sullivanhttp://ieeexplore.ieee.org/xpl/articleDetail
s.jsp?tp=&arnumber=315646&queryText%3D
Ans wer+Checker.

[9] Proceedings., Fourteenth Annual International

Computer Software and Applications Conference,
S. Ramanna,J.F. Peters Kansas State University,
Manhattan, KS, USA
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp
=&arnumber=139342&queryText%3DAns
wer+Checker