NLP - SMS Sentimental Analysis

Successfully created a model that accurately classifies text messages and provides insights into the sentiment present in the messages

About Project

This project endeavors to perform sentiment analysis on SMS chat messages by creating a precise text classification model for insights into message sentiments and investigating cross-country sentiment patterns

Project was executed in the following stages

Data Import & Overview
Planning Analysis & Choosing NLP Model/Technique
Text Preprocessing
Analysis
- Exploratory Data Analysis
- Sentiment Analysis
Feature Selection for classifier
Data Splitting & Model Selection
Message Classification/ Model Evaluation
Sentiment Distribution & Analysis by Country
Conclusion & Insights
Future Directions

Project was executed in the following stages

Data Import & Overview
Planning Analysis & Choosing NLP Model/Technique
Text Preprocessing
Analysis
- Exploratory Data Analysis
- Sentiment Analysis
Feature Selection for classifier
Data Splitting & Model Selection
Message Classification/ Model Evaluation
Sentiment Distribution & Analysis by Country
Conclusion & Insights
Future Directions

Project was executed in the following stages

Data Import & Overview
Planning Analysis & Choosing NLP Model/Technique
Text Preprocessing
Analysis
- Exploratory Data Analysis
- Sentiment Analysis
Feature Selection for classifier
Data Splitting & Model Selection
Message Classification/ Model Evaluation
Sentiment Distribution & Analysis by Country
Conclusion & Insights
Future Directions

Project was executed in the following stages

Data Import & Overview
Planning Analysis & Choosing NLP Model/Technique
Text Preprocessing
Analysis
- Exploratory Data Analysis
- Sentiment Analysis
Feature Selection for classifier
Data Splitting & Model Selection
Message Classification/ Model Evaluation
Sentiment Distribution & Analysis by Country
Conclusion & Insights
Future Directions

Project Stack

NLP, EDA, Python, ML

Project Stack

NLP, EDA, Python, ML

Role

Data Analyst, Data Scientist

Role

Data Analyst, Data Scientist

Services

EDA, viz, Language processing, Modeling

Services

EDA, viz, Language processing, Modeling

CodeBase

GitHub Repo

CodeBase

GitHub Repo

Jupyter Notebook

Nbviewer

Jupyter Notebook

Nbviewer

"Using a sentiment classification model, I assessed sentiments in the NUS SMS Corpus, revealing that 64.3% of messages were negative, and 35.8% were positive. Trinidad and Tobago ranked highest in positivity, while Lebanon scored highest in negativity. This analysis highlights sentiment patterns across countries."

Waqas Ahmad - Data Analyst

Take me back home