EXPLORATORY DATA ANALYSIS

PLAYSTORE USING
MACHINE LEARNING
SYNOPSIS
PROJECT

MASTER OF COMPUTER APPLICATIONS (III SEMESTER)

SUBMITTED BY
Lorens Mishra
Batch Year – 2023-2025
Enrollment No. – I2010307

PROJECT GUIDE – Dr. SARIKA YADAV

Centre of Computer Education & Training

Institute of Professional Studies
University of Allahabad, Prayagraj
Uttar Pradesh
 TABLE OF CONTENTS

Sr.No. TOPICS Page. NO.

1 INTRODUCTION 1

2 PROBLEM DEFINITION 2

3 MOTIVATION 3

4 OBJECTIVE 4

5 GOALS 5

6 REQUIREMENTS ANALYSIS &SPECIFICATION 6

7 ARCHITECTURE 7

9 DATA FLOW DIAGRAM 10

10 NAME OF ALGORITHMS 14

11 MILESTONES 16

12 MEETINGS WITH THE SUPERVISOR 17

13 REFERENCES 18

14 PLAGIARISM 19
 EXPLORATORY DATA ANALYSIS PLAYSTORE 1

INTRODUCTION

Provide background and context for the analysis .

Dataset Description: Details about the dataset including the source and structure. "The dataset

contains information on various apps from the Google Play Store, including attributes such as

app name, category, rating, review count, size, installs, type (free or paid), price, content rating,

and genre".

Context and Relevance: Explain the importance of analyzing this data. “Understanding app

ratings and their determinants is crucial for developers to improve their products and for

marketers to target their strategies effectively."

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 2

PROBLEM DEFINITION

Define the specific questions or problems being addressed.

Problem Statement: Articulate the issues to be explored. "The problem is to determine which

app features most significantly impact ratings and identify trends in app popularity across

different categories".

Scope and Boundaries: Define the scope of the analysis."The analysis will focus on apps with

at least 100 reviews and exclude those with missing or invalid ratings."

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 3

MOTIVATION

Exploratory Data Analysis (EDA) is essential for effective data analysis, aiming to reveal

underlying patterns and structures. It begins with “data understanding", where analysts explore

the dataset's patterns and layout. Next, a “quality check”cleans the data by addressing missing

values, inconsistencies, and outliers, ensuring accuracy. Analysts then seek “feature insights”

to uncover relationships and trends between variables. This leads to “hypothesis generation”,

where explanations for observed patterns are proposed and tested. Ultimately, EDA supports

“informed decision-making”, allowing organizations to make strategic choices and optimize

strategies based on clear, accurate, and insightful data analysis.

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 4

OBJECTIVE

Define the specific goals of the EDA(Exploratory Data Analysis).

Primary Objective: To uncover trends and patterns in app ratings and popularity.

Secondary Objectives: To analyze the relationship between app characteristics (e.g., size,

price, category…etc) and ratings, and to identify any anomalies or outliers.

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 5

GOALS

Outline specific targets for the EDA process.

• Understanding Data: Goals related to understanding the structure and quality of the

dataset.

• "Assess the completeness of the dataset and handle missing values".

• Data Visualization: Objectives for creating visual representations of the data.

• "Visualize rating distributions and the popularity of different app categories using

histograms and bar charts”.

• Statistical Analysis: Goals for performing statistical tests or deriving metrics.

• "Calculate correlation coefficients between app features and ratings".

• Data Cleaning: Targets for preparing the data for analysis.

• "Standardize the format of numerical columns and handle inconsistencies in

categorical data”.

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 6

REQUIREMENTS ANALYSIS & SPECIFICATION

Minimum Hardware Requirements:

Processor : Intel Core i3 or equivalent, or higher

RAM : 8 GB minimum (16 GB recommended for larger datasets)

Storage : 256 GB SSD or higher (more if handling large datasets)

Display : 1920x1080 resolution or higher (for better data visualization)

Graphics : Integrated graphics are usually sufficient; discrete GPU may help with

large datasets and visualizations

Minimum Software Requirements:

Operating System: Windows 10/11, macOS 10.15 etc

HTML/CSS/JS: Modern web browsers with developer tools (for creating and testing

interactive web applications)

Visual Studio Code: Visual Studio Code or Visual Studio (for coding and development)

Jupyter Notebook: Jupyter Notebook or JupyterLab (if needed for interactive analysis)

IDE: An Integrated Development Environment (IDE) like Visual Studio Code

or Visual Studio

Data Handling Tools: Excel or Google Sheets (for preliminary data handling)

Visualization Tools: Libraries such as D3.js, Chart.js, or other JavaScript-based visualization

libraries

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 7

ARCHITECTURE
1. Data Source

Play Store Data:

• Format: CSV, JSON, or Database.

• Data Elements: App Name, Category, Rating, Number of Reviews, Price,

Size, etc.

• Acquisition: Methods to obtain data (scraping, API, public datasets).

2. Backend (Optional)

Server:

• Technology: Node.js.

• Function: Handle requests, process data, and serve it to the frontend.

• Endpoints: Define API routes if applicable (e.g., /api/apps, /api/reviews).

3. Data Processing

Data Cleaning:

• Remove duplicates, handle missing values, standardize formats.

• Data Transformation:

• Aggregations, calculations (e.g., average ratings), and transformations.

• Data Storage:

• Temporary storage during processing or for a database.

4. Frontend

HTML/CSS/JS:

• Structure: Define layout and elements (e.g., navigation bar, filters).

• Style: Design with CSS for aesthetics and responsiveness.

• Functionality: Implement dynamic features with JavaScript (e.g., interactivity,

data fetching).

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 8

Visualization Libraries:

• D3.js: For complex, custom visualizations.

• Chart.js: For simple and interactive charts.

• Plotly: For interactive and advanced charts.

• Implementation: Integrate libraries to visualize data (e.g., bar charts, scatter

plots).

5. Deployment

Hosting Service:

• GitHub Pages: For static sites.

• Netlify: For static sites with continuous deployment.

• AWS: For scalable solutions (e.g., S3 for static files, EC2 for dynamic

servers).

• Heroku: For deploying Node.js applications and full-stack solutions.

6. User Interaction

UI/UX Design:

• User Experience: Ensure the site is intuitive and easy to navigate.

• User Interface: Design interactive elements and feedback mechanisms.

• Accessibility: Implement best practices to make the site usable for everyone.

7. Testing and Feedback

Testing:

• Unit Testing: Test individual components and functions.

• Integration Testing: Ensure components work together correctly.

• User Testing: Conduct usability tests to gather feedback on the user

experience.

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 9

Feedback:

• Collect: Use surveys, feedback forms, or direct user interviews.

• Analyze: Evaluate feedback to make improvements.

• Iterate: Implement changes based on feedback and retest.

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 10

DATA FLOW DIAGRAM

App: Name of the application.

Category: App classification (e.g., games, utilities).
Rating: Average user score.
Reviews: User feedback comments.
Size: Storage space needed.
Installs: Number of downloads. Attributes:
Type: Paid or free. Reviews, App, User,
Price: Cost of the app. Rating,
Content: App’s features and functionality. Translated_Review,
Genres: Content themes (e.g., action, puzzle).
Last Update: Most recent update date.
Sentiment
Current Version: App version number.
Android Version: Minimum required Android OS.

PLAY STORE DATA CSV REVIEWS DATA CSV

Import Data

Exploratory Data Analysis

Generate Reports

Outputs:
Insight, Charts&
Summary

User Reports

0 level data flow diagram

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 11

1 level data flow diagram

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 12

Play Store Data CSV Reviews Data CSV User’s reports

Handle Missing Values:
Identify Missing Values: Detect any missing or null values in the dataset.
Handle Missing Data:
Import data Impute: Fill in missing values using statistical measures like mean, median, or
mode.
Remove: Exclude records with missing values if imputation is not feasible.
Consistency Checks: Ensure data completeness after handling missing values.

Handle Missing Values

Normalize Data:
Normalize Numerical Data:
Scale Values: Adjust numerical data to a common scale (e.g., Min-Max scaling,
Z-score normalization).
Cleaned data Standardize Formats: Ensure consistent formatting for numerical data (e.g.,
decimal places).
Text Normalization: Standardize text fields (e.g., convert to lowercase, trim
spaces).

Normalize Data
Remove Duplicates:
Identify Duplicates: Detect duplicate records based on unique identifiers or key
attributes.
Remove Duplicates:
Remove Exact Duplicates: Eliminate identical duplicate records.
Normalized
Merge Near-Duplicates: Combine similar records if necessary for data
consistency.
Data Integrity: Validate that removing duplicates does not negatively impact
data quality.

Remove Dupdlicaatates

Final cleaned data

Cleaned data

DA
2 level data flow diagram

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 13

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 14

NAME OF ALGORITHMS

To analyze data in machine learning (ML), various algorithms can be employed depending on
the type of analysis you wish to perform. Here are some common types of analyses and the
corresponding ML algorithms:

1. Classification

Algorithm Name: Decision Tree Classifier

Pseudocode:

1. Load Data

2. Preprocess Data (e.g., handle missing values, encode categorical features)

3. Split Data into Training and Testing Sets

4. Initialize Decision Tree Classifier

5. Train the Model on Training Data

6. Evaluate Model on Testing Data

7. Make Predictions

8. Analyze Performance Metrics (e.g., accuracy, precision, recall)

2. Regression

Algorithm Name: Linear Regression

Pseudocode:

1. Load Data

2. Preprocess Data (e.g., handle missing values, normalize features)

3. Split Data into Training and Testing Sets

4. Initialize Linear Regression Model

5. Train the Model on Training Data

6. Evaluate Model on Testing Data

7. Make Predictions

8. Analyze Performance Metrics (e.g., Mean Absolute Error, R-squared)

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 15

3. Clustering

Algorithm Name: K-Means Clustering

Pseudocode:

1. Load Data

2. Preprocess Data (e.g., handle missing values, normalize features)

3. Initialize K-Means with number of clusters K

4. Assign Data Points to Closest Centroid

5. Update Centroids Based on Mean of Assigned Points

6. Repeat Steps 4-5 Until Convergence

7. Analyze Cluster Assignments and Centroids

4. Dimensionality Reduction

Algorithm Name: Principal Component Analysis (PCA)

Pseudocode:

1. Load Data

2. Preprocess Data (e.g., handle missing values, standardize features)

3. Initialize PCA with Number of Components

4. Fit PCA to Data

5. Transform Data to Reduced Dimension

6. Analyze Principal Components and Explained Variance

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 16

MILESTONES

Sr.No. Project Activity Estimated Start Date Estimated End Date

1 Project Allotment 26/07/2024 26/07/2024

2 Requirements Gathering 27/07/2024 09/08/2024

3 Design Phase 10/08/2024 30/08/2024

4 Development Phase 02/09/2024 25/09/2024

5 Testing Phase 26/09/2024 07/10/2024

6 End-User Validation 08/10/2024 19/10/2024

7 Final Deployment 20/10/2024 25/10/2024

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 17

MEETINGS WITH THE SUPERVISOR

Date of the Meet Comments by the Supervisor Signature of The Supervisor

30/07/2024 Reading Research Paper: Review

literature and identify gaps.

02/08/2024 Initial Proposal: Good structure; refine

objectives and methods.

14/20/27 Feedback Incorporated: Reviewed

revised objectives and methods;
ensure alignment with research goals.

20/08/2024 Design Phase Start

27/08/2024 Implementation Plan Start.

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 18

REFERENCES

Exploratory Data Analysis Playstore

1. Tukey, J. W. (1977). Exploratory Data Analysis. Addison-Wesley.
Data Visualization for Understanding
2. Cleveland, W. S., & McGill, R. (1984). Graphical perception: Theory,
experimentation, and application to the development of graphical methods. Journal of
the American Statistical Association, 79(387), 531-554.
https://doi.org/10.1080/01621459.1984.10478080
Statistical Data Analysis
3. Iglewicz, B., & Hoaglin, D. C. (1993). How to detect and handle outliers. Sage
Publications.
Machine Learning & Data Mining
4. Bishop, C. M. (2006). Pattern Recognition and Machine Learning. Springer.
5. Han, J., Kamber, M., & Pei, J. (2011). Data Mining: Concepts and Techniques.
Morgan Kaufmann.
6. Chen, M., Mao, S., & Liu, Y. (2014). Big data: A survey. Mobile Networks and
Applications, 19(2), 171-209. https://doi.org/10.1007/s11036-013-0489-0
7. Witten, I. H., Frank, E., & Hall, M. A. (2016). Data Mining: Practical Machine
Learning Tools and Techniques. Morgan Kaufmann.
Statistical Analysis with Missing Data
8. Little, R. J. A., & Rubin, D. B. (2002). *Statistical Analysis with Missing Data* (2nd
ed.). Wiley. https://doi.org/10.1002/9781119013563
Interactive Data Visualization
9. Heer, J., & Shneiderman, B. (2012). Interactive dynamics for visual analysis.
Communications of the ACM, 55(4), 45-54. https://doi.org/10.1145/2133806.2133821

CCET, IPS, University of Allahabad

 EXPLORATORY DATA ANALYSIS PLAYSTORE 19

PLAGIARISM REPORT

CCET, IPS, University of Allahabad