Volume 8, Issue 7, July – 2023 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Development of a Tool for Quick Result Analysis

Mohit Sunil Pansare1; Gladys Gince Skariah1; Rhea Nilesh Bhalekar1; Garima Tripathi (Professor)1
1
Department of Artificial Intelligence and Data Science, Mumbai University, Maharashtra, India
1
Fr. Conceicao Rodrigues College of Engineering, Bandra, India

Abstract:- This research paper introduces the Stakeholders at Mumbai University may make data-
“Development of a tool for Quick Result Analysis for driven decisions to improve the quality of instruction and
Mumbai University”, a system created to provide encourage student achievement by integrating these
detailed reports based on the outcomes of a given branch capabilities into the system. This technology makes it
for a certain semester. Currently there are four branches possible to evaluate class performance in great detail,
of Engineering included in our system. The analyzer making the process of evaluating results more effective and
provides rapid and precise insights into student efficient.
performance, grade distributions and comparative
assessments across the courses by utilizing modern data The existing approach requires instructors or other
processing techniques. Mumbai University’s Engineering staff members to manually note and enter the grades issued
programme will see continual academic progress as by Mumbai University. This work is time-consuming and
students, teachers and administrators will have access to prone to human mistakes. Additionally, keeping track of the
useful information that will help to improve academic extra points and any grade adjustments is challenging. The
performance and decision-making procedures. input from the university alone is now insufficient for the
teachers to assess the performance criteria of the class as a
Keywords:- Result Analysis, Python, Report Generation, whole. As a result, a system is required for the upkeep and
Template, Data Visualization. updating of records from the text file the institution has
submitted.
I. INTRODUCTION
This research paper explores the usage of the
In today’s digital age, data plays a vital role in decision Zamzar[2] website for converting a PDF file into a CSV
making, analysis, and various other domains. However, data (Comma-Separated Values) file format. PDF (Portable
comes in different formats and converting it into a more Document Format) files are widely used for sharing and
accessible and versatile format is often necessary. presenting documents due to their consistent formatting
across different platforms. However, extracting data from a
The result analyzer system[1] retrieves raw result data PDF file for further analysis can be challenging, as the
for a specific branch and semester from the Mumbai format is primarily designed for display rather than data
University database and applies statistical algorithms to manipulation. On the other hand, CSV files offer a
process and analyze the data. The generated reports provide structured, tabular format that is easily readable by
key information such as the name of the branch, year of spreadsheet software and can be readily used for data
graduation, the selected semester and the class strength. analysis and manipulation.
Additionally, the system presents graphical representations
including a CGPA graph which allows one to understand II. METHODOLOGY
visually, the distribution of students’ grades.

Moreover, the system provides student statistics for the

entire class, which includes student percentage ranges, top
three students based on CGPA, the maximum marks
obtained in theory subjects, the number of students with a 10
CGPA, the number of students with KT’s (failures), and the
number of students scoring 60 in theory subjects.

These insightful statistics allow students to assess their

performance and understand their relative position within
the class. Faculty members can utilize this information to
identify subjects which may require additional support and
tailor their teaching strategies accordingly. Administrators
benefit from the system by gaining a comprehensive
overview of a class performance and can identify trends and
patterns that may inform the curriculum and assessment
improvements.
Fig 1 Process of Result Analysis and Report Generation

IJISRT23JUL2018 www.ijisrt.com 2132

Volume 8, Issue 7, July – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The entire process of result analysis and report DOCX to PDF, MP4 to AVI, JPEG to PNG, and more. The
generation can be divided into six broad steps. The first and service is accessible through a web browser, eliminating the
second step involve the result document being received by need for software installations. Zamzar offers a user-friendly
the user and it being converted into a CSV document by interface, making it easy for individuals and businesses to
them. Once the file is converted, the user now uploads the convert files quickly and efficiently. It provides convenience
document into the Result Analyzer tool. The tool reads the and flexibility for managing file formats and is widely used
document and stores the data into a Pandas DataFrame. At for its reliable and comprehensive conversion capabilities.
this step, the data is now ready for cleaning and analysis
operations. The next step involves cleaning of the data using  Hardware:
different methods and storing it in a new DataFrame. This The hardware used for this project can be any
cleaned data is then put through numerous analysis methods computer or laptop with a minimum of 4GB RAM and 1
that extract useful insights from the data. Once the data is GHz processor.
analyzed, a report is generated using the insights that
summarizes the entire result document using parameters like III. DATA DESCRIPTION
average CGPA, average percentage, subject toppers, class
topper, etc.

The development of the project involved the utilization

of various materials and technologies to create an efficient
and effective system for result analysis.

 The following Tools/Technologies were used in the

Development Process:

 Programming Languages: Python 3.7 or Higher.

 Data Analysis Libraries:

 The following Data Analytics Libraries were used:

 Pandas
 Numpy
 Scikit-Learn
 Seaborn
 Statistics
 Werkzeug
 Flask Fig 2 Example of the Result Document Received from the
University of Mumbai
 IDE: Visual Studio Code.
 Data Description:
 Web Development Framework: Flask.
Flask[3][4] is a lightweight and flexible web  Seat (Roll) number, Candidate name.
framework for building web applications in Python. With its  PP#: each # column indicates the marks obtained by the
minimalist design, Flask provides developers with a simple student in each subject. (# indicates a number: 1,2,3,..)
and intuitive approach to web development. It offers a  TH/IA/Tot/…: Bifurcations of the subject marks.
routing system for handling different URLs and HTTP
methods, a templating engine for dynamic HTML Table 1 Data Description
generation, and supports the integration of various Abbreviation Marks obtained in
extensions for added functionality. Flask promotes modular TH Theory Exam
development through blueprints, allowing developers to IA Internal Assessment
organize their applications into reusable components. It also TW Term Work
includes a built-in development server for easy testing and OR Oral/Viva Exam
debugging. With its versatility and ease of use, Flask is a PR Project
popular choice among developers looking to create web Tot Total
applications efficiently.
 ∑C, ∑CG: Summation of ‘Credits’ and ‘Product of
 PDF to CSV Conversion Tool: Zamzar Credits & Grades’.
Zamzar[5] is an online file conversion service that  GPA: ∑C / ∑CG
allows users to convert files from one format to another.  Remark: Pass(P) or Fail(F)
With Zamzar, users can convert a wide range of files,
including documents, images, audio, and video files. It
supports conversions between popular file formats such as

IJISRT23JUL2018 www.ijisrt.com 2133

Volume 8, Issue 7, July – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
IV. DATA PREPROCESSING  Read .csv into python script: Use the pandas. Read
csv()[9] function to read the csv data.
 Remove headers and footers from the data
 Correct indices
 Rearrange/reorder columns
 Extract roll numbers
 Remove empty/incorrect rows and columns
 Reassign roll numbers to the dataframe
 Create new dataframe with roll number, name, subject
marks and CGPI
 Add subject codes as column heads.

 Explanation of Data Preparation functions:

 `Remove_rows_columns(result, branch, semester)`:

Removes the initial rows and specific columns from the
DataFrame `result` based on the given `branch` and
`semester` values, effectively removing unnecessary
information.
Fig 3 Flowchart for Data Cleaning
 `Correct_rows_columns(result)`: Sorts the columns of
 Data Preprocessing Steps[5]: the DataFrame `result` in ascending order, resets the
index, and reassigns new indices to the rows to ensure a
clean and consistent data structure.
 Convert .pdf to .csv: The result document generated
from the University is in a .pdf format. PDF documents  `Find_junk_indices(result)`: Identifies the indices of
can directly be read in a python script[6] but the process rows in the DataFrame `result` that contain junk data
is tedious and the results are not always as desired. Some based on a specific condition, helping locate irrelevant or
libraries like PyPDF[7], PyPDF2[8] do exist that can be inconsistent data points.
used to read data directly from a PDF file but for our  `Cleared_result(result_cleared, result)`: Extracts and
particular application, the results were not as expected. A concatenates the relevant data from the DataFrame
more reliable way of reading the data from the PDF `result` into the `result_cleared` DataFrame by excluding
document is to first convert it to CSV (Comma the junk rows, resulting in a cleaned dataset.
Separated Values) and then reading it into a Python  `Extract_rollnumbers(result)`: Extracts roll numbers
script. There exist various tools that can be used to from the first column of the DataFrame `result`,
convert .pdf to .csv, one of which is Zamzar. There are removing NaN values, longer strings, and the string
various other tools that do the same thing but it was 'No.', and converting the remaining values to integers.
observed that Zamzar gives the best results in terms of  `Remove_empty_rows(result_cleared)`: Removes rows
least loss of data and least change in the formatting. from the DataFrame `result_cleared` where certain
columns have missing values (NaN), further refining the
dataset by eliminating incomplete data.
 `Reassign_rollnos(roll_numbers, result_cleared)`:
Reassign the extracted roll numbers obtained from
`extract_rollnumbers` to the first column of the
DataFrame `result_cleared`, ensuring accurate and
consistent roll number representation.
 `Add_header(result_cleared, sem)`: Adds a header to the
DataFrame `result_cleared` based on the given semester
`sem`, assigning appropriate column names to the
DataFrame for better data interpretation.

Fig 4 Format of Data after Conversion

No Preprocessing done.

IJISRT23JUL2018 www.ijisrt.com 2134

Volume 8, Issue 7, July – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 `Find_sub_mark_ranges(cleared_result, semester)`:
Determines the count of students who have scored above
60 marks in each subject for a given semester, as
specified in the 'subjects.xlsx' file. It returns a dictionary
where the keys are subject names, and the values are the
counts of students who scored above 60 marks.

VI. RESULTS AND DISCUSSION

Fig 5 Format of the Data after Preprocessing Steps

V. DATA ANALYSIS

 After the data has gone through all of the

aforementioned functions, it is finally in a form that can
be read and used for analysis purposes. The analysis
process that we have employed in this project include
plotting graphs, calculating aggregate percentage,
Fig 6 Input form of the Toll. The users Need to Upload their
CGPI, etc. The functions that have been created are
Result Document in .csv Format Along with other Details
explained below.
The product of the entire cleaning and analysis process
 ‘Count_students(percentages)`: Counts the number of
of the result data is a report[10] that is generated from the
students falling within specific percentage ranges, such
tool. This report is a concise review of the data projected in
as '>60', '(50, 60]', '(40, 50]', '(12, 40]', and 'Fail/KT',
the result of the students. This report is a simplified
based on their percentages. It returns a dictionary with
visualization of this information. This report can prove to be
the counts for each range.
of great help for teachers to understand the trends that are
 `Get_avg_cgpa(cleared_result)`: Calculates the average seen in marks of a particular class over the eight semesters
CGPA (Cumulative Grade Point Average) from the or engineering or the marks obtained by students in a
'cgpa' column of the `cleared_result` DataFrame. It also particular subject in case students of continuous batches are
returns the list of CGPA values for further analysis. struggling with the subject.
 `Get_avg_percentage(CGPA)`: Calculates the average
percentage based on the CGPA values provided. The
function converts CGPA to percentage using specific
conversion formulas and returns the average percentage
along with the list of percentage values.
 `Find_toppers(cleared_result)`: Finds the top three
students based on their CGPA from the `cleared_result`
DataFrame. It returns a list of dictionaries where each
dictionary contains the roll number and CGPA of a
topper.
 `Find_sub_max(cleared_result, semester)`: Retrieves the
maximum marks obtained in each subject for a given
semester, as specified in the 'subjects.xlsx' file. It returns
a dictionary where the keys are subject names and the
values are the corresponding maximum marks.
 `Find_kts(cleared_result)`: Counts the number of
students who have KT (Keep Terms) in any subject,
identified by missing CGPA values in the
`cleared_result` DataFrame. It returns the count of KT
students.

IJISRT23JUL2018 www.ijisrt.com 2135

Volume 8, Issue 7, July – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 7 Output given by the Tool this is the ‘Print Preview’ View of the Report that is Generated

VII. CONCLUSION [3 ]. Flask tutorial, https://www.tutorialspoint.com

/flask/index.html , Jan 2022
In conclusion, we have successfully developed a [4 ]. Adding HTML CSS in Flask,
system to process and analyze university result documents https://thinkinfi.com/flask-adding-html-and-css/
from the University of Mumbai using Python. The system [5 ]. Huong Ngo, 'How to clean data in python
has been designed to convert the result documents into a ‘https://towardsdatascience.com/how-to-clean-your-
standard format that can be easily read and analyzed using data-in-python-8f178638b98d ', July 2022
data analysis techniques and libraries in Python. The system [6 ]. How to Extract Data from PDF Files with Python
has been tested with a sample set of result documents and https://www.freecodecamp.org/news/extract-data-
has been found to be effective in extracting meaningful from-pdf-files-with-python/
insights from the data. [7 ]. PyPDF https://pypi.org/project/pypdf/
[8 ]. PyPDF2 documentation https://pypdf2.readthedocs.
The website allows users to upload their result io/en/3.0.0/
documents in CSV format, and the system automatically [9 ]. Pandas.read_csv https://pandas.pydata.org/docs/
converts them into a format that can be analyzed by Python. reference/api/pandas.read_csv.html
The website displays the analysis results in an easy-to- [10 ]. Gjorgjevikj, Dejan & Madjarov, Gjorgji & Chorbev,
understand format, including summary statistics, trends and Ivan & Angelovski, Martin & Georgiev, Marjan &
patterns, and visualizations. Dikovski, Bojan. (2011). ASGRT – Automated
Report Generation System. 369-376. 10.1007/978-3-
REFERENCES 642-19325-5_38.
[11 ]. Baradwaj, Brijesh & Pal, Saurabh. (2011). Mining
[1 ]. Miss. S. I. Tamboli, Miss. P. V. Tate, Miss. S. P. Educational Data to Analyze Students' Performance.
Kagwade, Miss. H. T. Magdum, and Prof. S. S. International Journal of Advanced Computer Science
Chavan, “Result Analysis,” International Journal for and Applications. 2. 63-69. 10.14569/IJACSA. 2011.
Research in Applied Science and Engineering 020609.
Technology, vol. 10, no. 6. International Journal for [12 ]. Lei, Yujiao & Deng, Jiqiu & Lin, Jian & Dick,
Research in Applied Science and Engineering Jeffrey & Lessani, M.Naser & Liu, Chaoyue. (2020).
Technology (IJRASET), pp. 821–826, Jun. 30, 2022. Research of Automatic Generation for Engineering
doi: 10.22214/ijraset.2022.43760. Geological Survey Reports Based on a Four-
[2 ]. Zamzar documentation Dimensional Dynamic Template. ISPRS
https://developers.zamzar.com/docs International Journal of Geo-Information. 9. 496.
10.3390/ijgi9090496.

IJISRT23JUL2018 www.ijisrt.com 2136