SlideShare a Scribd company logo
1 of 11
Download to read offline
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 5
Automated Question Paper Generator and
Answer Checker Using Information Retrieval
Approach
Mansi Palav#1
, Pranjal Singh#2
, Pratik Vishwakarma#3
, Pooja Pandit#4
, Prof. Neelam Phadnis#5
B.E, Department of Computer Engineering, Mumbai University
Mumbai, Maharashtra, India
Abstract
This is a challenging span due to the growth in
the field of computer science and demand we are
facing today. Hence, examinations play a vital role in
testing studentโ€™s performance. That is why it is
important to have a smart development question
model for growth of students as well as to test their
learning skills thereby keeping a check on student
performance. Generating an effective question paper
is a task of great importance for any educational
institute. The traditional method, where lecturers
manually prepare question paper, it is very tedious
and challenging. Our System allows faculty to
generate question papers with random questions,
which covers the chapters selected by the faculty. In
this study, we propose an automated scoring
approach for descriptive answers by usingJaro-
Winkler
Keywords - Question Paper, Answer Checker,
Randomize Algorithm, Jaro-Winkler.
I. INTRODUCTION
Generating an effective question paper is a task of
great importance for any educational institute. The
traditional method, where lecturers manually prepare
question paper, is very tedious and
challenging.Automated Question Paper Generator and
Answer Checker System can reduce time
consumption by replacing the conventional method of
question paper generation.The system fully automates
the process of question paper generation and selective
answer checker. The advanced system generates
question paper based on database such that all types
of questions and answers such as (MCQs, Theory
based, and objectives.) are stored in database. The
questions are randomly selected by the system from
database and generates a question paper such that it
covers all the chapters, which are selected, and allows
student to attempt the examination and to score
accordingly.Answer Checker checks the paper and
generates marks/score for the student by finding the
difference between model answer and student answer
by usingJaro-Winkler
II. LITERATURE REVIEW
This section presents the significant approaches in
the field of information retrieval,the techniques used
for text similarity and to understand the need for
automatic generation of question paper and answer
checking. Different models used to evaluate the
results of these techniques are also reviewed.
A. Background Study
1. Question Paper Generation
A literature survey was started to understand
the need for automatic generation of question paper.
As mentioned in, many existing LMS support tagging
feature but users may not utilize this feature fully.
The comparative study shows that Moodle is the best
LMSfor any educational institutionand to support
large number of users. But it allows user to define
only question type. Hence the questions in the
repository may have only basic tags or no tags at all.
So it becomes overhead for teachers to tag these
questions before using them. Properly tagged
questions can be efficiently retrieved from repository.
Hence it is very much necessary to tag the questions
before adding them to repository. A system which
offers generation of question paper using user given
input parameters considers only fixed range of values.
Our system not only supports upper and lowers
bounds for inputs but also supports more granular
level of topics than chapters and more question types
as compared to only three types offered by this
system. We are using automatically tagged question
repository as input instead of untagged questions.[5]
2. Answer Checker System
Different models used to evaluate the results are
reviewed.
a. Intelligent Essay Assessor (IEA)
It uses statistical model to compare descriptive
answers and checks the semantic
similaritybetweentwo answer sets. It is also used to
analyse and to give score for the essay type of
answers.
b. E-rater
E-rater is use to analyse the essay type of answers
and specifies syntactical and lexical issues in the text.
[3]
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 6
c. C-rater
C-rater is primarily use for assigning the marks
as per the studentโ€™s correct answers. It also deploys
the similar kind of words used in the answer, spelling
errors, syntax variations, whichare checked
automatically. [2]
3. Text to Text Similarity Approaches
The primary methods of similarity are classified as
knowledge-based similarity, corpus-based similarity,
and string-based similarity measures. [4]
a. Knowledge-based similarity
It applies text-to-text similarity to determine the
shortest path of similarity by detecting lexical chains
between pairs in a text using the WordNet hierarchy.
b. Corpus-based Similarity
It is used to find similarity between words
according to the sets called as corpus. It checks the
occurrences of the word in the particular answer.
c. String-based Similarity
String-based similarity evaluates the measures of
similarity or dissimilarity between two text strings.
There are two different types of string-based
algorithms for evaluating similarity between the
student's answer (SA) and the model answer (MA).
c.1Character-based similarity
Character-based similarity is used to find out the
distance between two strings and perform minimum
operations. Operations such as insertion, deletion,
substitution and transposition of a single character.
c.2Term-based similarity
Term-based similarity is the distance between
two itemsand sum of the distances of their
corresponding items
D. Cosine Similarity
The SA and MA are represented as vectors, where
student's answer and model answer are a set of terms;
each term has a weight, which reflects its importance
on that MA or SA. There are several ways to
calculate this weight, such as the Term Frequency-
Inverse Document Frequency (TF-IDF), where the
(TF) refers to the term frequency in the model answer,
and the IDF represents the importance of a term with
respect to the entire corpus. It is calculated by the
number of answers in the corpus divided by the
number of answers containing a term.
The cosine similarity measure is based on term
weighting scheme, which is the TF-IDF.
It is usually used as a weighting factor in information
retrieval and text mining. The formulas of TF, IDF
and TF-IDF are illustrated bellow as follows:
๐‘‡๐น =
๐‘›๐‘ข๐‘š๐‘๐‘’๐‘Ÿ ๐‘œ๐‘“ ๐‘œ๐‘๐‘๐‘ข๐‘Ÿ๐‘’๐‘›๐‘๐‘’๐‘  ๐‘œ๐‘“ ๐‘ก๐‘’๐‘Ÿ๐‘š ๐‘–๐‘› ๐‘Ž๐‘›๐‘ ๐‘ค๐‘’๐‘Ÿ
๐‘›๐‘ข๐‘š๐‘๐‘’๐‘Ÿ ๐‘œ๐‘“ ๐‘ก๐‘’๐‘Ÿ๐‘š๐‘  ๐‘–๐‘› ๐‘Ž๐‘›๐‘ ๐‘ค๐‘’๐‘Ÿ
๐ผ๐ท๐น =
๐‘
log๐‘›๐‘—+1
Where N is the total number of answers, NJ is the
number of answers containing the term.
๐‘‡๐น๐ผ๐ท๐น = ๐‘‡๐น โˆ— ๐ผ๐ท๐น
The main idea behind this model is to calculate the
weight of each term in each answer with respect to
the entire corpus.
The TF-IDF is used to compare a studentโ€™s answer
vector with a model answer vector using cosine
similarity measure. Cosine similarity measures the
cosine of the angle between two vectors. Two vectors
of attributes, SA and MA, the cosine similarity and
cosine (ฮธ) are represented by using a dot product and
magnitude as follows:
๐‘๐‘œ๐‘ ๐‘–๐‘›๐‘’ ๐‘ ๐‘–๐‘š๐‘–๐‘™๐‘Ž๐‘Ÿ๐‘–๐‘ก๐‘ฆ(๐‘†๐ด, ๐‘€๐ด) = ๐ท๐‘œ๐‘ก ๐‘ƒ๐‘Ÿ๐‘œ๐‘‘๐‘ข๐‘๐‘ก (๐‘†๐ด ,๐‘€๐ด )
๐‘†๐ด โˆ—|๐‘€๐ด |
Where Dot Product is:
Dot Product (SA, MA) =SA [0] *MA [0] +โ€ฆ..+ SA
[n] *MA [n]
And Distances | | and |MA| is defined as:
| | = [0] + [1] + โ€ฆ.. + [ ]
And
|MA| = MA [0] + MA [1] + โ€ฆ.. + MA [n]
After the cosine similarity between the model
answers and student answers are calculated, marks are
assigned.
4. Automated Tagging
The following four tags were identified for the
automatic generation of the question paper based on
Blooms Taxonomy.
a. Cognitive level Identification
It is the process to Understand, Percept, Evaluate,
Analyse, Recall and Create the functioning of the
system
b. Question-type Identification
Question-type is to identify the objective or
subjective type of questions.
Tags Values
Cognitive Level Recall, Understand, Apply,
Analyse, Evaluate, Create
Question Type Fill in the blanks, Multiple
choice, Match the
following, True False,
Answer in one word,
Definition
Content Topics and subtopic from
syllabus
Difficulty Level Low, Medium, High
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 7
c. Content Identification
It is to identify the topics as well as subtopics.
d. Difficulty- Level Identification
The difficulty level of a question depends on
concept involved, type of question, cognitive level
and decide whether it is low, high or medium.
III. PROPOSED SYSTEM
We are presenting an Automated Question Paper
Generator System and Answer Checker System that
can reduce time consumption by replacing the
traditional method of question paper generation.
System Architecture
B. Question paper generation system
The examiner will input the questions as per
their need in their respective database record if they
want to manipulate (Add, Delete or Change) the data
through a GUI. Questions will get updated as per the
function obtained. As this is a web-based application
Faculty can set the difficulty level, structure of the
answer, Chapters that they want to add according to
their subject, Exam Score for how many marks they
want to set their papers. The system will generate
random questions of the chapters specified by
extracting it from database by using algorithm the
questions will be well organized
1. Randomized Algorithm
Randomized Algorithm checks the duplicate
questions and use to display random questions. The
algorithm is as followed, N = total no. of questions
in the database. The randomized algorithm randomly
generates these questions
Step 1: Create an array of N locations.
Step 2: Generate random number.
Step 3: if (lock==0)
Store generated number.
Else
Compare the generated number with previous
number in array.
If matching value found, go to step 2;
Else
Store the no in next location.
Step 4: Repeat step 2 for N numbers.
Step 5: Select questions from DB matching with
values from array location one by one [1].
B. Answer Checker System
The Student appearing for the test needs to login
and appear for the test (Test will be of Subjective as
well as objectivepattern).The more accurate the
answers the more he/she will earn the marks.The
difference between
model answer and student answer is done by using
Jaro-Winkler and checks all the possible mistakes
that the student has made and gives the final score of
the test.
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 8
Answer Checking Process
1. Pre-processing
Pre-processing plays a very important role in answer
checking. The pre-processing operations needed are
segmentation, stop-words removal, normalization,
finding synonyms and extracting roots.
a. Tokenization
Tokenization is needed to identify the end of
each sentences. The sentences are ended with
various punctuation marks that can be dot (.),
comma (,), colons (:) etc.
b. Stop-words Removal
Stop-words are the most frequent words, which
we use in our answer such as prepositions, articles,
conjunctionsand are not much useful for automatic
scoring. Removal of these stop-words will improve
the performance of the system.
c. Normalization
Normalization is needed to modify the text to
make it definite as per needed by removing
unnecessary characters, non-alphanumeric characters
to improve the performance of the system.
d. Root Extraction
The sources of the keyword are extracted from
both student answer as well as model answer.
e. Jaro-Winkler distance
Jaro-Winkler is use to compare strings by
measuring the edit distance between two strings. The
minimum the distance, the more the strings are
similar. The score is normalize such that 0 is an
exact match and 1 is for similarity. Jaro-Winkler
similarity is 1- Jaro-Winkler distance.
The Jaro Similarity (simj) between two strings s1
and s2 .
sim ๐‘—
= 0 ๐‘–๐‘“๐‘š = 0
1
3
๐‘š
|๐‘ 1|
+
๐‘š
|๐‘ 2|
+
๐‘š โˆ’ ๐‘ก
๐‘š
๐‘œ๐‘กโ„Ž๐‘’๐‘Ÿ๐‘ค๐‘–๐‘ ๐‘’
Where |Si| = length of the strings
M = Number of matching characters
T = is half the number of transposition.
Two strings s1 and s2 are considered matching only
if it is same and not greater than
maxโก
( ๐‘ 1 , |๐‘ 2|)
2
โˆ’ 1
Jaro-Similarity uses prefix scale (p) which gives
ratings to the strings that are matched from very start.
So for this Jaro-Winkler similarity will be
Simw = simj + lp (1 โˆ’ simj)
Where
L is length of common prefix max up to 4 characters.
P is constant scaling factor to adjust upwards to have
common prefix. p should not be greater than 0.25
else the distance will become greater than 1.
Standard value of p = 0.1
Jaro-Winkler distance is dw.
dw = 1 โˆ’ simw
1.6 Automatic Scoring
The similarity measure value is converted into a
score using the following formula:
Mark = Similarity โ€“ Value * Mark
Similarityis calculated by JaroWinkler method.
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 9
IV. IMPLEMENTATION OF THE SYSTEM
1. Login Page
Login Page
In our system login page is common for admin, staff
as well as students. All the username and password
are by default inserted in the database uniquely for all
type of user (Admin, Faculty and Student).
The user no need to register for their session.Only
they need to authenticate themselves through login
and start with their session respectively.
2. Add Question and Answer
The faculty can insert additional questions in the
database if required.They only need to enter subject
name, Chapter Number of which the question belongs
to.Then enter the question as well answer as the
faculty need to give their model answer in the
database.They have to select marks (1, 2, and 5)for
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 10
that particular question that they need to
allocate.Select the difficulty level (Low, Medium, and
High) and the type of question (MCQ, FIB, Define,
and Brief).
3. Generate Question Paper
Faculty need to enter the Subject name that they need
to generate. Pattern of the paper (20, 40, 60,
80).Chapter names are displayed faculty need to
select at least twoChapters to generate the paper
pattern.
Exam Date is selected when the exam is
conducted.By clicking on generate question paper the
question paper is generated with random questions as
per the selected chapters.
3.1 Sample Paper Generated
Inputs
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 11
Output
4. Student- Start Exam
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 12
Student login and start its session when it needs to
appear for the exam.Student need to select the
question paper, which is hosted and then the exam
session gets started by clicking on Letโ€™s Begin.With
the help of marks of individual questions, total marks
is then calculated automatically and result is
displayed
5. Answer Sheet
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 13
5.1 Sample Answer Generated
Inputs
Output
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 14
After Submission
V CONCLUSIONS
We have implemented automatic question paper
generation using randomized algorithm. System can
generate 20, 40, 60 and 80 marks of question papers
with various types of questions like MCQโ€™s (Multiple
Choice Questions), Fill in the Blanks, Answer in one
sentence and Short answers also.The answerchecking
system generates the marks for all the individual
questions based on the similarity measure between
the student answer and the model answer. This
system can be used further in various colleges and
schools to reduce their work and effective time
utilization.
REFERENCES
[1] Kapil Naik, Shreyas Sule, Shruti Jadhav, Surya Pandey,
โ€œ Automatic Question Paper Generation System Using
Randomization Algorithmโ€, (IJETR) ISSN: 2321-
0869,Volume-2, Issue-12, December 2014.
[2] C. Leacock and M. Chodorow, โ€œC-Rater: Automated Scoring
of Short-Answer Question,โ€ Computers and the
Humanities, vol. 37,no. 4, pp. 389-405, 2003.
International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019
ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 15
[3] Y.Attali, J. Burstein, "Automated Essay Scoring with E-
Rater V.2.0", Paper presented at the Conference of the
International Association for Educational Assessment (IAEA)
Philadelphia, 13 - June.
[4] M.Mohler, R. Mihalcea, "Text-to-text semantic similarity for
automatic short answer grading", Proceedings of the 12th
Conference of the European Chapter of the Association for
Computational Linguistics Association for Computational
Linguistics, pp. 567-575, March.
[5] Rababah, H., & Al-Taani, A. T. (2017). An automated
scoring approach for Arabic short answers essay questions.
2017 8th International Conference on Information
Technology (ICIT).
[6] Noor Hasimah Ibrahim Teo, Nordin Abu Bakar and Moamed
Rezduan Abd Rashid, โ€œRepresenting Examination Question
Knowledge into Genetic Algorithmโ€, IEEE Global
Engineering Education Conference (EDUCON), 2014.
[7] A Statistical approach for Automatic Text Summarization by
Extraction Munesh Chandra Vikrant Gupta Santosh Kr. Paul
2011 International Conference on Communication Systems
and Network Technologies
[8] Deshpande, R., Vaze, K., Rathod, S., & Jarhad, T, (2014).
"Comparative Study of Document Similarity Algorithms and
Clustering Algorithms for Sentiment Analysis", International
Journal of Emerging Trends & Technology in Computer
Science, Volume 3, Issue 5, September-October 2014.
[9] A New Fuzzy Information Retrieval System Based on User
Preference Model Dae-Won Kim and Kwang H. Lee
Department of Computer Science Korea Advanced Institute
of Science and Technology Kusung-dong, Yusung-gu, 305-
701, Daejon, Korea.

More Related Content

Similar to Automated Question Paper Generator And Answer Checker Using Information Retrieval Approach

DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
ijcsit
ย 
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
AIRCC Publishing Corporation
ย 
An Adaptive Evaluation System to Test Student Caliber using Item Response Theory
An Adaptive Evaluation System to Test Student Caliber using Item Response TheoryAn Adaptive Evaluation System to Test Student Caliber using Item Response Theory
An Adaptive Evaluation System to Test Student Caliber using Item Response Theory
Editor IJMTER
ย 
A scoring rubric for automatic short answer grading system
A scoring rubric for automatic short answer grading systemA scoring rubric for automatic short answer grading system
A scoring rubric for automatic short answer grading system
TELKOMNIKA JOURNAL
ย 

Similar to Automated Question Paper Generator And Answer Checker Using Information Retrieval Approach (20)

4 de47584
4 de475844 de47584
4 de47584
ย 
An Adaptive Approach for Subjective Answer Evaluation
An Adaptive Approach for Subjective Answer EvaluationAn Adaptive Approach for Subjective Answer Evaluation
An Adaptive Approach for Subjective Answer Evaluation
ย 
IRJET- Implementation of Automatic Question Paper Generator System
IRJET- Implementation of Automatic Question Paper Generator SystemIRJET- Implementation of Automatic Question Paper Generator System
IRJET- Implementation of Automatic Question Paper Generator System
ย 
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
ย 
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
DEVELOPING A FRAMEWORK FOR ONLINE PRACTICE EXAMINATION AND AUTOMATED SCORE GE...
ย 
MACHINE LEARNING TOOLBOX
MACHINE LEARNING TOOLBOXMACHINE LEARNING TOOLBOX
MACHINE LEARNING TOOLBOX
ย 
AI Based Question Answering System
AI Based Question Answering SystemAI Based Question Answering System
AI Based Question Answering System
ย 
IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)
IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)
IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)
ย 
IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)
IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)
IRJET- Personalized E-Learning using Learnerโ€™s Capability Score (LCS)
ย 
Vikalp - Automatic multiple choice questions generator
Vikalp - Automatic multiple choice questions generatorVikalp - Automatic multiple choice questions generator
Vikalp - Automatic multiple choice questions generator
ย 
An Adaptive Evaluation System to Test Student Caliber using Item Response Theory
An Adaptive Evaluation System to Test Student Caliber using Item Response TheoryAn Adaptive Evaluation System to Test Student Caliber using Item Response Theory
An Adaptive Evaluation System to Test Student Caliber using Item Response Theory
ย 
A New Active Learning Technique Using Furthest Nearest Neighbour Criterion fo...
A New Active Learning Technique Using Furthest Nearest Neighbour Criterion fo...A New Active Learning Technique Using Furthest Nearest Neighbour Criterion fo...
A New Active Learning Technique Using Furthest Nearest Neighbour Criterion fo...
ย 
IRJET- Sentimental Analysis for Studentsโ€™ Feedback using Machine Learning App...
IRJET- Sentimental Analysis for Studentsโ€™ Feedback using Machine Learning App...IRJET- Sentimental Analysis for Studentsโ€™ Feedback using Machine Learning App...
IRJET- Sentimental Analysis for Studentsโ€™ Feedback using Machine Learning App...
ย 
Data Clustering in Education for Students
Data Clustering in Education for StudentsData Clustering in Education for Students
Data Clustering in Education for Students
ย 
K0176495101
K0176495101K0176495101
K0176495101
ย 
A Study on Learning Factor Analysis โ€“ An Educational Data Mining Technique fo...
A Study on Learning Factor Analysis โ€“ An Educational Data Mining Technique fo...A Study on Learning Factor Analysis โ€“ An Educational Data Mining Technique fo...
A Study on Learning Factor Analysis โ€“ An Educational Data Mining Technique fo...
ย 
A017640107
A017640107A017640107
A017640107
ย 
AUTOMATIC QUESTION GENERATION USING NATURAL LANGUAGE PROCESSING
AUTOMATIC QUESTION GENERATION USING NATURAL LANGUAGE PROCESSINGAUTOMATIC QUESTION GENERATION USING NATURAL LANGUAGE PROCESSING
AUTOMATIC QUESTION GENERATION USING NATURAL LANGUAGE PROCESSING
ย 
Analysis of sms feedback and online feedback using sentiment analysis for ass...
Analysis of sms feedback and online feedback using sentiment analysis for ass...Analysis of sms feedback and online feedback using sentiment analysis for ass...
Analysis of sms feedback and online feedback using sentiment analysis for ass...
ย 
A scoring rubric for automatic short answer grading system
A scoring rubric for automatic short answer grading systemA scoring rubric for automatic short answer grading system
A scoring rubric for automatic short answer grading system
ย 

More from Sheila Sinclair

More from Sheila Sinclair (20)

Visual Medium Advertisement Analysis Es. Online assignment writing service.
Visual Medium Advertisement Analysis Es. Online assignment writing service.Visual Medium Advertisement Analysis Es. Online assignment writing service.
Visual Medium Advertisement Analysis Es. Online assignment writing service.
ย 
Personal Essay Template. The Per. Online assignment writing service.
Personal Essay Template. The Per. Online assignment writing service.Personal Essay Template. The Per. Online assignment writing service.
Personal Essay Template. The Per. Online assignment writing service.
ย 
Steps On How To Write An Essay. Steps To Writing An
Steps On How To Write An Essay. Steps To Writing AnSteps On How To Write An Essay. Steps To Writing An
Steps On How To Write An Essay. Steps To Writing An
ย 
Free Writing Paper Cliparts, Download Free Writing Pa
Free Writing Paper Cliparts, Download Free Writing PaFree Writing Paper Cliparts, Download Free Writing Pa
Free Writing Paper Cliparts, Download Free Writing Pa
ย 
Descriptive Paragraph On Nature. Essay On Nature. 2
Descriptive Paragraph On Nature. Essay On Nature. 2Descriptive Paragraph On Nature. Essay On Nature. 2
Descriptive Paragraph On Nature. Essay On Nature. 2
ย 
.Mickey And Minnie Stationary Writing Paper
.Mickey And Minnie Stationary Writing Paper.Mickey And Minnie Stationary Writing Paper
.Mickey And Minnie Stationary Writing Paper
ย 
Sample Seminar Report. Online assignment writing service.
Sample Seminar Report. Online assignment writing service.Sample Seminar Report. Online assignment writing service.
Sample Seminar Report. Online assignment writing service.
ย 
Writing A Speech For Your Presentation - Soalanrule
Writing A Speech For Your Presentation - SoalanruleWriting A Speech For Your Presentation - Soalanrule
Writing A Speech For Your Presentation - Soalanrule
ย 
Synthesis Journal Example. Synthesis Exa
Synthesis Journal Example. Synthesis ExaSynthesis Journal Example. Synthesis Exa
Synthesis Journal Example. Synthesis Exa
ย 
Self-Introduction Essay - 6 Examples, Form
Self-Introduction Essay - 6 Examples, FormSelf-Introduction Essay - 6 Examples, Form
Self-Introduction Essay - 6 Examples, Form
ย 
Owl Writing Paper Teaching Resources. Online assignment writing service.
Owl Writing Paper Teaching Resources. Online assignment writing service.Owl Writing Paper Teaching Resources. Online assignment writing service.
Owl Writing Paper Teaching Resources. Online assignment writing service.
ย 
Admission Essay Outline Of A Descriptive Essay
Admission Essay Outline Of A Descriptive EssayAdmission Essay Outline Of A Descriptive Essay
Admission Essay Outline Of A Descriptive Essay
ย 
Essay Size - Dimension, Inches, Mm, Cms, Pixel
Essay Size - Dimension, Inches, Mm, Cms, PixelEssay Size - Dimension, Inches, Mm, Cms, Pixel
Essay Size - Dimension, Inches, Mm, Cms, Pixel
ย 
How To Write An Introduction Paragraph For An Essay About A Boo
How To Write An Introduction Paragraph For An Essay About A BooHow To Write An Introduction Paragraph For An Essay About A Boo
How To Write An Introduction Paragraph For An Essay About A Boo
ย 
008 Self Introduction Essay Sample Help G
008 Self Introduction Essay Sample Help G008 Self Introduction Essay Sample Help G
008 Self Introduction Essay Sample Help G
ย 
30 High School Scholarship Essay Examples Exampl
30 High School Scholarship Essay Examples Exampl30 High School Scholarship Essay Examples Exampl
30 High School Scholarship Essay Examples Exampl
ย 
How To Write Explanation Paper Allsop Author
How To Write Explanation Paper Allsop AuthorHow To Write Explanation Paper Allsop Author
How To Write Explanation Paper Allsop Author
ย 
14 Financial Analysis Templates - AI, PSD, Google Docs, Apple Pages
14 Financial Analysis Templates - AI, PSD, Google Docs, Apple Pages14 Financial Analysis Templates - AI, PSD, Google Docs, Apple Pages
14 Financial Analysis Templates - AI, PSD, Google Docs, Apple Pages
ย 
Top 10 Best Term Paper Writing Services, Professional Paper Writing
Top 10 Best Term Paper Writing Services, Professional Paper WritingTop 10 Best Term Paper Writing Services, Professional Paper Writing
Top 10 Best Term Paper Writing Services, Professional Paper Writing
ย 
College Essay Persuasive Essay Conclusion Format
College Essay Persuasive Essay Conclusion FormatCollege Essay Persuasive Essay Conclusion Format
College Essay Persuasive Essay Conclusion Format
ย 

Recently uploaded

ๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝ
ๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝ
ๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝ
ไธญ ๅคฎ็คพ
ย 
Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...
Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...
Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...
Nguyen Thanh Tu Collection
ย 
The basics of sentences session 4pptx.pptx
The basics of sentences session 4pptx.pptxThe basics of sentences session 4pptx.pptx
The basics of sentences session 4pptx.pptx
heathfieldcps1
ย 

Recently uploaded (20)

INU_CAPSTONEDESIGN_แ„‡แ…ตแ„†แ…ตแ†ฏแ„‡แ…ฅแ†ซแ„’แ…ฉ486_แ„‹แ…ฅแ†ธแ„…แ…ฉแ„ƒแ…ณแ„‹แ…ญแ†ผ แ„‡แ…กแ†ฏแ„‘แ…ญแ„Œแ…กแ„…แ…ญ.pdf
INU_CAPSTONEDESIGN_แ„‡แ…ตแ„†แ…ตแ†ฏแ„‡แ…ฅแ†ซแ„’แ…ฉ486_แ„‹แ…ฅแ†ธแ„…แ…ฉแ„ƒแ…ณแ„‹แ…ญแ†ผ แ„‡แ…กแ†ฏแ„‘แ…ญแ„Œแ…กแ„…แ…ญ.pdfINU_CAPSTONEDESIGN_แ„‡แ…ตแ„†แ…ตแ†ฏแ„‡แ…ฅแ†ซแ„’แ…ฉ486_แ„‹แ…ฅแ†ธแ„…แ…ฉแ„ƒแ…ณแ„‹แ…ญแ†ผ แ„‡แ…กแ†ฏแ„‘แ…ญแ„Œแ…กแ„…แ…ญ.pdf
INU_CAPSTONEDESIGN_แ„‡แ…ตแ„†แ…ตแ†ฏแ„‡แ…ฅแ†ซแ„’แ…ฉ486_แ„‹แ…ฅแ†ธแ„…แ…ฉแ„ƒแ…ณแ„‹แ…ญแ†ผ แ„‡แ…กแ†ฏแ„‘แ…ญแ„Œแ…กแ„…แ…ญ.pdf
ย 
factors influencing drug absorption-final-2.pptx
factors influencing drug absorption-final-2.pptxfactors influencing drug absorption-final-2.pptx
factors influencing drug absorption-final-2.pptx
ย 
2024_Student Session 2_ Set Plan Preparation.pptx
2024_Student Session 2_ Set Plan Preparation.pptx2024_Student Session 2_ Set Plan Preparation.pptx
2024_Student Session 2_ Set Plan Preparation.pptx
ย 
Danh saฬch HSG Boฬฃฬ‚ moฬ‚n caฬ‚ฬp truฬ›oฬ›ฬ€ng - Caฬ‚ฬp THPT.pdf
Danh saฬch HSG Boฬฃฬ‚ moฬ‚n caฬ‚ฬp truฬ›oฬ›ฬ€ng - Caฬ‚ฬp THPT.pdfDanh saฬch HSG Boฬฃฬ‚ moฬ‚n caฬ‚ฬp truฬ›oฬ›ฬ€ng - Caฬ‚ฬp THPT.pdf
Danh saฬch HSG Boฬฃฬ‚ moฬ‚n caฬ‚ฬp truฬ›oฬ›ฬ€ng - Caฬ‚ฬp THPT.pdf
ย 
Bแป˜ LUYแป†N NGHE TIแบพNG ANH 8 GLOBAL SUCCESS Cแบข Nฤ‚M (Gแป’M 12 UNITS, Mแป–I UNIT Gแป’M 3...
Bแป˜ LUYแป†N NGHE TIแบพNG ANH 8 GLOBAL SUCCESS Cแบข Nฤ‚M (Gแป’M 12 UNITS, Mแป–I UNIT Gแป’M 3...Bแป˜ LUYแป†N NGHE TIแบพNG ANH 8 GLOBAL SUCCESS Cแบข Nฤ‚M (Gแป’M 12 UNITS, Mแป–I UNIT Gแป’M 3...
Bแป˜ LUYแป†N NGHE TIแบพNG ANH 8 GLOBAL SUCCESS Cแบข Nฤ‚M (Gแป’M 12 UNITS, Mแป–I UNIT Gแป’M 3...
ย 
Capitol Tech Univ Doctoral Presentation -May 2024
Capitol Tech Univ Doctoral Presentation -May 2024Capitol Tech Univ Doctoral Presentation -May 2024
Capitol Tech Univ Doctoral Presentation -May 2024
ย 
size separation d pharm 1st year pharmaceutics
size separation d pharm 1st year pharmaceuticssize separation d pharm 1st year pharmaceutics
size separation d pharm 1st year pharmaceutics
ย 
ๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝ
ๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝ
ๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝๆœƒ่€ƒ่‹ฑ่ฝ
ย 
Dementia (Alzheimer & vasular dementia).
Dementia (Alzheimer & vasular dementia).Dementia (Alzheimer & vasular dementia).
Dementia (Alzheimer & vasular dementia).
ย 
Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...
Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...
Tแป”NG HแปขP Hฦ N 100 ฤแป€ THI THแปฌ T๏ปฟแปT NGHIแป†P THPT VแบฌT Lร 2024 - Tแปช CรC TRฦฏแปœNG, TRฦฏ...
ย 
Behavioral-sciences-dr-mowadat rana (1).pdf
Behavioral-sciences-dr-mowadat rana (1).pdfBehavioral-sciences-dr-mowadat rana (1).pdf
Behavioral-sciences-dr-mowadat rana (1).pdf
ย 
An Overview of the Odoo 17 Discuss App.pptx
An Overview of the Odoo 17 Discuss App.pptxAn Overview of the Odoo 17 Discuss App.pptx
An Overview of the Odoo 17 Discuss App.pptx
ย 
UNIT โ€“ IV_PCI Complaints: Complaints and evaluation of complaints, Handling o...
UNIT โ€“ IV_PCI Complaints: Complaints and evaluation of complaints, Handling o...UNIT โ€“ IV_PCI Complaints: Complaints and evaluation of complaints, Handling o...
UNIT โ€“ IV_PCI Complaints: Complaints and evaluation of complaints, Handling o...
ย 
The basics of sentences session 4pptx.pptx
The basics of sentences session 4pptx.pptxThe basics of sentences session 4pptx.pptx
The basics of sentences session 4pptx.pptx
ย 
Post Exam Fun(da) Intra UEM General Quiz - Finals.pdf
Post Exam Fun(da) Intra UEM General Quiz - Finals.pdfPost Exam Fun(da) Intra UEM General Quiz - Finals.pdf
Post Exam Fun(da) Intra UEM General Quiz - Finals.pdf
ย 
philosophy and it's principles based on the life
philosophy and it's principles based on the lifephilosophy and it's principles based on the life
philosophy and it's principles based on the life
ย 
Application of Matrices in real life. Presentation on application of matrices
Application of Matrices in real life. Presentation on application of matricesApplication of Matrices in real life. Presentation on application of matrices
Application of Matrices in real life. Presentation on application of matrices
ย 
Removal Strategy _ FEFO _ Working with Perishable Products in Odoo 17
Removal Strategy _ FEFO _ Working with Perishable Products in Odoo 17Removal Strategy _ FEFO _ Working with Perishable Products in Odoo 17
Removal Strategy _ FEFO _ Working with Perishable Products in Odoo 17
ย 
REPRODUCTIVE TOXICITY STUDIE OF MALE AND FEMALEpptx
REPRODUCTIVE TOXICITY  STUDIE OF MALE AND FEMALEpptxREPRODUCTIVE TOXICITY  STUDIE OF MALE AND FEMALEpptx
REPRODUCTIVE TOXICITY STUDIE OF MALE AND FEMALEpptx
ย 
Navigating the Misinformation Minefield: The Role of Higher Education in the ...
Navigating the Misinformation Minefield: The Role of Higher Education in the ...Navigating the Misinformation Minefield: The Role of Higher Education in the ...
Navigating the Misinformation Minefield: The Role of Higher Education in the ...
ย 

Automated Question Paper Generator And Answer Checker Using Information Retrieval Approach

  • 1. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 5 Automated Question Paper Generator and Answer Checker Using Information Retrieval Approach Mansi Palav#1 , Pranjal Singh#2 , Pratik Vishwakarma#3 , Pooja Pandit#4 , Prof. Neelam Phadnis#5 B.E, Department of Computer Engineering, Mumbai University Mumbai, Maharashtra, India Abstract This is a challenging span due to the growth in the field of computer science and demand we are facing today. Hence, examinations play a vital role in testing studentโ€™s performance. That is why it is important to have a smart development question model for growth of students as well as to test their learning skills thereby keeping a check on student performance. Generating an effective question paper is a task of great importance for any educational institute. The traditional method, where lecturers manually prepare question paper, it is very tedious and challenging. Our System allows faculty to generate question papers with random questions, which covers the chapters selected by the faculty. In this study, we propose an automated scoring approach for descriptive answers by usingJaro- Winkler Keywords - Question Paper, Answer Checker, Randomize Algorithm, Jaro-Winkler. I. INTRODUCTION Generating an effective question paper is a task of great importance for any educational institute. The traditional method, where lecturers manually prepare question paper, is very tedious and challenging.Automated Question Paper Generator and Answer Checker System can reduce time consumption by replacing the conventional method of question paper generation.The system fully automates the process of question paper generation and selective answer checker. The advanced system generates question paper based on database such that all types of questions and answers such as (MCQs, Theory based, and objectives.) are stored in database. The questions are randomly selected by the system from database and generates a question paper such that it covers all the chapters, which are selected, and allows student to attempt the examination and to score accordingly.Answer Checker checks the paper and generates marks/score for the student by finding the difference between model answer and student answer by usingJaro-Winkler II. LITERATURE REVIEW This section presents the significant approaches in the field of information retrieval,the techniques used for text similarity and to understand the need for automatic generation of question paper and answer checking. Different models used to evaluate the results of these techniques are also reviewed. A. Background Study 1. Question Paper Generation A literature survey was started to understand the need for automatic generation of question paper. As mentioned in, many existing LMS support tagging feature but users may not utilize this feature fully. The comparative study shows that Moodle is the best LMSfor any educational institutionand to support large number of users. But it allows user to define only question type. Hence the questions in the repository may have only basic tags or no tags at all. So it becomes overhead for teachers to tag these questions before using them. Properly tagged questions can be efficiently retrieved from repository. Hence it is very much necessary to tag the questions before adding them to repository. A system which offers generation of question paper using user given input parameters considers only fixed range of values. Our system not only supports upper and lowers bounds for inputs but also supports more granular level of topics than chapters and more question types as compared to only three types offered by this system. We are using automatically tagged question repository as input instead of untagged questions.[5] 2. Answer Checker System Different models used to evaluate the results are reviewed. a. Intelligent Essay Assessor (IEA) It uses statistical model to compare descriptive answers and checks the semantic similaritybetweentwo answer sets. It is also used to analyse and to give score for the essay type of answers. b. E-rater E-rater is use to analyse the essay type of answers and specifies syntactical and lexical issues in the text. [3]
  • 2. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 6 c. C-rater C-rater is primarily use for assigning the marks as per the studentโ€™s correct answers. It also deploys the similar kind of words used in the answer, spelling errors, syntax variations, whichare checked automatically. [2] 3. Text to Text Similarity Approaches The primary methods of similarity are classified as knowledge-based similarity, corpus-based similarity, and string-based similarity measures. [4] a. Knowledge-based similarity It applies text-to-text similarity to determine the shortest path of similarity by detecting lexical chains between pairs in a text using the WordNet hierarchy. b. Corpus-based Similarity It is used to find similarity between words according to the sets called as corpus. It checks the occurrences of the word in the particular answer. c. String-based Similarity String-based similarity evaluates the measures of similarity or dissimilarity between two text strings. There are two different types of string-based algorithms for evaluating similarity between the student's answer (SA) and the model answer (MA). c.1Character-based similarity Character-based similarity is used to find out the distance between two strings and perform minimum operations. Operations such as insertion, deletion, substitution and transposition of a single character. c.2Term-based similarity Term-based similarity is the distance between two itemsand sum of the distances of their corresponding items D. Cosine Similarity The SA and MA are represented as vectors, where student's answer and model answer are a set of terms; each term has a weight, which reflects its importance on that MA or SA. There are several ways to calculate this weight, such as the Term Frequency- Inverse Document Frequency (TF-IDF), where the (TF) refers to the term frequency in the model answer, and the IDF represents the importance of a term with respect to the entire corpus. It is calculated by the number of answers in the corpus divided by the number of answers containing a term. The cosine similarity measure is based on term weighting scheme, which is the TF-IDF. It is usually used as a weighting factor in information retrieval and text mining. The formulas of TF, IDF and TF-IDF are illustrated bellow as follows: ๐‘‡๐น = ๐‘›๐‘ข๐‘š๐‘๐‘’๐‘Ÿ ๐‘œ๐‘“ ๐‘œ๐‘๐‘๐‘ข๐‘Ÿ๐‘’๐‘›๐‘๐‘’๐‘  ๐‘œ๐‘“ ๐‘ก๐‘’๐‘Ÿ๐‘š ๐‘–๐‘› ๐‘Ž๐‘›๐‘ ๐‘ค๐‘’๐‘Ÿ ๐‘›๐‘ข๐‘š๐‘๐‘’๐‘Ÿ ๐‘œ๐‘“ ๐‘ก๐‘’๐‘Ÿ๐‘š๐‘  ๐‘–๐‘› ๐‘Ž๐‘›๐‘ ๐‘ค๐‘’๐‘Ÿ ๐ผ๐ท๐น = ๐‘ log๐‘›๐‘—+1 Where N is the total number of answers, NJ is the number of answers containing the term. ๐‘‡๐น๐ผ๐ท๐น = ๐‘‡๐น โˆ— ๐ผ๐ท๐น The main idea behind this model is to calculate the weight of each term in each answer with respect to the entire corpus. The TF-IDF is used to compare a studentโ€™s answer vector with a model answer vector using cosine similarity measure. Cosine similarity measures the cosine of the angle between two vectors. Two vectors of attributes, SA and MA, the cosine similarity and cosine (ฮธ) are represented by using a dot product and magnitude as follows: ๐‘๐‘œ๐‘ ๐‘–๐‘›๐‘’ ๐‘ ๐‘–๐‘š๐‘–๐‘™๐‘Ž๐‘Ÿ๐‘–๐‘ก๐‘ฆ(๐‘†๐ด, ๐‘€๐ด) = ๐ท๐‘œ๐‘ก ๐‘ƒ๐‘Ÿ๐‘œ๐‘‘๐‘ข๐‘๐‘ก (๐‘†๐ด ,๐‘€๐ด ) ๐‘†๐ด โˆ—|๐‘€๐ด | Where Dot Product is: Dot Product (SA, MA) =SA [0] *MA [0] +โ€ฆ..+ SA [n] *MA [n] And Distances | | and |MA| is defined as: | | = [0] + [1] + โ€ฆ.. + [ ] And |MA| = MA [0] + MA [1] + โ€ฆ.. + MA [n] After the cosine similarity between the model answers and student answers are calculated, marks are assigned. 4. Automated Tagging The following four tags were identified for the automatic generation of the question paper based on Blooms Taxonomy. a. Cognitive level Identification It is the process to Understand, Percept, Evaluate, Analyse, Recall and Create the functioning of the system b. Question-type Identification Question-type is to identify the objective or subjective type of questions. Tags Values Cognitive Level Recall, Understand, Apply, Analyse, Evaluate, Create Question Type Fill in the blanks, Multiple choice, Match the following, True False, Answer in one word, Definition Content Topics and subtopic from syllabus Difficulty Level Low, Medium, High
  • 3. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 7 c. Content Identification It is to identify the topics as well as subtopics. d. Difficulty- Level Identification The difficulty level of a question depends on concept involved, type of question, cognitive level and decide whether it is low, high or medium. III. PROPOSED SYSTEM We are presenting an Automated Question Paper Generator System and Answer Checker System that can reduce time consumption by replacing the traditional method of question paper generation. System Architecture B. Question paper generation system The examiner will input the questions as per their need in their respective database record if they want to manipulate (Add, Delete or Change) the data through a GUI. Questions will get updated as per the function obtained. As this is a web-based application Faculty can set the difficulty level, structure of the answer, Chapters that they want to add according to their subject, Exam Score for how many marks they want to set their papers. The system will generate random questions of the chapters specified by extracting it from database by using algorithm the questions will be well organized 1. Randomized Algorithm Randomized Algorithm checks the duplicate questions and use to display random questions. The algorithm is as followed, N = total no. of questions in the database. The randomized algorithm randomly generates these questions Step 1: Create an array of N locations. Step 2: Generate random number. Step 3: if (lock==0) Store generated number. Else Compare the generated number with previous number in array. If matching value found, go to step 2; Else Store the no in next location. Step 4: Repeat step 2 for N numbers. Step 5: Select questions from DB matching with values from array location one by one [1]. B. Answer Checker System The Student appearing for the test needs to login and appear for the test (Test will be of Subjective as well as objectivepattern).The more accurate the answers the more he/she will earn the marks.The difference between model answer and student answer is done by using Jaro-Winkler and checks all the possible mistakes that the student has made and gives the final score of the test.
  • 4. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 8 Answer Checking Process 1. Pre-processing Pre-processing plays a very important role in answer checking. The pre-processing operations needed are segmentation, stop-words removal, normalization, finding synonyms and extracting roots. a. Tokenization Tokenization is needed to identify the end of each sentences. The sentences are ended with various punctuation marks that can be dot (.), comma (,), colons (:) etc. b. Stop-words Removal Stop-words are the most frequent words, which we use in our answer such as prepositions, articles, conjunctionsand are not much useful for automatic scoring. Removal of these stop-words will improve the performance of the system. c. Normalization Normalization is needed to modify the text to make it definite as per needed by removing unnecessary characters, non-alphanumeric characters to improve the performance of the system. d. Root Extraction The sources of the keyword are extracted from both student answer as well as model answer. e. Jaro-Winkler distance Jaro-Winkler is use to compare strings by measuring the edit distance between two strings. The minimum the distance, the more the strings are similar. The score is normalize such that 0 is an exact match and 1 is for similarity. Jaro-Winkler similarity is 1- Jaro-Winkler distance. The Jaro Similarity (simj) between two strings s1 and s2 . sim ๐‘— = 0 ๐‘–๐‘“๐‘š = 0 1 3 ๐‘š |๐‘ 1| + ๐‘š |๐‘ 2| + ๐‘š โˆ’ ๐‘ก ๐‘š ๐‘œ๐‘กโ„Ž๐‘’๐‘Ÿ๐‘ค๐‘–๐‘ ๐‘’ Where |Si| = length of the strings M = Number of matching characters T = is half the number of transposition. Two strings s1 and s2 are considered matching only if it is same and not greater than maxโก ( ๐‘ 1 , |๐‘ 2|) 2 โˆ’ 1 Jaro-Similarity uses prefix scale (p) which gives ratings to the strings that are matched from very start. So for this Jaro-Winkler similarity will be Simw = simj + lp (1 โˆ’ simj) Where L is length of common prefix max up to 4 characters. P is constant scaling factor to adjust upwards to have common prefix. p should not be greater than 0.25 else the distance will become greater than 1. Standard value of p = 0.1 Jaro-Winkler distance is dw. dw = 1 โˆ’ simw 1.6 Automatic Scoring The similarity measure value is converted into a score using the following formula: Mark = Similarity โ€“ Value * Mark Similarityis calculated by JaroWinkler method.
  • 5. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 9 IV. IMPLEMENTATION OF THE SYSTEM 1. Login Page Login Page In our system login page is common for admin, staff as well as students. All the username and password are by default inserted in the database uniquely for all type of user (Admin, Faculty and Student). The user no need to register for their session.Only they need to authenticate themselves through login and start with their session respectively. 2. Add Question and Answer The faculty can insert additional questions in the database if required.They only need to enter subject name, Chapter Number of which the question belongs to.Then enter the question as well answer as the faculty need to give their model answer in the database.They have to select marks (1, 2, and 5)for
  • 6. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 10 that particular question that they need to allocate.Select the difficulty level (Low, Medium, and High) and the type of question (MCQ, FIB, Define, and Brief). 3. Generate Question Paper Faculty need to enter the Subject name that they need to generate. Pattern of the paper (20, 40, 60, 80).Chapter names are displayed faculty need to select at least twoChapters to generate the paper pattern. Exam Date is selected when the exam is conducted.By clicking on generate question paper the question paper is generated with random questions as per the selected chapters. 3.1 Sample Paper Generated Inputs
  • 7. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 11 Output 4. Student- Start Exam
  • 8. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 12 Student login and start its session when it needs to appear for the exam.Student need to select the question paper, which is hosted and then the exam session gets started by clicking on Letโ€™s Begin.With the help of marks of individual questions, total marks is then calculated automatically and result is displayed 5. Answer Sheet
  • 9. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 13 5.1 Sample Answer Generated Inputs Output
  • 10. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 14 After Submission V CONCLUSIONS We have implemented automatic question paper generation using randomized algorithm. System can generate 20, 40, 60 and 80 marks of question papers with various types of questions like MCQโ€™s (Multiple Choice Questions), Fill in the Blanks, Answer in one sentence and Short answers also.The answerchecking system generates the marks for all the individual questions based on the similarity measure between the student answer and the model answer. This system can be used further in various colleges and schools to reduce their work and effective time utilization. REFERENCES [1] Kapil Naik, Shreyas Sule, Shruti Jadhav, Surya Pandey, โ€œ Automatic Question Paper Generation System Using Randomization Algorithmโ€, (IJETR) ISSN: 2321- 0869,Volume-2, Issue-12, December 2014. [2] C. Leacock and M. Chodorow, โ€œC-Rater: Automated Scoring of Short-Answer Question,โ€ Computers and the Humanities, vol. 37,no. 4, pp. 389-405, 2003.
  • 11. International Journal of Computer Trends and Technology ( IJCTT ) - Volume 67 Issue 4 โ€“ April 2019 ISSN: 2231 โ€“ 2803 http://www.ijcttjournal.org Page 15 [3] Y.Attali, J. Burstein, "Automated Essay Scoring with E- Rater V.2.0", Paper presented at the Conference of the International Association for Educational Assessment (IAEA) Philadelphia, 13 - June. [4] M.Mohler, R. Mihalcea, "Text-to-text semantic similarity for automatic short answer grading", Proceedings of the 12th Conference of the European Chapter of the Association for Computational Linguistics Association for Computational Linguistics, pp. 567-575, March. [5] Rababah, H., & Al-Taani, A. T. (2017). An automated scoring approach for Arabic short answers essay questions. 2017 8th International Conference on Information Technology (ICIT). [6] Noor Hasimah Ibrahim Teo, Nordin Abu Bakar and Moamed Rezduan Abd Rashid, โ€œRepresenting Examination Question Knowledge into Genetic Algorithmโ€, IEEE Global Engineering Education Conference (EDUCON), 2014. [7] A Statistical approach for Automatic Text Summarization by Extraction Munesh Chandra Vikrant Gupta Santosh Kr. Paul 2011 International Conference on Communication Systems and Network Technologies [8] Deshpande, R., Vaze, K., Rathod, S., & Jarhad, T, (2014). "Comparative Study of Document Similarity Algorithms and Clustering Algorithms for Sentiment Analysis", International Journal of Emerging Trends & Technology in Computer Science, Volume 3, Issue 5, September-October 2014. [9] A New Fuzzy Information Retrieval System Based on User Preference Model Dae-Won Kim and Kwang H. Lee Department of Computer Science Korea Advanced Institute of Science and Technology Kusung-dong, Yusung-gu, 305- 701, Daejon, Korea.